US3865894A - Process for paraffin-olefin alkylation - Google Patents

Abstract

Olefin-paraffin alkylate is prepared by contacting C3-C9 monoolefin with C4-C6 iso-paraffin (which can, if desired, be prepared in situ from other paraffin isomers) in liquid phase with a substantially anhydrous acidic crystalline aluminosilicate zeolite, and stopping such contacting after substantial alkylation (which can include self-alkylation of the isoparaffin) has occurred but before the weight rate of production of unsaturated hydrocarbon becomes greater than the weight rate of production of saturated hydrocarbon. The degree of conversion of olefins and paraffins to saturate products can be increased by use of halide adjuvants containing bromine, chlorine or fluorine.

Description

United States Patent Kirsch et a1.

14 1 Feb. 11,1975

[ 1 PROCESS FOR PARAFFIN-OLEFIN 3,308,069 3/1967 Wadlinger et a1. 208/120 I 3,312,615 4/1967 Cramer et a1 260/683.43 ALKYLAT ON 3,352,796 11/1967 Kimberlin, Jr. et al 208/120 [75] In entors: Francis illi m r y 3,354,078 11/1967 Miale et a1 208/120 David S. Barmby, Media; John D. Potts sprmgfield' an of Primary Examiner-De1bert E. Gantz [73] Assignee: Sun Oil Company of Pennsylvania, 145mm"! Crasanakis Phil d l hi P;; Attorney, Agent, or FirmGeorge L. Church; .1. Filed: Mar. 1968 Edward Hess; Barry A. B1sson Related Application Data Olefin-paraffin alkylate is prepared by contacting [63] Continuation-impart of Ser. No. 581,129, Aug. 25, C -C monoolefin with C -C iso-paraffin (which can.

19661 abandonedif desired, be prepared in situ from other paraffin isomers) in liquid phase with a substantially anhydrous [52] [1.8. C1. 260/683.43 idi ry talline alumino-silicate zeolite, and stopping [5 Cl. 6 3 uch contacting after substantia alkylation can 1 Field of Search 260/683-43 633-4, 683-64; include self-alkylation of the isoparaffin) has occurred 208/120; 252/455 but before the weight rate of production of unsaturated hydrocarbon becomes greater than the weight 1 References Clted rate of production of saturated hydrocarbon. The de- UNITED STATES PATENTS gree of conversion of olefins and paraffins to saturate 2,834,818 5/1958 Schmerling et al. 260/683.43 P F can beiincreased by use Of halide adjuvams 3,236,762 2/1966 R8130 et a1. 260/683.4 conta mg m ne, hlor ne or fluorme. 3,251,902 5/1966 Garwood et a1. 260/683.43 3,264,208 8/1966 Plank et a] 208/120 3 Chm, 5 Drawmg Figures kc B YIELD C PARAFFlN/% C -ENE CHARGE ISOBUTANE-BUTENE-ZALKYLATION WITH Ac1D1c "Y" ZEOLITE CATALYST 2O QTIRRED AUTOCLAVEI ao=c 25o PSIG. rec ADDED TIME, Hr.

YIELD C PARAFFlN/"la C '-ENE CHARGE YIELD C UNSATURATES/WT C -ENE CHG.

PATENTEU 3.865.894

SHEET 4 F I62 a 0.28 E 0: LL] 7 8 0.24 g

Y C| u 6 A I 0.20 P (5 4 A I 0.: 6 Q H 2 A" u 2 2 I 0|2 m I 0 4 B 0.08 o

o 2 3 4 s a I FIG.1

ISOBUTANE-BUTENE-ZALKYLATION WITH ACIDIC "Y" ZEOLITE CATALYST 2O STIRRED AUTOCLAVEI C,

250 Plsle. TBC ADDED O o I 2 3 4 5 6 7 TIME, Hr.

INVENTORS we. MWMMQr ATTORNEY MEMEU H915 3.865.894

SHEET 26F 4 FEED INERT FEED PARAFFIN GAS OLEFIN v 5 2 FEED SECTION /l7 Ml do co 4 3 X m v 22 Egza 2s FIG 3 INVENTORS FRANCIS WILLIAM KIRSCH BY DAVID S. BARMBY JOHN D. POTTS I ATTORNEY FQIENIEU I I575 1.885.894

SHEET 3 OF 4 REACTOR SECTION HIGH LOW P P 57 LOW I U PRESSURE ADJUVANTS 62 63 (HCl,H O,ETC.I

- 34' L r\Z- 33 l REACTION FEED PRODUCT 55 To 35 Y REACTOR 39 36 I 40 42 HIGH U PRESSURE 3 47 INERT GAS 48 FIG. 4

INVENTORS FRANCIS WILLIAM KIRSCI' Y DAVID S. BARMBY JOHN D.POTTS uLQ. MWQA ATTORNEY PIIIEIIIEII EB 1 I915 3.865.894

SHEEI LI, 0F 4 PRODUCT RECOVERY SECTION 37 REACTION PRODUCT so r 82 W R479 83 r r r $98 LIQUID. PRODUCT (ALKYLATE) INVENTORS FRANCIS WILLIAM KIRSCH DAVID s. BARMBY BY JOHN D. POTTS ATTORNEY 1 PROCESS FOR PARAFFIN-OLEFIN ALKYLATION CROSS REFERENCE TO RELATED APPLICATION This application is a continuation-in-part of Ser. No. S8l,l29, filed Aug. 25, 1966 by the present inventors, now abandoned and assigned to the Sun Oil Company to whom the present application is also assigned.

BACKGROUND OF THE INVENTION This invention relates to the production of normally liquid, saturated hydrocarbons, useful in gasoline blending, by reacting isoparaffins with olefins in liquid phase in the presence of a substantially anhydrous cystalline alumino-silicate zeolite. These zeolites, in bydrated form, are chemically characterized by the empirical formula, xM. X(AlO2).y(SiO ).zH O, wherein M is H and/or an equivalent valence of metalcontaining cations and x, y and z are integers, the ratio x/y being usually (but not necessarily) from 1 .0 to 0.2.

The invention also comprises the use of halide catalyst adjuvants to increase the degree of conversion of olefins and paraffins to alkylate. Novel catalysts and novel alkylate products are also within the scope of our invention and are described hereinafter.

The invention will be described more particularly in connection with a process for the preparation of an olefin-paraffin alkylate comprising:

a. contacting C -C monoolefin with C -C isoparaffin and with a substantially anhydrous acidic crystalline alumino-silicate zeolite, at a temperature below the critical temperature of the lowest boiling hydrocarbon reactant and at a pressure such that the reactants are at least partially in liquid phase, and

b. stopping such contacting after substantial alkylation has occurred but before the weight rate of production of unsaturated hydrocarbon becomes greater than the weight rate of production of saturated hydrocarbon.

It is usual in the laboratory to effect paraffin-olefin alkylation by means of strong acids, such as AlCl HF, and H SQg however, in petroleum refining, aluminum trichloride catalysis is accompanied by equipment corrosion, cracking, sludge formation, and other side reactions and has not proven economical, except for ethylene-isoparaffin alkylation (for which there is no other satisfactory catalytic process). Commerical isoparaffin alkylation with C -C olefins to produce high octane components for gasoline utilizes either H 80 or HF as the catalyst. These acids, although very effective as alkylation catalysts are highly corrosiye and are potentially hazardous to workmen; therefore, strict safety procedures must be adhered to in their use. In addition, as for example in U.S. 2,359,119, alkylation processes utilizing sulfuric acid normally require reaction temperatures from about l5C.; therefore, costly cooling is required.

The art has long sought to find a process for paraffinolefin alkylation utilizing catalyst which does not have the above-mentioned disadvantages possessed by AlCl HF, or H 804. In particular, a heterogeneous process utilizing a solid alkylation catalyst has been sought by the art since processes using a liquid catalyst require that the acid and feed hydrocarbons, which are mutually immiscible, be kept in homogeneous suspension. Such homogenization requires expensive agitation devices and consumes much power. In addition,

emulsions can be formed which are different and costly to break."

Heretofore, attempts to effect paraffin-olefin alkylation utilizing a solid catalyst have had but little success. In all such published attempts, the bromine numbers of the reported products have been high, indicating that olefin polymerization (or some other competing reaction) has occurred to a substantial extent rather than the hoped-for paraffin-olefin alkylation.

Aromatic-olefin alkylation utilizing a crystalline, alumino-silicate catalyst has been reported (e.g., U.S. Pat. No. 2,904,607). However, the art has long recognized that olefin-aromatic alkylation and olefinparaffin alkylation are very different chemical reactions and that there is no equivalency between processes for these two dissimilar combinations.

Although the chemistry of the alkylation reactions is complex and not completely understood, it is very probable that one major reason for the nonequivalency of olefin-aromatic alkylation and olefinparaffin alkylations is that aromatic-olefin alkylation is dependent upon proton ejection from the intermediate carbonium ion whereas proton ejection in paraffin olefin reactions yields unsaturated products. In contrast, the production of saturated paraffin-olefin alkylate requires hydride transfer to the intermediate carbonium ion and, thus, a catalyst and process condition which favor hydride transfer over proton ejection.

In U.S. Pat. No. 3,251,902, claims are directed to the alkylation of C, and C isoparaffins with C,,-C olefins. The examples, however, show only ethylene or propylene as feed olefins. Ethylene-paraffin reactions do not teach how to alkylate paraffins with C -C olefins. With propylene and isobutane, the reported characteriszations of the products in the examples of U.S. Pat. No.

3,251,902 are what would be expected in view of the above-discussed prior art, particularly the art dealing with olefin-paraffin alkylations. That is, with propylene and isobutane as the feed hydrocarbons, the reported products of the alkylation process of U.S. Pat. No. 3.25l,902 are highly unsaturated, and, in fact, the inventors admit that such unsaturation indicates that polymerization of the olefin is more pronounced than alkylation. They attribute such polymer production in their process to the thermal stability of the feed olefin and the high concentration of acid sites in certain of their zeolite catalysts. In no case do they recognize the above-discussed importance of hydride transfer in the production of saturated rather than unsaturated alkylate.

SUMMARY OF THE INVENTION As is further disclosed herein, we have discovered a process for the production of highly saturated alkylate from C3-C monoolefins which requires not only a catalyst with a large number of acid sites of sufficient strength for hydride transfer but which also utilizes conditions which favor hydride transfer, such as introducing the olefin to the reactor in the liquid phase and in intimate admixture with C -C isoparaffin and, preferably, controlling the addition of the feed olefin such that the unreacted olefin in the hydrocarbon-catalyst reaction mixture is maintained at less than 12 mole percent (and most preferably less than 7 mole percent) based on the total paraffin content of the reaction mixture. In contrast to the process U.S. Pat. No. 3,251,902, we have also discovered that better yields of superior products can be obtained in our process with the more highly acid catalysts than with the less acid catalysts.

The use of dilute olefin feed streams has been suggested in conjunction with processes for olefinaromatic alkylation (e.g., US. Pat. No. 3,251,897); however, there is no prior art suggestion or teaching of our liquid phase, acid zeolite-catalyzed paraffin-olefin alkylation process wherein C -C feed monoolefins are intimately premixed with feed paraffin, nor of our control of the concentration of unreacted olefin in the reaction mixture with acidic zeolite catalysts in order to obtain paraffin-olefin alkylation rather than polymerization.

We have further discovered that the production of saturated alkylation (and self-alkylation) products rather than unsaturated hydrocarbons is effected when the mean residence time (or retention or holding time) of the reaction mixture with the catalyst is in the range of 0.05 to 0.5 hours per (gram of hydrocarbon per gram of catalyst). More preferably, the mean residence time is 0.1-0.4 hours.

Preferably, there is present in the reactor mixture a halide adjuvant containing fluorine, chlorine or bromine. Particularly favorable alkylation conditions involve a temperature in the range of 25 to 120C. (more preferably 50l00C.) and sufficient pressure to maintain a substantial part of each reactant in liquid phase (since mixed gas-liquid phase conditions are more likely to result in poor mixture of the feed paraffin and feed olefin, thus promoting olefin homopolymerization).

In general, unsaturated reaction products are indicative of olefin homopolymerization rather than paraffinolefin alkylation. True alkylation, including isoparaffin self-alkylation, produces saturated hydrocarbons, However, apart from these primary reactions, i.e., polymerization and alkylation, the acidic zeolites catalyze many secondary reactions, such as cracking, disproportionation, and aromatization. These reactions can transform unsaturated polymer to saturated hydrocarbons and can cause unsaturated hydrocarbons to be formed.

In our process we desire to maximize the production of saturated hydrocarbons, and particularly the trimethylpentanes, since, as is shown by C. R. Cupit, et al., Petrol. Chem. Eng., Dec., 1961 at pages 204-5, these have the more desirable antiknock characteristics, such as low sensitivity. With C olefins and isobutane, therefore, we desire to maximize the percentage yield, based on the weight of olefin charged, of C; saturates and the yield of trimethylpentanes.

One means of maximizing this percent yield of C saturates is to prepare suitably active acid catalysts which favor hydride transfer and to conduct the process under conditions (as described herein) such that, as primary reactions, paraffin-olefin alkylation and isoparaffin self-alkylation are favored over polymerization. The reaction should also be controlled in a manner which reduces the occurrence of undesirable reactions.

Another method of increasing C saturate yield is to use reaction conditions which favor the secondary reaction of octanes with isobutane to form C C and C paraffins.

We have found that with isobutane and butene-2 feeds, even when such secondary reactions have occurred to some extent, the molecular ratio of trimethylpentanes/dimethylhexanes (TMP/DMH in the reaction mixture indicates the relative degree to which the primary reaction was alkylation or polymerization. That is, dimethylhexanes arise from olefin dimerization followed by hydride transfer; whereas, trimethylpentane formation is largely dependent upon paraffinolefin combination to form a carbonium ion species followed by hydride abstraction from the isobutane. Therefore, the higher the ratio TMP/DMI-I the greater the effect of alkylation reactions, in contrast to olefin homopolymerization.

In general, the greater the tendency for the catalyst to initiate hydride transfer, the greater the ratio TMPIDMI-I For example, with isobutane-butene-Z feeds, AlCl catalyst (at 30C.) produces reaction products where TMP/DMH is about 2/l. With l-IF or with H2804, alkylates can be obtained with TMP DMl-I ratios about 8/1. commercial H 50, alkylates have TlViP/DIVIPI ratios between 3/1 and 6/1.

Prior art publications, such as those previously referred to, fail to teach how to use solid catalyst to obtain a C -C isoparaffin C -C olefin reaction product in which saturated products predominate rather than unsaturate. They also do not teach how to obtain products having high TMP/DMH ratios. In contrast, we have discovered, and disclose herein, a paraffin-olefin alkylation process which utilizes acidic cystalline zeolite catalyst to obtain a predominantly saturated product in which the TMP/DMI-I, ratio can be greater than 7, We further teach how our process can be used to produce a paraffin-olefin alkylate which contains only negligible amounts of unsaturated reaction products. We have also discovered that maximum conversion of olefin to saturated products can be obtained with this process under conditions where the reaction mixture contains but a minor amount of unsaturated reaction products.

We have further discovered, unexpectedly (in view of prior art), that the conversion of olefin to saturated products can also be increased by the use of halide catalyst adjuvants. That is, the incorporation in the reaction mixture of small amounts of certain halides containing bromine, chlorine or fluorine allows the production of as much as percent more saturated hydrocarbons from the same quantity of olefin reactant than can be produced under the same reaction conditions in the absence of the halide adjuvant.

BRIEF DESCRIPTION OF THE DRAWINGS In the attached drawings, FIG. 1 illustrates the variation in the yield of C paraffins based on the olefin reactant as the reaction time is increased in our process.

FIG. 2 illustrates (by the solid curve) the weight percent of C; unsaturates produced, based on the olefin reactant (here, butene-Z), as the reaction time is increased. Also illustrated in FIG. 2, by the broken curve, is the moles of n-butane produced per mole of butene-Z converted, as a function of time.

FIGS. 3, 4 and 5 illustrate an apparatus which is particularly useful for effecting the continuous production of alkylate from a paraffin-olefin feed, utilizing the process of the present invention. This apparatus comprises three sections, a feed-mixing section, FIG. 3, a stirred, slurry reactor section, FIG. 4, and a product recovery section, FIG. 5.

In the feed section (FIG. 3), C -C monoolefin is admixed with C -C isoparaffin and transported, as by pumping, to the reactor section. The reactor section comprises a pressure vessel with means for maintaining the catalyst in suspension, such as a turbine mixer and baffles, means for introducing feed hydrocarbon and adjuvants such as a halide promoter, and means for separating a catalyst-free portion of the reaction mixture and transporting it from the reactor to the product recovery section.

The reactor section (FIG. 4), also includes means for maintaining sufficient pressure in the reaction vessel to insure that the reactants and the reaction mixture are in liquid phase, means (such as a differential pressure cell) for maintaining the liquid inside the reactor at a desired level and means for maintaining the reactor at the desired temperature (such as by a water jacket and heating coils).

The product recovery section (FIG. 5), comprises means for cooling the reaction mixture, means for separating gases (such as unreacted feed isoparaffin) from the desired liquid alkylate, and means for recycling unreacted feed hydrocarbons.

FURTHER DESCRIPTION OF THE INVENTION Although our paraffin-olefin alkylation process requires the control of many inter-related process variables, such as reaction temperature, mixing rate, catalyst selection, concentration and preparation, the more critical conditions in our process are control of the maximum ratio of unreacted C -C olefin to C -C isoparaffin in the reaction mixture, that all the feed components (whether olefin or paraffin) must be well intermixed and kept at least partially (preferably predominantly) in the liquid phase, and most important, contact time of the olefin-containing reaction mixture with the catalyst must be controlled closely. A key measure of such contact is the mean residence time of the olefin-containing reaction mixture with the catalyst, in such units as mean hours per (gram of hydrocarbon per gram of catalyst). The criticality of these conditions with respect to obtaining a predominantly saturated alkylate has not been taught or suggested by the prior art.

The major reaction conditions which must be controlled in order to obtain a satisfactory product are inter-related in that they all influence the probability that a given molecule of reactant will collide with an active site of the zeolite catalyst and form a desirable carbonium ion. These major variables are the initial concentrations of isoparaffin and catalyst, the paraffin/olefin feed ratio, the intimacy of the paraffin-olefin premixing, the feed rate, agitation rate, temperature, pressure, catalyst type and particle size, and the contact time (or for a continuous stirred reactor, the mean residence time).

We have discovered that, within the operable ranges discussed herein, if the other variables are fixed (especially those which determine the probability that a given molecule of reactant will collide with an active site of the catalyst), a certain minimum contact time or induction period is required for substantial paraffinolefin alkylation to occur. There is also a maximum contact time beyond which the quantity of saturated alkylate in the reaction mixture no longer increases, and unsaturated reaction products start to build up. We have found, in our process, that the production of saturated alkylation products is effected when the mean residence time of the reaction mixture with the catalyst is in the range of 0.05 to 0.5 hours.

The published art contains no disclosure, suggestion or even speculation of such criticality in a zeolitecatalyzed paraffin-olefin reaction system. The best mode of practice of our process utilizes this discovery to maximize conversion of our prarftin-olefin feed to desirable saturatedproducts and to minimize, or effectively eliminate, the production of unsaturated hydrocarbons.

A preferred embodiment of our process for the preparation of an olefin-paraffin alkylate comprises the following steps:

1. contacting C -C monoolefin in admixture with C -C isoparaffin having a tertiary carbon atom, at a temperature below the critical temperature of the lowest boiling reactant and at a pressure such that the reactants are in liquid phase, with a substantially anhydrous acidic crystalline alumino-silicate zeolite;

2. controlling the addition of the said olefin reactant such that the amount of unreacted feed olefin in the reaction mixture is maintained at less than 12 mole percent and preferably less than about 7 percent, based on the total paraffin content and of the reaction mixture (or, more preferably, based on the unreacted C -C isoparaffin);

3. stopping such contacting after substantial alkylation has occurred but before the weight rate of formation of unsaturated hydrocarbon products exceeds the weight rate of formation of saturated hydrocarbons.

Preferably, in Step 3 above, the mean residence (or contact) time of the olefin-containing reaction mixture with the catalyst is in the range of 0.05 to 5 hours per (gram of hydrocarbon in the mixture per gram of catalyst). More preferred, is a mean residence time of 0.1 to 0.4 hours. I

When our process is compared with the prior art processes for the alkylation of aromatic hydrocarbons with olefins, it becomes apparent that the prior art does not teach our process (particularly in view of the abovenoted differences with regard to proton abstraction and hydride transfer between such reactions and paraffinolefin reactions). In fact, it is evident that the prior art teachings relating to olefin-aromatic alkylation would lead the ordinary man skilled in the art (particularly one who attempted to substitute a C -C isoparaffin for the aromatic hydrocarbon in prior art examples) to conclude that paraffin-oletin alkylation catalyzed by crystalline zeolites results primarily in the production of unsaturated hydrocarbons.

Surprisingly, in view of the prior art reports of the ability of acidic crystalline zeolites, particularly the protonated or decationized zeolites, to strongly catalyze the polymerization of olefins, we have found a process, catalyzed by protonated and other acidic zeolites, whereby polymerization and/or other reactions which tend to produce unsaturate materials can be virtually eliminated, or held to .a well-controlled minimum, while producing a saturated paraffin-olefin alkylate which is useful as a high octane component for gasoline.

We have further prefected our process so that with constant product removal and reactant'recycle, saturated hydrocarbons may be produced continuously or continually in high yields. We can also control the molecular weight distribution of these hydrocarbons so that the desirable high anti-knock paraffins predominate. We have also discovered how to obtain a high degree of conversion of a C olefin-isobutane feed into alkylate in which the C paraffin portion contains 80 percent or more of trimethylpentanes and which has TMP/DMl-I ratios greater than 5 and even as high as As is further illustrated herein, our process can, by appropriate selection of catalyst and conditions, be used to produce novel paraffin-olefin alkylates, useful as motor fuels and as gasoline blending components, comprising at least 60 mole percent C paraffms and less than 1 weight percent unsaturates and wherein the C paraffins consist of from 5 to 20 mole percent dimethylhexanes, from 0 to 1.5 mole percent methylheptanes, from 80 to 95 mole percent of trimethylpentanes, and wherein less than about 30 mole percent of the trimethylpentanes is 2,2,4-trimethylpentane.

Paraffin-olefin alkylates which contain a high proportion of trimethylpentanes have high octane ratings and good blending values and, therefore, are highly desirable components of gasoline. In contrast, olefinic hydrocarbons are less desirable as gasoline components because they are gurn formers, are sensitive to oxidation, and when highly branched have poorer motor octanes than the corresponding paraffin hydrocarbons. Aromatic hydrocarbons, although they are generally good gasoline blending components with high octane ratings, are not desired in our process since they can decrease the activity of the catalyst.

The catalysts of our process are those substantially anhydrous acidic crystalline alumino-silicate zeolites which, in hydrated form, are chemically characterized by the empirical formula M (AlO );(iO2) (H2Q);, where M is H* and/or an equivalent valence of metal cations and x, y and z are integers, the ratio x/y being usually (but not necessarily from 1.0 to 0.2. A 10 percent aqueous suspension of the acidic zeolite catalyst will have a pH less than 7, preferably less than 5. For our process we prefer that the critical pore diameter of the zeolite be at least large enough to permit adsorption of benzene. We also prefer those acidic zeolites which contain both I-I and polyvalent metal cations (including metal cations in which part of the charge is balanced by oxide or hydroxyl groups), although catalysts having only H or polyvalent metal hydroxides (e.g., Ce(Ol-I) or Ce(OI-I) are effective in catalyzing paraffin-olefin alkylation.

These catalysts are normally prepared from alkali metal-containing zeolites (which in 10 percent aqueous suspension will have a pH greater than 7, and usually greater than 9) by ionexchanging the alkali metal ions for I'I and/or polyvalent metal cations. Hydorgen-ion (or proton) exchange can be effected by exchange from aqueous or non-aqueous medium with mineral acids, such as dilute aqueous HCl, or by exchange with solutions of acids and polyvalent metal ions (such as aqueous HNO3 and Ce(NO3) For zeolites, such as the faujasites, which can be degraded by direct acid exchange, we prefer (as our exchange media) aqueous solutions containing, as at least one component, ammonium salts. Polyvalent metal exchange can be effected with solutions of salts of the metals, such as their nitrates.

Our preferred catalysts are prepared by such ammonium ion exchange, followed by polyvalent metal cation exchange, of an alkali metal faujasite (such as sodium type Y zeolite) having an Sim/A1 0 molar ratio in the range of 4.0 to 5.0.

ILLUSTRATIVE EXAMPLES The present invention may be further illustrated by the following specific examples. Examples I-VI illustrate the preparation of acidic, or potentially acidic, solvated crystalline zeolites by aqueous cation exchange. Example Vll illustrates the activation" of the solvated zeolites by removing solvent from the zeolite.

The remaining examples, excepting Example XVlll, illustrate the use of such substantially anhydrous acidic crystalline alumino-silicate zeolites as alkylation catalysts. Of these, Examples VllI-Xll show the effect of reaction temperature on the yield and product distribution. Examples X, XI and XIII show the unexpected, large increase in product yield effected by the use of various halide adjuvants. Example XII, when compared with Example IX, shows that at a given temperature the more highly exchanged catalyst does not necessarily produce the greatest product yield. When Example XII is compared with Examples X and XI, it is seen that, for a given catalyst and a given contact time, a large increase in reaction temperature will not necessarily give the large increase in yield which is obtainable by the use of a halide adjuvant.

Example XIV illustrates the influence of the contact time (or the time that the feed is in contact with the catalyst) and, in particular, shows the criticality, when maximizing saturate production based on olefin feed, of applicants process step of stopping such contact after substantial alkylation has occurred but before the weight rate of production of unsaturated hydrocarbon becomes greater than the weight rate of production of saturated hydrocarbon. Further shown is the percent consumption of feed olefin to produce an alkylate containing less than 0.5 percent of unsaturates. The importance of thorough premixing of the feed and incorporation of these steps into continuous process schemes is also disclosed.

Example XV illustrates the effect that catalyst composition has on the yield and product distribution. Example XVI shows the effect of various feed olefins on yield and product distribution. Example XVII shows the use of a fixed bed of catalyst in our process and compares the results with our preferred process using a catalyst slurry in a stirred reactor.

Example XVIII shows that highly unsaturated products are obtained when process conditions analogous to those of the prior art (in particular, the process conditions in U.S. Pat No. 3,251,902) are used to attempt to react butene-2 with isobutane in liquid phase in the presence of a substantially anhydrous acidic zeolite (similar to that of Example III). Example XVIII, when compared with other examples such as Example VIII or Example XIX, shows the importance of our requirement that the addition of the feed olefin be controlled such that the amount of unreacted olefin in the reaction mixture is maintained at less than 12 mole percent (preferably less than 7 percent) based on the unreacted C -C isoparaffin.

Example XIX illustrates a preferred embodiment of our invention wherein we produce a highly saturated, novel alkylate, highly desirable as a solvent and as a component of blended fuel for high compression automobile engines, containing less than l percent of unsaturates and comprising at least 60 mole percent of C paraffins, said C paraffins consisting of less than 1 mole percent methylheptanes, 5-10 percent dimethylhexanes, and at least 90 percent trimethylpentanes, said trimethylpentanes comprising less than 20 percent 2,2,4-trimethylpentane.

Example XX illustrates the effect of the gas used in catalyst activation on the paraffin yield, per weight of olefin charged, obtained from the resulting catalyst. Example XXI illustrates the practice of our process in a continuous production operation wherein fresh portions of the hydrocarbon reactants (feed paraffin and feed olefin) are constantly added to the reaction mixture and a catalyst-free alkylate is constantly separated from the reaction mixture and withdrawn from the reactor.

Example XXII illustrates the effect, on yield and product quality, of the catalyst/olefin ratio.

Example XXIII illustrates the correlation between alkylate yield and the ESR measurements of total spin count when aromatic hydrocarbons are absorbed on the Cel-IY catalyst.

EXAMPLE I This example describes the ammonium exchange of a crystalline, alkali metal alumino-silicate zeolite which can be heated to remove loosely bound water and to decompose the ammonium ion to produce a substantially anhydrous acidic crystalline alumino-silicate zeolite which can be used as a catalyst in our process. Preferably, before such decomposition or decationizing, such ammonium-exchanged zeolites are further exchanged with polyvalent metal cations, as is shown in EXAMPLE lll hereinafter.

A kilogram of a commercially available hydrated crystalline alumino-silicate zeolite, identified as sodium zeolite Y, was dried in air at 125C. for 18 hrs., broken up into particles of 100 mesh or less, redried in air at 125C. for 18 hrs., and suspended with stirring, in 1.7 liters of a 9.1% by weight aqueous solution of ammonium chloride at 80C. After 30 minutes the resulting ammonium-exhanged Y zeolite. was separated from the liquid by filtration and recontacted at 80C. in a similar manner with a second 1.7 liter portion of fresh NH Cl solution.

After 6 more such 30-minute exchange cycles, the filtered zeolite was washed with distilled water (pH 6.5) at 20C. until no chloride ion could be detected in the spent wash liquid with acidic silver nitrate reagent.

The washed ammonium-exchanged zeolite was dried for about 18 hours in air at 125C., then ground to about 200 mesh and stored. The dried ammoniumexchanged zeolite produced by the above series of eight ammonium-exchanges analyzed 1.34 percent Na and 4.6 percent N, and had a loss on ignition of 26.5 percent. After the first ammonium-exchange cycle, a similarly washed and dried portion of the zeolite analyzed 5.5 percent Na and 2.3 percent N, and had a loss on ignition of 25.6 percent.

The sodium Y zeolite before this ammonium exchange had a pore size sufficiently large to enable it to absorb benzene and analyzed 7.5% sodium and 8.86 percent aluminum, and had an Al/Si atomic ratio of 0.40. The sieve had a loss on ignition at 1,800F. of 23.8%. All ignition losses referred to hereinafter were run at 1800F.

EXAMPLE n This example illustrates the preparation of more highly ammonium-exchanged zeolites than that of Example l. Example I was repeated except that the sodium Y zeolite was subjected to 8 additional hot NH C1 exchange cycles before it was washed chloride free. The washed, dried ammonium-exchanged zeolite, resulting from this total of 16 ammonium-exchange cy cles, contained 0.77 percent sodium and 4.14 percent nitrogen, and had 29.8 percent loss on ignition.

A similar exchange for a total of 32 cycles produced a washed, dried zeolite containing 0.21 percent Na and 4.64 percent N and having 28.7 percent loss on ignition.

Ammonium exchange of alkali metal zeolites can also be accomplished by suspending the zeolite in'a vessel containing the exchange solution and maintaining a flow of fresh exchange solution into the vessel while withdrawing an equal volume of catalyst-free liquid from the vessel. Removal of catalyst-free liquid from the vessel can be effected by forcing the liquid with pressure or suction through a pleated microporous, woven stainless steel screen l0 filter. In such continuous flow processing, the flow rate is preferably regulated so as to maintain a relatively constant pH in the exchange vessel. Hydrochloric acid ir nitric acid addition can also be used for pH control. With 10 percent ammonium chloride solutions it is preferred to maintain a pH of about 4.5 i 0.3 (at C.). Ammonium exchange can also be effected by percolating the exchange solution through a fixed bed of zeolite.

EXAMPLE III This example illustrates the further exchange of an ammonium-exchanged zeolite with a solution containing polyvalent metal ions in order to produce a zeolite containing both polyvalent metal ions and ammonium ions. A portion of the dried, ammonium-exchanged zeolite of Example I was contacted, with stirring, for 30 minutes at 80C. with 1.7 parts by weight of a 1.3 percent solution Ce(NO '6H O, then separated from the exchange solution by filtration and recontacted for 30 minutes at 80C. with 1.7 parts by weight of fresh cerium nitrate solution. After 6 more such exchange cycles (or a total of 8 exchanges), the filtered Ceexchanged/ammonium-exchanged zeolite was washed with water until no nitrate ion could be detected in the spent liquor by diphenylamine reagent. The washed Ce-NHUY zeolite was dried for 18 hours at C., ground, redried for 18 hours at 125C., and stored in a moisture-tight container. The dried Ce -NI-Lfexchanged zeolite analyzed 6.18% Ce, 1.25% Na, and 1.43% N. It had a 25.6 percent weight loss on ignition.

EXAMPLE IV This example illustrates the use of additional cerium exchange cycles and a more highly ammoniumexchanged base zeolite in order to obtain zeolites with a greater cerium content and a lower sodium content than the zeolite of Example 111. A portion of the washed, dried l6 cycle" NHf-exchanged zeolite of Example 11 was contacted according to the procedure of Example 111 for a total of 16 Ce(NO exchange cycles, then similarly washed and dried. The resulting Ce -NH -exchanged zeolite analyzed 10.1% Ce,

0.69% N, 0.68% Na, and had a loss on ignition of 24.4%.

A similar series of 16 cerium exchanges performed on the 32 cycle ammonium-exchanged zeolite of Example ll produced a washed, dried Ce -NH exchanged zeolite which analyzed 0.23% Na, 10.3% Ce, 0.8% N, and had a loss on ignition of 24.7%.

Hereinafter, sometimes, a catalyst will be identified according to the number and type of such exchange cycles according to the code: number of ammonium exchange cycles/number of polyvalent metal exchange cycles. That is, the above zeolite propared by 6 cerium exchange cycles of a 32 cycle ammonium is, by this code, a 32/16 zeolite, (or, after activation, a 32/16 catalyst).

EXAMPLE V This example illustrates the preparation of a C exchanged sodium Y zeolite. A portion of the commercial sodium Y zeolite of Example I was ground and exchanged for 16 exchange cycles with Ce(NO solution in a manner similar to the exchange of Example Ill, then washed and dried. The resulting Ce -eXchanged Na Y zeolite containted 9.6 percent certium, 1.69 percent sodium, and had a loss on ignition of 25.1 percent.

The cerium exchanges of Examples III, IV and this example can be effected in a continuous manner, similar to that described in Example 11 for ammonium exchanges. Preferably the pl-l should be about 4.5. The particular polyvalent metal salt chosen and the pH of the exchange solution will determine whether the cationic exchange species is the metal or a hydroxylated complex ion of the metal. Other polyvalent metal ions, such as those referred to hereinafter, and in particular cations of the polyvalent rare earth metals and mixtures thereof, may be similarly exchanged with alkali metalcontaining and/or ammonium containing crystalline zeolites. Especially preferred catalysts can be obtained from crystalline alumino-silicate zeolites which have been so exchanged with aqueous solutions of salts of gadolinium, such as Gd( N09 or with mixtures of salts of Gd and Ce. In this specifiction the term rare earth metals includes lanthanum, that is, the term rare earth" herein is used as a synonym for lanthanon. The lanthonons include La, Ce, Pr, Nd Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb and Lu.

EXAMPLE VI This example illustrates the ammonium exchange of a Ce -exchanged sodium Y zeolite. A portion of a washed, dried cerium-exchanged zeolite prepared similar to that of Example V and containing 7.7% Ce, 0.63% Na and with 26.4% loss on ignition was exchanged with hot aqueous ammonium chloride, for a total of 8 cycles, using the procedure of Example I. The washed, dried NH;, -Ce -exchanged Y zeolite analyzed 0.22% Na, 2.8 N, 5.3 Ce and had a loss on ignition of 27.6%. Therefore, about 30 percent of the cerium was removed from the cerium-exchanged Na Y zeolite during the ammonium-exchange cycle.

An alkali metal zeolite which was exchanged by the reverse procedure, that is, 8 ammonium exchanges followed by 16 cerium exchanges, contained 87 percent more cerium (it analyzed 9.9% Ce, 1.3% Na and 0.26% N, and had a loss on ignition of 22.5 percent).

EXAMPLE vu This example illustrates a preferred method of activation of hydrous crystalline alumino-silicate zeolites prior to their use as catalysts in our paraffin-olefin alkylation process. In general, hydrous crystalline zeolites are activated by controlled heating under vacuum or in a stream of a gas, such as air, hydrogen, nitrogen, helium or oxygen, to remove water. In the case of ammonium-exchanged zeolites, not only is loosely bound water removed but also the ammonium ion is decomposed to obtain a substantially anhydrous, decationized or protonated" zeolite. Such zeolites are highly acidic and are similar catalytically to those prepared by direct exchange with an aqueous acid.

When the hydrous ammonium zeolite also contains polyvalent metal ions, the resulting activated zeolite will be partially protonated or cation deficient. Such zeolites are not only highly acidic, but are more resistant to the detrimental effects of the activation procedure.

The heating rate and temperatures of such activation will depend to a great extent on the type of zeolite, that is, an Al/Si atomic ration, and the type and percent of polyvalent cations and monovalent ions such as hydrogen or ammonium ion. In any event the hydrated zeolite is first heated at a temperature sufficiently high to remove the bulk of the uncombined or uncomplexed water from the pores of the zeolite. At atmospheric pressures this temperature is preferably from l25-300C., most preferably from 240C.

In the case of an ammonium-exchanged zeolite the temperature is then raised to a higher temperature than that used for such water removal and such temperature is maintained for a sufficient time to remove a substantial amount of the ammonium ion from the zeolite as N11 This removal may also involve decomposition of the ammonium ion by such reactions as oxidation of ammonia to nitrogen oxides or nitrogen and water.

At atmospheric pressure, with ammoniumexchanged zeolites which also contain appreciable quanitites of exchanged polyvalent metal cations, this higher temperature is preferably 320 -500C.

With ammonium-exchanged zeolites which contain no polyvalent metal cations or have a low content of polyvalent cations, it is important that the activation temperature be kept below about 400C., since at higher temperatures the intensity of the X-ray diffraction peaks of the zeolite decreases greatly (due to a degradation of crystalline structure) and the resulting catalyst is less active for paraffin-olefin alkylation. In US. Pat. No. 3,130,007 a similar intensity measurement is used to determine the percent zeolite, and appears to relate to crystallinity of the zeolite.

We have also found that, if an ammonium-exchanged crystalline alkali metal zeolite is further exchanged with polyvalent metal cations, the resulting polyvalent metal Nl-lf-exchangeti zeolite retains a much greater proportion of its X-ray peak intensity after activation than does the base NHf-exchanged zeolite. Although small quantities of polyvalent cations will be of some benefit in this respect, for our catalysts it is preferable that the zeolite contain at least the following quantity of polyvalent metal cations (or a combination thereof of equivalent valence):

l. at least one tetravalent metal, metal oxide or metal hydroxide for every 16 atoms of aluminum in the alumino-silicate tetrahedra of the zeolite, or

2. at least one trivalent metal, metal oxide or metal hydroxide for every 12 atoms of aluminum in the alumino-silicate tetrahedra, or

3. at least one divalent metal, metal oxide or metal hydroxide for every 8 atoms of aluminum in the alumino-silicate tetrahedra.

In addition, for optimum activity, the polyvalent cation should be selected from classes 1, 2-and 3 above (and mixtures thereof) when the atomic ratio Al/Si of the alumino-silicate tetrahedra comprising the zeolite is greater than 0.65, or from classes 2 and 3 above (and mixtures thereof) when the atomic ratio Al/Si is from 0.65 to 0.35, or from class 3 above when the atomic ration Al/Si is less than 0.35. For example, the cation of our zeolite catalyst is preferably selected from the following:

1. at least one cation selected from the class consisting of V, Mo, W, Pa, U VOH, Cr(OH) CrO, MnO, Mn(Ol-l) NbOl-l, MoOl-l, Mo- (Ol-lh, M, RuO Ru(Ol-l) RuO, Ru- (OH) SbOH, OW, W(OH) WOH, Re- (Ol-UJ, Re(Ol-l) ReO, Os(Ol-l OsOf, 00s, Os(OH) lrO, MOI-1);, BiOH, Biol-1 PaOl-l, U0, U(OH) and UOH, when Al/Si is from 1.0 to 0.65,

2. at least one cation selected from the group consisting of A1, Ni, TiOH, V, VOH, V0, V(OH)g CT(OH)3+3, MTl(OH).; MUD-3+3, Mn(OH) Mn(Ol-1) Mn, GeOH", ZrOH, Nb(Ol-1) NbO, Mo(Ol-l) Mo(OH) M00 MoOl-l, Mo, Ru, RuOH, Ru(QH) Ru- (O1-1) Rh,- RhO'l-I, PdOH, SnOH, Sb, Sb(OH) SbO, La, HfOl-l, Ta(Ol-l) TaO,

Thomsonite, levynite, and the Type X zeolite of U.S. Pat. No. 2,822,244 are crystalline zeolites having an Al/Si atomic ratio greater than 0.65. Analcite, chabazite, phillipsite, and the Type Y zeolite of U.S. Pat. No. 3,130,007 have Al/Si ratios between 0.65 and 0.35. l-leulandite and the Type L zeolite of U.S. Pat. No. 3,013,984 have Al/Si ratios less than 0.35. Mordenite has an Al/Si ratio in the range of 0.2 and some mordenites have been reported to have an Al/Si ratio appreciably less than 0.2 (e.g., 0.13). Such low Al content mordenites, when exchanged and activated by the procedures taught herein, have some catalytic activity in our process but are not among our preferred catalysts.

As catalysts in our process we further prefer substan tially anhydrous protonated alumino-silicates which are capable of adsorbing benzene, wherein the ratio Al/Si in the tetrahedra is from 0.65 to 0.35 and which contain at least one rare earth metal cation for every 9 aluminum atoms in the tetrahedra since such catalysts have high alkylation activity and retain a high degree of X-ray peak intensity on activation or regeneration.

For example, in illustration of our preferred method of activation of a preferred species of hydrous zeolite. the 16-cycle Ce -exchanged/l6-cycle NH -exchanged zeolite of Example 1V was heated at 230C. in a rotating kiln in a stream of flowing air for about one hour to remove water. No loss of ammonium ions was detected during this heating period. The temperature of the kiln was then raised at the rate of about 10C. per minute to a temperature of 400C. During this heating, ammonia could be detected, by MnSo -AgNO reagent, in. the exhaust gases from the kiln. The kiln was maintained at 400C. for 2 hours, at which point no ammonia could be detected in the exhaust gases. The heat was then removed from the kiln and the kiln was cooled rapidly in a flowing stream of dry air. The activated catalyst was maintained overnight in a slowly flowing stream of dry air. The resulting, substantially anhydrous, protonated crystalline alumino-silicate had a loss on ignition of 3.7 percent.

Summation of the intensity of the significant X-ray diffraction peaks of the hydrous zeolite before activation and of an activated sample showed no decrease in intensity for the activated zeolite. in contrast, a similarly activated portion of the base 16-cycle ammoniumexchanged zeolite showed an intensity decrease of 64 percent.

To illustrate the stabilizing effect of even small quantities of polyvalent metal ions, a sample of the base 16 cycle ammonium-exchanged zeolite was submitted to a 16-cycle Ce exchange using one-tenth the usual cerium salt concentration to produce a dried, washed zeolite which analyzed 1.23% Ce (ignited basis). After activation according to the above procedure, the activated zeolite showed an intensity decrease of 47.4 percent.

The bulk density in g/ml of the dry (at C.) hydrated zeolite is about 0.71 for sodium Y zeolite, 0.78 for highly ammonium-exchanged sodium Y zeolite (NH Y), 0.90 for highly cerium-exchanged NH Y (CeNH Y) and 0.89 for highly Gd exchanged NH Y (GdNHqY). If one assumes no significant volume is largely in the form of NH ions, was removed more readily than the water which remains after 1 hour at 450F. Furthermore, ammonia removal releases protons (NH, NH H*). When an activated Nl-lrY zeolite is ignited at 1800F., OH groups are destroyed and H 0 is evolved in an amount equivalent to one molecule of H20 for every two OH groups.

Uytterhoeven, Christner and Hall, J. PHYS. CHEM. 69, 21 17-26 (1965), have proposed the following stoichiometry to account for protonation and dehydroxylation:

sometimes, bound," or combined or "complexed water to distinguish it from that water which is readily evolved from the exchanged zeolite below 300C. Equilibrated zeolite is a zeolite which has been exposed to air of about 50 percent relative humidity, at about 68C. for about 12 hours.

We have further established that our preferred substantially anhydrous, acidic crystalline zeolite catalyst, containing polyvalent metal ions and, more preferably, having some degree of protonation sometimes termed cation deficiency, will evolve substantially no bound water when heated for about 1 hour at 300C. but when ignited at 1800F. will evolve about A to 2 mole of water for each atom of exchanged polyvalent metal. In particular, in our novel, activiated, cerium-containing catalysts, for each atom of cerium in the catalyst, 0.8 to 1.2 molecules of water will be evolved upon ignition at 1,800F.

We have concluded that in the catlyst this water is present, in the form Ce(OI-l) To understand the basis for this finding, one must first consider the' behavior on activation of the hydrated NaY and NH Y zeolites. Behavior of NH Y catalyst during activation at temperatures from l50to 1292F. (65 to 700C.) and for times up to 4 hours, as is summarized in Table 1. Two series of experiments were performed. Experiment A being at different times at constant temperature and Experiment B being at different temperatures at constant time. Total water (that is, sorbed and combined) appeared to be retained by this catalyst more firmly than ammonia.

Table 1 shows that about two-thirds of the total water present on dried NH Y zeolite had been removed after 1 hour at 450F. This water removal is an endothermic reaction and probably represents loosely held water that is in molecular form when sorbed. Data from DTA-EGA measurements agree with this observation.

Water removable only at 750F. and higher temperatures is more firmly bound and is chemically combined in a form other than molecular H O. Ammonia, which Comparing H O contents of activated NH, zeolite with half the lattice NH, on the basis of Uytterhoevens stoichiometry reveals the following differences between calculated and measured H O:

Experiment A-activation for zero time at each temperature H2O Difference (gv mole) Temp. "F (Experimental-Calculated) Experiment B-activation at 750 F.

Time, min. H O Difference (g. mole) (Experimental-Calculated) It is interesting to note that in most instances the excess H O in the preceding table is about numerically equivalent to the residual sodium value of about 0.044 g. ion all on the basis of 100 g. anhydrous base. Carter, Lucchesi and Yates, .l. PHYS. CHEM., 68, 1385-1391 (1964), described IR bands on NaX zeolite at 3400 and 1655 cm which presisted up to 450C. (842F.) and which they concluded were due to residual hydrogenbonded polymeric water." It is probable that the H 0 measured over and above that produced from intact and deaminated NH, sites is this hydrogen-bonded water structurally related to residual Na species in the lattice.

In fact, the NaY zeolite itself may contain more than one kind of H 0. For example, if it is assumed that the 0.414 g. ion Na had one mole H O associated and that this represents strongly bound H O then the H equivalent is 7.46 wt. H O. This firmly bound H O would represent 31 percent of the total 24.32 wt. H O found (Table 1). With NI-I Y, about 33 percent of the H 0 was firmly bound enough to remain after one hour at 450F.

Activation studies of two ammonium-, ceriumexchanged catalysts were made in a manner similar to those for NI-I Y.

Cerous nitrate exchange of NH 'Y catalysts replaced most of the Nl-L, ions with cermium but removed only -25% of the small residual sodium (Table 2). The Ce -exchanged product, therefore, contained residues of NH.,* and Na*. In contrast to NI-I Y catalyst, the subsequently Ce -exchanged material was able to lose NH, during activation down to a level of 0.01 rnole/lOO g. or less at 750F. The NH Y had required temperatures above 750F. to accomplish this degree of removal.

The sum of chemical equivalents for Na, NH, and Ce after exchange was always noticeably less than the 0.414 g. ion Na 100 g. anhydrous base found with the original NaY zeolite. One explanation for this cation deficiency of the exchanged catalyst is that some protons are structurally incorporated during exchange but not directly measured by analysis. Chemically this incorporation is possible because the pH of the cerous nitrate solution was about 4.5, and favored cerous salt hydrolysis.

An increase in SiOl-l groups and a growing cation deficiency was observed as the catalyst became progressively deaminated during activation. However, the SiOl-I groups and intact NI-I ions were not enough to account for all of the H 0 measured by ignition loss at 1800F. Total I-I O measured was 0.16 to 0.10 mole/l00 g. anhydrous base, but SiOH and NI-If could not have produced more than 0.05 mole H O on ignition.

The NH Y study showed that residual sodium ions were complexed at a l-I O/Na ratio of about 1. Continuation of this behavior in the Ce -exchanged catalyst could at most produce 0.04 mole H O on ignition. Therefore, the water not derivable from SiOH, NH. and Na amounted to nearly half the measured H O. Another source was obviously contributing to the total.

Calculation of (H O)/Ce ratios, are shown for two 750F. activations in Table 3. Calculation of H 0 measured at 1800F. does not imply the existence of associated molecular water but is equally capable of interpretation as ce(OH) species according to the equation:

This correspondence of ineasured water to the Ca H2O 750F. did not lower the I-I O/Ce (or Ce(OH) /total Ce) ratio effectively below 1. NFL? removal occurred during this time, and, thus SiOH increased accordingly as evidenced by the growing cation deficiency. Ce- (Ol-I), however, was far more stable, and that behavior in itself is more indicative of Ce(OH) thaifof Ce(I-I O) Only by increasing temperature above 750F. could the fraction of total Ce in the Ce(OI-I) state he reduced.

Isoparaffin-olefin alkylations with NH Ce exchanged Type Y gave maximum alkylate yields and selectivity when the catalyst had been activated at about 750F. rather than at lower or higher temperatures, Possibly, maximal Ce(OH) establishes the sites needed for isoparaffin-olefin alkylation. Also, Ceexchanged Type Y catalysts have been far more stable toward temperatures above 400C. than NH Y, as measured by X-ray diffraction. When Ce(OH) sites become dehydroxylated, Ce

7 I V V v V and Al Si species can re-form. 0n the other hand, dehydrcxyla- 9 cure of Al Si and species, which could be less stable.

When we state that Ce(OI'I) sites are preferred for alkylation, we do not means that these sites per se are the sole locus of activity. Rather, these sites form an essential part of a newtowrk or complex of sites, including HO A]. and Si species. Very possibly the entire complex is required to achieve the carbonium ion-olefin combinations accompanied by good hydride transfer which are vital for a highly paraffinic alkylate. Structural considerations further suggest that only a portion of these complexes may be effective for alkylation, even though enough H O for total Ce(OI-I) formation is a necessary ingredient of catalyst composition.

As is shown hereinafter, catalytic activity for paraffin-olefin alkylation is related to cerium content. ESR total spin counts of these catalysts with aromatics (such as benzene, perylene, anthracene, etc.) sorbed on them revealed a dependence of electron withdrawl ability upon cerium content. Calculations reveal about 5 percent of the cerium ions to be on the external catalyst surface and the data indicated a 1:1 numerical correspondence between these ceriums and the total spin count.

As is illustrated hereinafter in Example XXIII, there is a nearly linear relation between the total ESR spin count of adsorbed aromatics on our cerium catalyst (when activated at temperatures below about 450C.)

amount needed for Ce(OI-I) formation indicates that and the alkylate yield under a given set of reaction consubstantially all of the cerium in Type Y zeolite activated at 750F. (400C.) is in this form or in an equivalent combination of such forms as Ce, Ce(OI-I) Ce- (OH) Ce and Ce(OH)* r A 750F. activation with a sample of the same hydrated CeNI-I Y zeolite in another experiment revealed in H OlCe ratio of 1.041 (Table 4). Continuing this experiment at a series of temperatures up to I292F. (700C.) showed a steady decline of the ratio to 0.370 with increasing temperature. Heating for up to 4 hr. at

ditions. In general the more preferred ceriumcontaining catalysts have total ESR spin counts above 3 X 1O /g. with anthracene. Thus, there is a correlation between alkylation yield with its essential dependence upon hydride transfer and the electron withdrawal from anthrancene by our catalyst.

EXAMPLE VIII This example illustrates the use of substantially anhydrous acidic crystalline alumino-silicate zeolite as a paraffin-olefin alkylation catalyst. The activated l6- cycle Ce /16-cycle NHf-exchanged zeolite of Example VII was charged in amount of 23.3 g. into a l-liter, stirred autoclave containing a four-member baffle to diminish vortex formation. Then 444 milliliters of liquid isobutane was added. The stirring rate (of 'a sixmember, flat-blade turbine) was adjusted such that substantially all of the zeolite was suspended in the liquid isobutane (about 550 rpm). The temperature in the reactor was raised to 80C. using sufficient nitrogen to produce a total pressure of 250 p.s.i.g. Under these conditions nearly all of the hydrocarbon is in the liquid phase. Then a liquid mixture of one part by volume of butene-Z and five volumes of isobutane was charged from a .lerguson gauge via a needle valve and dip tube into the isobutane-catalyst slurry (and near the bottom of the reactor) at the rate of one milliliter of mixture per minute for a period of 220 minutes. Nearly all of the hydrocarbon was maintained in liquid phase. At the end of this time the reaction was stopped by cooling the reactor to 17C., then separating the reaction mixture from the catalyst by first removing the normally gaseous hydrocarbons at room temperature and atmospheric pressure, and then separating the liquid product from the catalyst by filtration. The used catalyst analyzed 0.9% coke (non-volatile residue). Some propane and n-butane but no methane, ethane, ethylene or propylene were found in the normally gaseous hydrocarbons. The C; paraffin yield of the reaction mixture, based on the weight of olefin charged, was 71.4% and the C unsaturate yield was 0.24 percent on the same basis. Hereinafter all yield data are reported as based on the weight of olefin charged.

EXAMPLE IX when the reaction of Example VIII was repeated except that the temperature was 120C. and the pressure 475 p.s.i.g., the Cy paraffin yield was 129.4% and the unsaturated C hydrocarbon yield was 4.3%.

Table 5 further characterizes the C5 product ob-- tained in the reactions of Examples VIII and IX.

EXAMPLE X This example illustrates the unexpectedly large increase in degree of conversion of olefin reactant to saturated C hydrocarbon product when a small amount of a halide adjuvant is present in the reaction mixture. The reaction of Example VIII was repeated at 60C. except that the catalyst used was the 32/ l 6 zeolite of Example IV which had been prepared by 32 NI -1 exchange cycles followed by 16 Ge -exchange cycles. The catalyst was activated by the procedure of Example VII. Without halide addition at 60C., the yield of C; paraffins was 90 percent and the yield of C unsaturates was 1 1.5 percent. On a mole basis this amounted to 0.44 mole of C; paraffins per mole of C olefin charged. In contrast, when 2.4 X mole of tertiary butyl chloride (hereinafter, sometimes, TBC) was added to the reactor for each mole of initial isobutane, the yield of C paraffins was 120 percent and C unsaturates 6 percent. The TMP/DMH ratio was 6.6. The used catalyst had no measurable coke content.

EXAMPLE XI With the same proportion of t-butyl chloride the reaction of Example X was repeated at 40C. (125 p.s.i.g.) and at 25C. (125 p.s.i.g.). At 25C. only 12 percent of C; paraffins was produced, and 0.49 percent of C unsaturates. At 40C. 120 percent of C;

paraffins was produced and 6.5 percent of C unsaturates. The TMP/DMH, ratio was 4.10 at 25C. and 7.86 at 40C.

EXAMPLE XII When Example X was repeated at 120C. (484 p.s.i.g.) without halide addition, percent of C; paraffins and 1.1 percent of C; unsaturates were produced. The TMP/DMI-I ratio was 3.16.

Of commercial importance is the finding that, by practice of our invention, we cannot only obtain good yields of alkylate which has a high TMP/DMH ratio and is high in trimethylpentanes but that in these trimethylpentanes there is a low proportion of the less desirable 2,2,4-trimethylpentane (regarding this undesirability, see US. Pat. No. 2,646,453). Table 6 compares the percent of the total trimethylpentanes which is 2,2,4trimethylpentane ( percent 2,2,4 in TMP) in the products of Examples X, X1 and X11 with similar distributions reported by Cupit, CR. (Ibid. p. 211) for I-I- SO alkylation and Kennedy, R.M. (Ibid. p. 30) for HF or AlCl alkylation of isobutane with butene-2. For the compositions of the products obtained with other .olefins (e.g., propylene, cyclohexane) and I-I SO catalyst see J. E. I-Iofmann, J. ORG. CHEM., 29 (Part II), 1497-1499 (1964).

As is further illustrated herein, our process can be used to directly product novel paraffin-olefin alkylates, useful in gasoline blending, comprising at least 60 mole percent C paraffins and less thann one weight percent unsaturates and wherein the C paraffins consist of from 5-20 mole percent dimethylhexanes, from 0-1 .5 mole percent methylheptanes, from -95 mole percent trimethylpentanes, and wherein less than 30 mole percent of the trimethylpentanes is 2,2,4- trimethylpentane. Such novel alkylates can also be produced using hydrogenation and/or adsorbents to reduce the unsaturates in such products of our process as that of Example X which is cited in Table 6.

In general, in the temperature range from about 25C. to l20C., the proportion of 2,2,4-TMP in the total TMPs decreases (and usually the proportion of 2,3,3-TMP increases) as the reaction temperature decreases.

Product distributions in the alkylates produced by our 32/16 catalyst (and in virtually all of the alkylates reported herein as produced by our process) are far removed from calculated equilibrium values. For example, the calculated equilibrium TMP among C paraf fins at about 60C. is only 12%1 whereas, in our process TMP usually constitute more than 80 percent of the C paraffins. Among the TMP there is a similar departure from calculated equilibrium, as shown by the following mole percent data at 60C.:

Experimental TMP With TBC No TBC Calculated Equilibrium l. TBC promoter with solid catalyst had an evident influence upon TMP distribution which is not characteristic of equilibrium control.

2. Equilibrium calculations predict that the fraction of 2,2,4-TMP decreases with rising temperature. With solid catalyst, it is increased as temperature went up.

Alkylation with solid zeolite catalyst appears to bethe result of a sensitive balance among competing kinetic paths. Our halide adjuvant, such as TBC, favorably alters one or more of those paths.

Surveying temperatures with a 16/16 catalyst (under conditions as in Examples VllI but with improved feed premixing) similarly points out the alkylate yield gain to be realized by operating at an intermediate temperature of about 80C. (Table 7). In a continuous reactor residence time must be taken into account; the same temperature with a given catalyst may not be preferred if residence time is changed.

It is interesting to note that 60C. with the 16/ 16 catalyst produced no more than half as much alkylate as 80C. with this catalyst. The 32/ 16 catalyst was not so responsive to temperature changes above 40C. Again, as temperature decreased from 80 to 60C., a shift toward a heavier product occurred (Cf), but the 2,2,4- TMP content was desirably low. When this isomer decreased, the largest gain was in 2,3,3-TMP, as it has been with the 32/16 catalyst.

Temperature is a useful device in elucidating catalyst differences. When the catalyst exchanged only with NH, (32/0) was tested at 120C. without TBC promoter, it was less active than a 16/16 catalyst (Table 8). The C paraffin yields were 109.5%, based on olefin charge, with 32/0 and 129.4 percent with 16/16.

Evaluating an NHfi-exchanged catalyst (16/0) with TBC relative to 16/16 at 80C. revealed a more dra-v matic difference in alkylate yield and tion.

The 16/() catalyst produced too little C paraffin and too much CJ and C These factors could also be used to understand the importance of a polyvalent metal, such as cerium, on the catalyst. But testing catalysts at milder conditions is even more effective in uncovering differences between them, as shown by the data from 32/0, 16/16, and 16/0. Therefore, low operating tempertures can be used as a research tool in distinguishing among alkylation catalysts that appear to be more similar at relatively high temperatures.

EXAMPLE XIII Table 5 reports the products obtained from similar runs at 8C. using the activated catalyst of Example VIII with t-butyl, chloride, n-propyl chloride or n-butyl chloride as adjuvants (at a level of 2.4 X l0 mole of adjuvant per mole of initial isobutane).

Table 9 reports the products obtained from similar runs (but with more intimate premixing of the feed olefin and feed paraffin) using CCl TBC and various other adjuvants and using either continuous or pulsed" addition of the adjuvant to the reaction mixture. In this table, the amount of adjuvant is reported as millimoles per mole of feed olefin charged (m.mole/m. 0C).

In run 606 of Table 9, the catalyst was preconditioned by contact with a solution of perylene in CCl The perylene was quantitatively adsorbed by the catalyst along with some CCl The catalyst developed a dark, intense, blue color upon contact with the perylene solution. Removal of residual CCl, by vacuumpumping at ambient temperature caused the catalyst product distribucolor to turn to black. This black catalyst was the catalyst used in run 606.

In run 600, the catalyst was preconditioned with carbon tetrachloride as a control experiment for 606. The catalyst developed an intense red color on contact with the CCl Upon vacuum pumping, the red color disappeared. It is this pumped catalyst which was used in run 600.

Potential catalyst adjuvants are those halides, both organic and inorganic (e.g., AlBr BF HBCI AsCl which are capable, under the reaction conditions, of sufficient polarization to promote carbonium ion reactions or to have carboniogenic properties. For precise control of the reaction product distribution (or alkylate quality) and to prolong catalyst life, we prefer to avoid adjuvants which contain atoms other than hydrogen, carbon, bromine, fluorine and chlorine (although as seen in run 632, oxyen, as in the form of alcoholic OH groups, can be present in reaction mixture. Water, C to C saturated alcohols (e.g., tertiary butyl alcohol, cyclohexanol) or mixtures thereof can be used, per se, as adjuvants or in combination with halides. To avoid accumulation of large organic molecules at the catalyst surface, we prefer to avoid those organic halides wherein the organic radical has a critical diameter greater than about 9A, such as the chlorinated naphthenic waxes. Note, however, in Table 9, that perylene presorbed on the catalyst from CCl, solution did not.

act as a poison but allowed about 10 r ative percent more C; paraffin yield than a control experiment with carbon tetrachloride alone. This carbon tetrachloride control experiment itself produced a better than 10 relative percent increase in C paraffin yield over a similar experiment with tertiary butyl chloride and without CCl In contrast, NH, presorbed on the catalyst acted as a poison, even when TBC was added continuously to the reactor.

Our preferred halide adjuvants, when present in solution in the reaction mixture at a level of from 1 X 10 to l X 10 mole per mole of C -C isoparaffin reactant, are HF, HCl, HBr and the saturated halohydrocarbons containing at least one atom per molecule of bromine. chlorine or fluorine. Mixtures of these substances can also be used as adjuvants. Of these adjuvants we prefer carbon tetrachloride and the aliphatic saturated monochlorides having no more than six carbon atoms. When the isoparaffin reactant is predominantly isobutane, we prefer touse, as halide adjuvants, the aliphatic saturated monochlorides having 3 or 4 carbon atoms.

The adjuvant can also be added to the catalyst after the final washing, in the exchange procedure but, more preferably, is added to catalyst after activation, as by passing gaseous HCl through the catalyst at the final stage of activation (or while cooling catalyst after activation). It can be important, especially in our continuous process, to control the amount of adjuvant present in the reactor vapor space since the vapor pressure of such adjuvant affects the adjuvant concentration in the reaction mixture.

EXAMPLE XIV This example illustrates, at a given feed rate, the influence ofreaction time or, more precisely, a combination of catalyst/feed contact time and residence time (of the paraffin product) on the yield of C5 saturates and C; unsaturates. This reaction time,which combines residence and contact time, reflects both kinetic influence and the age of the catalyst-reactant system.

The process of Example Vlll was repeated at C., 250 p.s.i.g., using the activated catalyst of Example VIII, with 2.4 X l mole of tertiary butyl chloride present in the reactor initially per mole of initial isobutane (444 ml.). As in Example VIII the rate of addition of the butene-2/isobutane feed was I milliliter per minute. Four runs were made at various contact times (60, 120, 220 and 280minutes). In each run contacting was stopped by rapidly cooling the samples to l-20C., then slowly reducing the pressure to atmospheric and simultaneously distilling off lighter gases. The remainder of the reaction mixture, which was liquid at that temperature, was separated from the solid catalyst by filtration.

FIG. 1 illustrates the variation in the yield of C paraffins based on the olefin reactant as the reaction time increased.

FIG. 2 illustrates, by the solid curve, the weight percent of C unsaturates produced, based on the olefin reactant, as the reaction time increased. The broken curve, of FIG. 2, show the weight percent of n-butane produced per mole of olefin converted, as the reaction time increased.

From FIGS. 1 and 2, it can be seen that after 2 hours only negligible amounts of C; unsaturates were found in the reaction mixture and the yield of C; paraffins was 96 percent of the weight of olefin charged. Of the C; paraffins, 60 mole percent was C and of the C paraffins there was 0 mole percent methylheptanes. The ratio TMP/DMI-I was 7.18. Of the trimethylpentanes, 24.7 percent was 2,2,4-TMP. At this point a total of 20 milliliters of butene-2 had been added to the reactor along with an additional 100 milliliters of isobutane. When added to the original 444 milliliters of isobutane this amounted to a total of 544 milliliters of isobutane which had been charged to the reactor at that time along with 20 milliliters of butene-Z (density 0.60), which produced 11.5 grams of C? paraffin.

After 220 minutes, 37 ml. of butene-2 had been charged to the reactor along with 183 ml. of isobutane to give a total hydrocarbon charge of 664 ml. The C; paraffin yield was 142 percent based on the weight of olefin charged (37 ml.) or a total C; paraffin yield of 3l.6 grams. Of the C paraffins, 60 mole percent was C Of the C paraffins, 1 mole percent was methylheptane. The ratio TMP/DMI-I was 4.98. Of the trimethylpentanes, 20.9 percent was 2,2,4-TMP.

After 280 minutes a total of 724 ml. of hydrocarbon had been charged to the reactor of which 47 ml. was butene-2. The C; paraffin yield was 1 10 percent of the weight of olefin charged or 31.1 grams. The ratio TMP/DMH was 5.44 and 21.0 percent of the trimethylpentanes was 2,2,4-TMP. At 220 minutes 0.84 grams of C unsaturates had been detected in the reaction mixture or 3 .75 percent based on the olefin charged. At 280 minutes 1.6 grams of 'C unsaturates were detected in the reaction mixture, which amounted to 5.5 percent of C; unsaturates based on the weight of olefin charged.

An inspection of FIGS. 1 and 2 shows that at point B of FIG. 1 and point B of FIG. 2, the net rate of production of unsaturated hydrocarbon is about 2 weight percent C unsaturates/olefin charged/hour but the rate of production of saturated hydrocarbon is about 100 weight percent C saturates/olefin charged/hour. At point C in FIG. 1 and C in FIG. 2, the net weight rate of production of unsaturated hydrocarbon becomes greater than the net weight rate of production of saturated hydrocarbon per weight of olefin charge.

The broken line in FIG. 2 shows that there is a high initial production of n-butane but that by point A" (corresponding in time to point A of the solid curve) the proportion of n-butane as a function oftime had become nearly constant. This behavior is understandable if n-butane is formed as a result of olefin protonation and subsequent hydride transfer (which is considered necessary for initiating alkylation).

If one wishes to maximize the production of C paraffin per olefin charged the reaction should be stopped at the point corresponding to the letter A in FIG. 1, at which point the C product of the reaction will contain about 3.6 percent of unsaturated hydrocarbons (see point A of FIG. 2).

However, if one wishes to have a substantially olefinfree C product, the reaction would be stopped in the vicinity of point B in FIG. 2. In the latter case, the catalyst life can be greatly prolonged in comparison with operation between points B and A of FIG. 1 or B and A of FIG. 2.

Similar relations were observed at 60C. and 120C. At 120C. (without halide addition) 28.7 grams of CH paraffin were obtained after 220 minutes 129.4 weight percent yield/olefin charged) while only 23.3 grams were obtained after 384 minutes of reaction time (Tmzii /yield). When the weight of catalyst? used at 120C. (without adjuvant was doubled, 22.5 grams of C; paraffin were obtained after 220 minutes( 101.4% yield).

Catalyst life can be greatly prolonged at conversion ratios approaching that at point B by constantly separating a catalyst-free alkylate and a concomitant amount of unreacted feed from the reaction zone whileconstantly adding an approximately equal volume of fresh portions of the hydrocarbon reactants. This constant separation nd withdrawal of alkylate and unreacted feed in conjunction with the addition of fresh reactants (including recycle of unreacted feed) can be accomplished continuously by adding a steady stream of reactants and withdrawing a steady stream of the mixture, as by utilizinga continuous stirred reactor system such that of FIGS. 3, 4 and 5.

Surprisingly, in view of US. Pat. No. 3,251,902, the degree of conversion of butene-2 to paraffins is very high in our process. For example, in the above runs, analysis of the reaction mixture for unreacted butene-2 showed that in the 60-minute and l20-minute runs the butene-2 conversion was In the 220-minute run 92.6 percent of the feed butene-Z was converted and 88.5% was converted in the 280-minute run.

The most important consideration in continuous operation is to coordinate the rate of olefin addition with the rates of feed olefin consumption and removal in order that the amount of unreacted olefin in the reaction mixture is maintained at less than 12 mole percent based on the unreacted C -C isoparaffin, and preferably less than about 7 percent. Also, the preferred mean residence time of the hydrocarbons in the reaction mixture, with the catalyst, is in the range of 0.05-0.5 hour per gram of catalyst per gram of hydrocarbon in the reaction mixture.

In this respect, the preferred procedure is to thoroughly premix the feed olefin and feed paraffin. The uniformity and intimacy of such premixing can greatly influence the character of the C product. In Table 10, for example, two runs are shown which were identical except for the feed premixing technique. In one run the feed was introduced through the bottom of the Jerguson gauge. This increased the uniformity and intimacy of the premixing (by reducing charge segregation or layering). The resulting C product contained only 0.26 percent unsaturates and was also lower in C,,* paraffins, higher in pentanes and had a higher TMP/DMH, ratio than the product from the corresponding run where the feed was introduced at the top of the mixing buret (which, unless otherwise noted, was the technique used in all the other examples reported herein). Note that the product from the run with the feed introduced at the top of the mixing buret had a 685 percent greater concentration of unsaturates than the product from the former run where the premixing was more uniform and more complete. In these examples the runs utilizing the improved feed premixing (from the bottom of the buret) are so noted or are numbered in the range of 502-698 and 804-898.

For any given type of activated catalyst and feed hydrocarbon, the rate of olefin consumption and, correspondingly, the rate of olefin addition, will be a function of the reaction temperature, the mean retention time of feed olefin in the reactor, the mixing rate, the particle size and concentration of catalyst, and the rate of product removal. One method of controlling such a continuous" process, under conditions of good feed olefin mixing, is to control the rate of feed addition and the rate of reaction mixture withdrawal such that the C; component of the reaction mixture contains substantially no C; unsaturates and, preferably, such that there is little unreacted feed olefin in the withdrawn portion.

Another important variable to be considered in our process is the proportion of the reactants (particularly of the feed olefin) which can be present in the vapor space of the reactor. The proportion of olefin in the vapor space is a function of the vapor pressure of the olefin at the reaction temperature and of the degree of reactor filling (that is, the vapor space in the reactor). In our continuous stirred reactor system of FIGS. 3, 4 and the degree of reactor filling can be precisely controlled, as by means of the differential pressure cell.

One convenient means of stopping the contact of the olefin-isoparaffin feed with the zeolite catalyst and effecting the concomitant removal of a portion of the C product from the reaction mixture is to submerse a line, the opening of which is covered by a filter device, such as a very fine screen, into the reaction mixture and to constantly withdraw a catalyst-free portion of the reaction mixture from the reactor to a zone where unreacted feed olefin (if present) and feed isoparaffin are separated by distillation from the C; hydrocarbons, and recycled to the reactor. This means can be utilized in the reactor section (FIG. 4) of our continuous stirred reactor system (of FIGS. 3, 4 and 5) for producing an olefin-paraffin alkylate.

In the reaction illustrated by FIGS. 1 and 2, the mean residence time in hours (per gram of catalyst per gram of hydrocarbon) of the hydrocarbons in the reaction mixture with the catalyst was 0.089 after the first 60 minutes, 0.167 after 120 minutes, 0.236 after 180 minutes and 0.297 after 240 minutes. This is to be con-' trasted with 0.62 hours in Example II, Table II, Column 3, 1.25 hours in Example VII and 1.47 hours in Example I, Table I, Column 1 of U.S. Pat. No. 3,251,902, previously cited herein.

In our process the preferred mean residence time is in the range of 0.05-0.45 hour, more preferably 0.1 to 0.4 hour.

An illustration of the calculation of mean residence time, for the first 60 minutes in the reaction illustrated by FIGS. 1 and 2 herein, follows:

(444 ml. i-butane)(0.5543) 246.11 g. isobutane for entire time i 23.3 g. of catalyst Change 1 vol. butene-2 (density =0.5988 g./ml.)

5 vol. isobutane (density =0.5543 g./ml.) (6)(D) (5)(0.5543) (l)(0.5988) 2.7715 0.5988 3.3703 D* density of hydrocarbon mixture 0.5617 For 60 min.

(1 h0ur)(23.3 grams catalyst) 246.11 (60 min.)(l

ml./min.)(0.56l7 g.ml.) 0.08861 hr./(g. Hydrocarbon) (g. Catalyst) EXAMPLE XV This example illustrates the effect that catalyst composition has on the yield of C reaction product and on the product distribution, in particular with regard to the proportion of C paraffins and the distribution of these C paraffins into trimethylpentanes and dimethylhexanes.

The process of Example VIII was repeated except that the reaction temperature was 120C. (which was close to the critical temperature of the reaction mixture), the reaction pressure was 500 p.s.i.g., and the reaction time was 3.67 hours. Separate runs were made with equal weights (activated basis) of zeolites of varied Na, H and polyvalent metal contents, which were prepared similarly to the catalysts of Examples 11, III, IV and V.

Runs were also made, at C., 250 p.s.i.g., and 2.4 X 10" moles t-butyl chloride per mole of initial ibutane, with catalysts prepared from the following: the 1.72 percent (ignited) Ce zeolite of Example VII; a 16- cycle ammonium-exchanged NaY zeolite which was further exchanged with 16 cycles of a 13.3 g./1. aqueous solution of La(NO -6H O; a 16-cycle ammoniumexchanged NaY zeolite which was further exchanged with 16 cycles ofa 13.3 g./l. aqueous solution of hydrated mixed rare earth nitrates (approximate salt analysis, 48% Ce O 24% La O 17% Nd O 5% Pr O 3% Sm O 2% Gd O and, a l6-cycle ammoniumexchanged NaY zeolite which was further exchanged with 16 cycles of aqueous Ce(NO 6I-I O (as in Example IV).

All of these catalysts were activated by the procedure of Example VII.

The yields of the C paraffin and C unsaturates, based on the weight percent of olefin charged, the C; paraffin distribution and the C paraffin distribution of the products are shown in Table 11.

The yields and product distributions shown in Table 11 indicate that, in substantially anhydrous acidic crystalline alumino-silicate zeolites which have been prepared by ammonium exchange of sodium zeolites with ammonium ions and polyvalent metal ions, the catalytic activity and selectivity in paraffinolefin alkylation are dependent upon the amount and type of exchanged polyvalent metal and the degree of protonation or cationic deficiency (which is related to the nitrogen content before activation). Therefore, when other reaction variables are fixed, an appropriate selection of the catalyst can be used to vary the yield' and product distribution in our process.

Table 12 and Table 13 illustrate the effect on the ultimate catalyst of the type of salt used in the exchange solution.

It is evident from Table 13 that the yield differences are not determined only by the total amount of rare olite with salts of Gd, Dy" and Sm.

One precaution to be taken with data from Table 13 concerns the apparent gain in selectivity for C paraffins with the La(NO and CeCl catalysts. In fact, this gain is more in line with the selectivity gain which is typical when our process is operated at a relatively low degree of reactant conversion or product yield. In other words, if the Ce(NO catalyst had been used to produce only 68 to 73% CJ paraffin yield (the range for La(NO and CeCl the molar C paraffin content of the C parafiins would have increased to about 80 per cent instead of remaining at the 69.0 percent actually observed at 132.0 percent C paraffin yield.

As shown by these data, the anion in the exchange solution exerts an influence on catalyst performance. The effect is related to the condition of metal cations in aqueous solution as a function of anion, cation concentratlon, pH and temperature. An effect such as the following is the probable cause:

Other cations which can affect the catalyst are the alkali metals, such as lithium, sodium, potassium and cesium. As shown in Table 14, at comparable sodium levels, C paraffin yield progressed from 26 to 132 wt. olefin charge for an increase of cerium from 2.0 to 13.5%. Even at 8.3% cerium, the CJ paraffin yield was only 62.7% on the same basis. The probability that the 1.68% sodium content did not have the principal deleterious effect upon this 62.7 yield is supported by the 1 18.9% yield for a catalyst containing 2.8% sodium but 13.0% cerium and by the 1 15.4% yield for another catalyst with a 1.68% sodium and a 12.8% cerium content.

Some gain in C; paraffin yield (1 15.4 to 132.0) can be inferred for a reduction in sodium content from 1.68 to 0.76%.

Selectivity effects of cerium are illustrated by the relatlvely high Cf unsaturate production with catalysts containing less than about 12 percent cerium. Trimethylpentane/dimethylhexane (TMP/DMX ratios were also comparatively low for those catalysts, and relatively undesirable C paraffins constituted as much as 27.2 mole of the total C{* paraffins for the lowest cerium catalyst. These data show that with less than about 12 percent cerium, alkylate will be not only lower in yield but also poorer in quality.

A series of NH.,*-, Ce -exchanged catalysts having very similar sodium levels clarified the essential role of cerium in producing favorable yields of high quality alkylate.

When cerium replaced ammonium on a Type Y zeolite at constant sodium level, the following effects were observed:

1. Appreciable gains were realized in C5 paraffin yield, in relative proportion of Cg paraffins, and in selectivity for trimethylpentanes (TMP/DMH ratio).

2. Simultaneously, desirable decreases were found in Cf unsaturates and in the relative proportion of C paraftins.

3. The only undesirable trned was an increase in the relative amount of 2,2,4-TMP up to 26.4 mole of the total TMP. However, typical sulfuric acid alkylates have 2,2,4-TMP contents above 40 percent. This isomer has the lowest F-l octane number of all the TMP.

1 28 4. The largest gains in yield and selectivities occurred at values of (Ce /NHf') equivalent ratio below about 2.5. Higher ratios are desirable, but corresponding product improvements become smaller. These catalysts were prepared from the same common lot of NH exchanged Type Y zeolite. The following analytical data establish that Ce was exchanging for NH, and that no net loss of Na occurred from the NHf-exchanged zeolite:

Catalyst No. Analysis (g. equivalent/100 g. anhydrous residue Na NHU Ce'** 1 FXIO 0.064 0.389 0.543 FXl0 l-2 0.054 0.190 0.175 0.419 FX10'13 0.047 0.102 0.253 0.402 FX10 1-4 0.047 0.049 0.299 0.395

* Average of 3 analyses The original zeolite had a sodium content of 0.426 equiv./ g. anhydrous residue after correction for 1,800F. ignition loss. Residual sodium content was thus 11-13 percent of the original.

Another interesting but undecided aspect of these catalysts is their growing cation deficiency as cerium exchange increases. A deficiency is said to occur when the sum of residual sodium, ammonium and rare earth does not equal the positive charged initial sodium. The presence of protons-bound or solvated-can account for the apparent deficiency.

As has been shown in Examples I to VII, we prefer to prepare the substantially anhydrous acidic aluminosilicate zeolites by controlled activation of zeolites which are prepared from crystalline sodium zeolites by first exchanging the bulk of the sodium with ammonium ions and then exchanging the resulting zeolite, which is low in sodium and high in ammonium ions, with solutions of polyvalent metal cations. When the base zeolite is sodium Y, the ammonium-exchanged zeolite should contain, on an ignited basis, less than 3% Na and preferably less than 1.0% Na.

In our ammonium exchange we also prefer that the sodium content of the exchange solution be kept as low as is practicable. One means of removing sodium ions from ammonium salt solutions is by a separate cation exchange of the solution with a bed of ammoniumcontaining ion-exchange resins or noncrystalline ammonium zeolites. In this sodium-ion removal step, which is particularly advantageous in continuous ammonium exchange (as in the procedures of Example 11), the sodium ion in the ammonium-ion exchange solution exchanges with the ammonium ion in the resin and the resulting ammonium-rich solution is recycled to the vessel containing the crystalline zeolite for additional exchange with the sodium in the zeolite. The ionexchange resin bed (or noncrystalline zeolite bed) can be regenerated by contacting the ammonium-sodium equilibrium resin with an ammonium-rich stripping stream. The sodium-rich effluent from the regeneration is discarded after, if desired, residual ammonia has been recovered by flash distillation.

Products obtained from a preferred Gd catalyst and from two other, less preferred, catalyst types are shown in Table 15. One of the two less preferred catalysts was obtained by activation (as in Example Vll but with 8 hours at 400C. to insure good NH removal) of a highly (16 cycles) ammonium-exchanged type Y, zeolite (to produce HY catalyst). The HY catalyst produced only about one-fourth as much alkylate, together with more C and C and less C as its cerium counter part.

The other less preferred catalyst was prepared by activation of a 16-cycle cerium exchanged, 16-cycle ammonium-exchanged sodium X zeolite (to produce CeI-IX catalyst). In comparison with CeI-IY catalyst (run 664) the CeHX catalyst produced an appreciably smaller C paraffin yield. An Analysis of this paraffin product showed 23.9 mole to be isopentane (which is 2 to 4 times isopentane usually found in alkylate produced by CeI-IY catalyst). Accordingly, the C paraffin in the alkylate produced by the CeHX was only 59 mole compared wth about 70% for Cel-IY.

Runs 628 and 674 were made with catalysts prepared by an exchange procedure similar to that of Example IV and activated as in Example VII (except that for the run 674 catalyst helium was substituted for air), but wherein gadolinium nitrate was used instead of cerium nitrate in the exchange solution. The resulting novel Gd-alumino-silicate, upon activation, produced a novel catalyst which is very useful for hydrocarbon conversion reactions, particularly in our process for paraffinolefin alkylation.

EXAMPLE XVI This example shows the effect of feed olefins other than butane-2 on the yield of C; products and their distribution. Example VIII was repeated, with a similarly prepared catalyst, except that the feed olefin was butene-l. The C; paraffin yield and the C; unsaturate yield were about the same as those obtained in Example Vlll with butene-2 and the distribution of C paraffins and the C distributions (see Table were similar to those obtained in Example IX with butene-2.

We have found that butene-l, in the presence of nbutane, is readily isomerized to cis and trans butene-2 under our alkylation conditions with acidic zeolite catalysts. This ready isomerization provides the explanation for the similarity between the products obtained when isobutane is alkylated with butene-2 and the products obtained when butene-l is the feed olefin. Surprisingly, a significant quantity of highly saturated C liquid product is also obtained from this lilquid phase isomerization of butene-l in the presence of nbutane. A very small amount of isobutane was also detected. An analysis of the C liquid product of one such run is shown in Table 10. At least some of this C liquid appears to be the result of the combination of the n-butane and the C, olefin.

A similar run was made using 2-methylbutene-2 as the feed olefin and a catalyst, prepared in a manner similar to the catalyst of Example 111, prepared from a zeolite which before activation analyzed 5.0 percent cerium, 1.19 percent sodium, and had a loss on ignition of 25.65%. The catalyst was activated (final temperature 400C.) as in Example VII. The C paraffin yield was 28.6 percent and the C unsaturate yield was 31.2%, based on the weight of olefin charged. The molar ratio C IC of the CJ paraffins was 1.00.

A similar run with a somewhat more acidic catalyst produced a C5 paraffin yield of 49.0 percent and 15.4. percent C unsaturates based on the weight of olefin charged. Of the CH paraffins 29 mole percent were C paraffins and 36.6 mole percent were C paraffins (molar ratio C /C was 1.25). The presence of C paraffins indicates that self-alkylation of isobutane occurred during the reaction.

A similar run using a portion of the same catalyst (activated at 500C.) and a butene-2 feed resulted in a C;

paraffin yield of 51.8 percent of which 82.8 mole percent was C and 4.2 mole percent C paraffins. The

distributions of the C; paraffins in these two products are shown in Table 10.

10 Similarly when isobutylene or propylene or a BB refinery stream (i.e., a mixture of butanes and butenes containing a minor amount of propylene) is the feed olefin, good yields can be obtained (of a product in which the C saturates predominate over the C unsaturates) by stopping contact of the catalyst with the reaction mixture after substantial alkylation has occurred but before the weight rate of production of C olefins becomes greater than the weight rate of production of C paraffins.

In general, our process can be used to produce (either by alkylation, self-alkylation or both) good yields of C; saturated hydrocarbons from C -C isoparaffin (or mixtures thereof) and any monoolefin having from 3 to 9 carbon atoms, including the cyclic olefins, such as cyclohexene, and mixtures of such monoolefins.

Table 16 reports a typical product obtained from isobutane at 60C. in our process under reaction conditions similar to those of Example VII when propylene is the feed olefin, with and without halide adjuvant. Table 17 details similar products obtained in our process from isobutane and propylene under varied reaction conditions.

Note that under the conditions of Run 544 propylene-isobutane alkylation produced a C; alkylateyield of 123.3 wt. based on olefin charge. This number corresponds to 0.513 mole of C paraffins per mole of olefin charged to the reactor. Close examination of the product revealed 66.0 mole C among the C? paraffins, and 97.2% of the C, was 2,3-dimethylpentane.

From these results and other such studies the following observations can be made:

1. In such batch reactions alkylation at 60C. for 2 hr., with attendant increases in relative amounts of promoter and catalyst, produced the highest alkylate yield.

2. Raising or lowering temperature to or 40C., decreasing isoparaffin/olefin ratio (15/1 to 10 or 5/1), increasing contact time at 60C. (2 to 4 hr.), and allowing more olefin to be in the vapor phase-all had adverse effects upon yield and selectivity. Low alkylate yields invariably were accompanied by relatively high amounts of C; paraffin and low quantities of C 3. Although isobutane self-alkylation is strongly indi' cated by C8 paraffin in the product, the distribution of

Claims (8)

1. A PARAFFIN-OLEFIN ALKYLATION PROCESS WHICH COMPRISES CONTACTING MONOOLEFIN OF THE C2-C9 RANGE IN ADMIXTURE WITH PARAFFIN OF C4-C8 RANGE AND WITH A SUBSTANTIALLY ANHYDROUS ACIDIC CRYSATLLINE ALUMINO-SILICATE ZEOLITE UNER ALKYLATING CONDITIONS AND WHEREIN THERE IS PRESENT IN SOLUTION IN THE REACTION MIXTURE FROM 10**-5 TO 10**=1 MOLE PER MOLE OF C4-C6 PARAFFIN OF A HALIDE, SAID HALIDE BEING SELECTED FROM THE GROUP CONSISTING OF HCI, CARBON TETRACHLORIDE AND THE ALIPHATIC SATURATED MONOCHLORIDES HAVING NO MORE THAN 6 CARBON ATOMS.

2. An isoparaffin-olefin alkylation process wherein C3-C9 monoolefin in admixture with C4-C6 isoparaffin having a tertiary carbon atom is contacted with a substantially anhydrous, acidic crystalline alumino-silicate zeolite under alkylation conditions at a temperature below the critical temperature of the lowest boiling hydrocarbon reactant and at a pressure such that the reactants are substantially in liquid phase, and wherein there is present in solution in the reaction mixture from 10 5 to 10 1 mole per mole of C4-C6 isoparaffin of a halide, said halide being selected from the group consisting of HCl, carbon tetrachloride and the aliphatic saturated monochlorides having 1 to 4 carbon atoms.

3. A process for the preparation of an isoparaffin-olefin alkylate comprising contacting isobutane with monoolefin selected from the group consisting of isobutylene, butene 2 and butene-1 and with a substantially anhydrous acidic crystalline alumino-silicate zeolite, at a temperature in the range of 25-120*C. and at a pressure such that each of the reactants is substantially in liquid phase, i. said contacting being effected utilizing sufficient agitation so that substantially all of said zeolite is maintained in suspension in the liquid reaction mixture, the amount of unreacted olefin in the reaction mixture being maintained at less than 7 mole percent based on the unreacted isobutane, and iii. wherein the mean weight hourly space velocity of the hydrocarbons in the reaction mixture is in the range of 2-20 gram hydrocarbons per hour-gram catalyst; and wherein there is present in solution in the reaction mixture from 10 5 to 10 1 mole per mole of isobutane of a halide adjuvant selected from the group consisting of HCl, carbon tetrachloride and the aliphatic saturated monochlorides having 1 to 4 carbon atoms.

4. Process for the preparation of an isoparaffin-olefin alkylate comprising contacting isobutane with monoolefin selected from the group consisting of isobutylene, butene-2 and butene-1, and with a substantially anhydrous acidic crystalline alumino-silicate zeolite, at a temperature in the range of 40*-80*C. and at a pressure such that each of the reactants is substantially in liquid phase, i. said contacting being effected utilizing sufficient agitation so that substantially all of said zeolite is maintained in suspension in the liquid reaction mixture, ii. the amount of unreacted olefin in the reaction mixture being maintained at less than 7 mole percent based on the unreacted isobutane, iii. wherein the mean weight hourly space velocity of the hydrocarbons in the reaction mixture is in the range of 2-20 gram hydrocarbons per hour-gram catalyst; and iv. wherein said contacting is in the presence of an alkylation-promoting amount of a halide containing chlorine, bromine or fluorine.

5. Process according to claim 4 wherein there is present in solution in the reaction mixture from 10 5 to 10 1 mole per mole of isobutane of a halide, said halide being selected from the group consisting of HCl, carbon tetrachloride and the aliphatic saturated monochlorides having 1 to 4 carbon atoms.

6. An isoparaffin-olefin alkylation process which comprises contacting 2,3-dimethylbutene in admixture with C4-C6 isoparaffin having a tertiary carbon atom in the presence of an alkylation-promoting amount of a halide containing chlorine, bromine or fluorine with a substantially anhydrous acidic crystalline alumino-silicate zeolite under alkylating conditions, i. wherein said contacting is at a temperature below the critical temperature of the lOwest boiling hydrocarbon reactant and is at a pressure such that the reactants are substantially in liquid phase, and ii. wherein said product of said contacting comprises 2,3-dimethylbutane.

7. Process for the preparation of an olefin-paraffin alkylate comprising contacting butene-1 with n-butane and with substantially anhydrous acidic crystalline alumino-silicate zeolite, at a temperature in the range of 40*-80*C. and at a pressure such that each of the reactants is substantially in liquid phase, i. said contacting being effected utilizing sufficient agitation so that substantially all of said zeolite is maintained in suspension in the liquid reaction mixture, ii. the amount of unreacted olefin in the reaction mixture being maintained at less than 7 mole percent based on the unreacted n-butane, iii. wherein the mean weight hourly space velocity of the hydrocarbons in the reaction mixture with the catalyst is in the range of 2-20 gram hydrocarbon per hour-gram catalyst; and iv. wherein said contacting of step (i) is in the presence of an alkylation-promoting amount of a halide containing chlorine, bromine or fluorine.

8. Process for the preparation of an olefinparaffin alkylate comprising contacting butene-1 with n-butane and with a substantially anhydrous acidic crystalline alumino-silicate zeolite, at a temperature of 25-120*C. and at a pressure such that each of the reactants is substantially in liquid phase, i. said contacting being effected utilizing sufficient agitation so that substantially all of said zeolite is maintained in suspension in the liquid reaction mixture, ii. the amount of unreacted olefin in the reaction mixture being maintained at less than 7 mole percent based on the unreacted n-butane, iii. wherein the mean weight hourly space velocity of the hydrocarbons in reaction mixture with the catalyst is in the range of 2-20 gram hydrocarbon per hour-gram catalyst; and, iv. wherein said contacting of step (i) is in the presence of an alkylation-promoting amount of a halide containing chlorine, bromine or fluorine.

Images