BACKGROUND OF THE INVENTION Metallocene olefin polymerization catalysts generally use cyclopentadienyl (Cp), indenyl and fluorenyl groups z-bonded to Group III to VI metals as procatalysts and cocatalysts comprising primarily alumoxanes and boranes. These metallocene catalysts can be used to produce stereospecific polymers from olefins, which is complicated by the structure of these compounds. From the nature of their structure, some compounds of the metallocenes exist as two stereoisomers: a racemic modification and a meso form. The racemic modification includes a pair of enantiomers which is simply a pair of stereoisomers that are nonsuperimposable mirror images. The meso form is a stereoisomer which can be superimposed on its mirror image.

Polymerization employing the meso form has traditionally resulted in the production of primarily atactic polymer. See e.g., U.S. Patent Nos.

5,455,366; 5,329,033; and 5,145,819. See also, Wolfgang A. Herrmann, et al., The First Example of an Ethylene-Selective Soluble Ziegler Catalyst of the Zirconocene Class, 28 Angew. Chem. Int. Ed. Engl., 1511 (No. 11, 1989). Production of polymer primarily of the isotactic form has necessitated use of a catalyst consisting almost exclusively of the racemic modification.

Preparation of metallocene procatalysts has traditionally been conducted in polar solvent systems. See, for example, in European Patent Publication 0 351 392 A. However, due to the aggressive nature of these solvent systems, metallocene procatalysts produced in polar solvents are impure, sensitive to air and moisture, and are susceptible to the decomposition of reaction intermediates, particularly the lithiated ligand, which is generally indicated by a dark colored reaction solution. Additionally, procatalyst yields are also generally low for processes involving a polar solvent system.

Metallocene procatalysts prepared in a polar solvent suffer another difficulty if their intended use is as a catalyst for the production of substantially isotactic polyolefins because the use of these metallocene procatalysts to make stereospecific polymers has typically dictated an additional step in which the racemic modification is separated from the meso form. See e.g., U.S. Patent Nos. 5,455,366; 5,329,033; and 5,145,819. See also, Herrmann, et al., 28 angel. Chem.

Int. Ed. Engl., 1511 (No. 11, 1989). Additionally, separation ofthe two isomers is inefficient because of the low yield of procatalyst, as well as the additional purification step required. Alternatively, processes have been described in which the meso form is defeated by substituting various groups onto the Cp groups. See, e.g., U.S. Patent No.

5,349,032.

To avoid the difficulty associated with the use of polar solvents, a process which employs non-polar solvent has been used, but this process is relatively slow and does not address those specific instances in which it is intended that the metallocene procatalyst be used in stereospecific polymerizations. See e.g., U.S. Patent No.

5,117,020.

A more desirable process would allow for the production of metallocene procatalysts at high yields but in relatively short reaction times. Additionally, such a process would allow for the production of metallocene pro catalysts that could be used for the stereospecific polymerization but would not entail preparative steps in which various stereoisomers of the procatalyst are separated and would not be restricted to procatalyst that, due to substitutions, do not have a meso form as a stereoisomer.

SUMMARY OF THE INVENTION In accordance with this invention, there is provided metallocene procatalysts known to be useful in the polymerization of olefins to make stereospecific polymers and a process for preparing same. The process for preparing the metallocene procatalysts of this invention entails reacting (a) a transition metal salt and (b) a reaction product obtained by reacting (i) an alkyl compound of an alkali metal with (ii) a ligand (including cyclopentadiene, indenyls or fluorenyls which may or may not be substituted and which may or may not be bridged) in the presence of (c) an accelerant used in a quantity insufficient to produce solution. The transition metal salt and the reaction product are reacted in the presence of an accelerant used in a quantity insufficient to produce a solution.

When this process of the present invention is used to form metallocene procatalysts which exist in both racemic and meso forms, the resultant metallocene procatalysts can be used in the stereospecific polymerization of olefins to make stereospecific polymers without separation of the stereoisomers.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a 1H NMR spectra of a ligand produced from a series in Example 1 of reactions employing indene, butyl lithium, and dimethydichlorosilane as reactants; Figure 2 is a 1H NMR spectra of a reaction product containing bisindenyldimethylsilyl lithium; Figure 3 is a 1H NMR spectra of the soluble portion of the reaction product produced in Example 1, section C, containing the dimethylsilyl-bisindenyl ZrCl2; and Figure 4 is a diagram of the racemic and meso stereoisomers of dimethylsilyl-bisindenyl ZrCl2 produced in Example 1.

DETAILED DESCRIPTION OF THE INVENTION Metallocenes are organometallic coordination complexes of one or more it-bonded moieties in association with a metal atom from groups III to VIII or the rare earth metals of the Periodic Table.

Bridged and unbridged mono-, di-, and tri-cycloalkadienyl/metal compounds are the most common metallocene catalysts, and generally are of the formula: (L)yR1z(L)MX(x-y-1) (I) wherein M is a metal from groups III to VIII or a rare earth metal of the Periodic Table; L and L' are the same or different and are z-bonded ligands coordinated to M, preferably cycloalkadienyl groups such as Cp, indenyl, or fluorenyl groups (optionally substituted with one or more hydrocarbyl groups containing 1 to 20 carbon atoms); R1 is selected from the group consisting of C1-C4 substituted or unsubstituted alkylene radicals, dialkyl or diaryl germanium or silicon groups, and alkyl or aryl phosphine or amine radicals bridging L and L'; each X is independently hydrogen, an aryl, alkyl, alkenyl, alkylaryl, or arylalkyl radical having 1-20 carbon atoms, a hydrocarboxy radical having 1-20 carbon atoms, a halogen, R2CO2-, or R22NCO2-, wherein each R2 is a hydrocarbyl group containing 1 to about 20 carbon atoms; y is 0, 1, or 2; x is 1, 2, 3, or 4 depending upon the valence state of M; z is 0 or 1 and is 0 when y is 0; and x-y # 1.

Illustrative, but non-limiting, examples of metallocenes are dialkyl and diaryl metallocenes [e.g., (Cp)2Ti(CH3)2, (Cp)2Ti(C6Hs)2, (Cp)2Zr(CH3)2, (Cp)2Zr(C6Hs)2, (Cp)2Hf(C6H5)2, (Cp)2Ti di-neopentyl, (Cp)2Zr di-neopentyl, (Cp)2Hf di-neopentyl, (Cp)2Ti di-benzyl, (Cp)2Zr di-benzyl,(Cp)2Hf di-benzyl]; mono alkyl metallocenes [e.g., (Cp)2M'(CH3)J, (Cp)2M' neopentyl J, (Cp)2M'C6HsJ, (where J = F, Cl, Br, or I, preferably Cl, and where M' is Zr, Hf or Ti, preferably Zr)]; trialkyl metallocenes [e.g., CpTi(CH3)3, CpTi(C6Hs)3, CpTi tri- neopentyl, CpZr(CH3)3, CpZr(C6Hs)3, CpZr tri-neopentyl, CpHf(CH3)3 CpHf(C6Hs)3, CpHf tri-neopentyl]; mono Cp metal tri-halide complexes [e.g.,CpM'J3, penta-methylCpM'J3, penta-ethylCpM'J]; bis-Cp metal di-halide complexes [e.g., Cp2M'J2, penta-methyl Cp2M'J2, penta- ethylCp2M'X2, bis( 1,3 dimethyl-Cp)M'J2]; dialkyl indenyl metallocenes [e.g., (indenyl)2M'(CH3)2, (indenyl)2M'(C6H5)2, (indenyl)2M' di- neopentyl, (indenyl)2M' di-benzyl]; mono alkyl indenyl metallocenes, [e.g., (indenyl)2M'(CH3)J, (indenyl)2M' neopentyl J, (indenyl)2M'C6H5J], trialkyl or triaryl indenyl metallocenes [e.g., (indenyl)M'(CH3)3, (indenyl)M'(C6H5)3, (indenyl)M' tri-neopentyl]; mono-indenyl metal tri- halide complexes [e.g., (indenyl)M'J3, hepta-methylindenyl M'J3, hepta- ethylindenylM'J3; bis-indenyl metal di-halide complexes [e.g., indenyl2M'J2, hepta-methylindenyl2M'J2, hepta-ethylindenyl2M'J2, bis( 1,3 dimethyl-indenyl)M'X2]; bisfluorenyl structures [e.g., bisfluorenylM'J2, bis-nona methyl fluorenylM'J2, bis - 1-methyl fluorenylM'J2]; the carbene represented by the formula bis(Cp)Ti=CH2 and derivatives of this reagent; silicon, phosphine, amine or carbon bridged Cp complexes [e.g., (CH3)2silyldiCp Ti(C6H5)2 or Cl2 CH Cl2 phosphine diCp Ti (C6Hs)2 or Cl2]; methylene diCpTiCl2 or (C6H5)2 and other dihalide complexes, and the like; as well as bridged metallocenes which are like those above with z=1 per the above formula I, except with the following bridging groups (i.e., R' in the above formula I): Me2Si, Et2Si, Ph2Si, MePhSi, MeEtSi, EtPhSi, Me2Ge, Et2Ge, Ph2Ge, MePhGe, MeEtGe, MeCH, Me2C, Et2C, Ph2C, MePhC, MeEtC, EtPhC, iPr2C, t-Bu2C, ethylene, tetramethylethylene, diphenyl ethylene, methyl ethylene, propylene, methylamine, butylene, and methyl phosphine.

The cocatalyst may be any of those known to one of ordinary skill in the art for activating the metallocene, and preferably, is one of the following: (a) branched or cyclic oligomeric poly(hydrocarbyl- aluminum oxide)s which contain repeating units of the general formula -(Al(R*)O)-, where R* is hydrogen, an alkyl radical containing from 1 to about 12 carbon atoms, or an aryl radical such as a substituted or unsubstituted phenyl or naphthyl group; (b) ionic salts of the general formula [A+] [BR**4-], where A+ is a cationic Lewis or Bronsted acid capable of abstracting an alkyl, halogen, or hydrogen from the metallocene catalysts, B is boron, and R* is a substituted aromatic hydrocarbon, preferably a perfluorophenyl radical; and (c) boron alkyls of the general formula By**3, where R** is as defined above.

Preferably, the activating cocatalyst is an aluminoxane such as methylaluminoxane (MAO) or modified methylaluminoxane (MMAO), or a boron alkyl. Aluminoxanes are preferred and their method of preparation is well known in the art. Aluminoxanes may be in the form of oligomeric linear alkyl aluminoxanes represented by the formula: or oligomeric cyclic alkyl aluminoxanes of the formula: wherein s is 1-40, preferably 10-20; p is 3-40, preferably 3-20; and R8** is an alkyl group containing 1 to 12 carbon atoms, preferably methyl or an isobutyl radical.

When the process of the present invention is used for the preparation of metallocene procatalysts which can be subsequently used in the formation of stereospecific polymers the choice of ligands is restricted. Acceptable ligands for this use of the process include Cp type ligands that when reacted with an alkyl compound of an alkali metal and subsequently with a transition metal salt result in the formation of metallocene compounds having as stereoisomers both a racemic modification and a meso form. When this is the desired use of the catalyst, the substituent is preferably of sufficient bulk to provide steric hindrance for the metal center. For this reason, bisindenyl and bis(substituted indenyls) groups are preferred. Generally, the two groups around the metal center are bridged to prevent rotation of the metallocene around the metal center; however, non-bridged groups can be employed as long as their size substantially minimizes rotation around the metal center. Ligands that are particularly preferred for use with the process in the production of iPP have a bridged-bisindenyl group and include dimethylsilyl-bisindenyl, (Indenyl)2Si(CH3)2MCl2, (Indenyl)2C(Ph)2 MCl2, where M = Zr, Ti, Hf.

Ligands which may be used with the process may be prepared by using any number of processes known in the art. See e.g., U.S. Patent 4,985,576; and 110 J. AM. CHEM. Soc., 976, 976-978 (1988).

The reaction of the above ligands with an alkyl compound of alkali metals is known in the art. See e.g., U.S. Patent No. 4,931,417; European Patent Publication 0 320 762 A2. Any alkali-alkyl can be used, although Li, Na and K compounds are preferred, with Li compounds being the most preferred. Typically, the lithium alkyl or other alkali-alkyl is added dropwise over a period of time to the ligand which has been previously slurried with a hydrocarbon solvent system.

Other solvents, such as ethers, also may be present. The resultant product can be isolated by any of the known processes, including filtering or evaporation of the solvent. The product can be, and preferably is, subjected to further steps such as washing in an inert hydrocarbon, recrystallization, or grinding. Typical lithium alkyls employed include CH3Li, n-C3H7Li, and n-C4HsLi. Particularly preferred is n-butyl lithium.

Metal halides and metal halide derivatives are typically employed as the transition metal salt which is reacted with the product of the ligand and alkali-alkyl to produce the metallocene.

Preferred salts for this invention are the tetrahalides of the previously indicated preferred transition metals, with tetrachlorides being particularly preferred.

The metal salt and the product of the reaction between the ligand and the alkali-alkyl are reacted in the presence of an accelerant.

Accelerants are compounds with a heteroatom having at least one lone pair of electrons in their outer valence shell that do not have a donateable or labile proton (that is, excluding compounds like primary and secondary amines, and primary and secondary phosphines). These accelerants include, but are not limited to ethers, such as diethyl ether, butyl ether and tetrahydrofuran (THF); tri, triaryl, or triaralkyl alkylphosphines, such as triethyl phosphine, trimethyl and triphenyl phosphine; tertiary alkyl, aryl, and aralkyl amines, such as trimethylamine, tripropylamine, and tributylamine; and pyridines, substituted or unsubstituted. Also included in this class are certain sulfur containing compounds. The accelerant may be present as a carryover material from a prior reaction step, such as the reaction between the ligand and the alkali-alkyl. Optionally, it can be added to the transition metal salt and the product of the reaction between the ligand and the alkali-alkyl once they are combined or to one of these reactants prior to the addition of the other.

When a solvent is not used, the amount of accelerant employed should be an amount greater than 0% and less than about 70% by weight of the combined mass of the transition metal salt and the reaction product of alkali-alkyl and ligand. Preferably the accelerant is used at an amount greater than 0% and less than about 55% by weight, and, more preferably, between about 40% and about 55% by weight. In all cases, the accelerant is used in an amount insufficient to form a solution in which the accelerant would be the solvent, i.e., the accelerant is not to be used as the solvent.

When a solvent is used, the amount of accelerant employed should be an amount greater than 0% and less than about 20% by weight of the combined mass of the transition metal salt, the reaction product of alkali-alkyl and ligand and the non-polar solvent.

Preferably, the accelerant is used at an amount greater than 0% (usually greater than 0.01%) and less than about 10% by weight, and, more preferably, between about 2% and about 10% by weight. In all cases, the accelerant is used in an amount insufficient to form a solution in which the accelerant is used in an amount insufficient to form a solution in which the accelerant would be the solvent, i.e., the accelerant is not to be used as the solvent.

In contrast, if the metallocene procatalyst is produced in a quantity of accelerant sufficient to create a somewhat polar solvent system (e.g., with the accelerant as the solvent), the reaction mixture turns dark even when the mixture is maintained at low temperatures (25"C), which indicates the decomposition of the reaction intermediates.

The reaction between the transition metal salt and the product of the reaction between the ligand and the alkali-alkyl in the presence of a small quantity of an accelerant may be conducted in a substantially non-polar solvent system, i.e., solvents with a dielectric constant less than 7. However, it is noted that it is not necessary for the reactants to be completely soluble in the chosen solvent.

Non-limiting examples of non-polar solvents for this purpose include aromatic and non-aromatic hydrocarbons such as waxes, mineral oils, benzene, toluene, alkanes, and alkenes. Chlorinated solvents, such as chlorobenzene, can also be used, although they are not preferred.

Preferred non-polar solvents are the alkanes and alkenes having 3 to 12 carbon atoms, more preferred are the alkanes having 3 to 8 carbon atoms, and most preferred is hexane.

The use of a non-polar solvent provides an advantage when metallocene procatalysts are produced which can be utilized in the stereospecific formation of polymers. When a procatalyst which can exist as both a racemic modification and a meso form (and which is therefore potentially useful in stereospecific polymerizations) is desired, the use of a substantially non-polar solvent produces a metallocene procatalyst which can be used without the additional isolation and separation of the racemic form which is required as compared to when a polar solvent is used in the reaction between transition metal salt and the product of the ligand and the alkali-alkyl reaction.

Because of the avoidance of these steps, the metallocene obtained herein is typically produced at higher yields than those previous processes in which a polar solvent was used.

The metal salt and the product of the reaction between the ligand and the alkali-alkyl, existing generally as powders, can be mixed together prior to solvent addition or, alternatively, one of the powders can be added to the solvent prior to the addition of the other powder. Likewise, the point in the reaction at which the accelerant can be added is not limited. These alternatives include adding the accelerant to the combined reactants prior to solvent addition, adding the accelerant to one of the reactants either before, with, or after solvent addition, or adding the accelerant after the two reactants have been added to the solvent. Optionally, the accelerant may be present as a carry over material from a previous process step, such as the reaction between the ligand and the alkali-alkyl. It is preferred that the accelerant be added to the suspension of the two reactants in an inert solvent.

The volume of accelerant used should not exceed 30% of the total solvent volume. The accelerant should be employed preferably in the range of about 2% to about 5% of the total solvent volume.

Additionally, the amount of accelerant employed should fall within the range previously indicated regarding the mass of accelerant used relative to the combined mass of the transition metal salt and the reaction product of alkali-alkyl and ligand.

The metal salt and the product of the reaction between the ligand and the alkali-alkyl should be mixed thoroughly to aid the reaction. Preferably, mixing is aided by ball milling or mortar and pestle, although other means known in the art for mixing to ensure completeness of reaction may be employed. This preparation process does not require the use of a solvent. Observed color changes when the reactants, typically existing as powders, are combined and mixed indicate that the reaction proceeds relatively rapidly toward the desired procatalyst. When Zr is used, the observed color of the procatalyst is typically orange. When Hf is used, the procatalyst is typically light yellow to light orange. It is not necessary to remove the alkali halide which is formed in the reaction.

Color changes are again observed when the transition metal salt and the product of the reaction between the ligand and the alkali- alkyl are combined in a non-polar solvent system containing a sufficient amount of accelerant. Conversely, when these materials are combined in a non-polar solvent system or with no accelerant present, no color change indicative of a reaction is observed within 48 hours.

The procatalyst formed by the presently disclosed method is distinct from previously known metallocenes in that they are insoluble in methylene dichloride at reflux conditions. In distinction, it is noted, that previously described metallocenes are soluble in methylene dichloride. By insoluble it is possible that a minute amount of metallocene does dissolve or is carried in the methylene dichloride, but such amount is at less than 0.5 grams per liter of methylene dichloride.

The reaction which results in the formation of the procatalyst can be carried out under a variety of conditions. However, temperatures and pressures employed should be selected to ensure that neither the solvent nor the accelerant escapes the reaction vessel.

The temperatures and pressures employed will therefore depend upon the boiling points of the compounds used as solvents and accelerants.

Preferably, the reaction is conducted in a temperature range of between about 10° to about 30"C under pressures sufficient to ensure that the contents of the reactor vessel do not escape. The reaction can be conducted with or without mixing. Mixing or increased temperature results in a shorter reaction time. Preparation and storage of the metallocene is generally conducted under moisture and oxygen free conditions, typically in a dry box.

The metallocene product prepared can be recovered by any known means, including evaporation, decantation, or filtration.

Filtration followed by an extended drying period is preferred. Drying may be done at about 10°C to 70"C, preferably about 25° to 40"C, and preferably under an inert gas, e.g., N2. The product may be washed with several aliquots of a non-polar solvent which may or may not be the same as the non-polar solvent previously used. Washing in a non- polar solvent is preferred because, among other things, washing aids in the removal of unreacted intermediates and accelerant which may interfere with catalyst activation.

Alternatively, the metallocene obtained from the process of this invention can be used without isolating it from the solvent system in which it was produced. The cocatalyst can be added to the procatalyst which has not been separated from the solvent. The cocatalyst and procatalyst may be preactivated, prepolymerized, or not combined until the reactor, whichever mode is desired. The catalyst system existing as a suspension of procatalyst and cocatalyst in the solvent can be used in a polymerization. Although the use of this method is viable, it may be necessary to ensure any excess reactants and accelerant are removed; otherwise, catalyst activity might be detrimentally affected.

The amount of activating cocatalyst and metallocene catalyst in the catalyst composition, whether the catalyst composition is formed in situ as it is being introduced into the reaction zone or formed prior to introduction into the reaction zone, can vary over a wide range.

When the cocatalyst is a branched or cyclic oligomeric poly(hydrocarbylaluminum oxide), the mole ratio of aluminum atoms contained in the poly(hydrocarbylaluminum oxide) to metal atoms contained in the metallocene catalyst is generally in the range of from about 2:1 to about 100,000:1, preferably in the range of from about 10:1 to about 10,000:1, and most preferably in the range of from about 50:1 to about 2,000:1. When the cocatalyst is an ionic salt of the formula [A+] [BR*4-] or a boron alkyl of the formula BR*3, the mole ratio of boron atoms contained in the ionic salt or the boron alkyl to metal atoms contained in the metallocene catalyst is generally in the range of from about 0.5:1 to about 10:1, preferably in the range of from about 1:1 to about 5:1.

The catalyst, cocatalyst, or entire catalyst composition may be introduced to the reaction zone of the polymerization system on a solid support, in unsupported, liquid form such as a solution or dispersion, spray dried, in the form of a prepolymer, or formed in-situ in the reaction zone. In the case of a supported catalyst composition, typical supports include, for example, silica, carbon black, polyethylene, polycarbonate porous crosslinked polystyrene, porous crosslinked polypropylene, alumina, thoria, zirconia, and magnesium halide (e.g., MgCl2) as well as other well known support materials and mixtures thereof. When supported, the catalyst and/or the activating cocatalyst are impregnated in or deposited on the surface of the inert support substrate, such that the impregnated material is between 1 and 90 percent by weight of the total weight of the impregnated material and the support.

Particularly preferred among these is a catalyst composition that is spray dried or in unsupported, liquid form. For example, the catalyst composition may be introduced into the reaction zone in unsupported, liquid form as described in U.S. Patent No. 5,317,036.

As used herein, "unsupported, liquid form" includes liquid catalyst, liquid cocatalyst, solution(s) or dispersions of catalyst and cocatalyst in the same or different solvent(s), and combinations thereof.

Unsupported, liquid form catalyst compositions have a number of practical benefits. Unsupported catalyst compositions avoid the costs associated with support material and its preparation, and the provide for the realization of a very high catalyst surface area to volume ratio.

Furthermore, unsupported catalyst compositions produce polymers having a much lower residual ash content than polymers produced using supported catalyst compositions.

According to the invention, polymer build up or fouling in an olefin polymerization system employing a catalyst composition described above may be inhibited by introducing into the polymerization system an antifouling agent. Preferably, the antifouling agent is an ether, alcohol, ketone, aldehyde, carboxylic acid, ester, carbonate, phosphine, phosphine oxide, phosphate, phosphite, amine, amide, nitrile, alkoxy silane, aluminum alkoxide, oxygen, nitric oxide, carbon oxide, and the like.

As stated above, the present invention is particularly suited to make stereospecific polymers. Stereospecificity can be measured by the amount of xylene solubles present in the resulting polymer.

Generally, it is desirable to have xylene solubles of less than about 7 wt % (21 CFR 177.1520), and more preferably less than about 4 wt. %, which would indicate stereospecific polymerization.

The invention is applicable to the polymerization of olefins by any suspension, solution, slurry, or gas phase process, using known equipment and reaction conditions, and is not limited to any specific type of polymerization system. Generally, olefin polymerization temperatures range from about 0°C to about 200"C at atmospheric, subatmospheric, or superatmospheric pressures. Slurry or solution polymerization systems may utilize subatmospheric or superatmospheric pressures and temperatures in the range of about 40"C to about llO"C. A useful liquid phase polymerization system is described in U.S. Patent 3,324,095. Liquid phase polymerization systems generally comprise a reactor to which olefin monomer and catalyst composition are added, and which contains a liquid reaction medium for dissolving or suspending the polyolefin. The liquid reaction medium may consist of the bulk liquid monomer or an inert liquid hydrocarbon that is nonreactive under the polymerization conditions employed. Although such an inert liquid hydrocarbon need not function as a solvent for the catalyst composition or the polymer obtained by the process, it usually serves as solvent for the monomers employed in the polymerization. Among the inert liquid hydrocarbons suitable for this purpose are isopentane, hexane, cyclohexane, heptane, benzene, toluene, and the like. Reactive contact between the olefin monomer and the catalyst composition should be maintained by constant stirring or agitation. The reaction medium containing the olefin polymer product and unreacted olefin monomer is withdrawn from the reactor continuously. The olefin polymer product is separated, and the unreacted olefin monomer and liquid reaction medium are recycled into the reactor.

The invention is, however, especially useful with gas phase polymerization systems, with superatmospheric pressures in the range of 1 to 1000 psi, preferably 50 to 400 psi, most preferably 100 to 300 psi, and temperatures in the range of 30 to 1300C, preferably 65 to llO"C. Stirred or fluidized bed gas phase polymerization systems are particularly useful. Generally, a conventional gas phase, fluidized bed process is conducted by passing a stream containing one or more olefin monomers continuously through a fluidized bed reactor under reaction conditions and in the presence of catalyst composition at a velocity sufficient to maintain a bed of solid particles in a suspended condition.

A stream containing unreacted monomer is withdrawn from the reactor continuously, compressed, cooled, optionally partially or fully condensed, and recycled into the reactor. Product is withdrawn from the reactor and make-up monomer is added to the recycle stream. As desired for temperature control of the polymerization system, any gas inert to the catalyst composition and reactants may also be present in the gas stream. In addition, a fluidization aid such as carbon black, silica, clay, or talc may be used, as disclosed in U.S. Patent No.

4,994,534.

The polymerization system may comprise a single reactor or in two or more reactors in series, and is conducted substantially in the absence of catalyst poisons. Conventional adjuvants, including scavenging agents, may be used in the process, provided they do not interfere with the operation of the catalyst composition in forming the desired polyolefin. Hydrogen may be used as a chain transfer agent in the process, in amounts up to about 10 moles of hydrogen per mole of total monomer feed.

Olefin polymers that may be produced according to the invention include, but are not limited to, ethylene homopolymers, homopolymers of linear or branched higher olefins containing 3 to about 20 carbon atoms, and interpolymers of such olefins, with densities ranging from about 0.86 to about 0.95. Suitable higher olefins include, for example, propylene, 1-butene, 1-pentene, 1-hexene, 4-methyl- 1-pentene, 1-octene, and 3 ,5,5-trimethyl-1-hexene. Though, because of the utility of the present invention with stereospecific polymerizations, it is preferred to use olefin momomers which take advantage of same, e.g., propylene. Olefin polymers according to the invention may also be based on or contain conjugated or non- conjugated dienes, such as linear, branched, or cyclic hydrocarbon dienes having from about 4 to about 20, preferably 4 to 12, carbon atoms. Preferred dienes include 1,4-pentadiene, 1,5-hexadiene, 5- vinyl-2-norbornene, 1,7-octadiene, vinyl cyclohexene, dicyclopentadiene, butadiene, isobutylene, isoprene, ethylidene norbornene and the like. Aromatic compounds having vinyl unsaturation such as styrene and substituted styrenes, and polar vinyl monomers such as acrylonitrile, maleic acid esters, vinyl acetate, acrylate esters, methacrylate esters, vinyl trialkyl silanes and the like may be polymerized according to the invention as well. Specific olefin polymers that may be made according to the invention include, for example, polyethylene, polypropylene, ethylene/propylene rubbers (EPR's), ethylene/propylene/diene terpolymers (EPDM's), polybutadiene, polyisoprene and the like.

EXAMPLE S EXAMPLE 1 A. Preparation of the Ligand 20.0 grams (g) indene (172 mmol) was dissolved in 30 mL diethyl ether. This was cooled to 0CC in an ice bath. 53.5 mL C4H9Li (133 mmol, 2.5 M in hexane) was added dropwise over 75 minutes.

The resulting orange solution of indenyl lithium was warmed to room temperature and stirred for 15 minutes. The solution was transferred to a dropping funnel attached to a flask containing (CH3)2SiCl2 (8.4 g, 65 mmol, in 20 mL of diethyl ether). This indenyl lithium solution was added to the (CH3)2SiCl2 solution dropwise over 2.5 hours. The resulting solution was stirred overnight. The solution was washed with a saturated solution of sodium chloride in water, and extracted with diethyl ether. The ether layer was dried twice over MgSO.

Ether was removed in vacuo, and the residue was warmed to 40° C under vacuum to remove any unreacted indene. The residue bisindenyl(CH3)2Si was crystallized from methanol, and the resulting solid was dried in vacuo (yield 14.84 g, 79%). 1H NMR of the product in CDCl3 is given in Figure 1. The 1H NMR is consistent with the presence of the two diastereomers.

B. Reaction of the Ligand and Alkyllithium 8.0 g bisindenyl(CH3)2Si (27.7 mmol) was slurried in 10 mL diethyl ether and 15 mL hexane. The resulting slurry was cooled to 0°C, and C4H9Li (55 mmol, 2.5 M in hexane) was added dropwise over 1 hour. The resulting solution was warmed to room temperature (about 25"C) and stirred for 2 hours. The precipitate was filtered and washed twice with hexane. The precipitate [bisindenyl(CH3)2Si]Li2 was dried under vacuum overnight (yield 9.93 g, 94%). 1H NMR of the product in purified THF-d8 is given in Figure 2.

C. Reaction of the Transition Metal Salt with the Lithium Salt of the Ligand 7.00 g [bisindenyl(CH3)2Si]Li2 (19.4 mmol) was added to 4.30g ZrCl4 (18.4 mmol) in the dry box, and the two solids were mixed thoroughly at 25° C. 75 mL hexane was added to the mixture and the resulting suspension was stirred. 2 mL diethyl ether was then added dropwise with a syringe over a period of 2 minutes. The reaction mixture gradually turned orange, and was stirred for 20 hours. The resulting mixture was filtered and washed with hexane. The solid [Reaction Product C] was dried in vacuo for 15 hours (yield 8.92 g, 90.39%). lH NMR of the extrable portion of the reaction product in CD2Cl2 (which was only a very small portion of the entire reaction product) is given in Figure 3. (CH3)2Si-bisindenylZrCl2 produced in this reaction exists as both the racemic modification and meso form.

The 1H NMR is consistent with the presence of both the racemic modification and the meso form. The product obtained from this reaction exhibits lower solubility in polar solvents such as THF than previously described metallocenes. Attempts to purify the reaction product by extracting it in CH2Cl2 at reflux were unsuccessful, even at reflux conditions. The compound was insoluble in CH2Cl2. This would indicate that the catalyst was different than those previously described in the art.

D. Polymerization of Propylene 22 mmol methylalumoxane was combined with 50 mL isooctane. The resultant solution was injected into a 3.8 L autoclave equipped with a mechanical stirrer. The autoclave contained 650 g of liquid propylene and was maintained at a temperature in the range of 60"-90"C. After 15 minutes, 0.011 mmol based on Zr of the Reaction Product C which had been dissolved in 50 mL of isooctane was injected into the autoclave. The contents of the reactor were allowed to polymerize in the absence of hydrogen for 60 minutes. After the 60 minutes, the reaction was stopped by venting off unreacted monomer.

The 338 g polymer produced was highly isotactic, exhibited by an isotacticity index of 95.3% and xylene solubles of 1.8%, and had a narrow MWD, exhibited by a Mw/Mn value of 2.3.

EXAMPLE 2 A metallocene procatalyst was prepared according to the steps indicated in Example 1, except that 18.4 mmol of HfCl4 were used rather than an equivalent molar amount of ZrCl4. A light yellow solid comprising (CH3)2Si-bisindenylHfCl2was obtained from the series of reactions.

For the polymerization of the procatalyst thus produced, 0.0192 mmol (based on Hf) of the light yellow reaction product and 9.8 mmol methylalumoxane were charged to a reactor previously containing 1300 g of liquid propylene to produce 22 g polymer having a xylene solubles of 7.0 %. The polymer had a Mw/Mn of 12.9. However, this broader than expected MWD can be attributed to the presence of ZrCl4 at a level of 5% in the HfCl4 employed.

COMPARATIVE EXAMPLE 1 The preparation of the metallocene procatalyst (CH3)2Si- bisindenylZrCl2 was prepared according to the steps indicated in Example 1, except that in step C, 75 mL THF, a polar solvent, was used rather than 75 mL of hexane. When the ZrCl4 was combined with the [bisindenyl(CH3)2Si]Li2 in the solution of THF, the reaction mixture immediately turned dark. The reaction was exothermic. As known in the art, the dark color suggested that a decomposition reaction, probably involving the lithiated ligand, had occurred. As is generally known in the art that any solid recovered from such a reaction will not be useful as a catalyst component for the production of stereospecific polyolefins, no additional efforts were taken with regards to this Example.

COMPARATIVE EXAMPLE 2 The preparation of the metallocene procatalyst (CH3)2Si- bisindenylZrCl2 was prepared according to the steps indicated in Example 1, except that in step C, no accelerant was employed. Only the 75 mL hexane was added to the mixture of [bisindenyl(CH3)2Si]Li2 and ZrCl4. After 4 hours of stirring, no color change indicative of a desired reaction between the two materials was observed.

EXAMPLE 3 A metallocene procatalyst, (CH3)2Si-bisindenylHfCl2, was prepared according to the steps indicated in Example 1, except that no solvent was used in step C and HfCl4.2THF was used instead of ZrCl4, and where the transition metal salt was reacted with the product of the reaction between the ligand and the alkalialkyl.

0.403 g [bisindenyl(CH3)2Si]Li2 was added to 0.5 g HfCl4.2THF in a mortar, located in a dry box maintained between 25°- 30C C. The materials were mixed thoroughly with a pestle. Upon the commencing of mixing, the contents of the mortar began to turn yellow within five minutes.

The production of the yellow (CH3)2Si-bisindenylHfCl2 indicates that the metallocenes of the present invention can be obtained by a solvent-free process. However, as previously indicated, use of a solvent in a subsequent step is preferred for the removal of the unreacted intermediates and the production of uniformly-sized metallocene particles. Therefore, in this example, 30 mL of hexane was added to the yellow solid obtained above. The combination was stirred overnight. Afterwards, the yellow solid comprising (CH3)2Si- bisindenylHfCl2 and LiCl was isolated from the solution by filtration.

The resultant solid was washed with hexane and dried under vacuo (yield 0.37 g) and then used in a polymerization process as indicated in Example 1.