Title of Invention	REFORMING USING A BOUND HALIDED ZEOLITE CATALYST
Abstract	A catalytic reforming process comprising the catalytic conversion of hydrocarbons to aromatics, said process comprising treating a halided zeolite catalyst (hiz-cat) containing a Group VIII metal at commercial startup conditions and then reforming hydrocarbons, wherein said catalyst is prepared by washing a bound zeolite catalyst base or a bound zeolite catalyst before halide addition. A preferred hiz-cat is a nonacidic Pt K L-zeolite catalyst prepared by a process that includes the steps of preparing a calcined silica-bound K L-zeolite catalyst base; washing said bound zeolite catalyst base with a liquid comprising water; and incorporating Pt and halogen-containing compound(s) comprising chlorine and fluorine into said washed catalyst base. Ammonium chloride and ammonium fluoride are preferred halide sources.

Title of Invention

REFORMING USING A BOUND HALIDED ZEOLITE CATALYST

Abstract

A catalytic reforming process comprising the catalytic conversion of hydrocarbons to aromatics, said process comprising treating a halided zeolite catalyst (hiz-cat) containing a Group VIII metal at commercial startup conditions and then reforming hydrocarbons, wherein said catalyst is prepared by washing a bound zeolite catalyst base or a bound zeolite catalyst before halide addition. A preferred hiz-cat is a nonacidic Pt K L-zeolite catalyst prepared by a process that includes the steps of preparing a calcined silica-bound K L-zeolite catalyst base; washing said bound zeolite catalyst base with a liquid comprising water; and incorporating Pt and halogen-containing compound(s) comprising chlorine and fluorine into said washed catalyst base. Ammonium chloride and ammonium fluoride are preferred halide sources.

Full Text	FORMING USING A BOUND HALIDED ZEOLITE CATALYST Related Patent Applications This application is a continuation-in-part of U.S. Serial. No. 09/004,216, filed January , 1998, which is incorporated herein by reference in its entirety Field of the Invention The present invention relates to improved hydrocarbon conversion processes sing bound and washed, halided zeolite catalysts. In a preferred embodiment the ivention relates to catalytic reforming with a bound and washed, halided Pt L-zeolite atalyst. More specifically, the invention relates to the discovery and control of catalyst ycle length problems that are associated with commercial startup of halided zeolite atalysts containing Group Vni metals. Background The advantages of monofimctional non-acidic platinum L-zeolite catalysts for atalytic reforming were discovered in the late 1970"s. US 4,104,320 to Bernard and Jury discloses that it is possible to dehydrocyclize paraffins to produce aromatics with igh selectivity using a monofimctional non-acidic L-zeolite catalyst. This L-zeolite atalyst has exchangeable cations of which at least 90% are sodium, lithium, potassium, abidium or cesium, and contains at least one Group VIII noble metal or tin or ermanium. Later, Buss and Hughes developed improved L-zeolite catalysts for reforming etroleum naphthas. These catalysts comprise a large pore zeolite exchanged with an Ikaline earth metal containing one or more Group VIII metals (preferably platinum), lee US PatentNos. 4,434,311; 4,435,283; M47^6, and4,517,306. Barium xchanged catalysts were found to provide even higher selectivities than the orresponding alkali exchanged L-zeolite catalysts disclosed in US 4,104,320. The igh selectivity catalysts of Bemard and Nury, and of Buss and Hughes, are all non-acidic" and are referred to herein as "monofimctional catalysts". These catalysts re highly selective for forming aromatics via dehydrocyclization of paraffins. After more than 10 years of extensive research and engineering, the use of h zeolite catalysts for reforming was commercialized. This commercialization required many additional discoveries. Two key discoveries were the criticality of ultra-low ulfur levels in the feed, and the impact of these ultra-low sulfur levels on reactor letallurgy, i.e., the discovery of the need to prevent coking, carburization and metal listing. While commercialization of ultra-low sulfur reforming was being pursued, a econd generation of platinum L-zeolite catalysts was being developed. These new atalysts (referred to herein generically as "hiz-cats") are produced by treating L-eolites with halogen-containing compounds, such as halocarbons (e.g., US 5,091,351 3 Murakawa et al.) or ammonium salts, (e.g., EP 498,182A and US 5,354,933 to )hashi et al.) and Group Vni metals. It is disclosed that these fialided catalysts are seful for preparing aromatic hydrocarbons such as benzene, toluene and xylenes from "e-Cg aliphatic hydrocarbons in high yield. They allow operations at high severity, >lerate a wide range of hydrocarbon feeds, and show high activity and long life in creening studies. Other patents that disclose hiz-cats include US 4,681,865, US ,761,512 and US 5,073,652 to Katsuno et al.; and US 5,196,631 and US 5,260,238 to iurakawa et al. Yet, our attempts to commercialize one of these hiz-cats resulted in a catalyst ath poor performance. When the patented recipe for preparing a bound halided L-eolite catalysts was followed, we discovered that the resulting catalyst was nexpectedly sensitive to reforming startup conditions. The catalyst performed well in :reening studies using high gas velocity, rapid heat rate, "laboratory" startup (LSU) onditions; but surprisingly, its fouling rate was unacceptably high when low velocity, imulated "conmiercial" startup (SCSU) conditions were employed. Scale-up and commercialization of a new process, particularly one using a new ataiyst, entail numerous changes from laboratory procedures. For instance, limitations ue to commercial equipment impact scale-up designs and the way commercial rocesses are operated. For example, commercial gas compressors set limits on gas ow rates, and heat transfer limitations influence reactor configurations, designs, and perating conditions. Moreover, laboratory studies are generally done to get results in a ;asonable time firame, while commercial operations are intended to operate for much )nger times. One example of unpredictability in the process scale-up area is Heyse et I., as described in US 5, 675,376. Here, potentially catastrophic carburization and letal dusting problems were only first observed when low sultur retorming was being sted in commercial reforming equipment. The metallurgy, process configuration, and mperature of commercial reforming operations lead to coke plugging problems not bserved in laboratory studies. Commercialization and scale-up also requires working with quantities of laterials that are 100 to 1,000,000 times larger than used in laboratory studies. This creased scale typically requires that catalyst preparation procedures be modified. A •ocess as simple as drying is done quite differently in the laboratory than at a commercial plant. Often, these scale differences do not significantly affect catalyst erformance, but occasionally they do. Furthermore, the catalyst composition is often modified to meet the demands of commercial operations, e.g. to attain needed crush strength. Also, (unbound) zeolite ystals or powders are often used in laboratory studies. In contrast, bound zeolites are syually used in commercial operations. This is because bound zeolites reduce the ressure drop through large reactor, provide improved gas and liquid flow rates, and are isier to load and unload. Yet it is known that the performance of bound catalysts, ;pecially bound zeolites, can be quite different from that of unbound powders. Dmetimes, the type of binder or binding method affects catalyst performance, ametunes special steps that are only done for bound catalysts, such as calcination of be bound zeolite, affects catalyst performance. Thus, studies on powdered zeolites do not always predict the commercial erformance of bound zeolites. One example that shows the unpredictability of catalyst erformance associated with adding a binder is Mohr et al., in US 5,106,803. Mohr et ., disclose the criticality of the water sensitivity index of a bound L-zeolite catalyst, his property was not observed to be important for the unbound catalyst. Yet one must assxmie that scale-up will proceed satisfactorily, since it is not pssible to predict what factors will be critical for a particular new process or new utalyst. For halided zeolite catalysts, there is no suggestion in the art that there might e problems associated with preparing bound catalysts or with the use of these catalysts I commercial operations, particularly problems related to catalyst startup procedures, deed, there is nothing in the art that teaches the importance of washing or otherwise treating a catalyst base (i.e., a bound zeolite) or a bound zeolite catalyst before adding halide, as described hereinbelow. Treating conventional, non-halided zeolites —including L-zeolite~ with water or an aqueous solution is known. For example, Kao et al., in US 4,987,109 adjust the pH i of L-zeolite crystals by washing them so as to provide a zeolite of pH 9.4-10.0. These washed zeolites are then bound. Washing the zeolite crystals to this pH improves catalyst activity. Also, Poeppelmeir et al., in US 4,595,668, teach having the nobel metal of in a zeolite dispersed in particles having a diameter of less than 7 angstroms. In Example 6, they describe how to prepare a bound catalyst of their invention using an " oxychlorination step. Another example is Poeppelmeir et al., in US 4,568,656. Here a bound L-zeolite is contacted with an aqueous solution containing a Ft salt and a non- platinum salt (e.g., KCl). This mixture is aged to distribute the Ft uniformly (see Example 1.) Also, Buss and Hughes in US 4,721,694 teach preparing platinum Ba L- zeolite catalysts using a barium ion exchange before Ft impregnation. The ion exchange replaces potasshim.w4th barium in the catalyst and is taught to improve selectivity. For example in Col 12, they,discuss treating extruded pellets by ion exchange with a barium solution, followed by Ft impregnation. These patents are all incorporated herein by reference. It is also known to wash powdered zeolitcatalysts containing halides. For instance, Tatsumi et al., have done laboratory studies investigating powdered, unbound Ft K L-zeolite catalysts for aromatization of hexane. They have looked at the effect of adding KCl and the addition method on catalyst performance \Chem. Lett., 387-390 (1993); J. Cat. 147, 311-321 (1994); Bull. Chem. Soc. Jpn., 67, 1553-1559 (1994); and Cat. Lett.. 45 (1997) 107-112]. They have also investigated the effect of other halogen anions (F, Br, I) on hexane aromatization {Cat. Lett., 11 (1994) 289-295 and Sci. & Tech. in Cat., 117-122, (1994)]. They report that chloride and fluoride give the best results. Tatsumi et al., typically add a potassium halide (e.g., KCl) after Ft addition. In general, these authors add halide along with alkali, i.e., then add a refractory alkali halide salt to the catalyst. (Note: these catalysts are not hiz-cats, as defined herein.) In trying to vmderstand why KCl addition improves catalyst selectivity, the authors of the J. Cat., 147 (1994) article add ammonium chloride to a powdered Ft L-zeolite catalyst that had been prepared by ion exchar which is a hiz- cat, did not show the improved selectivity of catalysts with KCl added, leading the authors to discuss the importance of adding salts containing both alkali and halide. Sugimoto et al., in Appl. Cat. A: General 96 (1993) 201-216, have also studied I unbound catalysts. They prepared powdered hiz-cats from a variety of zeolites, including those prepared from an L-zeolite powder. These catalysts were made using a halocarbon treatment following either an alkaline soaking or ion exchange. In either case, the powdered zeolites were washed. Catalysts prepared from these washed powders were tested for n-hexane aromatization. As already noted, these studies were lone on powdered catalysts, not bound catalysts. None of the references discussed above address problems associated with commercial startup ofbound halided zeolite catalysts. Nor do they teach the mportance or desirability of washing bound zeolite bases or bound catalysts, especialli" ion-acidic zeolite bases or catalysts, before adding halide as discussed hereinbelow. " Description of the Figures Figure 1 compares the catalytic reforming performance of a bound halided Pt L-;eolite catalyst prepared in the manner of EP 498,182A or US 5.354^931 after a aboratory startup procedure (curve A) and after a simulated commercial startup SCSU) procedure (curve B). Figure 2 compares the reforming performance after a startup at SCSU conditions fa bound Pt L-zeolite catalyst (circles) prepared in the manner ofEP 498,182 A or US ,354,933 with the performance of a bound and washed Pt L-zeolite hiz-cat of the invention (squares). Description of the Invention One object of the present invention is to provide a reforming process using ound, halided zeolite catalysts (hiz-cats), especially bound, halided Pt L-zeolite italysts, where the catalysts have long cycle lengths after commercial startup. Another bject of the invention is to improve the reproducibility of catalytic operations using iz-cats. As discussed below, we have discovered that unless the catalyst base or the 3und catalyst is washed before halide addition, the catalyst will foul rapidly and have a lort cycle length. In one embodiment, the present invention is a catalytic reforming process using bound, halided zeolite catalyst (a hiz-cat) comprising the catalytic conversion of hydrocarbons to aromatics wherein the catalyst is treated at commercial startup conditions. These commercial startup conditions include "low" gas flow rates (e.g., GHSV of between 400 and 2000 hr"") and "slow" heat-up rates (e.g., between about 5 and 50 °F/hr, and more typically at between about 10-25 °F/hr) during catalyst 3 reduction. Commercial reformer startup conditions are well known to those skilled in the art, especially in contrast with laboratory startup conditions, which are generally used in experimental studies. More particularly, the invention is directed to reforming using halogen-added or halogen-impregnated zeolite catalysts containing a Group VIII metal that have long cycle lengths and maintain very low fouling rates after commercial lb startiup. According to one embodiment, the invention is a catalytic reforming process comprising the catalytic conversion of hydrocarbons to aromatics, said process comprising treating a halided zeolite catalyst containing a Group VIII metal at commercial startup conditions and then reforming hydrocarbons, wherein said catalyst is prepared by washing a bound zeolite catalyst base or a bound zeolite catalyst before halide addition. Adding halide and, if not present, a catalytic metal results in a washed hiz-cat, as the term is used herein. Preferred washed hiz-cats contain both chloride and fluoride. An especially preferred hiz-cat is a bound and washed, halided Pt L-zeolite catalyst, for example a catalyst prepared with about 0.9 wt % initial chloride, 0.9 wt % 3D initial fluoride and about 1 wt % added Pt. Among other factors, the invention is based on our discovery that the reforming performance of bound hiz-cats, especially bound non-acidic, L-zeolite hiz-cats, can be very sensitive to reforming startup conditions. We have discovered that bound hiz-cats that have low fouling rates after being treated under high gas flow rate startup 25 conditions, (e.g., conditions such as those typically used in lab testing imits) do not necessarily exhibit these same low fouling rates when commercial startup conditions are used. Rather, low catalyst activity and/or rapid fouling are often observed when the bound catalyst is started up at simulated conunercial startup conditions. Figure 1 compares the catalytic reforming performance of a bound halided Pt L-zeolite catalyst 30 prepared in the manner ofEP498,182AorUS 5,354,933 after a startup using laboratory startup (LSU) conditions (curve A) and simulated commercial startup (SCSU) conditions (curve B). As the figure shows, the impact of startup conditions on 7 can de very large. we nave unexpectedly discovered that catalyst performance following a startup at SCSU conditions can be significantly improved by washing the catalyst or catalyst base before halide addition. See Figure 2, which compares the performance of a washed hiz-cat with a hiz-cat that has not been / washed. Hiz-cats useful in this invention can be prepared by washing the bound catalyst or preferably the catalyst base (i.e., the bound zeolite) before halogen impregnation / addition. According to one preferred embodiment, the catalyst is prepared by a process that includes binding, extruding, calcining and washing the catalyst base before halide 1/5 addition. In another preferred embodiment, the hiz-cat is produced by a method that includes: a) preparing a calcined silica-bound L-zeolite catalyst base; b) washing said bound zeolite catalyst base with a liquid comprising water; and c) incorporating a Group VIII metal and halogen-containing compound(s) comprising chlorine and fluorine into said washed base to produce a halided zeolite catalyst. Preferably the washed base is calcined before step (c). Non-acidic L-zeolite is a preferred zeolite. In another embodiment of the invention, the catalyst base is ion-exchanged with potassium, wherein washing is part of the ion exchange procedure. In yet another embodiment, the invention is a method of preparing a bound hiz- cat having a cycle length >1200 hr following startup at SCSU conditions, and the catalyst prepared thereby. According to another embodiment, the invention is also the use of a boimd hiz-cat that has been washed before halide addition in a hydrocarbon conversion process, wherein the catalyst is treated at commercial startup conditions. Detailed Description of the Invention In one broad aspect, the present invention is a process that comprises catalytic reforming using a bound halided zeolite catalyst, especially an L-zeolite catalyst comprising chloride and fluoride, which is treated at commerci^ startup conditions. In a preferred embodiment, the invention is a process for producing aromatics coi^prising preparing a halided zeolite catalyst by washing a calcined, bound zeolite catalyst base 30 with a liquid comprising water; adding at least one halogen-containing compoimd and a Group VIII metal compound to said washed base; bringing said catalyst on stream using commercial startup conditions; and contacting said catalyst with paraffinic 8 hydrocarbons at catalytic reforming conditions to produce aromatics. More preferably, the reforming process contacts a stream comprising C6and C7 hydrocarbons and containmg less than 50 ppb sulfur with a halided Pt K L-zeolite catalyst. The halides comprise chloride and fluoride, and said catalyst is prepared by washing a bound zeolite catalyst base before halide addition. Although the terms "comprises" or "comprising" are used throughout this specification, these terms are intended to encompass both the terms "consisting essentially of, and "consisting of, as understood in the patent art, in various preferred aspects and embodiments of the present invention. The terms "comprises" or "comprising" themselves are terms of the patent art. They are intended to be inclusive terms. The terms thus include the specified component or components, or the specified step or steps, being discussed or claimed, as well as any additional components or steps. As used herein, the term "bound", is intended to describe a zeolite, binder combination that is formed into aggregates such as pellets, pills, extrudates and the like. The term "catalyst base", as used herein, refers to a bound zeolite. The term "catalyst" is used herein in a broad sense and includes the final catalyst as well as precursors of the final catalyst. Precursors of the final catalyst nclude, for example, the calcined form of the catalyst containing the catalytic metal md also the catalyst prior to activation by reduction. The term "catalyst" is thus used to efer to the activated catalyst in some contexts herein, and in other contexts to refer to )recursor forms of the catalyst, as will be understood by skilled persons from the context. The catalysts of this invention are bound and comprise a Group VIII metal. Vlthough often discussed herein in terms of Pt being the Group VIII metal, it is ntended that this discussion is exemplary rather than limiting. As used herein, the term "hiz-cat" is mtended to include zeolite catalysts that esult firom addition of halogen-containing compounds to, or from halide impregnation >f, zeolites, catalyst bases (i.e., bound zeolites) or zeolite catalysts (i.e., zeolites ontaining catalytic metal). Hiz-cats are also referred to herein as "halided catalysts" or halided zeolite catalysts". The halides of the hiz-cats are in addition to those that may »e incorporated into the catalyst from the catalytic metal source; that is, this halogen ddition or impregnation is not just that associated with conventional platinum loading nd impregnation, where platinum halides are often used. Nor does this halogen 9 treatment include conventional zeolite ion exchange operations. Zeolite ion exchange sometimes uses a halide salt, such as KCl, to replace the cations in the zeolite; this ion exchange can leave small amounts of halide on the catalyst. Moreover, the term hiz-cat :s not intended to include catalysts where halide is added using alkali halides (e.g., KCl) or alkaline earth halides. As is discussed hereinbelow, we believe that added alkali is letrimental to hiz-cats performance. It is not known in what form the halide is present in hiz-cats. It may be as ionic lalide, neutral halide, or it may be part of compound, such as a silica halide or Pt lalide. Thus, as used herein, the term "halide" is used in a broad sense, and is not itended to denote the ionic state of the added halogen or of the halogen associated with tie catalyst. The halide of the hiz-cat can be chloride, fluoride, bromide, or iodide, or a ombination thereof In a preferred embodiment, the hiz-cat contains chloride or uoride, more preferably both. Especially preferred hiz-cats can be prepared by •eating the bound zeolite or bound catalyst with halocarbons such as freons or with ther chlorine-containing and/or fluorine-containing compounds, e.g. by impregnation ith ammonium chloride and ammonium fluoride. Preferred hiz-cats useful in this invention have high total halide after calcination (this includes all halides, e.g. both ploride and fluoride), i.e., they contain at least about 0.9 wt % halide, preferably at ast about 1.1 wt % and more preferably at least about 1.3 wt % total combined halide. [ore than about 5 wt % halide does not appear to provide significant advantages, ^en chloride and fluoride are both present, the weight ratio of CI to F can vary, referably it is between 1:10 and 10:1. More preferably chloride and fluoride are added a weight ratio of about 1:1. As used herein, the terms "chloride retensivity" and "retained chloride" denotes e residual chloride content of the catalyst after dry-out, reduction and about 300 hr i-stream. Hiz-cats evolve chloride during these steps, especially during reduction, lereafter, the chloride level on the catalyst remains substantially constant as the talyst is brought on-stream and operated. After startup, some halide does continue to olve, but very slowly. Care should be taken not to overheat the catalyst, (i.e., mperatures above about 950°F) since additional chloride will evolve, and this "erheating is not desirable. Based on our studies, we found that bound hiz-cats prepared in the manner of EP 498,182A or US 5,354,933 gave good performance (activity and life) in LSU tests. LSU tests use very high gas velocity (hydrogen at 10,000 GHSV) and a fast heat-up rate (100°F/hr). But, when this catalyst was started-up at simulated commercial gas flow and heat-up rates, (SCSU conditions), it unexpectedly exhibited an unacceptably high fouling rate and short cycle length. See Figure 1, Curve B. Extensive variations in the simulated commercial startup did not overcome the cycle length problem. Indeed, we found that the patented recipe for making bound halided Pt L-zeolite catalysts did not necessarily produce a catalyst suitable for long-term commercial operation using a commercial startup. After much experimentation, we found that bound hiz-cat performance can be significantly improved by washing, e.g., with water, before halide addition. We also found that catalysts that performed well after a SCSU retained less of the added chloride than poor catalysts, even though the retained chloride, i.e. the "chloride retensivity", is independent of startup conditions. Among the many variables examined, and upon further investigation, we also discovered that the sodium content of catalysts that performed poorly was consistently, albeit by a small amount (0.5 wt % vs about 0.4 wt %), higher than that of catalysts with long cycle lengths. Moreover, sodium content seems to correlate with retained chloride. In a preferred embodiment, calcined catalyst base is washed before adding the halogen-containing compounds and the Group Vin metal, e.g., platinum. In this way, these added components are not washed out of the catalyst. It is preferred that the catalyst base is washed with one or more volumes of wash water. The washing desirably removes at least 20%, preferably at least 50%, of the readily removable alkali. Quite remarkably, after platinum and halogen impregnation, the washed catalysts of the invention are amenable to commercial startup. For example, see Figure 2, which compares an "as prepared" bound hiz-cat made in the manner of EP 498,182A or US 5,354,933 with a water-washed bound hiz-cat as described below. The washed hiz-cat shows a much lower fouling rate and an improved cycle length. We have also found that catalyst performance can be improved by some types of ion exchange processes. Cation exchange, such as with potassium and the like, often includes a wash step. When ion exchange comprises a wash step, as described in this invention, hiz-cat performance improves. Here also the wash step ilrpreferably done on the bound catalyst base before halide(s) (and platinum) is added. It is not known what causes the improvement in catalyst performance that is associated with washing. Washing (and the calcining, which is preferably done thereafter) results in numerous changes including the removal of Na, Si, K, S and small amounts of Al. It may be the removal of one or more of these elements or some other change that improves the catalyst performance. We have noted that reduced sodium levels seem to correlate with improved catalyst performance following a SCSU. The bound and washed hiz-cat of the invention has a long cycle length and low fouling rate following a SCSU. The catalyst may exhibit an initial transient period of 100-300 hr, before catalyst performance lines out. Thereafter, it settles down to roughly linear fouling performance. Then, cycle length and fouling rate are measured. Preferred washed hiz-cats have a cycle length of > 1200 hr, preferably >2000 hr and more preferably >4000 hr, following startup at simulated commercial startup (SCSU) conditions as described below. Catalyst fouling rates are preferably less than 0.02 °F/hr after stabilization following a SCSU, more preferably less than 0.01 and most preferably less than 0.003 °F/hr. It appears that fouling rate and cycle length are related to retained chloride after startup and hydrocarbon processing. For example, we started with as-impregnated catalysts that initially contained about 0.9 wt % total chloride and about 0.9 wt % total fluoride. After cedcination, these catalysts contained about 0.75 wt % chloride and about 0.75 wt % fluoride. After startup and reforming, preferred catalysts were found to contain less than about 0.5 wt % chloride, more preferably less than 0.45 wt % and most preferably less than 0.4 wt % chloride. The sodium level of these catalysts is preferably less than about 0.4 wt %. While not wishing to be bound by theory, it is believed that halide evolution luring catalyst startup, particularly HCl evolution, somehow improves hiz-cat performance. Gases, including water and ammonia, evolve from the catalyst during startup. With high-velocity lab startups, these evolving gases are only present for short periods and at low partial pressure; they are quickly removed. In contrast, during commercial startups, especially with bound catalysts, there is more time for evolved vater and ammonia to contact the catalytic metal (e.g., Pt), and these gases are present at higher partial pressures. This can result in Pt migration. Thus, commercial startup conditions can lead to imdesired Pt agglomerates and poor catalyst performance. Thus, it is believed that evolving halides produced during commercial startup of bound hiz-cats help disperse or redisperse Pt, perhaps via mobile Pt or Pt halide species. Catalysts that have not been washed appear to evolve less chloride during startup, and have lower cycle lengths. It appears that for bound hiz-cats it may be necessary to evolve a threshold amount of mobile or mobilizing halide to achieve long cycle lengths after a commercial startup. Generally, non-acidic zeolite catalysts have high loadings of alkali or alkaline earth metals, e.g., about 15 wt % K2O. Potassium is often the predominant metal for non-acidic L-zeolites, as in the examples described herein. Yet, even with so much alkali present, it appears to be desirable to reduce the alkali content of bound hiz-cats somewhat. While not wishing to be bound by theory, it is believed that the readily-emovable (i.e., soluble) alkali, both sodium and potassium, captures evolved chloride from the hiz-cats during startup and thus prevents the desired Pt dispersion. Removal )f this alkali by water washing allows for improved Pt dispersion and catalyst performance. What appears to be important for bound hiz-cats is to have sufficient volved chloride to (re)disperse the Pt in-situ. Thus, we believe that adding refi-actory alide salts, e.g., alkali halides such as KCl, will not provide the desired evolved halide. Refractory halide salts, when added, are not included in calculating the halide content f hiz-cats. We expect that the need to improve catalyst performance by washing before alide addition will apply to other halided, medium pore and large pore zeolite italysts, as well as to the use of these bound hiz-cats in processes other than catalytic iforming. For example, this washing should be advantageous for the halided zeolite vtalysts described in US 5,260,238 to Murakawa et al. Zeolites Catalysts usefiil in this invention comprise one or more zeolites or non-zeolitic olecular sieves and at least one catalytic metal, preferably a Group VIII metal. The italyst usually also includes a binder such as a refractory oxide, e.g., silica, alumina, ilorided alumina or silica-alumina. Preferred zeolites and/or molecular sieves are lected from those of the large and intermediate pore variety. 13 Representative of the large pore zeolites are ZSM-3, ZSM-4, ZSM-10, ZSM-12, ZSM-20, zeolite beta, zeolite omega, zeolite L, zeolite X, zeolite Y, REY, USY, RE¬US Y, mordenite, LZ-210, LZ-210-M, LZ-210-T, LZ-210-A, SSZ-24, SSZ-26, SSZ-31, SSZ-33, SSZ-35, SSZ-37, SSZ-41, SSZ-42, SSZ-44 and MCM-58. ZSM-3 is described in US 3,415,736; ZSM-4 is described in UK 1,117 ZSM-10 is described in US 3,692,470; ZSM-12 is described in US 3,832,449; ZSM-20 is described in US 3,972,983; Zeolite beta is described in US Re. 28,341 (of original US 3,308,069); Zeolite omega is described in US 4,241,036; Zeolite L is described in US 3,216,789; Zeolite X is described in US 2,882,244; Zeolite Y is described in US 3,130,007; LZ-210, LZ-210M, LZ-210-T, LZ-210-A and mixtures thereof are described in US 4,534,853; SSZ-24 is described ik US 4,834,977; „SSZ-26 is described in US 4,910,006; SSZ-31 is described in US 5,106,801; SSZ-33 is described in US 4,963,337; SSZ-35 is described in US 5,316,753; SSZ-37 is described in US 5,254,514; SSZ-41 is described in US 5,591,421; SSZ-42 is described in US Ser. No. 08/199,040; SSZ-44 is described m US 5,580,540; MCM-58 is described in US 5,437,855. The entire contents of all these patents and patent applications are incorporated herein by reference. In a preferred embodiment, the invention uses an L-zeolite or a zeolite having an L-zeolite-type channel structure and size, such as ECR-2, which is described in US 4,552,731, and ECR-31, which is described in US£,624,657 (Vaughan). Preferably the invention uses a monofunctional, non-acidic K L-zeolite, Acidity in the L-zeolite generally leads to poor performance in catalytic reforming. Examples of useful L-zeolites include those described in Uj 3,216,789 (Breck), 4,552,731 (Vaughan), 4,544,539 (Wprtel), 5,491,119 (Verduijn) and 4,530,824 (assigned to Tosoh Ltd.). The entire contents of all these patents are mcorporated herein by reference. One useful aon-acidic L-zeolite is manufactured by Union Oil Product (UOP), Mobile, AL. A preferred non-acidic L-zeolite is manufactured by Tosoh Ltd., Japan, and sold under the lame HSZ-500KOA, For these non-acidic zeolites, potassium is a preferred cation; a jreferred catalyst comprises K L-zeolite. Preferred catalysts used in the process of the present invention are nonofunctional. They do not have the acid function of conventional reforming ;atalysts. In contrast, conventional reforming catalysts are bifunctional, with an acid and a metal function. Examples of monofunctional catalysts include platinum on L-zeolite, wherein the L-zeolite has been exchanged with an alkali metal, as disclosed in US 4,104,320 to Bernard et al.; platinum on L-zeolite, wherein the L-zeolite has been exchanged with an alkaline earth metal, as disclosed in US 4,634,518 to Buss and Hughes; and platinum on L-zeolite as disclosed in US 4,456,527 to Buss, Field and Robinson. The entire contents of all these patents are incorporated herein by reference The term "non-acidic" is understood by those skilled in this area of art, particularly by the contrast between monofunctional (non-acidic) reforming catalysts and bifunctional (acidic) reforming catalysts. One method of achieving non-acidity is by replacing protons with alkali and/or alkaline earth metals in the zeolite. This is preferably achieved, along with other catalyst enhancements, by an ion exchange process on the synthesized zeolite. The composition of type L zeolite expressed in terms of mole ratios of oxides, may be represented by the following formula: (0.9-1.3)M2/nO:Al2O3(5.2-6.9)SiO2:yH2O In the above formula M represents a cation, n represents the valence of M, and y may be any value from 0 to about 9. Zeolite L, its X-ray diffraction pattern, its properties, and method for its preparation are described in detail in, for example, US Patent No. 3,216,789, the contents of which is hereby incorporated by reference. The actual formula may vary without changing the crystalline structure. For example, the mole ratio of silicon to aluminum (Si/Al) may vary from 1.0 to 3.5. As discussed above, one preferred embodiment of the invention uses monofunctional Pt L-zeolite catalysts that have been treated with halogen-containing compounds. This type of halogen treatment is known. For example, US 5jO£l J.51iQ Murakawa et al., discloses preparing a Pt L-zeolite catalyst, and then treating it with a halogen-containing compound. Other related patents that disclose halided L-zeolite catalysts include EP498J82ApjiJS-5,354,933, which discloses co-impregnation of an L-zeolite with NH4CI and NH4F; US 4,681,865,4,761,512 and 5,073,652 to Katsuno et al. These patents are all incorporated herein by reference. One preferred hiz-cat for catalytic reforming comprises halided platinum K L-zeolite catalyst, especially one containing both chloride and fluoride. IS Useful intermediate pore size zeolites are exemplified by ZSM-5, ZSM-11, ZSM-22, ZSM-23, ZSM-35; ZSM-48, ZSM-57, SUZ-4, SSZ-23; SSZ-25; SSZ-28, SSZ-32, and SSZ-36. ZSM-5 is described in US Re. 29,948 (of original US 3,702,886); ZSM-11 is described in US 3,709,979; ZSM-22 is described in US 4,556,477; ZSM-23 is described in US 4,076,842. ZSM-35 is described in US 4,016,245; ZSM-48 is described in US 4,585,747; SUZ-4 is described in EP 353,915; SSZ-23 is described in US 4,859,422; SSZ-25 is described in US 4,827,667 and 5,202,014; SSZ-28 is described in US 5,200,377; SSZ-32 is described in US 5,053,373; and SSZ-36 is described in US Serial No. 60/034,252. The entire contents )f all these patents and patent applications are incorporated herein by reference. In addition to silicon, the useful zeolites herein can contain one or more iramework elements other than, or in addition to, aluminum, e.g., the borosilicate ".eolites. Also, the zeolites can be modified to alter their as-synthesized framework ilica to alumina ratio. Catalysts useful in the present invention can also include non-zeolitic molecular ieves with intermediate or large size pores. Non-zeolitic molecular sieves are licroporous compositions that are formed from [AIO2] and [PO2] tetrahedra and have "ectrovalently neutral frameworks. See US 4,861,743. Also included among the seful zeolites are materials of similar structure or behavior, e.g., crystalline etallophosphates such as those described in US 4,440,871. Non-zeolitic molecular eves include aluminophosphates (AIPO4) as described for example in US 4,310,440, etalloaluminophosphates, as described in US 4,500,651; 4,567,029; 4,544,143; and 686,093 and nonmetal substituted aluminophosphates as described in US Pat. 973,785. Useful catalysts also include intermediate pore silicoaluminophosphates APO"s) as the non-zeolitic molecular sieve component. Intermediate pore SAPO"s ilude SAPO-11, SAPO-31, SAPO-41 and SM-3. US 4,440,871 describes SAPO"s nerally and SAPO-11, SAPO-31, and SAPO-41 specifically. The preparation of SM-md its unique characteristics are described in US 5,158,665. All these patents are jorporated herein by reference. Binders in prepanng catalysts for use in the present invention, the zeolites and/or molecular sieves are bound. They are preferably composited with matrix materials resistant to the temperatures and other conditions employed in hydrocarbon conversion processes. Such matrix materials can include active and inactive materials. Frequently binders, such as naturally occurring clays and inorganic oxides, are added to improve the crush strength of the catalyst. The selection of binders and binding conditions depends on the zeolite and its intended use. The binder material can be selected from among synthetic or naturally occurring zeolites, as well as alumina, clays such as montmorillonite and kaolin, and the refractory oxides of metals of Groups IV A and IVB of the Periodic Table of the Elements. Particularly usefiil are the oxides of silicon, titanium and zirconium, with silica being preferred, especially low acidity silica. Combinations of such oxides with other oxides are also useful, for example, silica-alumina, silica-magnesia, silica-zirconia, silica-thoria, silica-beryllia, silica-titania, titania-zirconia, silica-alumina-thoria, silica-alimiina-zirconia, silica-alumina-magnesia and silica-magnesia-zirconia. These oxides can be crystalline or amorphous, or can be in the form of gelatinous precipitates, colloids, sols, or gels. Silica in the form of a silica sol is a preferred binder. A preferred silica sol has about 30 wt % silica and contains small particles (7-9 nm in diameter), which result in catalysts with good attrition resistance and excellent crush strengths. Forming pellets or extrudates from zeolites and/or molecular sieves generally involves using extrusion aids and viscosity modifiers in addition to the binders. These additives are typically organic compounds such as cellulose based materials, for example, Methocel® sold by Dow Chemical Co., ethylene glycol, and stearic acid. Many such compounds are known in the art. It is important these additives do not leave a detrimental residue, i.e., one -with undesirable reactivity or one that can block pores, after calcination. For this invention, it is especially desirable that such residues do not add significant amounts of alkali or alkaline earth ash to the catalyst. The above-described washing will remove low levels of these materials. The residue from the extrusion aid is preferably less than a few tenths of a percent, more preferably less than 0.1 wt%. Methods for preparing catalyst compositions are well known to those skilled in -V6- the art and include such conventional techniques as spray drying, pelletizing, extrusion, various sphere-making techniques and the like. The newly developed method of in-extrudate formation of the zeolite/binder as described in Miller, US 5,558,851 and US 5,514,362 can also be used. The entire contents of these patents are incorporated herein by reference. The relative proportions of zeolite to the binder/matrix can vary widely. Generally the zeolite content ranges from between about 1 to about 99 weight percent, and more usually in the range of from about 5 to about 95 weight percent, of the dry composite, more typically 50-85%. The invention requires a bound zeolite. It is preferred to use whole extrudates rather than crushed extrudates or unbound zeolites in commercial operations. Bound zeolites reduce the pressure drop through a reactor, provide improved flow rates, and are easier to load and unload. However, the difflission characteristics of whole extrudates are quite different from those of unbound powdered zeolites. The interaction of a diffusing gas, such as a halocarbon, is different for a powdered versus a bound zeolite. Diffusion differences would also be especially significant if the catalyst evolves materials such as gases or solids, for example during startup. Moreover, the act of binding itself— including selecting binding materials and the binding method ~ as well as calcining can affect catalyst performance. For example, the binder can interact with the sieve by simply blocking internal and external sites or by chemical interaction e.g., alkali from a basic sieve like the preferred L-zeolite of the present invention and silica are known to react imder thermal and hydrothermal calcination conditions). Moreover, the distribution of impregnated materials may vary considerably on zeolite (owder versus bound zeolites. Thus, studies on powders alone cannot be used to iredict commercial performance of bound zeolite catalysts. Group Vm Metal It is preferred that the catalyst herein comprises at least one Group VIII metal, referably a noble metal (Pt, Pd, Rh, Ir, Ru, Os) and more preferably platinum, referred amounts of metal are 0.1 to 5 wt %, more preferably 0.1 to 3 wt %, and most referably 0.3 to 1.5 wt %, based on the L-zeolite. Platinum compounds that form psitively charged platinum complex ions in solution are the preferred source of latinum. Platinum tetraammine chloride and nitrate are especially preferred. To these Group VIII metals can also be added one or more non-platinum group metals such as tin, indium and metals of Group VIIB such as rhenium. Examples include Pt/Sn, Pt/Pd, Pt/Ni, and Pt/Re. These metals can be readily introduced into the composite employing a variety of known and conventional techniques, e.g., ion-exchange, incipient wetness, pore fill, impregnation, etc. Care should be taken so that the Group VIII metal, e.g., platinum, is incorporated in a manner that results in excellent and uniform dispersion. The incipient wetness impregnation method is preferred. Washing and Adding Halides According to one embodiment, the catalyst used in this invention is prepared by process that includes extruding and then washing either the bound catalyst or the atalyst base before halide addition. Preferably, the catalyst is prepared by a process lat includes the steps of: (a) preparing a calcined silica-bound zeolite catalyst base; (b) cashing said bound zeolite catalyst base with a liquid comprising water, and (c) icorporating a group VIII metal and halogen-containing compound(s) comprising ilorine and fluorine into said washed base to produce a halided zeolite catalyst, referably, the catalyst is prepared using a low alkali, e.g., low sodium, extrusion aid in ep (a). The Pt and halogen-containing compounds can be incorporated sequentially or, eferably, simultaneously. The term washing, as used herem, is meant to include any process where water excess of the material"s pore volume is contacted with the catalyst base. Preferably, ; washing is done with clean water that is roughly neutral (pH 5 to 9,) low in sulfur g., sulfate-free,) and low in alkali and alkaline earth content ( It is preferred that the calcined catalyst base or calcined catalyst is washed "ore halide addition and that at least 20%, preferably at least 50%, of the readily lovable alkali is removed. The "readily removable" alkali is the alkali that can be shed out of the catalyst base after 5 washings (1:1 volume of liquid to wt of rudate) at ambient temperature. However, the temperature, amoimt of water and itact time in the washing steps are not critical, as long as the washing improves cycle 5th following startup at SCSU conditions. The wash water may contain additives, ti as alcohol, although this is not preferred. There are many ways to wash the catalyst, or preferably the catalyst base, that will be apparent to those skilled in the art. These include, for example, cover and drain, dipping, and belt washing. The type ofwasher is not believed to be critical. The washed base is then dried. Although not required, it is preferred to dry and calcine the washed base produced in step (b) prior to incorporating the halogen-containing compound(s). As noted earlier, we have found that washing the bound catalyst or the catalyst base before halide addition is important. As will be appreciated by those skilled in the art, there are various ways to perform the washing step; this is especially true as one moves from laboratory to flill commercial scale manufacturing. For example, the washmg can be static or stirred. The solution, or the solution and the extrudate may be stirred. The wash water may be recirculated, with or without fresh water makeup, in a recycle mode. A bleed stream of effluent wash water may be removed and make-up or added fresh wash water may be used. The wash process may be operated in a variety of batch, semi-continuous or continuous modes. Generally, the details of the wash step will vary someA\^t, depending on the equipment used for washing. It is not necessary or possible to specify all possible wash conditions here, but these will be evident to the skilled artisan. Three key wash variables are contact time, number of washes and wash temperature. A longer contact time (between the calcined catalyst base and the water) allows for more efficient removal of the silica and alkali. The contact time is preferably long enough to dissolve the maximum amount of material (detritus), but not so long as to be imeconomical due to slow throughput. Conversely, too short a contact time could mean that all the detritus was not dissolved or removed. Residual detritus can cause pore blockage and/or loss of surface area, with resultant poorer catalyst performance. More than one washing step is advantageous. Preferably the number of washes is between 2 and 20, more preferably between 3 and 8, and most preferably between 3 and 5. We have found that the temperature of the wash water is important for facilitating efficient and sufficient washing. Washing at a temperatiu-e above ambient, for example around 140°F and 180°F, has been found to give superior recovery of iorosity and higher surface areas for the resulting washed catalyst base. Wash amoieni and 210F, more preferably between 50 and 200°F, and most preferably between 100 and 190°F. In one preferred embodiment, the calcined catalyst base is washed at least twice with wash water having a temperature above about 100°F. More preferably, the calcined catalyst base is washed at least three times with wash water having a temperature above about 140°F. A well-washed catalyst base will demonstrate good catalytic performance after the catalytic metals are loaded. We have found that after room temperature washings, the surface area and tnicroporosity of the resulting washed calcined base is not always equal to (or greater Jian) that of the starting calcined base before washing. While not wishing to be bound )y a theory, we believe that this lower surface area and porosity are due to incomplete emoval of the alkali and silica during the washing steps. We have solved this problem by using higher wash temperatures and/or increased number of washing steps. We have bund that with five washes with water in the 140°F to 180oF range, the resulting cashed calcined base generally has equal or superior surface areas and microporosities s compared to the starting calcined base. Preferred wash conditions thus include evated temperatures and multiple washings. Washing has also been found to increase the surface area and porosity of the :truded base, as measured by the micropore volume. Preferably, the washing is done itil the surface area of the washed catalyst base after calcination returns to at least 80%, 3re preferably to at least 90%, and most preferably to at least 100% or more, of the I rface area of the calcined catalyst base before washing. I Although hiz-cats can contain only one halide, it is preferred that they contain )re than one halide, e.g., chloride and fluoride. Halogens, when there are more than 3, can be added separately or, preferably, at the same time. The halogen(s) can be led during platinum addition or in a separate pre- or post-addition step. In a preferred bodiment, they are added during Pt impregnation. The preferred amount of halide, ., chloride and/or fluoride, in the calcined catalyst depends on the type of zeolite, the alytic metal and its loading, and the type of hydrocarbon conversion. For L-zeolite ilysts used for catalytic reforming, which contain about 1.0 wt % Pt, the chloride il after calcination is preferably between 0.4 and 1.0 wt %, more preferably between and 0.8 wt %. Preferably, fluoride is also present, preferably at levels between 0.5 and 0.9 wt %, after calcination, more preferably between 0.6 and 0.8 wt %. The resulting bound hiz-cats are especially useful for preparing aromatic hydrocarbons such as benzene, toluene and xylenes from C6-C8 aliphatic hydrocarbons. Bound hiz-cats can be prepared using a variety of halide sources; preferred hiz-cats are prepared using amine or ammonium salts. The halides are preferably added by contacting the washed catalyst base with one or more of these halide salts. For L-zeolite reforming, preferred salts are ammonium chloride and ammonium fluoride. These halide salts and the catalytic metal(s) can be added sequentially in any order, or the salts can be added along with the metal, which is preferred. When determining the halide on the catalyst, it comprises both the halide added from the (amine) halide salt and any halide associated with the catalytic metal. When halide is added after Pt addition, it is preferred to add the halide before metal reduction. For example, one preferred halided Pt L-zeolite catalyst is prepared by impregnating a washed and calcined bound zeolite as described above with an impregnating solution of ammonium chloride, ammonium fluoride and platinum tetraammine chloride. Preferably, in amounts added yield an intermediate having 0.9 ivt % each for fluoride and chloride and 1.0 wt % Pt. This intermediate is vacuum dried It about 40-100°C and then calcined, preferably at a temperature from 260°C to SSO"C, !.g., at about 300EC. When amine halides are used, the calcining should be done so as 0 decompose these salts slowly. Care should be taken not to dry or calcine under onditions that would disturb the even distribution of the ingredients. A uniform istribution, i.e., a flat profile across the bound catalyst, is preferred for both Pt and the alides. If desired, an ion-exchange step, for example with potassium ions, e.g., KCl or NO3, can be done before the wash step to reduce acidity in the catalyst. As will be ■cognized by those skilled in the art, those ion exchange reactions that would be usefiil the present invention should be carried out within controlled pH parameters, e.g., stween about 4 to 10.8. This will avoid damage to the catalyst or binder and, for talysts that are intended to be non-acidic, will avoid adding any significant acidity. Commercial Start-up and Cvcle Length As noted earlier, commercial startup conditions differ substantially from >oratory startup conditions. These differences are control temperatures, and heat transfer considerations, and 2) the ability to achieve practical flow rates associated with use of commercial recycle compressors in conventional catalytic reformers. Thus, as used herein, commercial startup conditions for reforming encompass: low gas flow rates. For example a GHSV of between 400 and 2000 hr*", more preferably between 600 and 1500, and most preferably around about 1000 hr". A commercial start-up also means having "slow" heat-up rates, for example, between 5 and 50°F/hr, more preferably between 10 and 25°F/hr during catalyst reduction above about 500°F. This slow heat-up minimizes catalyst damage. Commercial startups may also include holding the temperature at various levels for several hours or days. This can, for example, facilitate complete reduction of the ;atalyst metal. Test startup conditions for SCSU and LSU are shown in Table 1. Hommercial startups for hiz-cats used for reforming generally include the following iteps: 1) dryout of the catalyst, 2) reduction of the catalyst, and 3) introduction of the eed into the reformers. These steps are usually done in-situ after loading the catalyst ato the reformer(s), but some or all of these steps can be done outside the reactor, i.e., x-situ. Preferred catalysts useful in this invention have good activity and maintain low )uling rates over many weeks after startup at SCSU conditions. Preferred catalysts ave a cycle length > 1200 hr, preferably >2000 hr and more preferably >4000 hr. The rm "cycle length", as used herein, is the time (oftentimes extrapolated) from start of m (begiiming with feed addition) until the temperature reaches 920°F. Temperature is Ijusted throughout the run so that the catalyst produces a specific target wt % omatics in the C5+ product. This target depends on the composition of the raffinate ed and is typically between 45 and 50 wt % aromatics. Transient temperature cursions are sometimes seen during the first 100-300 hr of a run. When present, ;se temperature excursions (but not the time) are ignored in determining the cycle igth. Cycle length depends on many factors, includmg for example, the H2 to drocarbon ratio, the composition and type of feed, the space velocity (LHSV,) which related to the feed to catalyst ratio, and the severity (e.g., operating temperature or smatively, degree of conversion). To better enable one skilled in the art to ierstand the invention and determine its scope, we have developed a simple test to asure the cycle length of a catalyst (see Example 9). The bound and washed hiy-cats useful in this invention have a longer cycle length after a commercial startup at SCSU conditions than catalysts prepared in a manner similar to EP 498,182A or US 5,354,933, Ex. 4, as described in Example 1 hereinbelow. Conversion Processes According to one embodiment, the invention is a catalytic reforming process, ;.g., low sulfur reforming using a bound, halided Pt L-zeolite catalyst. Catalytic eforming is well known. For example, it is described in the book, Catalytic leforming, by D. M. Little, PennWell Books (1985), which is incorporated herein by eference in its entirety. Preferred reforming process conditions include a temperature etween 700 and 1100°F, more preferably between 800 and 1050°F; a pressure between and 400 psig, more preferably between 15 and 150 psig; a recycle hydrogen rate ifiicient to yield a hydrogen to hydrocarbon mole ratio for the feed to the reforming action zone between 0.1 and 20, more preferably between 0.5 and 10 and most ■eferably between 2 and 6; and a liquid hourly space velocity (LHSV) for the drocarbon feed over the reforming catalyst of between 0.1 and 10 hr"1, more eferably between 0.5 and 5 hr"". Reforming produces hydrogen. Thus, additional drogen is not needed except when the catalyst is reduced upon startup, and when the ed is first introduced. Once reforming is underway, part of the hydrogen that is educed is preferably recycled over the catalyst. The feed to the reforming process of the present invention typically is or is ived from naphtha that contains at least some acyclic hydrocarbons or ylcyclopentanes. Particularly preferred feeds include rafiinates from an aromatics raction unit, such as a Udex raffmate; C6-C9, C7-C9, Cg-Cp, Ce-Cg, C7-C8, C7 and Cg ;tions, especially fractions containing mixtures of five and six-member saturated gs. Preferably, these fractions are obtained by distilling wider boiling range naphtha, by distilling a raffinate from an aromatics extraction unit. The feed should be stantially free of sulfur, nitrogen, metals and other known poisons. These poisons be removed by first using conventional hydrofining techniques, then using sorbents emove the remaining sulfur compounds and water. In a preferred embodiment, the rocarbons contacting the catalyst are substantially dry and free of sulfur, i.e. sulfur Is are preferably maintained below about 50 ppb, preferably below about 25 ppb more preferably below about 10 ppb. Sulfur removal systems are well known in art. iney include mild reforming using hydrogen and Pt on alumina, followed by sorption. Sorbents are well known in the art; examples include K on alumina, nickel oxide, and manganese oxide. The feed can be contacted with the catalyst in a fixed bed system, a moving bed system, a fluidized system, or a batch system. Either a fixed bed system or a moving bed system is preferred. In a fixed bed system, the preheated feed is passed into at least one reactor that contains a fixed bed of the catalyst. The flow of the feed can be upward, downward or radial. The effluent from the catalytic reforming reaction zone can be separated into the desired streams or fractions. Benzene, toluene and C8 iromatic streams can be recovered using conventional techniques such as distillation md extraction. Although discussed above in terms of catalytic reforming, the invention is pplicable to a variety of catalytic conversion processes that use bound, halided zeolite atalysts to convert feed to products. Preferred hydrocarbon conversion processes iclude dehydrocyclization, especially dehydrocyclization of C6 to Cg paraffins to romatics; catalytic reforming; non-oxidative and oxidative dehydrogenation of ydrocarbons to olefins and dienes; dehydrogenation of ethylbenzene to styrene and/or ehydrogenation of isobutane to isobutylene; conversion of light hydrocarbons to omatics; transalkylation of toluene to benzene and xylenes; hydrodealkylation of kylaromatics to aromatics; alkylation of aromatics to alkylaromatics; production of els and chemiceils from syngas (H2 and CO); steam reforming of hydrocarbons to H2 d CO; production of phenylamine fi-om aniline; methanol alkylation of toluene to lenes; and dehydrogenation of isopropyl alcohol to acetone. More preferred drocarbon conversion processes include dehydrocyclization, catalytic reforming, hydrogenation, isomerization, hydrodealkylation, and conversion of light drocarbon to aromatics, e.g., Cyclar-type processing. These processes, their immercial startup conditions and their usefiil range of process operating conditions are well known to those skilled in the art. These processes can be carried out in a single ctor or in a series of reactors, at least one of which contains the bound and washed -cat described hereinabove. In a preferred embodiment, the invention is a hydrocarbon conversion process iprising the catalytic conversion of a hydrocarbon-containing feed to product under 25 conversion conditions that include a commercial-type catalyst startup (at a low gas flow rate and a slow heat-up rate), said process using a bound and washed halided zeolite catalyst containing a Group VIII metal, wherein said halided catalyst has a cycle length of>1200hr following said startup. In another embodiment, the invention is a lydrocarbon conversion process, comprising the catalytic conversion of a hydrocarbon-ontaining feed to product under conversion conditions that include a catalyst startup at low gas flow rate and a slow heat-up rate, said process using a halided zeolite catalyst ontaining a Group VIII metal, said halided catalyst prepared by a process comprising ashing a bound zeolite catalyst base or catalyst before halide addition and before duction. According to another embodiment, the invention is the washed halided zeolite talyst. A preferred catalyst has a cycle length >1200 hr following startup at SCSU nditions. As discussed earlier, one preferred catalyst is prepared by a process that :ludes the steps of preparing a calcined silica-bound zeolite catalyst base; washing d bound zeolite catalyst base with a liquid comprising water; and either [lultaneously or sequentially incorporating a group VIII metal and one or more ogen-containing compoimd(s) comprising chlorine and fluorine into said washed ie to produce a halided zeolite catalyst. Preferably, the catalyst is prepared using a ca sol that is low in alkali, and a low-alkali extrusion aid. The extruded base is lined and then washed before halide addition. The washed base can again be lined (for example, at 950-1150E F, preferably between 1000-11OOE F) prior to Drporating the halogen-containing compound(s). Preferably, the halides are added ig with the Pt. Examples To obtain a more complete understanding of the present invention, the following emples illustrating certain aspects of the invention are set forth. It should be erstood, however, that the invention is not intended to be limited in any way to the secific details of the examples. Example 1 - Preparing a Halided Platinum L-zeolite Catalvst (Comparative) A halided platinum L-zeolite catalyst was prepared in a manner similar to nple 4 of EP 498,182A or US 5,354,933,. To 100 parts by weight of K L-type ■^- zeolite (Tosoh, Ltd., Japan), 20 parts by weight of a silica binder (SI-350 silica sol, sold by Cat. & Chem. Ind. Inc. [CCIC], Japan) were added with mixing. This mixture was kneaded and molded, and then air-calcined at 500°C (932°F) for 2 hr to produce a molded calcined silica-bound L-zeoIite extrudate. An impregnation liquid comprising 0.166 g of ammonium fluoride, 0.075 g of ammonium chloride, 0.171 g of platinum tetraammine chloride and 4.8 g of ion exchange water was prepared. This liquid was slowly dropped into 10 g of the molded L- zeolite with stirring. The resulting solid was vacuum dried and then treated at 300°C (572°F) for 3 hr in dry air. The resulting hiz-cat contained about 0.7 wt % F and 0.7 wt % CI. The sodium was 0.56 wt %, as measured by atomic adsorption. Example 2A ~ Large Scale Preparation of Washed Base A calcined L-zeolite extrudate was prepared on a large scale in a maimer similar to Example 1 using a Tosoh L-zeolite, SI-350 silica sol, and methyl cellulose (Methocel"^ as an extrusion aid. The extrudate (1800 lb) was placed in a large tank and covered with wash water (1.5 g/g base) that was low in alkali (4 ppm Na2O and 1 ppm K2O), alkaline earths, and sulfur species. The water, which had a pH of 7.3, was added over about 30 min and then drained from the tank over about 30 min. The extrudate was washed twice more in the same way. The resultant material was then dried and calcined at 950°F in a rotary calciner. The Na content of the final wash water was less than 100 ppm and the K content was less than 150 ppm. The washing reduced the Na content of the catalyst base from 0.48 to 0.38 wt %. This triple wash procedure was repeated on another catalyst base sample. The filtrate water contained the following alkali and silicon levels (ppm). Example 2B - Preparing a Finished Catalyst To make a finished catalyst, 160 g (on a volatile-free basis) of the washed and calcined extrudate of Ex. 2A was impregnated by pore fill at about 100% of the pore 2-1 volume. The pore fill solution comprised 2.86 g of tetraammine platinum chloride (56% Pt), 2.8 g of ammonium fluoride, and 1.3 g of ammonium chloride dissolved in water. Afler completion of the pore fill, the impregnated base was allowed to stand for 2hr. It was then put on a rotaiy evaporator under 22 inches of vacuum for three hr at 212°F. The resultant dry impregnated extnidate was heated fi-om 75° to 575""F over a one-hr period and then calcined at 570°F for 1 hr with 10 SCFH of dry air flowing through the bed. Analysis of the catalyst showed it contained 1 wt % Pt, 0.74 wt % CI and 0.79 wt % F. Example 3A - Single Washed Catalyst A calcined extnidate was prepared comprising 83% Tosoh L-zeolite (0.28 wt % Na) bound with 17% SI-350 silica. It had a Na content of about 0.5 wt %. It was washed with deionized water (2.5 cc H20/g of base) at room temperature with stirring for 15 minutes. This wash time was sufficient to achieve good contacting of the solids and liquids. The wash water contained 230 ppm Na and 450 ppm K as analyzed by ICP. The washed extnidate was then dried and calcined in air at 1000 °F. This wash and calcine procedure reduced the sodium content of the base to between 0.37-0.41 wt%. Example 3B -- Impregnation Procedure and Resulting Catalvst Properties The procedure of Ex. 2B was followed usmg the base of Ex. 3 A. The resulting catalyst made fi-om the washed base of Ex. 3 A contained 0.81 wt % CI and 0.72 wt %°F and nominally (i.e., as added) 1 wt % Pt. Example 4A - Triple Washed Base The procedure of Ex, 3 A was repeated except that the wash step was repeated 3 imes. The sodium content of the catalyst base was 0.38 wt % after the third washing, lie filtrate water contained the following alkali levels (ppm in solution) as analyzed by CP. 3rd wash water 26 41 A repeat of this triple wash procedure resulted in filtrate water containing the following materials (ppm). One must take care to avoid over-washing the catalyst. Over-washing can lead 3 excess acidity, which will result in decreased selectivity. Generally, washing with bout 1-3 volumes of water per weight of catalyst is adequate. We have used 2.5. Example 4B — Impregnation and Catalyst Properties The procedure of Ex. 2B was followed using the base of Ex. 4 A. The resulting italyst made firom the washed base of Ex. 4A was nominally loaded to 1.0 wt % Pt and mtained 0,69 wt % CI and 0.65 wt % F. Example 5 —Pilot Plant Configuration Catalyst performance was assessed in pilot plants employing a 1-inch ID type 6 stainless steel in a vertical tubular reactor. Between 40 and 130 cc of whole trudate catalyst (about 0.07 inch, screened to L/D of 2 to 3) was loaded into the ictor. Six thermocouples were equally spaced vertically in the catalyst bed. The term Italyst temperature" as used herein is the catalyst average temperature (CAT). It is culated firom the six thermocouple temperatures as follows: CAT = 1/10 (TCi + TC6) + 1/5 (TC2 + TC3 +TC4 + TC5) ere TC is the temperature of thermocouple Cx, and the thermocouples are arranged I numbered in order starting at the top of the catalyst bed. Example 6 -- Start-up Procedure for Lab Test Units The LSU was done in the pilot plant of Ex. 5. The gases were not recycled and evolved gases, including any halides, were vented. The procedure included drying 29 «..«ivuu».iug the catalyst in H2 from room temperature to 1000°F at a ramp rate of 100°F/hr, and then maintaining the catalyst at about 1000oF for 24 hr. GHSV was maintained at 10,000 hr"" during drying and reduction. After completing the LSU, the catalyst was cooled to approximately 800°F and then feed was introduced. The temperature was then adjusted to achieve the desired total wt % aromatics in the C5+ product. Although this LSU procedure is not commercially practical, it is much easier to perform and much quicker than a commercial type startup. Simulated commercial startups are more complex and can take a week or more to complete. Example 7 - Commercial Start-up A commercial reformer startup procedure was simulated using the reactor of Ex. . The gases, including evolved gases, were recycled. The startup included drying the atalyst in N2 from room temperature to 500°F for 79 hr, then heating the catalyst in a tiixture of 10% H2 in N2 from 500 to 932°F at a rate of 10°F/hr over a period of about 3 hr, and then maintaining the catalyst at about 932°F for 24 hr. The gas hourly space elocity (GHSV, which is volume of gas at STP per volume of catalyst) was maintained 11300 hr"" for the drying and reduction periods. The total elapsed time between italyst loading and feed introduction was 146 hr. Table 1 compares LSU and SCSU conditions. Example 8 -- Feed and Conditions Used for Tests The feeds used for the performance tests were C6-C7 UDEX raffinates from a conventional catalytic reformer that were dried and hydrotreated to Feed 1 was a lighter Udex raffinate than Feed 2. It contained 10.5 wt % C5 hydrocarbons and less C7"s than C6"s. Feed 2 contained less than 3 wt % C5. Both feeds contained less than 1 wt % C9. 31 The feeds were compared in side-by-side experiments at the following test conditions: a H2/HC ratio of 3; a LHSV of 1.6; and a pressure of 100 psig. To achieve equivalent perfonnance over a given catalyst, the wt % aromatics in the C5+ product was set at 46.5 wt % for Feed 1 and at 48.5 wt % for Feed 2. At these conditions usmg these feeds, good catalyst activity meant that the catalysts lined out at a CAT of between about 820 and 860oF, the lower the temperature, the better the activity. Example 9 - Determining the Cycle Length After completing the SCSU, the catalyst was cooled to approximately 800°F and then feed was introduced. The reactor temperature was adjusted to maintain an aromatics content of the liquid product as noted above. This was measured by gas chromatographic analysis. The plot of CAT vs time was used to determine the cycle length. The term "cycle length" as used herein, is the number of hr to reach 920oF from the start-of-run (SOR). It is calculated as 920°F-TSOR / FR, where TSOR is the catalyst average temperature (CAT) extrapolated to time = 0. FR is the fouling rate in hr; FR is the slope of CAT vs time, excluding any initial temperature excursions following startup. SOR, time = 0, is when hot feed is provided to the catalyst. Example 10 — Perfonnance of Catalysts Using Lab Start-up (LSU) Conditions The perfonnance of catalysts from Ex. 1 and Ex. 2 were determined. Approximate 40 cc of catalyst was loaded into a pilot plant and reduced using the LSU conditions of Ex. 6. The test was continued for about 500 hr, during which both catalysts showed good activity and cycle lengths of > 1200 hr. Example 11 — Performance of Catalysts Using SCSU Conditions The comparative catalyst of Ex. 1 was tested at the same reforming conditions as in Ex. 10, except that the SCSU conditions of Ex. 7 were used instead of the LSU conditions of Ex. 6. Feed 1 was used for this comparison. Figure 1 shows that the catalyst displayed about the same initial activity after a SCSU (D) as with the LSU (o). however, the catalyst having the SCSU fouled surprisingly rapidly starting at about 200 32. hr onstream; see curve B. It had a cycle length of about 1100 hr. The figure shows that the catalyst prepared in the manner of EP 498,182A or US 5,354,933 performs well after a laboratory startup, but fouls much more rapidly when a simulated commercial startup is used. This difference in catalyst performance as a function of startup conditions was unexpected. Moreover, modifications in the simulated startup conditions did not lead to significant improvements in catalyst performance. Example 12 ~ Impact of Washing the Catalyst Base Catalysts prepared according to the wash procedures of Examples 2B, 3B and 4B were tested for reforming in pilot plants using SCSU conditions. They all showed good performance, e.g., long cycle lengths and low fouling rates after a simulated commercial startup. Figure 2 shows a comparison of the reforming performance of two catalysts. Comparative catalyst 12A (o) was prepared in a manner similar to that in Ex. 1. This catalyst contained 0.56 wt % Na. Catalyst 12B (D), useful in the present invention, was prepared in a manner similar to Ex. 4B. This washed catalyst contained 0.34 wt % Na. The performance of these catalysts was compared using a SCSU and Feed 2. The benefit of washing the bound hiz-cat is evident. Catalyst 12B had a cycle length of >2000 hr, while comparative catalyst 12A had a cycle length of about 550 hr. Example 13 — Low Sodium Catalyst via Ion Exchange and Washing A calcined extrudate was prepared comprising 83% Tosoh L-zeolite (0.28 wt % Ma) bound with 17% SI-350 silica. It had a Na content of about 0.5 wt %. It was ion exchanged with 0.3 molar KN03, 10 volume per weight at 180°F for 3 hr, and washed times with deionized water (2.5 cc H20/g of base) at room temperature with stirring for 15 minutes. The washed extrudate was then dried and calcined in air at 1000°F for l hr. This ion exchange, wash and calcine procedure reduced the sodium content of the latalyst base to less than 0.09 wt %. Pt impregnation and addition of ammonium halides results in a catalyst that lerforms well after a simulated commercial startup. 33 Example 14 - Hot Water Washing of Extrudates Table 2 below demonstrates the advantage of multiple washes at elevated temperatures. The experiments with calcined extrudate A, a silica bound L-zeolite catalyst base, were performed as described in Example 4A except that 140oF water was used. The experiments with calcined extrudate B, also a silica-bound L-zeolite catalyst base, used a cover and drain wash process. They were performed in a manner similar to that described in Example 2 except that the wash water temperature was 180°F. The accessibility of the pore system, i.e., porosity, and surface area were neasured by conventional techniques. The surface area was measured by the standard nitrogen B.E.T. method (using 5 points). The zeolite micropore volume was measured )y the "t-plot" method. These methods are well known in the art and are merely llustrative; other pore blockage and porosity measurements can also be used. The vashed extrudates were dried at about 250°F and calcined at about 920°F for about an our before surface area and micropore volume were measured. Example 15 - Catalyst Performance from Hot-Water Washed Extrudates Catalysts were prepared from some of the materials described in Table 2, secifically from the 2 washes at 140°F base(14-A2), the 5 washes at 140°F base (14- 34 At)) and the 5 washes at 180°F base (14-B5). The catalyst preparation followed the method of Example 2B. The pilot plant testing of these three catalysts was performed as in Example 12 using SCSU conditions. The catalyst was run for about 500 hours. Catalyst performance is shown in Table 3. As can be seen, catalysts having an improved activity and fouling rates were obtained after additional washings (15B and 15C) at elevated temperatures. While the invention has been described above in terms of preferred embodiments, it is to be understood that variations and modifications may be used as will be appreciated by those skilled in the art. Indeed, there are many variations and modifications to the above embodiments, which will be readily evident to those skilled in the art, and which are to be considered within the scope of the invention as defined WE CLAIM : 1. A catalytic reforming process comprising the catalytic conversion of hydrocarbons to aromatics, said process comprising treating a halided zeolite catalyst containing a Group VIII metal at commercial startup conditions and then reforming hydrocarbons, wherein said catalyst is prepared by washing a bound zeolite catalyst base or a bound zeolite catalyst before halide addition. A catalytic reforming process according to claim 1 wherein the halides of the halided zeolite catalyst comprise both chloride and fluoride. A catalytic reforming process acco^lmgjo claim 1 wherein said halided zeolite catalyst comprises a halided platinum L-zeolite catalyst. A catalytic reforming process according to claim 2 wherein ammonium halides are used to prepare the halided zeolite catalyst. A catalytic reforming process according to claim 1 wherein said catalyst is prepared by a process that includes binding, extruding, calcining and washing the bound catalyst base before halide addition. A catalytic reforming process according to claim 1 wherein said catalyst is produced by a method that includes the steps of: a) preparing a calcined silica-bound zeolite catalyst base; b) washing said bound zeolite catalyst base with a liquid comprising water; and c) incorporating a group VIII metal and halogen-containing compound(s) comprising chlorine and fluorine into said washed base to produce a halided zeolite catalyst. A catalytic reforming process according to claim 6 wherein said catalyst is prepared by a process that utilizes an extrusion aid in step (a) and further includes calcining the washed base of step (b) prior to incorporating the halogen-containing compound. A catalytic reforming process according to claim 5, 6, or 7 wherein said halided zeolite catalyst comprises a halided platinum L-zeolite catalyst. 36 catalyst was prepared using ammonium chloride and fluoride. A catalytic reforming process according to claim I wherein said process comprises contacting a hydrocarbon stream containing less than 50 ppb sulfur with said halided zeolite catalyst. A halided zeolite catalyst prepared according to the catalyst preparation method ofClaim 5, 6, or 7. A halided zeolite catalyst according to Claim 11 wherein said halided catalyst comprises an L-zeolite. A halided zeolite catalyst according to Claim 11 wherein said halided catalyst comprises an L-zeolite manufactured By Tosoh, Ltd., Japan. A halided zeolite catalyst according to Claim 11 wherein said catalyst base is washed until the surface area of the washed catalyst base after calcination returns to at least 90% of the surface area of the calcmed catalyst base before washing. A halided zeolite catalyst according to Claim 11 wherein said catalyst base is washed at least twice with wash water having a temperature above about 100°F. A catalytic reforming process according to claim 1 wherein a stream containing at least C6 and C7 hydrocarbons containing less than 50 ppb sulfur is contacted with a non-acidic Pt K L-zeolite catalyst comprising chloride and fluoride, said catalyst having been washed with water before halide addition. A process for producing aromatics comprising preparing a halided zeolite catalyst by washing a calcined bound zeolite catalyst base with a liquid comprising water; adding at least one halogen-containing compound and a Group VIII metal compound to said washed base; bringing said catalyst on stream using commercial startup conditions; and contacting the catalyst with paraffinic hydrocarbons at catalytic reforming conditions to produce aromatics. The process of claim 17 wherein the catalyst comprises both chloride and fluoride, and the zeolite is an L-zeolite. \ hydrocarbon conversion process comprising the catalytic conversion of a lydrocarbon-containing feed to product under conversion conditions that process using a halided zeolite catalyst containing a Group VIII metal, said halided catalyst prepared by a process comprising washing a bound zeolite catalyst base or catalyst before halide addition. 20. A hydrocarbon conversion process according to claim 19 wherein the halides of the halided zeolite catalyst comprise both chloride and fluoride. 21. A hydrocarbon conversion process according to claim 19 wherein said halided zeolite catalyst was prepared using ammonium chloride and fluoride. 22. A catalytic reforming process, substantially as hereinabove described and illustrated with reference to the accompanying drawings.

Full Text

FORMING USING A BOUND HALIDED ZEOLITE CATALYST
Related Patent Applications
This application is a continuation-in-part of U.S. Serial. No. 09/004,216, filed January , 1998, which is incorporated herein by reference in its entirety
Field of the Invention
The present invention relates to improved hydrocarbon conversion processes
sing bound and washed, halided zeolite catalysts. In a preferred embodiment the
ivention relates to catalytic reforming with a bound and washed, halided Pt L-zeolite
atalyst. More specifically, the invention relates to the discovery and control of catalyst
ycle length problems that are associated with commercial startup of halided zeolite
atalysts containing Group Vni metals.
Background
The advantages of monofimctional non-acidic platinum L-zeolite catalysts for atalytic reforming were discovered in the late 1970"s. US 4,104,320 to Bernard and Jury discloses that it is possible to dehydrocyclize paraffins to produce aromatics with igh selectivity using a monofimctional non-acidic L-zeolite catalyst. This L-zeolite atalyst has exchangeable cations of which at least 90% are sodium, lithium, potassium, abidium or cesium, and contains at least one Group VIII noble metal or tin or ermanium.
Later, Buss and Hughes developed improved L-zeolite catalysts for reforming etroleum naphthas. These catalysts comprise a large pore zeolite exchanged with an Ikaline earth metal containing one or more Group VIII metals (preferably platinum), lee US PatentNos. 4,434,311; 4,435,283; M47^6, and4,517,306. Barium xchanged catalysts were found to provide even higher selectivities than the orresponding alkali exchanged L-zeolite catalysts disclosed in US 4,104,320. The igh selectivity catalysts of Bemard and Nury, and of Buss and Hughes, are all non-acidic" and are referred to herein as "monofimctional catalysts". These catalysts re highly selective for forming aromatics via dehydrocyclization of paraffins. After more than 10 years of extensive research and engineering, the use of h zeolite catalysts for reforming was commercialized. This commercialization required

many additional discoveries. Two key discoveries were the criticality of ultra-low ulfur levels in the feed, and the impact of these ultra-low sulfur levels on reactor letallurgy, i.e., the discovery of the need to prevent coking, carburization and metal listing.
While commercialization of ultra-low sulfur reforming was being pursued, a econd generation of platinum L-zeolite catalysts was being developed. These new atalysts (referred to herein generically as "hiz-cats") are produced by treating L-eolites with halogen-containing compounds, such as halocarbons (e.g., US 5,091,351 3 Murakawa et al.) or ammonium salts, (e.g., EP 498,182A and US 5,354,933 to )hashi et al.) and Group Vni metals. It is disclosed that these fialided catalysts are seful for preparing aromatic hydrocarbons such as benzene, toluene and xylenes from "e-Cg aliphatic hydrocarbons in high yield. They allow operations at high severity, >lerate a wide range of hydrocarbon feeds, and show high activity and long life in creening studies. Other patents that disclose hiz-cats include US 4,681,865, US ,761,512 and US 5,073,652 to Katsuno et al.; and US 5,196,631 and US 5,260,238 to iurakawa et al.
Yet, our attempts to commercialize one of these hiz-cats resulted in a catalyst ath poor performance. When the patented recipe for preparing a bound halided L-eolite catalysts was followed, we discovered that the resulting catalyst was nexpectedly sensitive to reforming startup conditions. The catalyst performed well in :reening studies using high gas velocity, rapid heat rate, "laboratory" startup (LSU) onditions; but surprisingly, its fouling rate was unacceptably high when low velocity, imulated "conmiercial" startup (SCSU) conditions were employed.
Scale-up and commercialization of a new process, particularly one using a new ataiyst, entail numerous changes from laboratory procedures. For instance, limitations ue to commercial equipment impact scale-up designs and the way commercial rocesses are operated. For example, commercial gas compressors set limits on gas ow rates, and heat transfer limitations influence reactor configurations, designs, and perating conditions. Moreover, laboratory studies are generally done to get results in a ;asonable time firame, while commercial operations are intended to operate for much )nger times. One example of unpredictability in the process scale-up area is Heyse et I., as described in US 5, 675,376. Here, potentially catastrophic carburization and

letal dusting problems were only first observed when low sultur retorming was being sted in commercial reforming equipment. The metallurgy, process configuration, and mperature of commercial reforming operations lead to coke plugging problems not bserved in laboratory studies.
Commercialization and scale-up also requires working with quantities of laterials that are 100 to 1,000,000 times larger than used in laboratory studies. This creased scale typically requires that catalyst preparation procedures be modified. A •ocess as simple as drying is done quite differently in the laboratory than at a commercial plant. Often, these scale differences do not significantly affect catalyst erformance, but occasionally they do.
Furthermore, the catalyst composition is often modified to meet the demands of commercial operations, e.g. to attain needed crush strength. Also, (unbound) zeolite ystals or powders are often used in laboratory studies. In contrast, bound zeolites are syually used in commercial operations. This is because bound zeolites reduce the ressure drop through large reactor, provide improved gas and liquid flow rates, and are isier to load and unload. Yet it is known that the performance of bound catalysts, ;pecially bound zeolites, can be quite different from that of unbound powders. Dmetimes, the type of binder or binding method affects catalyst performance, ametunes special steps that are only done for bound catalysts, such as calcination of be bound zeolite, affects catalyst performance.
Thus, studies on powdered zeolites do not always predict the commercial erformance of bound zeolites. One example that shows the unpredictability of catalyst erformance associated with adding a binder is Mohr et al., in US 5,106,803. Mohr et ., disclose the criticality of the water sensitivity index of a bound L-zeolite catalyst, his property was not observed to be important for the unbound catalyst.
Yet one must assxmie that scale-up will proceed satisfactorily, since it is not pssible to predict what factors will be critical for a particular new process or new utalyst. For halided zeolite catalysts, there is no suggestion in the art that there might e problems associated with preparing bound catalysts or with the use of these catalysts I commercial operations, particularly problems related to catalyst startup procedures, deed, there is nothing in the art that teaches the importance of washing or otherwise

treating a catalyst base (i.e., a bound zeolite) or a bound zeolite catalyst before adding halide, as described hereinbelow.
Treating conventional, non-halided zeolites —including L-zeolite~ with water or an aqueous solution is known. For example, Kao et al., in US 4,987,109 adjust the pH
i of L-zeolite crystals by washing them so as to provide a zeolite of pH 9.4-10.0. These washed zeolites are then bound. Washing the zeolite crystals to this pH improves catalyst activity. Also, Poeppelmeir et al., in US 4,595,668, teach having the nobel metal of in a zeolite dispersed in particles having a diameter of less than 7 angstroms. In Example 6, they describe how to prepare a bound catalyst of their invention using an
" oxychlorination step. Another example is Poeppelmeir et al., in US 4,568,656. Here a
bound L-zeolite is contacted with an aqueous solution containing a Ft salt and a non-
platinum salt (e.g., KCl). This mixture is aged to distribute the Ft uniformly (see
Example 1.) Also, Buss and Hughes in US 4,721,694 teach preparing platinum Ba L-
zeolite catalysts using a barium ion exchange before Ft impregnation. The ion
exchange replaces potasshim.w4th barium in the catalyst and is taught to improve
selectivity. For example in Col 12, they,discuss treating extruded pellets by ion
exchange with a barium solution, followed by Ft impregnation. These patents are all
incorporated herein by reference.
It is also known to wash powdered zeolitcatalysts containing halides. For instance, Tatsumi et al., have done laboratory studies investigating powdered, unbound Ft K L-zeolite catalysts for aromatization of hexane. They have looked at the effect of adding KCl and the addition method on catalyst performance \Chem. Lett., 387-390 (1993); J. Cat. 147, 311-321 (1994); Bull. Chem. Soc. Jpn., 67, 1553-1559 (1994); and Cat. Lett.. 45 (1997) 107-112]. They have also investigated the effect of other halogen anions (F, Br, I) on hexane aromatization {Cat. Lett., 11 (1994) 289-295 and Sci. & Tech. in Cat., 117-122, (1994)]. They report that chloride and fluoride give the best results. Tatsumi et al., typically add a potassium halide (e.g., KCl) after Ft addition. In general, these authors add halide along with alkali, i.e., then add a refractory alkali halide salt to the catalyst. (Note: these catalysts are not hiz-cats, as defined herein.) In trying to vmderstand why KCl addition improves catalyst selectivity, the authors of the J. Cat., 147 (1994) article add ammonium chloride to a powdered Ft L-zeolite catalyst that had been prepared by ion exchar which is a hiz-

cat, did not show the improved selectivity of catalysts with KCl added, leading the authors to discuss the importance of adding salts containing both alkali and halide. Sugimoto et al., in Appl. Cat. A: General 96 (1993) 201-216, have also studied I
unbound catalysts. They prepared powdered hiz-cats from a variety of zeolites,

including those prepared from an L-zeolite powder. These catalysts were made using a
halocarbon treatment following either an alkaline soaking or ion exchange. In either
case, the powdered zeolites were washed. Catalysts prepared from these washed
powders were tested for n-hexane aromatization. As already noted, these studies were
lone on powdered catalysts, not bound catalysts.
None of the references discussed above address problems associated with
commercial startup ofbound halided zeolite catalysts. Nor do they teach the
mportance or desirability of washing bound zeolite bases or bound catalysts, especialli" ion-acidic zeolite bases or catalysts, before adding halide as discussed hereinbelow. "
Description of the Figures Figure 1 compares the catalytic reforming performance of a bound halided Pt L-;eolite catalyst prepared in the manner of EP 498,182A or US 5.354^931 after a aboratory startup procedure (curve A) and after a simulated commercial startup SCSU) procedure (curve B).
Figure 2 compares the reforming performance after a startup at SCSU conditions fa bound Pt L-zeolite catalyst (circles) prepared in the manner ofEP 498,182 A or US ,354,933 with the performance of a bound and washed Pt L-zeolite hiz-cat of the invention (squares).
Description of the Invention One object of the present invention is to provide a reforming process using ound, halided zeolite catalysts (hiz-cats), especially bound, halided Pt L-zeolite italysts, where the catalysts have long cycle lengths after commercial startup. Another bject of the invention is to improve the reproducibility of catalytic operations using iz-cats. As discussed below, we have discovered that unless the catalyst base or the 3und catalyst is washed before halide addition, the catalyst will foul rapidly and have a lort cycle length.
In one embodiment, the present invention is a catalytic reforming process using bound, halided zeolite catalyst (a hiz-cat) comprising the catalytic conversion of

hydrocarbons to aromatics wherein the catalyst is treated at commercial startup conditions. These commercial startup conditions include "low" gas flow rates (e.g., GHSV of between 400 and 2000 hr"") and "slow" heat-up rates (e.g., between about 5 and 50 °F/hr, and more typically at between about 10-25 °F/hr) during catalyst 3 reduction. Commercial reformer startup conditions are well known to those skilled in the art, especially in contrast with laboratory startup conditions, which are generally used in experimental studies. More particularly, the invention is directed to reforming using halogen-added or halogen-impregnated zeolite catalysts containing a Group VIII metal that have long cycle lengths and maintain very low fouling rates after commercial lb startiup.
According to one embodiment, the invention is a catalytic reforming process comprising the catalytic conversion of hydrocarbons to aromatics, said process comprising treating a halided zeolite catalyst containing a Group VIII metal at commercial startup conditions and then reforming hydrocarbons, wherein said catalyst
is prepared by washing a bound zeolite catalyst base or a bound zeolite catalyst before halide addition. Adding halide and, if not present, a catalytic metal results in a washed hiz-cat, as the term is used herein. Preferred washed hiz-cats contain both chloride and fluoride. An especially preferred hiz-cat is a bound and washed, halided Pt L-zeolite catalyst, for example a catalyst prepared with about 0.9 wt % initial chloride, 0.9 wt %
3D initial fluoride and about 1 wt % added Pt.
Among other factors, the invention is based on our discovery that the reforming performance of bound hiz-cats, especially bound non-acidic, L-zeolite hiz-cats, can be very sensitive to reforming startup conditions. We have discovered that bound hiz-cats that have low fouling rates after being treated under high gas flow rate startup
25 conditions, (e.g., conditions such as those typically used in lab testing imits) do not necessarily exhibit these same low fouling rates when commercial startup conditions are used. Rather, low catalyst activity and/or rapid fouling are often observed when the bound catalyst is started up at simulated conunercial startup conditions. Figure 1 compares the catalytic reforming performance of a bound halided Pt L-zeolite catalyst
30 prepared in the manner ofEP498,182AorUS 5,354,933 after a startup using laboratory startup (LSU) conditions (curve A) and simulated commercial startup (SCSU) conditions (curve B). As the figure shows, the impact of startup conditions on
7

can de very large. we nave unexpectedly discovered that catalyst performance following a startup at SCSU conditions can be significantly improved by washing the catalyst or catalyst base before halide addition. See Figure 2, which compares the performance of a washed hiz-cat with a hiz-cat that has not been / washed.
Hiz-cats useful in this invention can be prepared by washing the bound catalyst or preferably the catalyst base (i.e., the bound zeolite) before halogen impregnation / addition. According to one preferred embodiment, the catalyst is prepared by a process that includes binding, extruding, calcining and washing the catalyst base before halide 1/5 addition. In another preferred embodiment, the hiz-cat is produced by a method that includes:
a) preparing a calcined silica-bound L-zeolite catalyst base;
b) washing said bound zeolite catalyst base with a liquid comprising water; and
c) incorporating a Group VIII metal and halogen-containing compound(s) comprising
chlorine and fluorine into said washed base to produce a halided zeolite catalyst.
Preferably the washed base is calcined before step (c). Non-acidic L-zeolite is a preferred zeolite. In another embodiment of the invention, the catalyst base is ion-exchanged with potassium, wherein washing is part of the ion exchange procedure.
In yet another embodiment, the invention is a method of preparing a bound hiz-
cat having a cycle length >1200 hr following startup at SCSU conditions, and the
catalyst prepared thereby. According to another embodiment, the invention is also the use of a boimd hiz-cat that has been washed before halide addition in a hydrocarbon conversion process, wherein the catalyst is treated at commercial startup conditions.
Detailed Description of the Invention
In one broad aspect, the present invention is a process that comprises catalytic
reforming using a bound halided zeolite catalyst, especially an L-zeolite catalyst comprising chloride and fluoride, which is treated at commerci^ startup conditions. In a preferred embodiment, the invention is a process for producing aromatics coi^prising preparing a halided zeolite catalyst by washing a calcined, bound zeolite catalyst base
30 with a liquid comprising water; adding at least one halogen-containing compoimd and a Group VIII metal compound to said washed base; bringing said catalyst on stream using commercial startup conditions; and contacting said catalyst with paraffinic
8

hydrocarbons at catalytic reforming conditions to produce aromatics. More preferably, the reforming process contacts a stream comprising C6and C7 hydrocarbons and containmg less than 50 ppb sulfur with a halided Pt K L-zeolite catalyst. The halides comprise chloride and fluoride, and said catalyst is prepared by washing a bound zeolite catalyst base before halide addition.
Although the terms "comprises" or "comprising" are used throughout this specification, these terms are intended to encompass both the terms "consisting essentially of, and "consisting of, as understood in the patent art, in various preferred aspects and embodiments of the present invention. The terms "comprises" or "comprising" themselves are terms of the patent art. They are intended to be inclusive terms. The terms thus include the specified component or components, or the specified step or steps, being discussed or claimed, as well as any additional components or steps.
As used herein, the term "bound", is intended to describe a zeolite, binder combination that is formed into aggregates such as pellets, pills, extrudates and the like. The term "catalyst base", as used herein, refers to a bound zeolite.
The term "catalyst" is used herein in a broad sense and includes the final catalyst as well as precursors of the final catalyst. Precursors of the final catalyst nclude, for example, the calcined form of the catalyst containing the catalytic metal md also the catalyst prior to activation by reduction. The term "catalyst" is thus used to efer to the activated catalyst in some contexts herein, and in other contexts to refer to )recursor forms of the catalyst, as will be understood by skilled persons from the context. The catalysts of this invention are bound and comprise a Group VIII metal. Vlthough often discussed herein in terms of Pt being the Group VIII metal, it is ntended that this discussion is exemplary rather than limiting.
As used herein, the term "hiz-cat" is mtended to include zeolite catalysts that esult firom addition of halogen-containing compounds to, or from halide impregnation >f, zeolites, catalyst bases (i.e., bound zeolites) or zeolite catalysts (i.e., zeolites ontaining catalytic metal). Hiz-cats are also referred to herein as "halided catalysts" or halided zeolite catalysts". The halides of the hiz-cats are in addition to those that may »e incorporated into the catalyst from the catalytic metal source; that is, this halogen ddition or impregnation is not just that associated with conventional platinum loading nd impregnation, where platinum halides are often used. Nor does this halogen
9

treatment include conventional zeolite ion exchange operations. Zeolite ion exchange sometimes uses a halide salt, such as KCl, to replace the cations in the zeolite; this ion exchange can leave small amounts of halide on the catalyst. Moreover, the term hiz-cat :s not intended to include catalysts where halide is added using alkali halides (e.g., KCl) or alkaline earth halides. As is discussed hereinbelow, we believe that added alkali is letrimental to hiz-cats performance.
It is not known in what form the halide is present in hiz-cats. It may be as ionic lalide, neutral halide, or it may be part of compound, such as a silica halide or Pt lalide. Thus, as used herein, the term "halide" is used in a broad sense, and is not itended to denote the ionic state of the added halogen or of the halogen associated with tie catalyst.
The halide of the hiz-cat can be chloride, fluoride, bromide, or iodide, or a ombination thereof In a preferred embodiment, the hiz-cat contains chloride or uoride, more preferably both. Especially preferred hiz-cats can be prepared by •eating the bound zeolite or bound catalyst with halocarbons such as freons or with ther chlorine-containing and/or fluorine-containing compounds, e.g. by impregnation ith ammonium chloride and ammonium fluoride. Preferred hiz-cats useful in this invention have high total halide after calcination (this includes all halides, e.g. both ploride and fluoride), i.e., they contain at least about 0.9 wt % halide, preferably at ast about 1.1 wt % and more preferably at least about 1.3 wt % total combined halide. [ore than about 5 wt % halide does not appear to provide significant advantages, ^en chloride and fluoride are both present, the weight ratio of CI to F can vary, referably it is between 1:10 and 10:1. More preferably chloride and fluoride are added a weight ratio of about 1:1.
As used herein, the terms "chloride retensivity" and "retained chloride" denotes e residual chloride content of the catalyst after dry-out, reduction and about 300 hr i-stream. Hiz-cats evolve chloride during these steps, especially during reduction, lereafter, the chloride level on the catalyst remains substantially constant as the talyst is brought on-stream and operated. After startup, some halide does continue to olve, but very slowly. Care should be taken not to overheat the catalyst, (i.e., mperatures above about 950°F) since additional chloride will evolve, and this "erheating is not desirable.

Based on our studies, we found that bound hiz-cats prepared in the manner of EP 498,182A or US 5,354,933 gave good performance (activity and life) in LSU tests. LSU tests use very high gas velocity (hydrogen at 10,000 GHSV) and a fast heat-up rate (100°F/hr). But, when this catalyst was started-up at simulated commercial gas flow and heat-up rates, (SCSU conditions), it unexpectedly exhibited an unacceptably high fouling rate and short cycle length. See Figure 1, Curve B. Extensive variations in the simulated commercial startup did not overcome the cycle length problem. Indeed, we found that the patented recipe for making bound halided Pt L-zeolite catalysts did not necessarily produce a catalyst suitable for long-term commercial operation using a commercial startup.
After much experimentation, we found that bound hiz-cat performance can be significantly improved by washing, e.g., with water, before halide addition. We also found that catalysts that performed well after a SCSU retained less of the added chloride than poor catalysts, even though the retained chloride, i.e. the "chloride retensivity", is independent of startup conditions. Among the many variables examined, and upon further investigation, we also discovered that the sodium content of catalysts that performed poorly was consistently, albeit by a small amount (0.5 wt % vs about 0.4 wt %), higher than that of catalysts with long cycle lengths. Moreover, sodium content seems to correlate with retained chloride.
In a preferred embodiment, calcined catalyst base is washed before adding the halogen-containing compounds and the Group Vin metal, e.g., platinum. In this way, these added components are not washed out of the catalyst. It is preferred that the catalyst base is washed with one or more volumes of wash water. The washing desirably removes at least 20%, preferably at least 50%, of the readily removable alkali. Quite remarkably, after platinum and halogen impregnation, the washed catalysts of the invention are amenable to commercial startup. For example, see Figure 2, which compares an "as prepared" bound hiz-cat made in the manner of EP 498,182A or US 5,354,933 with a water-washed bound hiz-cat as described below. The washed hiz-cat shows a much lower fouling rate and an improved cycle length. We have also found that catalyst performance can be improved by some types of ion exchange processes. Cation exchange, such as with potassium and the like, often includes a wash step. When ion exchange comprises a wash step, as described in this invention, hiz-cat

performance improves. Here also the wash step ilrpreferably done on the bound catalyst base before halide(s) (and platinum) is added.
It is not known what causes the improvement in catalyst performance that is associated with washing. Washing (and the calcining, which is preferably done thereafter) results in numerous changes including the removal of Na, Si, K, S and small amounts of Al. It may be the removal of one or more of these elements or some other change that improves the catalyst performance. We have noted that reduced sodium levels seem to correlate with improved catalyst performance following a SCSU.
The bound and washed hiz-cat of the invention has a long cycle length and low fouling rate following a SCSU. The catalyst may exhibit an initial transient period of 100-300 hr, before catalyst performance lines out. Thereafter, it settles down to roughly linear fouling performance. Then, cycle length and fouling rate are measured. Preferred washed hiz-cats have a cycle length of > 1200 hr, preferably >2000 hr and more preferably >4000 hr, following startup at simulated commercial startup (SCSU) conditions as described below. Catalyst fouling rates are preferably less than 0.02 °F/hr after stabilization following a SCSU, more preferably less than 0.01 and most preferably less than 0.003 °F/hr.
It appears that fouling rate and cycle length are related to retained chloride after startup and hydrocarbon processing. For example, we started with as-impregnated catalysts that initially contained about 0.9 wt % total chloride and about 0.9 wt % total fluoride. After cedcination, these catalysts contained about 0.75 wt % chloride and about 0.75 wt % fluoride. After startup and reforming, preferred catalysts were found to contain less than about 0.5 wt % chloride, more preferably less than 0.45 wt % and most preferably less than 0.4 wt % chloride. The sodium level of these catalysts is preferably less than about 0.4 wt %.
While not wishing to be bound by theory, it is believed that halide evolution luring catalyst startup, particularly HCl evolution, somehow improves hiz-cat performance. Gases, including water and ammonia, evolve from the catalyst during startup. With high-velocity lab startups, these evolving gases are only present for short periods and at low partial pressure; they are quickly removed. In contrast, during commercial startups, especially with bound catalysts, there is more time for evolved vater and ammonia to contact the catalytic metal (e.g., Pt), and these gases are present

at higher partial pressures. This can result in Pt migration. Thus, commercial startup conditions can lead to imdesired Pt agglomerates and poor catalyst performance.
Thus, it is believed that evolving halides produced during commercial startup of bound hiz-cats help disperse or redisperse Pt, perhaps via mobile Pt or Pt halide species. Catalysts that have not been washed appear to evolve less chloride during startup, and have lower cycle lengths. It appears that for bound hiz-cats it may be necessary to evolve a threshold amount of mobile or mobilizing halide to achieve long cycle lengths after a commercial startup.
Generally, non-acidic zeolite catalysts have high loadings of alkali or alkaline earth metals, e.g., about 15 wt % K2O. Potassium is often the predominant metal for non-acidic L-zeolites, as in the examples described herein. Yet, even with so much alkali present, it appears to be desirable to reduce the alkali content of bound hiz-cats somewhat. While not wishing to be bound by theory, it is believed that the readily-emovable (i.e., soluble) alkali, both sodium and potassium, captures evolved chloride from the hiz-cats during startup and thus prevents the desired Pt dispersion. Removal )f this alkali by water washing allows for improved Pt dispersion and catalyst performance. What appears to be important for bound hiz-cats is to have sufficient volved chloride to (re)disperse the Pt in-situ. Thus, we believe that adding refi-actory alide salts, e.g., alkali halides such as KCl, will not provide the desired evolved halide. Refractory halide salts, when added, are not included in calculating the halide content f hiz-cats.
We expect that the need to improve catalyst performance by washing before alide addition will apply to other halided, medium pore and large pore zeolite italysts, as well as to the use of these bound hiz-cats in processes other than catalytic iforming. For example, this washing should be advantageous for the halided zeolite vtalysts described in US 5,260,238 to Murakawa et al.
Zeolites Catalysts usefiil in this invention comprise one or more zeolites or non-zeolitic olecular sieves and at least one catalytic metal, preferably a Group VIII metal. The italyst usually also includes a binder such as a refractory oxide, e.g., silica, alumina, ilorided alumina or silica-alumina. Preferred zeolites and/or molecular sieves are lected from those of the large and intermediate pore variety.
13

Representative of the large pore zeolites are ZSM-3, ZSM-4, ZSM-10, ZSM-12, ZSM-20, zeolite beta, zeolite omega, zeolite L, zeolite X, zeolite Y, REY, USY, RE¬US Y, mordenite, LZ-210, LZ-210-M, LZ-210-T, LZ-210-A, SSZ-24, SSZ-26, SSZ-31, SSZ-33, SSZ-35, SSZ-37, SSZ-41, SSZ-42, SSZ-44 and MCM-58. ZSM-3 is described in US 3,415,736; ZSM-4 is described in UK 1,117 ZSM-10 is described in US 3,692,470; ZSM-12 is described in US 3,832,449; ZSM-20 is described in US 3,972,983; Zeolite beta is described in US Re. 28,341 (of original US 3,308,069); Zeolite omega is described in US 4,241,036; Zeolite L is described in US 3,216,789; Zeolite X is described in US 2,882,244; Zeolite Y is described in US 3,130,007; LZ-210, LZ-210M, LZ-210-T, LZ-210-A and mixtures thereof are described in US 4,534,853; SSZ-24 is described ik US 4,834,977; „SSZ-26 is described in US 4,910,006; SSZ-31 is described in US 5,106,801; SSZ-33 is described in US 4,963,337; SSZ-35 is described in US 5,316,753; SSZ-37 is described in US 5,254,514; SSZ-41 is described in US 5,591,421; SSZ-42 is described in US Ser. No. 08/199,040; SSZ-44 is described m US 5,580,540; MCM-58 is described in US 5,437,855. The entire contents of all these patents and patent applications are incorporated herein by reference.
In a preferred embodiment, the invention uses an L-zeolite or a zeolite having an L-zeolite-type channel structure and size, such as ECR-2, which is described in US 4,552,731, and ECR-31, which is described in US£,624,657 (Vaughan). Preferably the invention uses a monofunctional, non-acidic K L-zeolite, Acidity in the L-zeolite generally leads to poor performance in catalytic reforming. Examples of useful L-zeolites include those described in Uj 3,216,789 (Breck), 4,552,731 (Vaughan), 4,544,539 (Wprtel), 5,491,119 (Verduijn) and 4,530,824 (assigned to Tosoh Ltd.). The entire contents of all these patents are mcorporated herein by reference. One useful aon-acidic L-zeolite is manufactured by Union Oil Product (UOP), Mobile, AL. A preferred non-acidic L-zeolite is manufactured by Tosoh Ltd., Japan, and sold under the lame HSZ-500KOA, For these non-acidic zeolites, potassium is a preferred cation; a jreferred catalyst comprises K L-zeolite.
Preferred catalysts used in the process of the present invention are nonofunctional. They do not have the acid function of conventional reforming ;atalysts. In contrast, conventional reforming catalysts are bifunctional, with an acid

and a metal function. Examples of monofunctional catalysts include platinum on L-zeolite, wherein the L-zeolite has been exchanged with an alkali metal, as disclosed in US 4,104,320 to Bernard et al.; platinum on L-zeolite, wherein the L-zeolite has been exchanged with an alkaline earth metal, as disclosed in US 4,634,518 to Buss and Hughes; and platinum on L-zeolite as disclosed in US 4,456,527 to Buss, Field and Robinson. The entire contents of all these patents are incorporated herein by reference
The term "non-acidic" is understood by those skilled in this area of art, particularly by the contrast between monofunctional (non-acidic) reforming catalysts and bifunctional (acidic) reforming catalysts. One method of achieving non-acidity is by replacing protons with alkali and/or alkaline earth metals in the zeolite. This is preferably achieved, along with other catalyst enhancements, by an ion exchange process on the synthesized zeolite.
The composition of type L zeolite expressed in terms of mole ratios of oxides, may be represented by the following formula:
(0.9-1.3)M2/nO:Al2O3(5.2-6.9)SiO2:yH2O
In the above formula M represents a cation, n represents the valence of M, and y may be any value from 0 to about 9. Zeolite L, its X-ray diffraction pattern, its properties, and method for its preparation are described in detail in, for example, US Patent No. 3,216,789, the contents of which is hereby incorporated by reference. The actual formula may vary without changing the crystalline structure. For example, the mole ratio of silicon to aluminum (Si/Al) may vary from 1.0 to 3.5.
As discussed above, one preferred embodiment of the invention uses monofunctional Pt L-zeolite catalysts that have been treated with halogen-containing compounds. This type of halogen treatment is known. For example, US 5jO£l J.51iQ Murakawa et al., discloses preparing a Pt L-zeolite catalyst, and then treating it with a halogen-containing compound. Other related patents that disclose halided L-zeolite catalysts include EP498J82ApjiJS-5,354,933, which discloses co-impregnation of an L-zeolite with NH4CI and NH4F; US 4,681,865,4,761,512 and 5,073,652 to Katsuno et al. These patents are all incorporated herein by reference. One preferred hiz-cat for catalytic reforming comprises halided platinum K L-zeolite catalyst, especially one containing both chloride and fluoride.
IS

Useful intermediate pore size zeolites are exemplified by ZSM-5, ZSM-11, ZSM-22, ZSM-23, ZSM-35; ZSM-48, ZSM-57, SUZ-4, SSZ-23; SSZ-25; SSZ-28, SSZ-32, and SSZ-36. ZSM-5 is described in US Re. 29,948 (of original US 3,702,886); ZSM-11 is described in US 3,709,979; ZSM-22 is described in US 4,556,477; ZSM-23 is described in US 4,076,842. ZSM-35 is described in US 4,016,245; ZSM-48 is described in US 4,585,747; SUZ-4 is described in EP 353,915; SSZ-23 is described in US 4,859,422; SSZ-25 is described in US 4,827,667 and 5,202,014; SSZ-28 is described in US 5,200,377; SSZ-32 is described in US 5,053,373; and SSZ-36 is described in US Serial No. 60/034,252. The entire contents )f all these patents and patent applications are incorporated herein by reference. In addition to silicon, the useful zeolites herein can contain one or more iramework elements other than, or in addition to, aluminum, e.g., the borosilicate ".eolites. Also, the zeolites can be modified to alter their as-synthesized framework ilica to alumina ratio.
Catalysts useful in the present invention can also include non-zeolitic molecular ieves with intermediate or large size pores. Non-zeolitic molecular sieves are licroporous compositions that are formed from [AIO2] and [PO2] tetrahedra and have "ectrovalently neutral frameworks. See US 4,861,743. Also included among the seful zeolites are materials of similar structure or behavior, e.g., crystalline etallophosphates such as those described in US 4,440,871. Non-zeolitic molecular eves include aluminophosphates (AIPO4) as described for example in US 4,310,440, etalloaluminophosphates, as described in US 4,500,651; 4,567,029; 4,544,143; and 686,093 and nonmetal substituted aluminophosphates as described in US Pat. 973,785.
Useful catalysts also include intermediate pore silicoaluminophosphates APO"s) as the non-zeolitic molecular sieve component. Intermediate pore SAPO"s ilude SAPO-11, SAPO-31, SAPO-41 and SM-3. US 4,440,871 describes SAPO"s nerally and SAPO-11, SAPO-31, and SAPO-41 specifically. The preparation of SM-md its unique characteristics are described in US 5,158,665. All these patents are jorporated herein by reference.
Binders

in prepanng catalysts for use in the present invention, the zeolites and/or molecular sieves are bound. They are preferably composited with matrix materials resistant to the temperatures and other conditions employed in hydrocarbon conversion processes. Such matrix materials can include active and inactive materials. Frequently binders, such as naturally occurring clays and inorganic oxides, are added to improve the crush strength of the catalyst. The selection of binders and binding conditions depends on the zeolite and its intended use.
The binder material can be selected from among synthetic or naturally occurring zeolites, as well as alumina, clays such as montmorillonite and kaolin, and the refractory oxides of metals of Groups IV A and IVB of the Periodic Table of the Elements. Particularly usefiil are the oxides of silicon, titanium and zirconium, with silica being preferred, especially low acidity silica. Combinations of such oxides with other oxides are also useful, for example, silica-alumina, silica-magnesia, silica-zirconia, silica-thoria, silica-beryllia, silica-titania, titania-zirconia, silica-alumina-thoria, silica-alimiina-zirconia, silica-alumina-magnesia and silica-magnesia-zirconia. These oxides can be crystalline or amorphous, or can be in the form of gelatinous precipitates, colloids, sols, or gels. Silica in the form of a silica sol is a preferred binder. A preferred silica sol has about 30 wt % silica and contains small particles (7-9 nm in diameter), which result in catalysts with good attrition resistance and excellent crush strengths.
Forming pellets or extrudates from zeolites and/or molecular sieves generally involves using extrusion aids and viscosity modifiers in addition to the binders. These additives are typically organic compounds such as cellulose based materials, for example, Methocel® sold by Dow Chemical Co., ethylene glycol, and stearic acid. Many such compounds are known in the art. It is important these additives do not leave a detrimental residue, i.e., one -with undesirable reactivity or one that can block pores, after calcination. For this invention, it is especially desirable that such residues do not add significant amounts of alkali or alkaline earth ash to the catalyst. The above-described washing will remove low levels of these materials. The residue from the extrusion aid is preferably less than a few tenths of a percent, more preferably less than 0.1 wt%.
Methods for preparing catalyst compositions are well known to those skilled in
-V6-

the art and include such conventional techniques as spray drying, pelletizing, extrusion, various sphere-making techniques and the like. The newly developed method of in-extrudate formation of the zeolite/binder as described in Miller, US 5,558,851 and US 5,514,362 can also be used. The entire contents of these patents are incorporated herein by reference.
The relative proportions of zeolite to the binder/matrix can vary widely. Generally the zeolite content ranges from between about 1 to about 99 weight percent, and more usually in the range of from about 5 to about 95 weight percent, of the dry composite, more typically 50-85%.
The invention requires a bound zeolite. It is preferred to use whole extrudates rather than crushed extrudates or unbound zeolites in commercial operations. Bound zeolites reduce the pressure drop through a reactor, provide improved flow rates, and are easier to load and unload. However, the difflission characteristics of whole extrudates are quite different from those of unbound powdered zeolites. The interaction of a diffusing gas, such as a halocarbon, is different for a powdered versus a bound zeolite. Diffusion differences would also be especially significant if the catalyst evolves materials such as gases or solids, for example during startup. Moreover, the act of binding itself— including selecting binding materials and the binding method ~ as well as calcining can affect catalyst performance. For example, the binder can interact with the sieve by simply blocking internal and external sites or by chemical interaction e.g., alkali from a basic sieve like the preferred L-zeolite of the present invention and silica are known to react imder thermal and hydrothermal calcination conditions). Moreover, the distribution of impregnated materials may vary considerably on zeolite (owder versus bound zeolites. Thus, studies on powders alone cannot be used to iredict commercial performance of bound zeolite catalysts.
Group Vm Metal It is preferred that the catalyst herein comprises at least one Group VIII metal, referably a noble metal (Pt, Pd, Rh, Ir, Ru, Os) and more preferably platinum, referred amounts of metal are 0.1 to 5 wt %, more preferably 0.1 to 3 wt %, and most referably 0.3 to 1.5 wt %, based on the L-zeolite. Platinum compounds that form psitively charged platinum complex ions in solution are the preferred source of latinum. Platinum tetraammine chloride and nitrate are especially preferred.

To these Group VIII metals can also be added one or more non-platinum group metals such as tin, indium and metals of Group VIIB such as rhenium. Examples include Pt/Sn, Pt/Pd, Pt/Ni, and Pt/Re. These metals can be readily introduced into the composite employing a variety of known and conventional techniques, e.g., ion-exchange, incipient wetness, pore fill, impregnation, etc. Care should be taken so that the Group VIII metal, e.g., platinum, is incorporated in a manner that results in excellent and uniform dispersion. The incipient wetness impregnation method is preferred.
Washing and Adding Halides According to one embodiment, the catalyst used in this invention is prepared by process that includes extruding and then washing either the bound catalyst or the atalyst base before halide addition. Preferably, the catalyst is prepared by a process lat includes the steps of: (a) preparing a calcined silica-bound zeolite catalyst base; (b) cashing said bound zeolite catalyst base with a liquid comprising water, and (c) icorporating a group VIII metal and halogen-containing compound(s) comprising ilorine and fluorine into said washed base to produce a halided zeolite catalyst, referably, the catalyst is prepared using a low alkali, e.g., low sodium, extrusion aid in ep (a). The Pt and halogen-containing compounds can be incorporated sequentially or, eferably, simultaneously.
The term washing, as used herem, is meant to include any process where water excess of the material"s pore volume is contacted with the catalyst base. Preferably, ; washing is done with clean water that is roughly neutral (pH 5 to 9,) low in sulfur g., sulfate-free,) and low in alkali and alkaline earth content ( It is preferred that the calcined catalyst base or calcined catalyst is washed "ore halide addition and that at least 20%, preferably at least 50%, of the readily lovable alkali is removed. The "readily removable" alkali is the alkali that can be shed out of the catalyst base after 5 washings (1:1 volume of liquid to wt of rudate) at ambient temperature. However, the temperature, amoimt of water and itact time in the washing steps are not critical, as long as the washing improves cycle 5th following startup at SCSU conditions. The wash water may contain additives, ti as alcohol, although this is not preferred.

There are many ways to wash the catalyst, or preferably the catalyst base, that will be apparent to those skilled in the art. These include, for example, cover and drain, dipping, and belt washing. The type ofwasher is not believed to be critical. The washed base is then dried. Although not required, it is preferred to dry and calcine the washed base produced in step (b) prior to incorporating the halogen-containing compound(s).
As noted earlier, we have found that washing the bound catalyst or the catalyst base before halide addition is important. As will be appreciated by those skilled in the art, there are various ways to perform the washing step; this is especially true as one moves from laboratory to flill commercial scale manufacturing. For example, the washmg can be static or stirred. The solution, or the solution and the extrudate may be stirred. The wash water may be recirculated, with or without fresh water makeup, in a recycle mode. A bleed stream of effluent wash water may be removed and make-up or added fresh wash water may be used. The wash process may be operated in a variety of batch, semi-continuous or continuous modes. Generally, the details of the wash step will vary someA\^t, depending on the equipment used for washing. It is not necessary or possible to specify all possible wash conditions here, but these will be evident to the skilled artisan.
Three key wash variables are contact time, number of washes and wash temperature. A longer contact time (between the calcined catalyst base and the water) allows for more efficient removal of the silica and alkali. The contact time is preferably long enough to dissolve the maximum amount of material (detritus), but not so long as to be imeconomical due to slow throughput. Conversely, too short a contact time could mean that all the detritus was not dissolved or removed. Residual detritus can cause pore blockage and/or loss of surface area, with resultant poorer catalyst performance.
More than one washing step is advantageous. Preferably the number of washes is between 2 and 20, more preferably between 3 and 8, and most preferably between 3 and 5.
We have found that the temperature of the wash water is important for facilitating efficient and sufficient washing. Washing at a temperatiu-e above ambient, for example around 140°F and 180°F, has been found to give superior recovery of iorosity and higher surface areas for the resulting washed catalyst base. Wash

amoieni and 210F, more preferably between 50
and 200°F, and most preferably between 100 and 190°F. In one preferred embodiment, the calcined catalyst base is washed at least twice with wash water having a temperature above about 100°F. More preferably, the calcined catalyst base is washed at least three times with wash water having a temperature above about 140°F. A well-washed catalyst base will demonstrate good catalytic performance after the catalytic metals are loaded.
We have found that after room temperature washings, the surface area and tnicroporosity of the resulting washed calcined base is not always equal to (or greater Jian) that of the starting calcined base before washing. While not wishing to be bound )y a theory, we believe that this lower surface area and porosity are due to incomplete emoval of the alkali and silica during the washing steps. We have solved this problem by using higher wash temperatures and/or increased number of washing steps. We have bund that with five washes with water in the 140°F to 180oF range, the resulting cashed calcined base generally has equal or superior surface areas and microporosities s compared to the starting calcined base. Preferred wash conditions thus include evated temperatures and multiple washings.
Washing has also been found to increase the surface area and porosity of the
:truded base, as measured by the micropore volume. Preferably, the washing is done
itil the surface area of the washed catalyst base after calcination returns to at least 80%,
3re preferably to at least 90%, and most preferably to at least 100% or more, of the I
rface area of the calcined catalyst base before washing. I
Although hiz-cats can contain only one halide, it is preferred that they contain )re than one halide, e.g., chloride and fluoride. Halogens, when there are more than 3, can be added separately or, preferably, at the same time. The halogen(s) can be led during platinum addition or in a separate pre- or post-addition step. In a preferred bodiment, they are added during Pt impregnation. The preferred amount of halide, ., chloride and/or fluoride, in the calcined catalyst depends on the type of zeolite, the alytic metal and its loading, and the type of hydrocarbon conversion. For L-zeolite ilysts used for catalytic reforming, which contain about 1.0 wt % Pt, the chloride il after calcination is preferably between 0.4 and 1.0 wt %, more preferably between and 0.8 wt %. Preferably, fluoride is also present, preferably at levels between 0.5

and 0.9 wt %, after calcination, more preferably between 0.6 and 0.8 wt %. The resulting bound hiz-cats are especially useful for preparing aromatic hydrocarbons such as benzene, toluene and xylenes from C6-C8 aliphatic hydrocarbons.
Bound hiz-cats can be prepared using a variety of halide sources; preferred hiz-cats are prepared using amine or ammonium salts. The halides are preferably added by contacting the washed catalyst base with one or more of these halide salts. For L-zeolite reforming, preferred salts are ammonium chloride and ammonium fluoride. These halide salts and the catalytic metal(s) can be added sequentially in any order, or the salts can be added along with the metal, which is preferred. When determining the halide on the catalyst, it comprises both the halide added from the (amine) halide salt and any halide associated with the catalytic metal. When halide is added after Pt addition, it is preferred to add the halide before metal reduction.
For example, one preferred halided Pt L-zeolite catalyst is prepared by impregnating a washed and calcined bound zeolite as described above with an impregnating solution of ammonium chloride, ammonium fluoride and platinum tetraammine chloride. Preferably, in amounts added yield an intermediate having 0.9 ivt % each for fluoride and chloride and 1.0 wt % Pt. This intermediate is vacuum dried It about 40-100°C and then calcined, preferably at a temperature from 260°C to SSO"C, !.g., at about 300EC. When amine halides are used, the calcining should be done so as 0 decompose these salts slowly. Care should be taken not to dry or calcine under onditions that would disturb the even distribution of the ingredients. A uniform istribution, i.e., a flat profile across the bound catalyst, is preferred for both Pt and the alides.
If desired, an ion-exchange step, for example with potassium ions, e.g., KCl or NO3, can be done before the wash step to reduce acidity in the catalyst. As will be ■cognized by those skilled in the art, those ion exchange reactions that would be usefiil the present invention should be carried out within controlled pH parameters, e.g., stween about 4 to 10.8. This will avoid damage to the catalyst or binder and, for talysts that are intended to be non-acidic, will avoid adding any significant acidity.
Commercial Start-up and Cvcle Length As noted earlier, commercial startup conditions differ substantially from >oratory startup conditions. These differences are

control temperatures, and heat transfer considerations, and 2) the ability to achieve practical flow rates associated with use of commercial recycle compressors in conventional catalytic reformers. Thus, as used herein, commercial startup conditions for reforming encompass: low gas flow rates. For example a GHSV of between 400 and 2000 hr*", more preferably between 600 and 1500, and most preferably around about 1000 hr". A commercial start-up also means having "slow" heat-up rates, for example, between 5 and 50°F/hr, more preferably between 10 and 25°F/hr during catalyst reduction above about 500°F. This slow heat-up minimizes catalyst damage. Commercial startups may also include holding the temperature at various levels for several hours or days. This can, for example, facilitate complete reduction of the ;atalyst metal. Test startup conditions for SCSU and LSU are shown in Table 1. Hommercial startups for hiz-cats used for reforming generally include the following iteps: 1) dryout of the catalyst, 2) reduction of the catalyst, and 3) introduction of the eed into the reformers. These steps are usually done in-situ after loading the catalyst ato the reformer(s), but some or all of these steps can be done outside the reactor, i.e., x-situ.
Preferred catalysts useful in this invention have good activity and maintain low )uling rates over many weeks after startup at SCSU conditions. Preferred catalysts ave a cycle length > 1200 hr, preferably >2000 hr and more preferably >4000 hr. The rm "cycle length", as used herein, is the time (oftentimes extrapolated) from start of m (begiiming with feed addition) until the temperature reaches 920°F. Temperature is Ijusted throughout the run so that the catalyst produces a specific target wt % omatics in the C5+ product. This target depends on the composition of the raffinate ed and is typically between 45 and 50 wt % aromatics. Transient temperature cursions are sometimes seen during the first 100-300 hr of a run. When present, ;se temperature excursions (but not the time) are ignored in determining the cycle igth. Cycle length depends on many factors, includmg for example, the H2 to drocarbon ratio, the composition and type of feed, the space velocity (LHSV,) which related to the feed to catalyst ratio, and the severity (e.g., operating temperature or smatively, degree of conversion). To better enable one skilled in the art to ierstand the invention and determine its scope, we have developed a simple test to asure the cycle length of a catalyst (see Example 9). The bound and washed hiy-cats

useful in this invention have a longer cycle length after a commercial startup at SCSU conditions than catalysts prepared in a manner similar to EP 498,182A or US 5,354,933, Ex. 4, as described in Example 1 hereinbelow.
Conversion Processes According to one embodiment, the invention is a catalytic reforming process, ;.g., low sulfur reforming using a bound, halided Pt L-zeolite catalyst. Catalytic eforming is well known. For example, it is described in the book, Catalytic leforming, by D. M. Little, PennWell Books (1985), which is incorporated herein by eference in its entirety. Preferred reforming process conditions include a temperature etween 700 and 1100°F, more preferably between 800 and 1050°F; a pressure between and 400 psig, more preferably between 15 and 150 psig; a recycle hydrogen rate ifiicient to yield a hydrogen to hydrocarbon mole ratio for the feed to the reforming action zone between 0.1 and 20, more preferably between 0.5 and 10 and most ■eferably between 2 and 6; and a liquid hourly space velocity (LHSV) for the drocarbon feed over the reforming catalyst of between 0.1 and 10 hr"1, more eferably between 0.5 and 5 hr"". Reforming produces hydrogen. Thus, additional drogen is not needed except when the catalyst is reduced upon startup, and when the ed is first introduced. Once reforming is underway, part of the hydrogen that is educed is preferably recycled over the catalyst.
The feed to the reforming process of the present invention typically is or is ived from naphtha that contains at least some acyclic hydrocarbons or ylcyclopentanes. Particularly preferred feeds include rafiinates from an aromatics raction unit, such as a Udex raffmate; C6-C9, C7-C9, Cg-Cp, Ce-Cg, C7-C8, C7 and Cg ;tions, especially fractions containing mixtures of five and six-member saturated gs. Preferably, these fractions are obtained by distilling wider boiling range naphtha, by distilling a raffinate from an aromatics extraction unit. The feed should be stantially free of sulfur, nitrogen, metals and other known poisons. These poisons be removed by first using conventional hydrofining techniques, then using sorbents emove the remaining sulfur compounds and water. In a preferred embodiment, the rocarbons contacting the catalyst are substantially dry and free of sulfur, i.e. sulfur Is are preferably maintained below about 50 ppb, preferably below about 25 ppb more preferably below about 10 ppb. Sulfur removal systems are well known in

art. iney include mild reforming using hydrogen and Pt
on alumina, followed by sorption. Sorbents are well known in the art; examples include K on alumina, nickel oxide, and manganese oxide.
The feed can be contacted with the catalyst in a fixed bed system, a moving bed system, a fluidized system, or a batch system. Either a fixed bed system or a moving bed system is preferred. In a fixed bed system, the preheated feed is passed into at least one reactor that contains a fixed bed of the catalyst. The flow of the feed can be upward, downward or radial. The effluent from the catalytic reforming reaction zone can be separated into the desired streams or fractions. Benzene, toluene and C8 iromatic streams can be recovered using conventional techniques such as distillation md extraction.
Although discussed above in terms of catalytic reforming, the invention is pplicable to a variety of catalytic conversion processes that use bound, halided zeolite atalysts to convert feed to products. Preferred hydrocarbon conversion processes iclude dehydrocyclization, especially dehydrocyclization of C6 to Cg paraffins to romatics; catalytic reforming; non-oxidative and oxidative dehydrogenation of ydrocarbons to olefins and dienes; dehydrogenation of ethylbenzene to styrene and/or ehydrogenation of isobutane to isobutylene; conversion of light hydrocarbons to omatics; transalkylation of toluene to benzene and xylenes; hydrodealkylation of kylaromatics to aromatics; alkylation of aromatics to alkylaromatics; production of els and chemiceils from syngas (H2 and CO); steam reforming of hydrocarbons to H2 d CO; production of phenylamine fi-om aniline; methanol alkylation of toluene to lenes; and dehydrogenation of isopropyl alcohol to acetone. More preferred drocarbon conversion processes include dehydrocyclization, catalytic reforming, hydrogenation, isomerization, hydrodealkylation, and conversion of light drocarbon to aromatics, e.g., Cyclar-type processing. These processes, their immercial startup conditions and their usefiil range of process operating conditions are well known to those skilled in the art. These processes can be carried out in a single ctor or in a series of reactors, at least one of which contains the bound and washed -cat described hereinabove.
In a preferred embodiment, the invention is a hydrocarbon conversion process iprising the catalytic conversion of a hydrocarbon-containing feed to product under
25

conversion conditions that include a commercial-type catalyst startup (at a low gas flow rate and a slow heat-up rate), said process using a bound and washed halided zeolite catalyst containing a Group VIII metal, wherein said halided catalyst has a cycle length of>1200hr following said startup. In another embodiment, the invention is a lydrocarbon conversion process, comprising the catalytic conversion of a hydrocarbon-ontaining feed to product under conversion conditions that include a catalyst startup at low gas flow rate and a slow heat-up rate, said process using a halided zeolite catalyst ontaining a Group VIII metal, said halided catalyst prepared by a process comprising ashing a bound zeolite catalyst base or catalyst before halide addition and before duction.
According to another embodiment, the invention is the washed halided zeolite
talyst. A preferred catalyst has a cycle length >1200 hr following startup at SCSU
nditions. As discussed earlier, one preferred catalyst is prepared by a process that
:ludes the steps of preparing a calcined silica-bound zeolite catalyst base; washing
d bound zeolite catalyst base with a liquid comprising water; and either
[lultaneously or sequentially incorporating a group VIII metal and one or more
ogen-containing compoimd(s) comprising chlorine and fluorine into said washed
ie to produce a halided zeolite catalyst. Preferably, the catalyst is prepared using a
ca sol that is low in alkali, and a low-alkali extrusion aid. The extruded base is
lined and then washed before halide addition. The washed base can again be
lined (for example, at 950-1150E F, preferably between 1000-11OOE F) prior to Drporating the halogen-containing compound(s). Preferably, the halides are added ig with the Pt.
Examples To obtain a more complete understanding of the present invention, the following emples illustrating certain aspects of the invention are set forth. It should be erstood, however, that the invention is not intended to be limited in any way to the secific details of the examples.
Example 1 - Preparing a Halided Platinum L-zeolite Catalvst (Comparative)
A halided platinum L-zeolite catalyst was prepared in a manner similar to nple 4 of EP 498,182A or US 5,354,933,. To 100 parts by weight of K L-type
■^-

zeolite (Tosoh, Ltd., Japan), 20 parts by weight of a silica binder (SI-350 silica sol, sold by Cat. & Chem. Ind. Inc. [CCIC], Japan) were added with mixing. This mixture was kneaded and molded, and then air-calcined at 500°C (932°F) for 2 hr to produce a molded calcined silica-bound L-zeoIite extrudate. An impregnation liquid comprising 0.166 g of ammonium fluoride, 0.075 g of ammonium chloride, 0.171 g of platinum tetraammine chloride and 4.8 g of ion exchange water was prepared. This liquid was slowly dropped into 10 g of the molded L- zeolite with stirring. The resulting solid was vacuum dried and then treated at 300°C (572°F) for 3 hr in dry air. The resulting hiz-cat contained about 0.7 wt % F and 0.7 wt % CI. The sodium was 0.56 wt %, as measured by atomic adsorption.
Example 2A ~ Large Scale Preparation of Washed Base A calcined L-zeolite extrudate was prepared on a large scale in a maimer similar to Example 1 using a Tosoh L-zeolite, SI-350 silica sol, and methyl cellulose (Methocel"^ as an extrusion aid. The extrudate (1800 lb) was placed in a large tank and covered with wash water (1.5 g/g base) that was low in alkali (4 ppm Na2O and 1 ppm K2O), alkaline earths, and sulfur species. The water, which had a pH of 7.3, was added over about 30 min and then drained from the tank over about 30 min. The extrudate was washed twice more in the same way. The resultant material was then dried and calcined at 950°F in a rotary calciner. The Na content of the final wash water was less than 100 ppm and the K content was less than 150 ppm. The washing reduced the Na content of the catalyst base from 0.48 to 0.38 wt %.
This triple wash procedure was repeated on another catalyst base sample. The filtrate water contained the following alkali and silicon levels (ppm).

Example 2B - Preparing a Finished Catalyst To make a finished catalyst, 160 g (on a volatile-free basis) of the washed and calcined extrudate of Ex. 2A was impregnated by pore fill at about 100% of the pore
2-1

volume. The pore fill solution comprised 2.86 g of tetraammine platinum chloride (56% Pt), 2.8 g of ammonium fluoride, and 1.3 g of ammonium chloride dissolved in water. Afler completion of the pore fill, the impregnated base was allowed to stand for 2hr. It was then put on a rotaiy evaporator under 22 inches of vacuum for three hr at 212°F. The resultant dry impregnated extnidate was heated fi-om 75° to 575""F over a one-hr period and then calcined at 570°F for 1 hr with 10 SCFH of dry air flowing through the bed. Analysis of the catalyst showed it contained 1 wt % Pt, 0.74 wt % CI and 0.79 wt % F.
Example 3A - Single Washed Catalyst A calcined extnidate was prepared comprising 83% Tosoh L-zeolite (0.28 wt % Na) bound with 17% SI-350 silica. It had a Na content of about 0.5 wt %. It was washed with deionized water (2.5 cc H20/g of base) at room temperature with stirring for 15 minutes. This wash time was sufficient to achieve good contacting of the solids and liquids. The wash water contained 230 ppm Na and 450 ppm K as analyzed by ICP. The washed extnidate was then dried and calcined in air at 1000 °F. This wash and calcine procedure reduced the sodium content of the base to between 0.37-0.41 wt%.
Example 3B -- Impregnation Procedure and Resulting Catalvst Properties The procedure of Ex. 2B was followed usmg the base of Ex. 3 A. The resulting
catalyst made fi-om the washed base of Ex. 3 A contained 0.81 wt % CI and 0.72 wt %°F
and nominally (i.e., as added) 1 wt % Pt.
Example 4A - Triple Washed Base The procedure of Ex, 3 A was repeated except that the wash step was repeated 3 imes. The sodium content of the catalyst base was 0.38 wt % after the third washing, lie filtrate water contained the following alkali levels (ppm in solution) as analyzed by CP.

3rd wash water 26 41
A repeat of this triple wash procedure resulted in filtrate water containing the following materials (ppm).

One must take care to avoid over-washing the catalyst. Over-washing can lead 3 excess acidity, which will result in decreased selectivity. Generally, washing with bout 1-3 volumes of water per weight of catalyst is adequate. We have used 2.5.
Example 4B — Impregnation and Catalyst Properties The procedure of Ex. 2B was followed using the base of Ex. 4 A. The resulting italyst made firom the washed base of Ex. 4A was nominally loaded to 1.0 wt % Pt and mtained 0,69 wt % CI and 0.65 wt % F.
Example 5 —Pilot Plant Configuration Catalyst performance was assessed in pilot plants employing a 1-inch ID type 6 stainless steel in a vertical tubular reactor. Between 40 and 130 cc of whole trudate catalyst (about 0.07 inch, screened to L/D of 2 to 3) was loaded into the ictor. Six thermocouples were equally spaced vertically in the catalyst bed. The term Italyst temperature" as used herein is the catalyst average temperature (CAT). It is culated firom the six thermocouple temperatures as follows:
CAT = 1/10 (TCi + TC6) + 1/5 (TC2 + TC3 +TC4 + TC5) ere TC is the temperature of thermocouple Cx, and the thermocouples are arranged I numbered in order starting at the top of the catalyst bed.
Example 6 -- Start-up Procedure for Lab Test Units The LSU was done in the pilot plant of Ex. 5. The gases were not recycled and evolved gases, including any halides, were vented. The procedure included drying
29

«..«ivuu».iug the catalyst in H2 from room temperature to 1000°F at a ramp rate of 100°F/hr, and then maintaining the catalyst at about 1000oF for 24 hr. GHSV was maintained at 10,000 hr"" during drying and reduction.
After completing the LSU, the catalyst was cooled to approximately 800°F and then feed was introduced. The temperature was then adjusted to achieve the desired total wt % aromatics in the C5+ product.
Although this LSU procedure is not commercially practical, it is much easier to perform and much quicker than a commercial type startup. Simulated commercial startups are more complex and can take a week or more to complete.
Example 7 - Commercial Start-up A commercial reformer startup procedure was simulated using the reactor of Ex. . The gases, including evolved gases, were recycled. The startup included drying the atalyst in N2 from room temperature to 500°F for 79 hr, then heating the catalyst in a tiixture of 10% H2 in N2 from 500 to 932°F at a rate of 10°F/hr over a period of about 3 hr, and then maintaining the catalyst at about 932°F for 24 hr. The gas hourly space elocity (GHSV, which is volume of gas at STP per volume of catalyst) was maintained 11300 hr"" for the drying and reduction periods. The total elapsed time between italyst loading and feed introduction was 146 hr. Table 1 compares LSU and SCSU conditions.

Example 8 -- Feed and Conditions Used for Tests The feeds used for the performance tests were C6-C7 UDEX raffinates from a conventional catalytic reformer that were dried and hydrotreated to
Feed 1 was a lighter Udex raffinate than Feed 2. It contained 10.5 wt % C5 hydrocarbons and less C7"s than C6"s. Feed 2 contained less than 3 wt % C5. Both feeds contained less than 1 wt % C9.
31

The feeds were compared in side-by-side experiments at the following test conditions: a H2/HC ratio of 3; a LHSV of 1.6; and a pressure of 100 psig. To achieve equivalent perfonnance over a given catalyst, the wt % aromatics in the C5+ product was set at 46.5 wt % for Feed 1 and at 48.5 wt % for Feed 2. At these conditions usmg these feeds, good catalyst activity meant that the catalysts lined out at a CAT of between about 820 and 860oF, the lower the temperature, the better the activity. Example 9 - Determining the Cycle Length
After completing the SCSU, the catalyst was cooled to approximately 800°F and then feed was introduced. The reactor temperature was adjusted to maintain an aromatics content of the liquid product as noted above. This was measured by gas chromatographic analysis.
The plot of CAT vs time was used to determine the cycle length. The term "cycle length" as used herein, is the number of hr to reach 920oF from the start-of-run (SOR). It is calculated as 920°F-TSOR / FR, where TSOR is the catalyst average temperature (CAT) extrapolated to time = 0. FR is the fouling rate in hr; FR is the slope of CAT vs time, excluding any initial temperature excursions following startup. SOR, time = 0, is when hot feed is provided to the catalyst.
Example 10 — Perfonnance of Catalysts Using Lab Start-up (LSU) Conditions The perfonnance of catalysts from Ex. 1 and Ex. 2 were determined. Approximate 40 cc of catalyst was loaded into a pilot plant and reduced using the LSU conditions of Ex. 6. The test was continued for about 500 hr, during which both catalysts showed good activity and cycle lengths of > 1200 hr.
Example 11 — Performance of Catalysts Using SCSU Conditions The comparative catalyst of Ex. 1 was tested at the same reforming conditions as in Ex. 10, except that the SCSU conditions of Ex. 7 were used instead of the LSU conditions of Ex. 6. Feed 1 was used for this comparison. Figure 1 shows that the catalyst displayed about the same initial activity after a SCSU (D) as with the LSU (o). however, the catalyst having the SCSU fouled surprisingly rapidly starting at about 200

32.

hr onstream; see curve B. It had a cycle length of about 1100 hr. The figure shows that the catalyst prepared in the manner of EP 498,182A or US 5,354,933 performs well after a laboratory startup, but fouls much more rapidly when a simulated commercial startup is used. This difference in catalyst performance as a function of startup conditions was unexpected. Moreover, modifications in the simulated startup conditions did not lead to significant improvements in catalyst performance.
Example 12 ~ Impact of Washing the Catalyst Base Catalysts prepared according to the wash procedures of Examples 2B, 3B and 4B were tested for reforming in pilot plants using SCSU conditions. They all showed good performance, e.g., long cycle lengths and low fouling rates after a simulated commercial startup.
Figure 2 shows a comparison of the reforming performance of two catalysts. Comparative catalyst 12A (o) was prepared in a manner similar to that in Ex. 1. This catalyst contained 0.56 wt % Na. Catalyst 12B (D), useful in the present invention, was prepared in a manner similar to Ex. 4B. This washed catalyst contained 0.34 wt % Na. The performance of these catalysts was compared using a SCSU and Feed 2. The benefit of washing the bound hiz-cat is evident. Catalyst 12B had a cycle length of >2000 hr, while comparative catalyst 12A had a cycle length of about 550 hr.
Example 13 — Low Sodium Catalyst via Ion Exchange and Washing A calcined extrudate was prepared comprising 83% Tosoh L-zeolite (0.28 wt % Ma) bound with 17% SI-350 silica. It had a Na content of about 0.5 wt %. It was ion exchanged with 0.3 molar KN03, 10 volume per weight at 180°F for 3 hr, and washed times with deionized water (2.5 cc H20/g of base) at room temperature with stirring for 15 minutes. The washed extrudate was then dried and calcined in air at 1000°F for l hr. This ion exchange, wash and calcine procedure reduced the sodium content of the latalyst base to less than 0.09 wt %.
Pt impregnation and addition of ammonium halides results in a catalyst that lerforms well after a simulated commercial startup.
33

Example 14 - Hot Water Washing of Extrudates Table 2 below demonstrates the advantage of multiple washes at elevated temperatures. The experiments with calcined extrudate A, a silica bound L-zeolite catalyst base, were performed as described in Example 4A except that 140oF water was used.
The experiments with calcined extrudate B, also a silica-bound L-zeolite catalyst base, used a cover and drain wash process. They were performed in a manner similar to that described in Example 2 except that the wash water temperature was 180°F.
The accessibility of the pore system, i.e., porosity, and surface area were neasured by conventional techniques. The surface area was measured by the standard nitrogen B.E.T. method (using 5 points). The zeolite micropore volume was measured )y the "t-plot" method. These methods are well known in the art and are merely llustrative; other pore blockage and porosity measurements can also be used. The vashed extrudates were dried at about 250°F and calcined at about 920°F for about an our before surface area and micropore volume were measured.

Example 15 - Catalyst Performance from Hot-Water Washed Extrudates
Catalysts were prepared from some of the materials described in Table 2,
secifically from the 2 washes at 140°F base(14-A2), the 5 washes at 140°F base (14-

34

At)) and the 5 washes at 180°F base (14-B5). The catalyst preparation followed the method of Example 2B. The pilot plant testing of these three catalysts was performed as in Example 12 using SCSU conditions. The catalyst was run for about 500 hours. Catalyst performance is shown in Table 3.

As can be seen, catalysts having an improved activity and fouling rates were obtained after additional washings (15B and 15C) at elevated temperatures.
While the invention has been described above in terms of preferred embodiments, it is to be understood that variations and modifications may be used as will be appreciated by those skilled in the art. Indeed, there are many variations and modifications to the above embodiments, which will be readily evident to those skilled in the art, and which are to be considered within the scope of the invention as defined

WE CLAIM :
1. A catalytic reforming process comprising the catalytic conversion of
hydrocarbons to aromatics, said process comprising treating a halided zeolite catalyst containing a Group VIII metal at commercial startup conditions and then reforming hydrocarbons, wherein said catalyst is prepared by washing a bound zeolite catalyst base or a bound zeolite catalyst before halide addition.
A catalytic reforming process according to claim 1 wherein the halides of the halided zeolite catalyst comprise both chloride and fluoride.
A catalytic reforming process acco^lmgjo claim 1 wherein said halided zeolite catalyst comprises a halided platinum L-zeolite catalyst.
A catalytic reforming process according to claim 2 wherein ammonium halides are used to prepare the halided zeolite catalyst.
A catalytic reforming process according to claim 1 wherein said catalyst is prepared by a process that includes binding, extruding, calcining and washing the bound catalyst base before halide addition.
A catalytic reforming process according to claim 1 wherein said catalyst is produced by a method that includes the steps of:
a) preparing a calcined silica-bound zeolite catalyst base;
b) washing said bound zeolite catalyst base with a liquid comprising water; and
c) incorporating a group VIII metal and halogen-containing compound(s) comprising chlorine and fluorine into said washed base to produce a halided zeolite catalyst.
A catalytic reforming process according to claim 6 wherein said catalyst is prepared by a process that utilizes an extrusion aid in step (a) and further includes calcining the washed base of step (b) prior to incorporating the halogen-containing compound.
A catalytic reforming process according to claim 5, 6, or 7 wherein said halided zeolite catalyst comprises a halided platinum L-zeolite catalyst.
36

catalyst was prepared using ammonium chloride and fluoride.
A catalytic reforming process according to claim I wherein said process comprises contacting a hydrocarbon stream containing less than 50 ppb sulfur with said halided zeolite catalyst.
A halided zeolite catalyst prepared according to the catalyst preparation method ofClaim 5, 6, or 7.
A halided zeolite catalyst according to Claim 11 wherein said halided catalyst
comprises an L-zeolite.
A halided zeolite catalyst according to Claim 11 wherein said halided catalyst
comprises an L-zeolite manufactured By Tosoh, Ltd., Japan.
A halided zeolite catalyst according to Claim 11 wherein said catalyst base is
washed until the surface area of the washed catalyst base after calcination returns
to at least 90% of the surface area of the calcmed catalyst base before washing.
A halided zeolite catalyst according to Claim 11 wherein said catalyst base is washed at least twice with wash water having a temperature above about 100°F.
A catalytic reforming process according to claim 1 wherein a stream containing at least C6 and C7 hydrocarbons containing less than 50 ppb sulfur is contacted with a non-acidic Pt K L-zeolite catalyst comprising chloride and fluoride, said catalyst having been washed with water before halide addition.
A process for producing aromatics comprising preparing a halided zeolite catalyst by washing a calcined bound zeolite catalyst base with a liquid comprising water; adding at least one halogen-containing compound and a Group VIII metal compound to said washed base; bringing said catalyst on stream using commercial startup conditions; and contacting the catalyst with paraffinic hydrocarbons at catalytic reforming conditions to produce aromatics.
The process of claim 17 wherein the catalyst comprises both chloride and fluoride, and the zeolite is an L-zeolite.
\ hydrocarbon conversion process comprising the catalytic conversion of a lydrocarbon-containing feed to product under conversion conditions that

process using a halided zeolite catalyst containing a Group VIII metal, said halided catalyst prepared by a process comprising washing a bound zeolite catalyst base or catalyst before halide addition.
20. A hydrocarbon conversion process according to claim 19 wherein the halides of the halided zeolite catalyst comprise both chloride and fluoride.
21. A hydrocarbon conversion process according to claim 19 wherein said halided zeolite catalyst was prepared using ammonium chloride and fluoride.
22. A catalytic reforming process, substantially as hereinabove described and illustrated with reference to the accompanying drawings.

Documents:

0036-mas-1999 abstract-duplicate.pdf

0036-mas-1999 abstract.pdf

0036-mas-1999 claims-duplicate.pdf

0036-mas-1999 claims.pdf

0036-mas-1999 correspondence-others.pdf

0036-mas-1999 correspondence-po.pdf

0036-mas-1999 description (complete)-duplicate.pdf

0036-mas-1999 description (complete).pdf

0036-mas-1999 drawings-duplicate.pdf

0036-mas-1999 drawings.pdf

0036-mas-1999 form-13.pdf

0036-mas-1999 form-19.pdf

0036-mas-1999 form-26.pdf

0036-mas-1999 form-4.pdf

0036-mas-1999 form-8.pdf

0036-mas-1999 other-document.pdf

0036-mas-1999 others.pdf

0036-mas-1999 petition.pdf

« Previous Patent

Next Patent »

Patent Number

216638

Indian Patent Application Number

36/MAS/1999

PG Journal Number

17/2008

Publication Date

25-Apr-2008

Grant Date

17-Mar-2008

Date of Filing

07-Jan-1999

Name of Patentee

CHEVRON CHEMICAL COMPANY LLC

Applicant Address

555 MARKET STREET, SAN FRANCISCO, CALIFORNIA 94105,

Inventors:

#	Inventor's Name	Inventor's Address
1	DENNIS L HOLTERMANN	55 PENNINGTON COURT, CROCKETT, CALIFORNIA 94525,
2	CHARLES R WILSON	1101 DIAMOND STREET, SAN FRANCISCO, CALIFORNIA 94114,
3	WAYNE MASAMURA	753 PERALTA AVENUE, BERKELEY, CALIFORNIA 94707,
4	MICHIO SUGIMOTO	1280 KAMIZUMI SODEGAURA, CHIBA 299-02,
5	HIROSHI OHASHI	1280 KAMIZUMI SODEGAURA, CHIBA 299-02,

PCT International Classification Number

C10G 35/04

PCT International Application Number

N/A

PCT International Filing date

PCT Conventions:

#	PCT Application Number	Date of Convention	Priority Country
1	09/004,216	1998-01-08	U.S.A.