Olefin Isomerization Process

1. Field of the Invention

This invention relates to an improved process for the conversion of hydrocarbons, and more specifically for the catalytic isomerization of olefinic hydrocarbons.

2. General Background

Olefinic hydrocarbons are feedstocks for a variety of commercially important addition reactions to yield fuels, polymers, oxygenates and other chemical products. The specific olefin isomer, considering the position of the double bond or the degree of branching of the hydrocarbon, may be important to the efficiency of the chemical reaction or to the properties of the product. The distribution of isomers in a mixture of olefinic hydrocarbons is rarely optimum for a specific application. It is often desirable to isomerize olefins to increase the output of the desired isomer.

Butenes are among the most useful of the olefinic hydrocarbons having more than one isomer. A high-octane gasoline component is produced from a mixture of butenes in many petroleum refineries principally by alkylation with isobutane; 2-butenes (cis- and trans-) generally are the most desirable isomers for this application. Secondary-butyl alcohol and methylethyl ketone, as well as butadiene, are other important derivatives of 2-butenes. Demand for 1-butene has been growing rapidly based on its use as a comonomer for linear low-density polyethylene and as a monomer in polybutene production. Isobutene finds application in such products as methyl methacrylate, polyisobutene and butyl rubber. The most important derivative influencing isobutene demand and butene isomer requirements, however, is methyl t-butyl ether (MTBE) which is experiencing rapid growth in demand as a gasoline component.

Pentenes also are valuable olefinic feedstocks for fuel and chemical products. Isoprene, which may be produced by dehydrogenation of isopentene, is an important monomer in the production of elastomers. To an increasing extent, pentenes obtained from refinery cracking units are alkylated with isobutane to obtain a high-octane gasoline component. The principal influence on trends in isopentene demand and pentene isomer requirements, however, is the rapid growth in demand for methyl t-amyl ether (TAME) as a gasoline component. This derivative is of increasing interest as restrictions on gasoline olefins and volatility reduce the utility of pentenes as a gasoline component and as ethers and alcohols are needed for reformulated gasolines with higher oxygen content. This interest may extend to hexenes and higher olefins having tertiary carbons which could be reacted to yield high-octane ethers.

Olefin isomers rarely are obtained in a refinery or petrochemical product in a ratio matching product demand. In particular, there is a widespread need to increase the proportion of isobutene, isopentene and other tertiary-carbon olefins for production of MTBE, TAME and other ethers. Catalytic isomerization to alter the ratio of isomers is one solution to this need. Since ethers must be supplied at lower cost to find widespread use as a fuel product and since isomerization competes with increased feedstock processing as a source of desired isomers, an isomerization process must be efficient and relatively inexpensive. In one aspect, a catalytic isomerization process must recognize olefin reactivity: isobutene in particular readily forms oligomers which could require a reconversion step to yield monomer if produced in excess. The principal problem facing workers in the art therefore is to isomerize olefins to increase the concentration of the desired isomer while minimizing product losses to heavier or lighter products.

Processes for the isomerization of olefinic hydrocarbons are widely known in the art. Many of these use catalysts comprising phosphate. U.S. Pat. No. 2,537,283 (Schaad), for example, teaches an isomerization process using an ammonium phosphate catalyst and discloses examples of butene and pentene isomerization. U.S. Pat. No. 3,211,801 (Holm et al.) discloses a method of preparing a catalyst comprising precipitated aluminum phosphate within a silica gel network and the use of this catalyst in the isomerization of butene-1 to butene-2. U.S. Pat. Nos. 3,270,085 and 3,327,014 (Noddings et al.) teach an olefin isomerization process using a chromium-nickel phosphate catalyst, effective for isomerizing 1-butene and higher alpha-olefins. U.S. Pat. No. 3,304,343 (Mitsutani) reveals a process for double-bond transfer based on a catalyst of solid phosphoric acid on silica, and demonstrates effective results in isomerizing 1-butene to 2-butenes. U.S. Pat. No. 3,448,164 (Holm et al.) teaches skeletal isomerization of olefins to yield branched isomers using a catalyst containing aluminum phosphate and titanium compounds. U.S. Pat. No. 4,593,146 teaches isomerization of an aliphatic olefin, preferably 1-butene, with a catalyst consisting essentially of chromium and amorphous aluminum phosphate. None of the above references disclose the olefin-isomerization process using the non-zeolitic molecular sieve (NZMS) of the present invention.

The art also contains references to the related use of zeolitic molecular sieves. U.S. Pat. No. 3,723,564 (Tidwell et al.) teaches the isomerization of 1-butene to 2-butene using a zeolitic molecular sieve. U.S. Pat. No. 3,751,502 (Hayes et al.) discloses the isomerization of mono-olefins based on a catalyst comprising crystalline aluminosilicate in an alumina carrier with platinum-group and Group IV-A metallic components. U.S. Pat. No. 3,800,003 (Sobel) discloses the employment of a zeolite catalyst for butene isomerization. U.S. Pat. No. 3,972,832 (Butler et al.) teaches the use of a phosphorus-containing zeolite, in which the phosphorus has not been substituted for silicon or aluminum in the framework, for butene conversion. None of the above teach the use of NZMS for selective butene isomerization, and Butler et al. discloses high yields of heavier olefins from butenes at a range of temperatures with a phosphorus-containing zeolite.

U.S. Pat. No. 4,503,282 (Sikkenga) reveals a process for converting linear alkenes to isomerized alkenes using a crystalline borosilicate molecular sieve, with examples demonstrating the conversion of linear butenes to isobutene. U.S. Pat. No. 5,132,467 (Haag et al.), filed Mar. 6, 1991, teaches a combination of two-stage etherification followed by common fractionation and olefin isomerization; the isomerization is carried out over a medium-pore metallosilicate catalyst with a range of ZSMs and MCM-22 being disclosed. The isomerization of olefins using NZMS, containing tetrahedral aluminum, phosphorus and at least one other element, has not been disclosed in the above references.

U.S. Pat. No. 5,107,050 (Gaffney et al.), filed Dec. 28, 1990, discloses butene isomerization using a MgAPSO or SAPO molecular sieve at a temperature above 900.degree. F. U.S. Pat. No. 5,136,108 (Gaffney et al.), filed Mar. 6, 1991, teaches a combination process for producing TAME and/or TAA by reacting tertiary pentenes with methanol and/or water, distillation to separate reactants, and isomerization of C.sub.5 hydrocarbons with return of branched hydrocarbons to TAME/TAA production; preferred isomerization catalysts are SAPOs and MgAPSOs.

"Non-zeolitic molecular sieves" or "NZMSs" as referenced herein include the "SAPO" silicoaluminophosphates of U.S. Pat. No. 4,440,871 (Lok et al.), the "FAPO" ferroaluminophosphates of U.S. Pat. No. 4,554,143 (Messina et al.), and the metal aluminophosphates of U.S. Pat. No. 4,567,029 (Wilson et al.) wherein the metal is at least one of Mn, Co, Zn and Mg. The application of NZMS-containing catalyst to the isomerization of a C.sub.8 aromatics stream is revealed in U.S. Pat. No. 4,740,650 (Pellet et al.). U.S. Pat. No. 4,689,138 teaches a process for isomerizing normal arid slightly branched paraffins using a catalyst comprising SAPO molecular sieves. The use of MgAPSO compositions for hydrocarbon conversion is taught in U.S. Pat. No. 4,882,038. U.S. Pat. No. 5,158,665 (Miller) and PCT Application WO 91/13132 teach a crystalline silicoaluminophosphate having the X-ray diffraction pattern of SAPO-11 and characterized by having phosphorus-alumina-silica surface ratios which differ from the respective bulk ratios, as well as its use in hydrocarbon conversion. None of these references discloses or suggests the present catalyst and process for isomerization of olefins.

It is an object of the present invention to provide an improved process for the isomerization of olefinic hydrocarbons. A corollary objective of the invention is to minimize product losses in an olefin isomerization process.

This invention is based on the discovery that a catalytic isomerization process using a catalyst comprising at least one noncondensed-silica SAPO and having the absence of a platinum-group metal demonstrates surprising efficiency in converting 2-butene to isobutene or 1-butene in a butene-isomerization operation and in the skeletal isomerization of pentenes.

A broad embodiment of the present invention is directed to the catalytic isomerization of olefinic hydrocarbons using a catalyst containing at least one noncondensed-silica SAPO and having the absence of a hydrogenation promoter.

In a preferred embodiment, the feedstock to catalytic isomerization comprises principally butenes. In a highly preferred embodiment, the catalytic isomerization increases the concentration of isobutene in the product. An alternative preferred embodiment comprises the isomerization of linear pentenes to isopentene. Optionally, the feedstock to isomerization comprises a mixture of butenes and pentenes.

In an alternative embodiment, the catalyst comprises a MgAPSO sieve.

These as well as other objects and embodiments will become apparent from the detailed description of the invention.

FIG. 1 shows the approach to equilibrium isobutene concentration in the reactor product, relative to reaction temperature, when employing catalysts of the invention and of the prior art. The proportion of isobutene relative to total butenes for each test was divided by the equilibrium proportion of isobutene, calculated from API Research Project 44, at the respective reaction temperature of the test.

FIG. 2 shows the yield of heavy (high-boiling) material as a percentage of total product, relative to the reaction temperature of each test, for catalysts of the invention and of the prior art. This figure thus indicates the loss of isobutene to heavy product in an isomerization process directed to isobutene reproduction.

FIG. 3 shows conversion and selectivity for the isomerization of 1-pentene to isopentenes over a period of about 480 hours.

FIG. 4 shows .sup.29 Si NMR spectra of SAPO-11 samples of the present invention.

FIG. 5 shows a .sup.29 Si NMR spectrum of a SAPO-11 control sample not of the invention.

FIG. 6 is an illustration of the use of the computer program LINESIM to decompose the NMR spectrum of the SAPO-11 control sample in order to quantify framework silica types.

FIG. 7 compares the proportions of types of framework silica in SAPO-11 samples of the invention and of the control sample.

FIG. 8 compares catalyst selectivity for butene isomerization as a function of conversion for catalysts of the invention and one not of the invention.

FIG. 9 compares catalyst selectivity for pentene isomerization as a function of conversion for catalysts of the invention and not of the invention.

To reiterate, a broad embodiment of the present invention is directed to the catalytic isomerization of olefinic hydrocarbons using a catalyst containing at least one NZMS.

According to the process of the present invention, an olefinic hydrocarbon feedstock is contacted with a catalyst containing at least one NZMS in a hydrocarbon isomerization zone. Contacting may be effected using the catalyst in a fixed-bed system, a moving-bed system, a fluidized-bed system, or in a batch-type operation. In view of the potential attrition loss of the valuable catalyst and of the operational advantages, a fixed-bed system is preferred. The conversion zone may be in one reactor or in separate reactors with suitable means therebetween to ensure that the desired isomerization temperature is maintained at the entrance to each reactor. Reactants may contact the catalyst in the liquid phase, a mixed vapor-liquid phase, or a vapor phase. Preferably, the reactants contact the catalyst in the vapor phase. The contact may be effected in each reactor in either an upward, downward, or radial-flow manner.

The feedstock may contact the catalyst in the absence of hydrogen or in presence of hydrogen in a molar ratio to feedstock of from about 0.01 to about 10. Hydrogen may be supplied totally from outside the isomerization process, or tile outside hydrogen may be supplemented by hydrogen separated from reaction products and recycled to the feedstock. Inert diluents such as nitrogen, argon, methane, ethane and the like may be present. Although the principal isomerization reaction does not consume hydrogen, there may be net consumption of hydrogen in such side reactions as cracking and olefin saturation. In addition, hydrogen may suppress the formation of carbonaceous compounds on the catalyst and enhance catalyst stability.

Alternatively, and particularly when isomerizing pentenes, a stable operation may be maintained in the substantial absence of hydrogen. "Absence of hydrogen" means that free or molecular hydrogen is substantially absent in a combined feed to the present catalyst, thus avoiding the expense of separating unreacted hydrogen from reaction effluent and optionally recycling hydrogen to the feed. Nitrogen or other inert gases may be used for plant pressurization in this option. Hydrogen is considered to be substantially absent at a level below the saturation level in the hydrocarbon feedstock, more usually at a molar ratio of about 0.005 relative to the feedstock and especially at a molar ratio to the feedstock of about 0.001 or less.

In the group of olefinic hydrocarbons suitable as feedstock to the catalytic isomerization process of the present invention, mono-olefins having from 4 to 10 carbon atoms per molecule are preferred. The mono-olefins should be present in the feedstock in a concentration of from about 0.5 to 100 mass %, and preferably from about 5 to 100 mass %, with most of the balance usually comprising paraffins. Butenes are an especially preferred feedstock. The feedstock should be rich in one or more of the linear butenes, i.e., 1-butene, cis-2-butene and trans-2-butene, if isobutene is the desired product.

An advantageous alternative feedstock within the group of preferred olefins comprises pentenes, often designated amylenes. Optimally one or both of the linear pentenes 1-pentene and 2-pentene are isomerized to one or more of the isopentenes 2-methyl-2-butene, 2-methyl-1-butene, and 3-methyl-1-butene. A pentene-containing feedstock also may contain hexenes and, optionally, higher olefins.

The feedstock to the present process may advantageously contain both butenes and pentenes. The olefins in the feedstock may comprise any ratio of butenes to pentenes, but a mixed olefin feed preferably has a molar ratio of butenes to pentenes respectively of from about 10:1 to 1:20, and especially from about 5:1 to 1:9.

The feedstock olefins may be contained in product streams from petroleum-refining, synthetic-fuel, or petrochemical operations such as catalytic cracking, thermal cracking, steam pyrolysis, oligomerization, and Fischer-Tropsch synthesis. Often the feedstock contains paraffins such as butanes, pentanes, and C.sub.6 and higher paraffins. An advantageous feedstock for isobutene or isopentene production is raffinate from an etherification process. The derivation of the feedstock from an etherification process is well known and is described, inter alia, in a paper by Bruno Notari, et al., "Skeletal Isomerization of Olefins," at the 1980 NPRA Annual Meeting in New Orleans on Mar. 23-25, 1980.

These streams may require removal of polar contaminants such as sulfur, nitrogen or oxygen compounds by, e.g., extraction or adsorption to maintain isomerization-catalyst stability. Raffinate from an etherification process would beneficially be water-washed to remove methanol and other oxygenates present at levels which could affect the performance of the present catalyst. Removal of dienes and acetylenes, e.g., by selective hydrogenation or polymerization, also may be desirable.

Isomerization conditions include reaction temperatures generally in the range of about 50.degree. to 750.degree. C. For the isomerization of butenes to increase the concentration of isobutene temperatures in the range of 200.degree. to 600.degree. C. are preferred. Selective butene isomerization to produce 1-butene is effected preferably at temperatures of from 50.degree. to 300.degree. C. Pentene isomerization is advantageously performed at temperatures in the range of about 200.degree. to 500.degree. C. Reactor operating pressures usually will range from about atmospheric to 50 atmospheres. The amount of catalyst in the reactors will provide an overall weight hourly space velocity of from about 0.5 to 100 hr.sup.-1, and preferably from about 1 to 40 hr.sup.-1.

The particular product-recovery scheme employed is not deemed to be critical to the present invention; any recovery scheme known in the art may be used. Typically, the reactor effluent will be condensed and the hydrogen and inerts removed therefrom by flash separation. The condensed liquid product then is fractionated to remove light materials from the liquid product. The selected isomer, e.g., isobutene or 1-butene or the isopentene isomer mixture, may be separated from the liquid product by adsorption, fractionation, extraction or reaction. A preferred reaction for separation of either isobutene or isopentene is etherification for production of methyl or ethyl t-butyl ether or for methyl t-amyl ether, respectively. Production of ethers from tertiary olefins is known in the art and described, inter alia, in J. D. Chase, et al., "MTBE and TAME--a Good Octane Boosting Combo," Oil and Gas Journal, Apr. 9, 1979, pp. 149-152. The raffinate from the separation step may be returned to the isomerization zone for further conversion to the selected isomer. The combination of olefin isomerization and etherification in a process combination has been widely disclosed, e.g., in a paper by Bruno Notari, et al., "Skeletal Isomerization of Olefins," at the 1980 NPRA Annual Meeting in New Orleans on Mar. 23-25, 1980, U.S. Pat. Nos. 3,979,461 (Ancillotti et al.) and 4,554,386 (Groenveld et al.), and French Publication 2 614 297 (Gaillard et al. ), incorporated herein by reference for disclosure of the state of the art of the isomerization/etherification combination.

An essential component of the catalyst of the present invention is at least one non-zeolitic molecular sieve, also characterized as "NZMS" and defined in the instant invention to include molecular sieves containing framework tetrahedral units (TO.sub.2) of aluminum (AlO.sub.2), phosphorus (PC.sub.2) and at least one additional element (EL) as a framework tetrahedral unit (ELO.sub.2). "NZMS" includes the "SAPO" molecular sieves of U.S. Pat. No. 4,440,871, "ELAPSO" molecular sieves as disclosed in U.S. Pat. No. 4,793,984 and certain "MeAPO", "FAPO", "TAPO" and "MAPO" molecular sieves, as hereinafter described. Crystalline metal aluminophosphates (MeAPOs where "Me" is at least one of Mg, Mn, Co and Zn) are disclosed in U.S. Pat. No. 4,567,029, crystalline ferroaluminophosphates (FAPOs) are disclosed in U.S. Pat. No. 4,554,143, titanium aluminophosphates (TAPOs) are disclosed in U.S. Pat. No. 4,500,651, MAPO metal aluminophosphates wherein M is As, Be, B, Cr, Ga, Ge, Li or V are disclosed in U.S. Pat. No. 4,686,093, and binary metal aluminophosphates are described in Canadian Pat. No. 1,241,943. ELAPSO molecular sieves also are disclosed in patents drawn to species thereof, including but not limited to GaAPSO as disclosed in U.S. Pat. No. 4,735,806, BeAPSO as disclosed in U.S. Pat. No. 4,737,353, CrAPSO as disclosed in U.S. Pat. No. 4,738,837, CoAPSO as disclosed in U.S. Pat. No. 4,744,970, MgAPSO as disclosed in U.S. Pat. No. 4,758,419 and MnAPSO as disclosed in U.S. Pat. No. 4,793,833. The aforementioned patents are incorporated herein by reference thereto. The nomenclature employed herein to refer to the members of the aforementioned NZMSs is consistent with that employed in the aforementioned applications or patents. A particular member of a class is generally referred to as a "-n" species wherein "n" is an integer, e.g., SAPO-11, MeAPO-11 and ELAPSO-31. In the following discussion on NZMSs set forth hereinafter the mole fraction of the NZMS are defined as compositional values which are plotted in phase diagrams in each of the identified patents, published applications or copending applications.

The preferred NZMSs are the silicoaluminophosphate molecular sieves described in U.S. Pat. No. 4,440,871. The silicoaluminophosphate molecular sieves are disclosed as microporous crystalline silicoaluminophosphates, having a three-dimensional microporous framework structure of PO.sub.2.sup.+, AlO.sub.2.sup.- and SiO.sub.2 tetrahedral units, and whose essential empirical chemical composition on an anhydrous basis is:

wherein "R" represents at least one organic templating agent present in the intracrystalline pore system; "m" represents the moles of "R" present per mole of (Si.sub.x Al.sub.y P.sub.z)O.sub.2 and has a value of from 0.02 to 0.3; "x", "y" and "z" represent, respectively, the mole fractions of silicon, aluminum and phosphorus present in the oxide moiety, said mole fractions being within the compositional area bounded by points A, B, C, D and E on the ternary diagram which is FIG. 1 of U.S. Pat. No. 4,440,871, and represent the following values for "x", "y" and "z":

The silicoaluminophosphates of U.S. Pat. No.4,440,871 are generally referred to therein as "SAPO" as a class, or as "SAPO-n" wherein "n" is an integer denoting a particular SAPO such as SAPO-11, SAPO-31, SAPO-40 and SAPO-41. The especially preferred species SAPO- 11 as referred to herein is a silicoaluminophosphate having a characteristic X-ray powder diffraction pattern which contains at least the d-spacings set forth below:

Preferably the silica (SiO4) tetrahedra of the SAPO framework are primarily noncondensed silica structures, i.e., most of the silica tetrahedra are separated from each other by PO.sub.2.sup.+ or AlO.sub.2.sup.- tetrahedra. This preferred embodiment is in contrast to a SAPO having a high proportion of silica tetrahedra with one or more adjacent silica tetrahedra. SAPO-11 containing primarily noncondensed framework silica, or "noncondensed-silica" SAPO-11, demonstrates improved conversion and selectivity in the skeletal isomerization of olefins.

The preferred SAPO contains at least about 60% of the silica present as noncondensed silica, preferably about 70% or more, and most preferably about 80% or more of noncondensed silica in the calcined molecular sieve. The proportion of noncondensed silica is determined by .sup.29 Si nuclear magnetic resonance, or NMR, a technique whose implications are understood by the skilled routineer. Solids NMR spectra may be obtained on NZMS samples using a variety of instruments offered in the industry, such as the Bruker MSL-400, and optimally using magic angle spinning (MASNMR) with proton decoupling. Noncondensed silica shows primarily a single NMR resonance having a .sup.29 Si chemical shift of about -93 ppm, referenced herein as "narrow-range" noncondensed silica. Silica whose resonances are deshielded from the -93 ppm value are believed, without limiting the invention thereby, also to be noncondensed silica with typically one--OH moiety representing noncondensed silica at the surface of the sieve; therefore, noncondensed silica is defined for this invention as that having a .sup.29 Si NMR chemical shift of about -93 ppm plus silica whose resonances are deshielded from the -93 ppm value. Shifts shielded more than -93 ppm are characteristic of condensed silicas. In order to determine the proportion of narrow-range noncondensed silica relative to silica in other configurations, the spectra are decomposed into component peaks using the LINESIM computer program. The proportion of narrow-range noncondensed silica as determined from the peak areas is at least about 50% of the silica present, preferably about 60% or more, and most preferably about 70% or more of the silica in the calcined molecular sieve.

The noncondensed silica SAPO-11 generally has a higher proportion of weaker acid sites than the condensed-silica SAPO-11, i.e., the proportion of total acid sites to strong acid sites is higher for SAPO-11 of the invention. As determined by ammonia-temperature-programmed desorption (NH.sub.3 -TPD), although it is recognized that results can vary depending on individual equipment and procedures, the ratio of total acid sites to acid sites determined at 500.degree.-550.degree. C. preferably is at least about 25.

The noncondensed silica structure is prepared by any means suitable to obtain the silica proportions described hereinabove. Preferably, the procedures described in the examples are utilized commencing with an aqueous gel containing a relatively low proportion of silica precursor. Use of a methanol-containing synthesis gel with a higher proportion of silica precursor provides a structure containing condensed silica. Other preparation steps are as described in the referenced U.S. Pat. No. 4,440,871.

Condensed-silica SAPO-11 is prepared in accordance with the teachings of U.S. Pat. No. 5,158,665 (Miller). Miller '665 is incorporated herein by reference for its teachings with respect to preparation and properties of the control catalyst of the prior art.

MeAPO molecular sieves are crystalline microporous aluminophosphates in which the substituent metal is one of a mixture of two or more divalent metals of the group magnesium, manganese, zinc and cobalt and are disclosed in U.S. Pat. No. 4,567,029. Members of this novel class of compositions have a three-dimensional microporous crystal framework structure of MO.sup.-2.sub.2, AlO.sup.-.sub.2 and PO.sub.2 + tetrahedral units and have an essential empirical chemical composition, on an anhydrous basis, of:

wherein "R" represents at least one organic templating agent present in the intracrystalline pore system; "m" represents the moles of "R" present per mole of (M.sub.x Al.sub.y P.sub.z)O.sub.2 and has a value of from zero to 0.3, the maximum value in each case depending upon the molecular dimensions of the templating agent and the available void volume of the pore system of the particular metal aluminophosphate involved; "x", "y" and "z" represent the mole fractions of the metal "M" (i.e., magnesium, manganese, zinc and cobalt), aluminum and phosphorus, respectively, present as tetrahedral oxides, said mole fractions being such that they are within the following limiting values for "x", "y" and "z":

When synthesized the minimum value of "m" in the formula above is 0.02.

The MgAPSO molecular sieves of U.S. Pat. No. 4,758,419 have a framework structure of MgO.sub.2.sup.-2, AlO.sub.2.sup.-, PO.sub.2.sup.+, and SiO.sub.2 tetrahedral units having an empirical chemical composition on an anhydrous basis expressed by the formula:

wherein "R" represents at least one organic templating agent present in the intracrystalline pore system; "m" represents the molar amount of "R" present per mole of (Mg.sub.w Al.sub.x P.sub.y Si.sub.z)O.sub.2 and has a value of zero to about 0.3; and "w", "x", "y" and "z" represent the mole fractions of elemental magnesium, aluminum, phosphorus and silicon, respectively, present as tetrahedral oxides. The mole fractions "w", "x", "y" and "z" are generally defined as being within the limiting compositional values or points as follows:

It is within the scope of the invention that the catalyst comprises two or more NZMSs. Preferably the NZMSs are as a multi-compositional, multi-phase composite having contiguous phases, a common crystal framework structure and exhibiting a distinct heterogeneity in composition, especially wherein one phase comprises a deposition substrate upon which another phase is deposited as an outer layer. Such composites are described in U.S. Pat. No. 4,861,739, incorporated herein by reference thereto. In a highly preferred embodiment the layered catalyst comprises a crystalline aluminophosphate of U.S. Pat. No. 4,310,440 and a SAPO, especially ALPO-11 and SAPO-11.

The NZMS preferably is combined with a binder for convenient formation of catalyst particles. The binder should be porous, adsorptive support having a surface area of about 25 to about 500 m.sup.2 g, uniform in composition and relatively refractory to the conditions utilized in the hydrocarbon conversion process. By the term "uniform in composition," it is meant that the support be unlayered, have no concentration gradients of the species inherent to its composition, and be completely homogeneous in composition. Thus, if the support is a mixture of two or more refractory materials, the relative amounts of these materials will be constant and uniform throughout the entire support., It is intended to include within the scope of the present invention carrier materials which have traditionally been utilized in hydrocarbon conversion catalysts such as: (1) refractory inorganic oxides such as alumina, titanium dioxide, zirconium dioxide, chromium oxide, zinc oxide, magnesia, thoria, boria, silica-alumina, silica-magnesia, chromia-alumina, alumina-boria, silica-zirconia, etc.; (2) ceramics, porcelain, bauxite; (3) silica or silica gel, silicon carbide, clays and silicates including those synthetically prepared and naturally occurring, which may or may not be acid treated, for example attapulgite clay, diatomaceous earth, fuller's earth, kaolin, kieseiguhr, etc.; (4) crystalline zeolitic aluminosilicates, either naturally occurring or synthetically prepared such as FAU, MEL, MFI, MOR, MTW (IUPAC Commission on Zeolite Nomenclature), in hydrogen form or in a form which has been exchanged with metal cations, (5) spinels such as MgAl.sub.2 O.sub.4, FeAl.sub.2 O.sub.4, ZnAl.sub.2 O.sub.4, CaAl.sub.2 O.sub.4, and other like compounds having the formula MO-Al.sub.2 O.sub.3 where M is a metal having a valence of 2; and (6) combinations of materials from one or more of these groups.

The preferred binder to effect a selective finished catalyst is a form of amorphous silica. The preferred amorphous silica is a synthetic, white, amorphous silica (silicon dioxide) powder which is classed as wet-process, hydrated silica. This type of silica is produced by a chemical reaction in a water solution, from which it is precipitated as ultra-fine, spherical particles. It is preferred that the BET surface area of the silica is in the range from about 120 to 160 m.sup.2/ g. A low content of sulfate salts is desired, preferably less than 0.3wt. %. It is especially preferred that the amorphous silica binder be nonacidic, e.g., that the pH of a 5% water suspension be neutral or basic (pH about 7 or above).

NZMS and binder are combined to form an extrudable dough, having the correct moisture content to allow for the formation of extrudates with acceptable integrity to withstand direct calcination. Extrudability is determined from an analysis of the moisture content of the dough, with a moisture content in the range of from 30 to 50wt. % being preferred. Extrusion is performed in accordance with the techniques well known in the art. A multitude of different extrudate shapes are possible, including, but not limited to, cylinders, cloverleaf, dumbbell and symmetrical and asymmetrical polylobates. It is also within the scope of this invention that the extrudates may be further shaped to any desired form, such as spheres, by any means known to the art.

An optional component of the present catalyst is a platinum-group metal including one or more of platinum, palladium, rhodium, ruthenium, osmium, and iridium. Preferably the catalyst is substantially free of a hydrogenation promoter such as a Group VI or Group VIII metal which would result in economically significant losses of olefins to paraffins through hydrogenation. The preferred catalyst contains less than 100 mass parts per million (ppm) on an elemental basis of hydrogenation promoter, and optimally less than about 10 mass ppm. It is especially preferred that the catalyst be substantially free of a platinum-group metal such as platinum and palladium.

The catalyst of the present invention may contain a halogen component. The halogen component may be either fluorine, chlorine, bromine or iodine or mixtures thereof. Chlorine is the preferred halogen component. The halogen component is generally present in a combined state with the inorganic-oxide support. The halogen component is preferably well dispersed throughout the catalyst and may comprise from more than 0.2 to about 15wt. %, calculated on an elemental basis, of the final catalyst.

The optional halogen component may be incorporated in the catalyst in any suitable manner, either during the preparation of the inorganic-oxide support or before, while or after other catalytic components are incorporated. For example, the carrier material may contain halogen and thus contribute at least some portion of the halogen content in the final catalyst. The halogen component or a portion thereof also may be added to the catalyst during the incorporation of other catalyst components into the support. Also, the halogen component or a portion thereof may be added to the catalyst by contacting with the halogen or a compound, solution, suspension or dispersion containing the halogen before or after other catalyst components are incorporated into the support. Suitable compounds containing the halogen include acids containing the halogen, e.g., hydrochloric acid. The halogen component or a portion thereof may be incorporated by contacting the catalyst with a compound, solution, suspension or dispersion containing the halogen in a subsequent catalyst regeneration step. The catalyst composite is dried at a temperature of from about 100.degree. to about 320.degree. C. for a period of from about 2 to about 24 or more hours and calcined at a temperature of from 400.degree. to about 650.degree. C. in an air atmosphere for a period of from about 0.1 to about 10 hours. The optional halogen component may be adjusted by including a halogen or halogen-containing compound in the air atmosphere.

The following examples are presented to demonstrate the present invention and to illustrate certain specific embodiments thereof. These examples should not be construed to limit the scope of the invention as set forth in the claims. There are many possible other variations, as those of ordinary skill in the art will recognize, which are within the spirit of the invention.

Example 1 illustrates the conversion of a feedstock rich in 2-butenes over a catalyst of the invention with reaction temperature as the principal variable. The composition of the feedstock was as follows in mass %:

Product yields are expressed as mass % of the total products. The approach to equilibrium butene-isomer distribution was determined by reference to equilibrium values calculated from free energies of formation contained in "Selected Values of Physical and Thermodynamic Properties of Hydrocarbons and Related Compounds," API Research Project 44 (1953). This approach to equilibrium is expressed for isomerization to isobutene in FIG. 1 as % of isobutene equilibrium relative to reaction temperature.

Catalysts were evaluated using a 1/2-inch stainless-steel tube as a micro-reactor. One gram of catalyst as powder was placed in the reactor. Butene-rich feedstock was charged to the reactor at its vapor pressure at 70.degree. F. by a syringe pump. The reaction temperature was monitored by a thermocouple in the catalyst bed and controlled by heating the reactor in a fluidized sandbath. Sandbath temperature was controlled by a thermocouple. The liquid products were analyzed by vapor-phase chromatography.

Catalyst performance was compared using a Figure of Merit, or FOM. FOM is obtained by multiplying the mass percent of desired product (isobutene or 1butene) in the total reactor products by the ratio of (desired product)/(desired product plus light plus heavy byproducts). The first term is a measure of conversion as well as selectivity, while the second term reflects selectivity. Thus, FOM is a measure of conversion and selectivity with an emphasis on selectivity. FOM is reported hereinbelow at the reactor temperature at which it reaches its maximum value for each of the catalysts.

The process of the present invention was demonstrated by effecting isomerization of 2-butene to isobutene over SAPO-11 catalyst characterized as described hereinabove. The 2-butene-rich feedstock described above was charged to the micro-reactor, operating at atmospheric pressure. Reaction temperature was increased in a series of steps, and two or more product analyses were performed at each temperature. The results shown below represent average results at each temperature:

A control of the prior art was developed to contrast with the process of the present invention. A zeolitic molecular-sieve catalyst containing about 99.9 mass % (silica +alumina) was tested for the isomerization of 2-butene to isobutene.

Tests were performed and results measured using the feedstock and procedures of Example I. Results were as follows:

Temperatures were not increased above 320.degree. C., as the proportion of isomerization became irrelevant due to the increasing predominance of the reaction forming heavy components.

The test results used to develop Examples I and II were compared with equilibrium isomer values based on the aforementioned API Research Project 44, and the results were plotted in FIG. 1. Data based on individual tests are shown in FIG. 1, whereas the tables of Examples I and II are based on averages of test results at substantially equivalent temperatures.

The process based on the catalyst of the invention achieves a substantially higher conversion to isobutene than the catalyst of the prior art. The equilibrium data understate the advantage of the present invention; due to the high yield of heavy product when using the prior-art catalyst.

FIG. 2 shows the yield of heavy (high boiling) material relative to reaction temperature for processes using the catalysts of both the present invention and the prior art. At temperatures of 290.degree. C and above, where isomerization to isobutene becomes significant, the present invention avoids the predominant reaction to heavy product of the prior art.

The comparative maximum FOM and corresponding temperatures were as follows, with two data points presented to show a range of operating temperatures:

Isomerization of 1-pentene to isopentene, as a combination of 2-methyl pentene-1 and 2-methyl pentene-2, was demonstrated over SAPO-11 catalyst. The, SAPO-11 catalyst of Example I was extruded with a silica binder and used for pilot-plant testing of pentene conversion and product selectivity. The feedstock was a mixture of 1-pentene and isopentane containing about 38 mass % pentene. Hydrogen was present at a molar ratio to 1-pentene of about 8. Reactor pressure was about 18 atmospheres, and temperature was varied between 327.degree. and 357.degree. C. After a 50-hour line-out period, pentene conversion averaged about 70% and molar yields from converted pentene were as follows:

The pilot-plant run was maintained over a period of about 480 hours in order to establish catalyst stability. The results of the extended run are plotted in FIG. 3.

Further pilot-plant tests were performed on the Example IV feedstock mixture of 1-pentene and isopentane, using the same bound SAPO-11 catalyst, in order to assess the effect of operating temperature on conversion and selectivity. Operating pressure was about 10 atmospheres, and temperatures assessed were 327.degree. and 427.degree. C. Pentene conversion and molar yields from converted pentene were as follows:

The effect of operating temperature on the oligomerization of normal butenes was investigated. The feedstock consisted essentially of normal butenes and butanes in a 1:1 molar ratio. Pressure was about 18 atmospheres, and hydrogen was present at a molar ratio of about 4:1 relative to the C.sub.4 feedstock. Two operating temperatures as indicated below were investigated at a range of butene conversions. Within the range of 30 to 50% conversion, formation of dimer from converted butenes averaged as follows with variations of only around .+-.10%:

The co-processing of butenes and pentenes by isomerization was investigated in a pilot-plant isomerization facility using a silica-bound 65%-SAPO-11 catalyst. The feedstock had the following composition, after selective hydrogenation for saturation of highly unsaturated compounds, in mass %:

The mole ratio of butenes to pentenes in the feedstock thus was about 1:4.4.

Isomerization pressure was about 18 atmospheres, operating temperature was about 380.degree. C, liquid hourly space velocity was 12 hr.sup.-.sup., and hydrogen was present at a molar ratio of about 3:1 relative to the C.sub.4 /C.sub.5 feedstock. The pilot-plant test was carried out over a period of about 26 days, about half of which was effected with a hydrogenated and/or resin-treated feed. Initial conversion of linear pentenes was about 70%, declining to about 58-60% using treated feedstock at the end of the run. Isopentenes in total product pentenes declined from 67-70 mole % at the start of run to 57-58% at end-of-run.

A feedstock comprising butenes and pentenes was subjected to isomerization in a pilot-plant facility using a silica-bound 80%-SAPO catalyst. Isomerization pressure was about 18 atmospheres, operating temperature was about 327.degree. C., liquid hourly space velocity was 24 hr.sup.-1 , and hydrogen was present at a molar ratio of about 3:1 relative to the C.sub.4 /C.sub.5 feedstock. Feedstock and product composition were as follows in mass %:

The mole ratio of butenes to pentenes in the feedstock thus was about 1:3.8.

The pilot-plant isomerization test thus resulted in the following increases in isoolefin content:

Three samples of SAPO-11 of the invention were prepared as described hereunder and designated SAPO-11A, SAPO-11B, and SAPO-11C, respectively. Molar proportions of 50 moles water to 1.8 moles of orthophosphoric acid as 85 mass % H.sub.3 PO.sub.4 were combined. To each of the solutions for preparation of SAPO11A and SAPO-11B were added 1.0 moles of alumina as Versal 250 with stirring of the mixture. Silica then was added as HiSil 250 in a molar proportion of moles. The directing agent was di-n-propylamine, added in a proportion of 1 moles. The composition of the reaction mixtures therefore was as follows for both SAPO-11A and SAPO-11B:

SAPO-11 seed amounting to 1.0 mass % of the oxides was added and the reaction mixtures were gradually heated respectively to 190.degree. C. (SAPO-11A) and 195.degree. C. (SAPO-11B) and held at that temperature for 12 hours; the SAPO-11B mixture was steam sparged from 100.degree. C. to 150.degree. C. The solid reaction product was recovered by centrifugation, washed with water and dried in air at 100.degree. C. Certain of the synthesis powders were bound with silica as catalysts and sized to 20-40 mesh as described hereinafter. The synthesis powders and bound catalysts then were calcined at 650.degree. C. in flowing air to remove the synthesis template and set the silica binder of the bound catalyst samples.

SAPO-11C was prepared in the same manner except that the di-n-propylamine was incorporated into the mixture prior to the alumina. Following addition of 1.8 moles of the amine and 1.0 moles of alumina as Versal 250, silica then was added as HiSil 250 in a molar proportion of 0.2 moles to provide a SAPO-11C reaction mixture having the following composition:

SAPO-11 seed amounting to 1.0 mass % of the oxides was added and the reaction mixture was heated gradually, using steam sparging from 100.degree. C. to 150.degree. C., to 195.degree. C. and held at that temperature for 12 hours. The solid SAPO-11C reaction product was recovered by centrifugation, washed with water, and dried in air at room temperature.

A sample of SAPO-11 not of the invention was prepared as a control for testing of the catalysts of Example IX and designated SAPO-11X. The control catalyst was prepared in accordance with the teachings of U.S. Pat. No. 5,158,665.

The proportion of noncondensed silica was defined by .sup.29 Si nuclear magnetic resonance, or NMR, with the tests being carried out on a Bruker MSL-400. The spectra were observed using a Chemagnetics/Otsuka Electronics Jumbo probe with .sup.29 Si-NMR MAS Bloch decay methods. No .sup.1 H decoupling was used.

The NMR spectra of the SAPO-11s of the invention (SAPO-11A,-B,-C) were determined as shown in FIG. 4. Noncondensed silica shows primarily a single NMR resonance having a .sup.29 Si chemical shift of about-93 ppm. Silica whose resonances are deshielded from the -93 ppm value is believed, without limiting the invention thereby, also to be noncondensed silica with typically one--OH moiety representing noncondensed silica at the surface of the sieve. Shifts shielded more than -93 ppm are characteristic of condensed silica.

In order to determine the proportion of narrow-range noncondensed silica, the experimental spectra derived from FIG. 4 were decomposed into component peaks using the computer program LINESIM as discussed in Example XIII. The proportion of narrow-range noncondensed silica is at least about 50% of the silica present, preferably about 60% or more, and most preferably about 70% or more of the silica in the calcined molecular sieve.

The NMR spectrum of the control SAPO-11X not of the invention was determined as described in Example XI, and the results are shown in FIG. 5. A relatively high proportion of condensed silica is indicated by a broad NMR resonance pattern, with substantial peaks shielded more than -93 ppm.

In order to determine the proportion of narrow-range noncondensed silica, the spectra were decomposed into component peaks using the computer program LINESIM. The spectrum derived from FIG. 5 was decomposed by an initial manual selection of components subsequently adjusted by LINESIM using a simplex adjustment of peak position, area, and line width of all peaks to minimize the sum of the squares of the differences of the simulated spectrum from the observed. An example of this analysis is shown for the spectrum of the control SAPO-11X as FIG. 6. The result of the analysis is two major peaks, with .sup.29 Si chemical shifts of about -90 ppm and -105 ppm, and two smaller peaks.

The proportion of narrow-range noncondensed silica is represented by the area of the curve having a .sup.29 Si chemical shift of about -90 ppm. The area of curves with peaks deshielded from -90 ppm (.delta.>-90 ppm) also is considered to be noncondensed silica as discussed hereinabove. The curves having peaks shielded from -90 ppm (.delta.<-100 ppm) encompass areas representing the proportion of condensed silica.

The proportion of the types of framework silica also was determined for SAPO-11A, -B, and -C using LINESIM to decompose spectra in the manner described in Example XIII. The results of the analysis are shown in FIG. 7 for the calcined SAPO-11 samples described in Examples IX-XIII. Also shown are proportions of silica types for the SAPOs before calcination. Both the narrow-range and the total noncondensed silica are indicated in the chart. Each of SAPO11A, -B, and -C contain more than 70% noncondensed silica after calcination, and at least 60% before calcination. The control SAPO-11X contained around 50% noncondensed silica before calcination, and a surprisingly lower proportion after calcination. The control sample was the only SAPO which showed a drop in noncondensed silica content upon calcination.

Each of SAPO-11A, SAPO-11B and SAPO-11C of the invention and control SAPO-11X were formulated as catalysts for further testing. Each of the SAPOs was formed into a paste having a proportion of 80 mass% SAPO and 20 mass % silica provided as Ludox AS-40. The bound SAPOs were dried and sized to 20-40 mesh, then calcined in flowing air at 650.degree. C. to remove the synthesis template and set the SiO.sub.2 binder.

The distribution of acid sites on finished catalysts prepared from SAPO-11B, SAPO-11C and SAPO-11X was determined by NH.sub.3 -TPD. Although absolute results can vary depending on individual equipment and procedures, relative values between samples can indicate distinctive properties. The amount of total sites and strong 500.degree.-550.degree. C. sites were recorded as follows, in mol/g of NH.sub.3:

Catalysts formulated according to Example XV were tested for butene skeletal-isomerization performance. Operating conditions were established as a temperature of 482.degree. C., pressure of 9.5 atmospheres, liquid hourly space velocity of 30, and hydrogen to hydrocarbon ratio of 4.

Conversion of n-butenes over the common 36-hour initial period averaged just over 45% for catalysts formulated from SAPO-11A and -B of the invention, and just under 35% for control SAPO-11X catalyst. Comparative selectivity is shown in FIG. 8 as selectivity vs. conversion, and shows a clear advantage for the catalysts of the invention.

Catalysts formulated according to Example XV were tested for pentene skeletal-isomerization performance. Operating conditions were established as a temperature of 327.degree. C., pressure of 18 atmospheres, variable LHSV of 12-48, and hydrogen to hydrocarbon ratio of about 3. Conversion of n-pentenes averaged about 80% for all catalysts. Comparative stability is shown in FIG. 9 as selectivity vs. conversion, and shows a clear advantage for catalysts of the invention formulated with SAPO-11B and SAPO-11C in comparison to performance of the control catalyst.