WO1998037103A1

WO1998037103A1 - Production of polyolefins of controlled particle size

Info

Publication number: WO1998037103A1
Application number: PCT/US1998/003329
Authority: WO
Inventors: Clark Curtis Williams; Timothy Roger Lynn; Mark Gregory Goode; Robert Converse Brady, Iii
Original assignee: Union Carbide Chemicals & Plastics Technology Corporation
Priority date: 1997-02-19
Filing date: 1998-02-18
Publication date: 1998-08-27
Also published as: ID19921A; ZA981303B; AU6178798A

Abstract

A method is provided for controlling the average particle size of a polyolefin produced by contacting under gas phase polymerization conditions at least one monomer with liquid droplets containing an unsupported catalyst, which comprises adjusting at least one of: a) the size of the liquid droplets, and b) the concentration of catalyst in the liquid droplets.

Description

PRODUCTION OF PO YO EFINS OF CONTROLLED PARTICLE SIZE

The present invention relates to controlling the average particle size of polyolefϊns produced using unsupported catalysts in gas phase polymerization.

BACKGROUND

Unsupported, liquid catalysts offer many advantages over conventional solid-supported catalysts. Unsupported catalysts require less equipment and raw materials to make them and impart fewer impurities to the final polymer product. The activity of an unsupported catalyst is not adversely influenced by the surface area of a support material. Additional advantages are encountered when a high-activity, unsupported metallocene catalyst is used for polymerizations in a fluidized bed reactor.

The use of unsupported catalysts is disclosed in U.S. Patent No. 5,317,036. This patent is directed to the use of unsupported, soluble, olefin polymerization catalysts, particularly metallocenes in liquid form, in gas phase reactions. These catalysts have a droplet size in the range of about 1 to about 1,000 micrometers.

When a droplet of unsupported catalyst solution is introduced into a gas phase reactor such as a fluidized bed reactor, several simultaneous or rapid processes can occur, including droplet heating, solvent evaporation, solute (catalyst and/or cocatalyst) precipitation, monomer and comonomer diffusion, and polymerization. During these processes, the rate of heat generation due to polymerization can increase the particle temperature to above that of the average fluidized bed temperature, resulting in particle melting or softening to the point of excessive agglomerating or fusing of particles. This can often occur shortly after the droplet has been introduced into the reactor because of the rapid heat generation at the onset of polymerization. Polymer particles can grow so large that they cannot be fluidized thereby causing the reactor to be shut down.

The tendency for polyolefin particles to overheat and grow to an unacceptable size can be reduced in several ways, including decreasing the catalyst activity by reducing the temperature of the reactor, or decreasing the partial pressure of monomer in the reactor. However, effecting either of these changes is not a practical or feasible solution for controlling particle size, since each is accompanied by a decrease in production rate. High productivity is required in olefin polymerization so that a minimum amount of catalyst, which is often expensive, can be used.

There is a need, accordingly, for a method of controlling the average particle size of a polyolefin produced using an unsupported catalyst during gas phase polymerization that does not affect the production rate or otherwise upset the stability of the production process.

Copending U.S. Application Serial No. 08/659,764 filed June 6, 1996, relates to one method of controlling the growth of polyolefin polymer particles made using an unsupported catalyst during gas phase polymerization. The unsupported catalyst is introduced into the reactor into a particle lean zone substantially free of resin. This allows a brief period of time for droplets containing the unsupported catalyst to undergo evaporation before contacting the polymer particles already in the reactor, reducing the tendency for excessive agglomeration of polymer particles.

. It has now been discovered that, advantageously, the average particle size of a polyolefin produced using an unsupported catalyst in gas phase polymerization is also well controlled by adjusting at least one of: a) the size of the liquid droplets containing the catalyst introduced into the polymerization reactor, and b) the concentration of catalyst composition in the liquid droplets. This technique provides a relatively simple method, readily carried out using manual or computer operated process control, for controlling particle size in the reactor. Advantageously, neither reactor stability nor production rates are adversely effected.

The relationship among the liquid droplet size, the catalyst concentration in the liquid droplets, and the average particle size of the polyolefin is often not described by a geometric equation resulting from a simplistic mechanism in which a single liquid droplet polymerizes into a single polyolefin particle. More often, complex processes relating to the tendency of individual liquid droplets to agglomerate with one another and with polyolefin particles already in the reactor take place. The fact that despite these complex processes polyolefin average particle size can be controlled by adjusting liquid droplet size and catalyst concentration in the liquid droplets underscores the unexpected nature of the invention.

SUMMARY OF THE INVENTION

The invention provides a process for producing a polyolefin in a gas phase polymerization reactor, which comprises: (i) introducing monomer into the polymerization reactor;

(ii) introducing liquid droplets containing an unsupported catalyst into the polymerization reactor;

(iii) withdrawing polyolefin product in the form of particles from the polymerization reactor; and

(iv) withdrawing unreacted monomer from the polymerization reactor; wherein the average particle size of the polyolefin product is controlled by adjusting at least one of: a) the size of the liquid droplets, and b) the concentration of catalyst in the liquid droplets.

DETAILED DESCRIPTION OF THE INVENTION

Polymers. Illustrative of the polymers which can be produced in accordance with the invention are the following: ethylene homopolymers and ethylene copolymers employing one or more C3-C12 alpha olefins; propylene homopolymers and propylene copolymers employing one or more C4-C12 alpha olefins; polyisoprene; polystyrene; polybutadiene; polymers of butadiene copolymerized with styrene; polymers of butadiene copolymerized with acrylonitrile; polymers of isobutylene copolymerized with isoprene; ethylene propylene rubbers and ethylene propylene diene rubbers; polychloroprene, and the like.

Polymerization. The present invention is not limited to any specific type of gas phase polymerization reaction and can be carried out in a stirred or fluidized bed reactor. The invention can be carried out in a single reactor or multiple reactors (two or more reactors in series). In addition to well known conventional gas phase polymerizations processes, "condensed mode," including the so-called "induced condensed mode," and "liquid monomer" operation of a gas phase polymerization can be employed.

A conventional fluidized bed process for producing resins is practiced by passing a gaseous stream containing one or more monomers continuously through a fluidized bed reactor under reactive conditions in the presence of a polymerization catalyst. Product is withdrawn from the reactor. A gaseous stream of unreacted monomer is withdrawn from the reactor continuously and recycled into the reactor along with make-up monomer added to the recycle stream.

Condensed mode polymerizations are disclosed in U.S. Patent Nos. 4,543,399; 4,588,790; 5,352,749; and 5,462,999. Condensing mode processes are employed to achieve higher cooling capacities and, hence, higher reactor productivity. In these polymerizations a recycle stream, or a portion thereof, can be cooled to a temperature below the dew point in a fluidized bed polymerization process, resulting in condensing all or a portion of the recycle stream. The recycle stream is returned to the reactor. The dew point of the recycle stream can be increased by increasing the operating pressure of the reaction/recycle system and/or increasing the percentage of condensable fluids and decreasing the percentage of non-condensable gases in the recycle stream. The condensable fluid may be inert to the catalyst, reactants and the polymer product produced; it may also include monomers and comonomers. The condensable fluid can be introduced into the reaction/recycle system at any point in the system. In addition, condensable fluids of the polymerization process itself may be used or other condensable fluids inert to the polymerization can be introduced to "induce" condensing mode operation. Examples of suitable condensable fluids may be selected from liquid saturated hydrocarbons containing 2 to 8 carbon atoms (e.g., propane, n-butane, isobutane, n- pentane, isopentane, neopentane, n-hexane, isohexane, and other saturated C6 hydrocarbons, n-heptane, n-octane and other saturated Cη and Cg hydrocarbons, and mixtures thereof). Unsaturated hydrocarbons may also be used. Condensable fluids may also include polymerizable condensable comonomers such as olefins, alpha-olefins, diolefins, diolefins containing at least one alpha olefin, and mixtures thereof. In condensing mode, it desirable that the liquid entering the fluidized bed be dispersed and vaporized quickly.

Liquid monomer polymerization mode is disclosed in U.S. Patent No. 5,453,471, U.S. Serial No. 510,375, PCT WO95/09826 and PCT WO95/09827. When operating in the liquid monomer mode, liquid can be present throughout the entire polymer bed provided that the liquid monomer present in the bed is adsorbed on or absorbed in solid particulate matter present in the bed, such as polymer being produced or fluidization aids (e.g., carbon black) present in the bed, so long as there is no substantial amount of free liquid monomer present more than a short distance above the point of entry into the polymerization zone. Liquid mode makes it possible to produce polymers in a gas phase reactor using monomers having condensation temperatures much higher than the temperatures at which conventional polyolefins are produced. In general, liquid monomer process are conducted in a stirred bed or gas fluidized bed reaction vessel having a polymerization zone containing a bed of growing polymer particles. The process comprises continuously introducing a stream of one or more monomers and optionally one or more inert gases or liquids into the polymerization zone; continuously or intermittently introducing a polymerization catalyst into the polymerization zone; continuously or intermittently withdrawing polymer product from the polymerization zone; and continuously withdrawing unreacted gases from the zone; compressing and cooling the gases while maintaining the temperature within the zone below the dew point of at least one monomer present in the zone. If there is only one monomer present in the gas-liquid stream, there is also present at least one inert gas. Typically, the temperature within the zone and the velocity of gases passing through the zone are such that essentially no liquid is present in the polymerization zone that is not adsorbed on or absorbed in solid particulate matter.

Monomers. Monomers that can be employed in the process include one or more C2 to C12 alpha-olefins; dienes such as those taught in U.S. Patent No. 5,317,036 to Brady et al. such as hexadiene, dicyclopentadiene, norbornadiene, and ethylidene norbornene; readily condensable monomers such as those taught in U.S. Patent No. 5,453,471 including isoprene, styrene, butadiene, isobutylene, and chloroprene, acrylonitrile, and the like.

Inert Particulate Materials. The invention can optionally employ inert particulate materials as fluidization aids. These inert particulate materials can include carbon black, silica, talc, and clays, as well as inert polymeric materials. Carbon black has a primary particle size of about 10 to about 100 nanometers, an average size of aggregate of about 0.1 to about 10 micrometers, and a specific surface area of about 30 to about 1,500 m^/gm. Silica has a primary particle size of about 5 to about 50 nanometers, an average size of aggregate of about 0.1 to about 10 micrometers, and a specific surface area of about 50 to 500 m^/gm. Clay, talc, and polymeric materials have an average particle size of about 0.01 to about 10 micrometers and a specific surface area of about 3 to 30 m^/gm. These inert particulate materials are employed in amounts ranging about 0.3 to about 80%, preferably about 5 to about 50%, based on the weight of the final product. They are especially useful for the polymerization of sticky polymers as disclosed in U.S. Patent Nos. 4,994,534 and 5,304,588.

Chain Transfer Agents and Other Additives. Chain transfer agents, promoters, scavenging agents and other additives can be, and often are, employed in the polymerization process of the invention. Chain transfer agents are often used to control polymer molecular weight. Examples of these compounds are hydrogen and metal alkyls of the general formula M^R^g, where M^ is a Group IA, HA or IIIA metal, R^ is an alkyl or aryl, and g is 1, 2, or 3. Preferably, a zinc alkyl is employed; and, of these, diethyl zinc is most preferred. Typical promoters include halogenated hydrocarbons such as CHCI3, CFCI3, CH3CCI3, CF2CICCI3, and ethyl trichloroacetate. Such promoters are well known to those skilled in the art and are disclosed in, for example, U.S. Patent No. 4,988,783. Other organometallic compounds such as scavenging agents for poisons may also be employed to increase catalyst activity. Examples of these compounds include metal alkyls, such as aluminum alkyls, most preferably triisobutylaluminum. Some compounds may be used to neutralize static in the fluidized-bed reactor, others known as drivers rather than antistatic agents, may consistently force the static to from positive to negative or from negative to positive. The use of these additives is well within the skill of those skilled in the art. These additives may be added to the reaction zone separately or independently from the liquid catalyst if they are solids, or as part of the catalyst provided they do not interfere with the desired atomization. To be part of the catalyst solution, the additives should be liquids or capable of being dissolved in the catalyst solution.

Catalyst. Any type of polymerization catalyst may be used in the present process, provided it is stable and sprayable or atomizable when in liquid form. A single unsupported catalyst may be used, or a mixture of unsupported catalysts may be employed if desired. The catalysts are used with appropriate cocatalysts and promoters well known in the art, which cocatalysts and promoters may be introduced into the polymerization zone either together with the catalyst or separately. Examples of suitable catalysts include the following.

A. Ziegler-Natta catalysts, including titanium based catalysts such as those described in U.S. Patent Nos. 4,376,062 and 4,379,758. Ziegler-Natta catalysts are well known in the art, and typically are magnesium/titanium/electron donor complexes used in conjunction with an organoaluminum cocatalyst.

B. Chromium based catalysts such as those described in U.S. Patent Nos. 3,709,853; 3,709,954; and 4,077,904.

C. Vanadium based catalysts such as vanadium oxychloride and vanadium acetylacetonate, such as described in U.S. Patent No. 5,317,036.

D. Single site catalysts such as metallocene catalysts.

E. Cationic forms of metal halides, such as aluminum trihalides.

F. Cobalt catalysts and mixtures thereof such as those described in U.S. Patent Nos. 4,472,559 and 4,182,814. G. Nickel catalysts and mixtures thereof such as those described in U.S. Patent Nos. 4,155,880 and 4,102,817.

_.H. Rare Earth metal catalysts, i.e., those containing a metal having an atomic number in the Periodic Table of 57 to 103, such as compounds of cerium, lanthanum, praseodymium, gadolinium and neodymium. Especially useful are carboxylates, alcoholates, acetylacetonates, halides (including ether and alcohol complexes of neodymium trichloride), and allyl derivatives of such metals. Neodymium compounds, particularly neodymium neodecanoate, octanoate, and versatate, are the most preferred rare earth metal catalysts. Rare earth metal catalysts are used to produce polymers of butadiene or isoprene.

Preferred are single site catalysts such as metallocene catalysts in liquid form with an activating cocatalyst. The practice of this invention is not limited to any particular class or kind of metallocene catalyst, and any unsupported metallocene catalyst useful in slurry, solution, bulk, or gas phase olefin polymerization may be used. One or more than one metallocene catalyst may be employed. For example, as described in U.S. Patent No. 4,530,914, at least two metallocene catalysts may be used in a single catalyst composition to achieve a broadened molecular weight distribution polymer product.

Metallocene catalysts are organometallic coordination complexes of one or more π-bonded moieties in association with a metal atom from Groups IIIB to VIII or the rare earth metals of the Periodic Table.

Bridged and unbridged mono-, bis-, and tris- cycloalkadienyl/metal compounds are the most common metallocene catalysts, and generally are of the formula: (L)_yRl_z(L')MX(_x__y_i) (I)

wherein M is a metal from groups IIIB to VIII of the Periodic Table; L and L' are the same or different and are π-bonded ligands coordinated to M, preferably cycloalkadienyl groups such as cyclopentadienyl, indenyl, or fluorenyl groups optionally substituted with one or more hydrocarbyl groups containing 1 to 20 carbon atoms; Rl is a C1-C4 substituted or unsubstituted alkylene radical, a dialkyl or diaryl germanium or silicon, or an alkyl or aryl phosphine or amine radical 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 R^ 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 metallocene catalysts represented by formula I are dialkyl metallocenes such as bis(cyclopentadienyl)titanium dimethyl, bis(cyclopentadienyl)titanium diphenyl, bis(cyclopentadienyl)zirconium dimethyl, bis(cyclopenta- dienyl)zirconium diphenyl, bis(cyclopentadienyl)hafhium methyl and diphenyl, bis(cyclopentadienyl)titanium di-neopentyl, bis(cyclopentadienyl)zirconium di-neopentyl, bis(cyclopentadienyl)titanium dibenzyl, bis(cyclopentadienyl)zirconium dibenzyl, bis(cyclopentadienyl)vanadium dimethyl; mono alkyl metallocenes such as bis(cyclopentadienyl)titanium methyl chloride, bis(cyclopentadienyl)titanium ethyl chloride, bis(cyclopentadienyl)titanium phenyl chloride, bis(cyclopentadienyl)zirconium methyl chloride, bis(cyclopentadienyl)zirconium ethyl chloride, bis(cyclopentadienyl)zirconium phenyl chloride, bis(cyclopentadienyl)titanium methyl bromide; trialkyl metallocenes such as cyclopentadienyl titanium trimethyl, cyclopentadienyl zirconium tri phenyl, and cyclopentadienyl zirconium trineopentyl, cyclopentadienyl zirconium trimethyl, cyclopentadienyl hafnium triphenyl, cyclopentadienyl hafnium trineopentyl, and cyclopentadienyl hafnium trimethyl; monocyclopentadienyl titanocenes such as, pentamethylcyclopentadienyl titanium trichloride, pentaethylcyclopentadienyl titanium trichloride; bis(pentamethylcyclopentadienyl) titanium diphenyl, the carbene represented by the formula bis(cyclopentadienyl)titanium=CH2 and derivatives of this reagent; substituted bis(cyclopentadienyl)titanium (IV) compounds such as: bis(indenyl)titanium diphenyl or dichloride, bis(methylcyclopentadienyl)titanium diphenyl or dihalide; dialkyl, trialkyl, tetraalkyl and pentaalkyl cyclopentadienyl titanium compounds such as bis(l,2-dimethylcyclopentadienyl)titanium diphenyl or dichloride, bis(l,2-diethylcyclopentadienyl)titanium diphenyl or dichloride; silicon, phosphine, amine or carbon bridged cyclopentadiene complexes, such as dimethyl silyldicyclopentadienyl titanium diphenyl or dichloride, methyl phosphine dicyclopentadienyl titanium diphenyl or dichloride, methylenedicyclopentadienyl titanium diphenyl or dichloride and other dihalide complexes, and the like; as well as bridged metallocene compounds such as isopropyl(cyclopentadienyl)(fluorenyl)zirconium dichloride, isopropyl(cyclopentadienyl)(octahydrofluorenyl)zirconium dichloride, diphenylmethylene(cyclopentadienyl)(fluorenyl)zirconium dichloride, diisopropylmethylene (cyclopentadienyl)(fluorenyl)-zirconium dichloride, diisobutylmethylene(cyclopentadienyl)(fluorenyl) zirconium dichloride, ditertbutylmethylene (cyclopentadienyl)- (fluorenyl)zirconium dichloride, cyclohexylidene(cyclopentadienyl)- (fluorenyl)zirconium dichloride, diisopropylmethylene (2,5- dimethylcyclopentadienyl)(fluorenyl)zirconium dichloride, isopropyl(cyclopentadienyl)(fluorenyl)hafnium dichloride, diphenylmethylene (cyclopentadienyD(fluorenyl)hafhium dichloride, diisopropylmethylene(cyclopentadienyl)(fluorenyl)hafnium dichloride, diisobutylmethylene(cyclopentadienyl)(fluorenyl)hafnium dichloride, ditertbutylmethylene(cyclopentadienyl)(fluorenyl)hafnium dichloride, cyclohexylidene(cyclopentadienyl)(fluorenyl)hafhium dichloride, diisopropylmethylene(2,5-dimethyl cyclopentadienyl) (fluorenyl)- hafnium dichloride, isopropyl(cyclopentadienyl)(fluorenyl)titanium dichloride, diphenylmethylene(cyclopentadienyl)(fluorenyl)titanium dichloride, diisopropylmethylene(cyclopentadienyl)(fluorenyl)titanium dichloride, diisobutylmethylene(cyclopentadienyl) (fluorenyl)titanium dichloride, ditertbutylmethylene(cyclopentadienyl)(fluorenyl)titanium dichloride, cyclohexylidene(cyclopentadienyl)(fluorenyl)titanium dichloride, diisopropylmethylene(2,5 dimethylcyclopentadienyl fluorenyDtitanium dichloride, racemic-ethylene bis (1-indenyl) zirconium (IV) dichloride, racemic-ethylene bis (4,5,6,7-tetrahydro-l- indenyl) zirconium (IV) dichloride, racemic-dimethylsilyl bis (1- indenyl) zirconium (IV) dichloride, racemic-dimethylsilyl bis (4,5,6,7- tetrahydro-1-indenyl) zirconium (IV) dichloride, racemic-1, 1,2, 2- tetramethylsilanylene bis (1-indenyl) zirconium (IV) dichloride, racemic-1, 1, 2, 2-tetramethylsilanylene bis (4,5,6,7-tetrahydro-l- indenyl) zirconium (IV) dichloride, ethylidene (1-indenyl tetramethyl cyclopentadienyl) zirconium (IV) dichloride, racemic- dimethylsilyl bis (2-methyl-4-t-butyl-l-cyclopentadienyl) zirconium (IV) dichloride, racemic-ethylene bis (1-indenyl) hafnium (IV) dichloride, racemic-ethylene bis (4,5,6,7-tetrahydro-l-indenyl) hafnium (IV) dichloride, racemic-dimethylsilyl bis (1-indenyl) hafnium (IV) dichloride, racemic-dimethylsilyl bis (4,5,6,7-tetrahydro-l- indenyl) hafnium (IV) dichloride, racemic-1, 1,2,2- tetramethylsilanylene bis (1- indenyl) hafnium (IV) dichloride, racemic-1, 1,2,2- tetramethylsilanylene bis (4,5,6,7-tetrahydro-l- indenyl) hafnium (IV), dichloride, ethylidene (l-indenyl-2,3,4,5- tetramethyl- 1- cyclopentadienyl) hafnium (IV) dichloride, racemic- ethylene bis (1- indenyl) titanium (IV) dichloride, racemic-ethylene bis (4,5,6,7- tetrahydro-1-indenyl) titanium (IV) dichloride, racemic- dimethylsilyl bis (1-indenyl) titanium (IV) dichloride, racemic- dimethylsilyl bis (4,5,6,7-tetrahydro-l-indenyl) titanium (IV) dichloride, racemic-1, 1,2,2- tetramethylsilanylene bis (1-indenyl) titanium (IV) dichloride racemic- 1,1,2,2-tetramethylsilanylene bis (4,5,6,7-tetrahydro-l-indenyl) titanium (IV) dichloride, and ethylidene ( l-indenyl-2, 3,4,5- tetramethyl-1-cyclopentadienyl) titanium IV) dichloride.

Particularly preferred metallocene catalysts have one of the following formulas (II or III):

(ID

or

wherein:

M is a metal from groups IIIB to VIII, preferably Zr or Hf;

L is a substituted or unsubstituted, π-bonded ligand coordinated to M, preferably a substituted cycloalkadienyl ligand; each Q is independently selected from the group consisting of -O-

, -NR³-, -CR 2- and -S-, preferably oxygen;

Y is either C or S, preferably carbon;

Z is selected from the group consisting of -OR³, -NR³2, -CR³3, - SR³, -SiR³3, -PR³2, and -H, with the proviso that when Q is -NR³- then Z is selected from the group consisting of -OR³, -NR³2, -SR³, - SiR³3, -PR³2, and -H, preferably Z is selected from the group consisting of -OR³, -CR³3, and -NR³2; n is 1 or 2;

A is a univalent anionic group when n is 2 or A is a divalent anionic group when n is 1, preferably A is a carbamate, carboxylate or other heteroallyl moiety described by Q, Y and Z combination; and each R³ is independently a group containing carbon, silicon, nitrogen, oxygen, and/or phosphorus and one or more R³ groups may be attached to the L substituent, preferably R³ is a hydrocarbon group containing from 1 to 20 carbon atoms, most preferably an alkyl, cycloalkyl or an aryl group;

T is a bridging group selected from the group consisting of alkyl ene or arylene groups containing from 1 to 10 carbon atoms optionally substituted with carbon or heteroatoms, germanium, silicone and alkyl phosphine; and m is 2 to 7, preferably 2 to 6, most preferably 2 or 3.

The supportive substituent formed by Q, Y and Z is a unicharged polydentate ligand exerting electronic effects due to its high polarizability, similar to the cyclopentadienyl group. In the most preferred embodiments of this invention, the disubstituted carbamates,

\

N- C f( " M

O'^'

and the carboxylates — C - C T ( ^{^}: M

I ^ O'-^' are employed.

Examples of complexes according to formulas II and III include indenyl zirconium tris(diethylcarbamate), indenyl zirconium tris(trimethylacetate), indenyl zirconium tris(p-toluate), indenyl zirconium tris(benzoate), (l-methylindenyl)zirconium tris(trimethylacetate), (2-methylindenyl) zirconium tris(diethylcarbamate), (methylcyclopentadienyl) zirconium tris(trimethylacetate), cyclopentadienyl tris(trimethylacetate), tetrahydroindenyl zirconium tris(trimethylacetate), and (pentamethyl cyclopentadienyl) zirconium tris(benzoate). Preferred examples are indenyl zirconium tris(diethylcarbamate), indenyl zirconium tris(trimethylacetate), and (methylcyclopentadienyl) zirconium tris(trimethylacetate).

Another type of metallocene catalyst that can be used in accordance with the invention is a constrained geometry catalyst of the formula:

Cp Y _

(X')a (IV)

wherein: M is a metal of Group IIIB to VIII of the Periodic Table of the Elements:

• Cp is a cyclopentadienyl or substituted cyclopentadienyl group bound in an r bonded mode to M;

Z' is a moiety comprising boron, or a member of Group PVB of the Periodic Table of the Elements and optionally sulfur or oxygen, the moiety having up to 20 non-hydrogen atoms, and optionally Cp and Z¹ together form a fused ring system;

X' is an anionic ligand group or a neutral Lewis base ligand group having up to 30 non-hydrogen atoms; a is 0, 1, 2, 3 or 4 depending on the valance of M; and

Y is an anionic or non-anionic ligand group bonded to Z' and M comprising is nitrogen, phosphorus, oxygen or sulfur having up to 20 non-hydrogen atoms, and optionally Y and Z' together form a fused ring system.

Constrained geometry catalysts are well known to those skilled in the art and are disclosed in, for example, U.S. Patent Nos. 5,026,798 and 5,055,438 and published European Application No. 0 416 815 A2.

Illustrative but non-limiting examples of substituents Z', Cp, Y', X' and M in formula IV are:

Cp X' M dimethyl-silyl cyclopenta-dienyl t-butylamido chloride titanium methyl- fluorenyl phenylamido methyl zirconium phenylsilyl diphenyl-silyl indenyl cyclohexylamido hafnium tetramethyl- oxo ethylene ethylene tetramethyl- cyclopenta-dienyl diphenyl- methylene

The invention is also useful with another class of single site catalyst precursors, di(imine) metal complexes, as described in PCT Application No. WO 96/23010. Such di(imine) metal complexes are transition metal complexes of bidentate ligands selected from the group consisting of:

wherein said transition metal is selected from the group consisting of Ti, Zr, Sc, V, Cr, a rare earth metal, Fe, Co, Νi, and Pd; R² and R⁵ are each independently hydrocarbyl or substituted hydrocarbyl, provided that the carbon atom bound to the imino nitrogen atom has at least two carbon atoms bound to it;

R³ and R⁴ are each independently hydrogen, hydrocarbyl, substituted hydrocarbyl, or R³ and R⁴ taken together are hydrocarbylene or substituted hydrocarbylene to form a carbocyclic ring;

R⁴⁴ is hydrocarbyl or substituted hydrocarbyl, and R²⁸ is hydrogen, hydrocarbyl or substituted hydrocarbyl or R⁴⁴ and R²⁸ taken together form a ring;

R⁴⁵ is hydrocarbyl or substituted hydrocarbyl, and R²⁹ is hydrogen, substituted hydrocarbyl or hydrocarbyl, or R⁴⁵ and R²⁹ taken together form a ring; each R³⁰ is independently hydrogen, substituted hydrocarbyl or hydrocarbyl, or two of R³⁰ taken together form a ring; each R³¹ is independently hydrogen, hydrocarbyl or substituted hydrocarbyl;

R⁴⁶ and R⁴⁷ are each independently hydrocarbyl or substituted hydrocarbyl, provided that the carbon atom bound to the imino nitrogen atom has at least two carbon atoms bound to it;

R⁴⁸ and R⁴⁹ are each independently hydrogen, hydrocarbyl, or substituted hydrocarbyl;

R²⁰ and R²³ are independently hydrocarbyl or substituted hydrocarbyl;

R²¹ and R²² are independently hydrogen, hydrocarbyl or substituted hydrocarbyl; and n is 2 or 3; and provided that: said transition metal also has bonded to it a ligand that may be displaced by or added to the olefin monomer being polymerized; and when the transition metal is Pd, said bidentate ligand is (V), (VII) or (VIII).

Activating cocatalysts suitable for use with metallocene catalysts include 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 BR 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 R*** is an alkyl group containing 1 to 12 carbon atoms, preferably methyl or an aryl radical such as a substituted or unsubstituted phenyl or naphthyl radical. In the case of MAO, R*** is methyl, whereas in MMAO, R*** is a mixture of methyl and C2 to C12 alkyl groups wherein methyl comprises about 20 to 80 percent by weight of the R*** group.

The mole ratio of activating cocatalyst to metallocene catalyst usefully employed 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 unsupported catalyst is introduced into the polymerization reactor in the form of liquid droplets. The liquid droplets comprise a solution, dispersion, or emulsion of one or more catalysts, optionally one or more cocatalysts, and optionally promoters or other catalyst additives in one or more liquid solvents. Preferably, the liquid droplets comprise a solution or emulsion of a catalyst and a cocatalyst. More preferably, the liquid droplets comprise a solution of a catalyst and a cocatalyst.

In practice, the catalyst is typically dissolved or prepared in a solvent. The cocatalyst is also prepared and handled in a solvent, which may be the same or different than that used for the catalyst. These two solutions can be combined after being fed into the reactor, or preferably before they are fed into the reactor. The feed rate of the two solutions can be varied to adjust the mole ratio of cocatalyst to catalyst in the mixture, in order to change resin properties or to minimize the total cost of cocatalyst and catalyst.

Solvents that can be utilized to form solutions of the unsupported catalyst and/or cocatalyst are inert solvents, preferably non-functional hydrocarbon solvents, and may include aliphatic hydrocarbons such as butane, isobutane, ethane, propane, pentane, isopentane, hexane, heptane, octane, decane, dodecane, hexadecane, octadecane, and the like; alicyclic hydrocarbons such as cyclopentane, methylcyclopentane, cyclohexane, cyclooctane, norbornane, ethylcyclohexane and the like; aromatic hydrocarbons such as benzene, toluene, ethylbenzene, propylbenzene, butylbenzene, xylene, tetrahydrofuran and the like; and petroleum fractions such as gasoline, kerosene, light oils, and the like. Likewise, halogenated hydrocarbons such as methylene chloride, chlorobenzene, and the like may also be utilized. By "inert" is meant that the material being referred to is non- deactivating in the polymerization zone under the conditions of gas phase, polymerization and does not deactivate the catalyst in or out of the polymerization zone. By "non-functional" is meant that the solvents do not contain groups such as strong polar groups which can deactivate the active catalyst metal sites.

Control of Average Particle Size. Any atomization device can be used to introduce the liquid droplets into the reactor. For atomization devices that produce a controllable range of liquid droplet size, the liquid droplet size and catalyst concentration in the liquid droplets can be varied independently. For nozzles producing liquid droplets of fixed size, the flow rate of solvent can be adjusted so that only the catalyst concentration and number of liquid droplets are changed. Suitable nozzles include air- assisted or air-blast atomizers, pressure atomizers, rotary atomizers, ultrasonic, effervescent, and electrostatic nozzles. Particularly preferred nozzles are those described in copending U.S.

Application Serial No. entitled "Improved Control of

Solution Catalyst Droplet Size with an Effervescent Spray Nozzle" of Williams et al. (docket no. 17733) and U.S. Application Serial No. entitled "Improved Control of Solution Catalyst Droplet

Size with a Perpendicular Spray Nozzle" of Williams et al. (docket no. 17755), both filed concurrently herewith.

Nozzles can produce either a narrow or wide liquid droplet size distribution. A narrow distribution is preferred, but a wide distribution can be tolerated. Nozzles that produce larger liquid droplets may require higher amounts of liquid to provide the desired amount of catalyst per liquid droplet. According to the invention, the average particle size of the polyolefin produced in the presence of the unsupported catalyst is controlled by adjusting the size of the liquid droplets containing catalyst, or the concentration of catalyst in the liquid droplets, or both. If both the size of the liquid droplets and the catalyst concentration in the liquid droplets are adjusted, they may be adjusted simultaneously or in sequence.

The nature of both the catalyst and the activating cocatalyst determine the magnitude and direction in which the size of the liquid droplets and the catalyst concentration in the liquid droplets should be adjusted in order to achieve a given average polyolefin product particle size. Typically, for a catalyst system comprising liquid droplets of unsupported metallocene catalyst and liquid aluminoxane cocatalyst in a solvent (or solvent mixture) having a given density, used to produce an ethylene copolymer without severe agglomeration, the average particle size of the ethylene copolymer may be increased or decreased by about 10 % by adjusting the size of the liquid droplets by about 10 % or adjusting the catalyst concentration in the liquid droplets (i.e., in the total liquid feedstream of unsupported catalyst, cocatalyst, solvent(s), etc.) by about 33 %. Preferably, the average particle size of an ethylene copolymer so made may be increased or decreased by about 20% by adjusting the size of the liquid droplets by about 20 % or adjusting the catalyst concentration by about 40 %. For conditions where an increase in liquid droplet size leads to an increased rate of particle agglomeration, a 10 % increase in liquid droplet size can lead to a 50% or more increase in ethylene copolymer average particle size. Under such conditions, diluting the catalyst in the liquid feedstream by 33% can decrease the ethylene copolymer average particle size by 50% or more.

The average diameter of the liquid droplets is generally in the range of about 0.1 to about 1000 micrometers, preferably 1 to 300 micrometers, most preferably about 10 to 75 micrometers.

The size, i.e., average diameter, of the liquid droplets may be adjusted in one of several ways. For example, the flow rate of the liquid feedstream of unsupported catalyst, cocatalyst, solvent(s), etc. may be increased in order to increase the size of the liquid droplets, or decreased to decrease the size of the liquid droplets.

Alternatively, when the liquid droplets of unsupported catalyst are introduced into the reactor with the aid of an inert carrier gas such as nitrogen, argon, alkane, or mixtures thereof, the flow rate of the inert carrier gas into the polymerization reactor may be increased to break up the liquid into smaller sized droplets, which in turn decreases the average particle size of the polyolefin produced. Alternatively, the flow rate of the inert carrier gas may be decreased, allowing the size of the liquid droplets to increase, thereby increasing the average particle size of the polyolefin produced. This is a preferred method of adjusting the liquid droplet size, and thereby polyolefin average particle size.

In one embodiment of the invention, the size of the liquid droplets containing the catalyst is adjusted while using an effervescent spray nozzle, such as that described in copending U.S. Application

Serial No. for "Improved Control of Solution Catalyst

Droplet Size with an Effervescent Spray Nozzle" of Williams et al. (docket no. 17733) to spray the liquid feedstream containing the unsupported catalyst into the polymerization reactor. In such an effervescent nozzle, a stream of liquid or gas is passed through an inner tube, while a liquid or gas is passed cocurrently through an annular space defined by the inner tube and a concentric outer tube. The direction of flow of the liquid and gas is generally along the central axis of the tubes. The liquid feedstream containing the unsupported catalyst and atomization gas are fed through their respective inlets and exit through a common orifice at the spray tip. Towards the tip of the inner tube, though not necessarily at the end, there are holes (orifices) which allow the gas to enter the liquid. The gas is introduced into the cocurrent flowing liquid near the common exit orifice. In this way, liquid slugging is prevented and steady droplet formation occurs. Gas bubbles which are formed are forced through an orifice at the tip of the outer tube, forcing the concurrent flow of liquid along the outside edge of the orifice. The thin film of liquid on the orifice wall is ejected from the orifice in thin sheets which disintegrate into small droplets. The gas bubbles are thought to rapidly increase in volume as they emerge form the orifice, providing additional energy which shatters the liquid into small droplets. Using a mathematical model, the size of the liquid droplets containing the unsupported catalyst sprayed from the effervescent nozzle can be readily calculated and adjusted as desired.

In another embodiment of the invention, the size of the liquid droplets containing the catalyst is adjusted while using a perpendicular spray nozzle such as that described in copending U.S.

Application Serial No. entitled "Improved Control of

Solution Catalyst Droplet Size with a Perpendicular Spray Nozzle" of Williams et al. (docket no. 17755), to spray the liquid containing the unsupported catalyst into the polymerization reactor. Such a perpendicular nozzle comprises a tube for delivering the liquid feedstream containing the unsupported catalyst wherein there is an inlet end for the input of the liquid, and optionally, a gas. The other end of the tube (i.e., "distal end") wherein there is at least one exit hole (orifice) which is at least 10-20°, preferably more than 45°, and most preferably 60 to 90°, off from the direction of flow of the liquid within the nozzle (i.e., from the central axis of the tube), where the orifice is located towards the distal end of the nozzle. Said nozzle may have any number of orifices and may include a gas stream within the liquid feedstream. There is no need for a separate mixing chamber for the gas and liquid within the nozzle.

The distal end of the nozzle may be of any geometric configuration, e.g., bulbous, rounded, parabolic, conical, or semicircular, but to limit turbulence the nozzle preferably is tapered at about 5 to 15 degrees off horizontal (the central axis of the tube). Higher taper angles can be tolerated given that the taper from horizontal is gradual. A tapered tip also minimizes fouling because of the small area available for accumulation of catalyst and polymer.

For perpendicular spraying, the liquid feedstream may be atomized with an inert carrier gas, as is done with a gas-assisted perpendicular spray nozzle. Alternately, a perpendicular pressure nozzle could be used to deliver a perpendicular spray of high-pressure liquid in the absence of an atomizing gas. Additionally, the perpendicular feeding geometry can be used with effervescent gas- liquid contact in the spraying nozzle or with an ultrasonic nozzle, or could also be applied to other known atomization devices, such as electrostatic, sonic-whistle, or rotary, etc. nozzles.

The concentration of catalyst in the liquid droplets is in the range of about 0.01 to about 10,000 millimoles/liter, preferably 0.1 to about 1000 millimoles/liter, more preferably 1 to about 100 millimoles/liter.

. The concentration of catalyst in the liquid droplets can be adjusted within this range by changing the amount of solvent fed with the catalyst to the polymerization zone. It should be borne in mind, however, that if too much solvent is added, the reactor dew point is increased and the catalyst activity per volume of droplet is low. These factors slow the evaporation rate from the droplets so that the overall average polymer particle size may be too large. With a high activity catalyst, a low dew point solvent, and a concentrated catalyst solution, the polymer particles can overheat and fuse together. Depending on the type of agglomeration occurring, a change in the amount of solvent fed with the catalyst can shift the average polyolefin particle size into a desirable range. The opposite can also occur, that is, changes in catalyst dilution can improve a very small average particle size to an acceptable range.

During commercial operation, a gas phase polymerization reactor may be cycled through dozens of products, requiring different reactor process conditions and different catalysts. The concentration of catalyst and cocatalyst as mixed may not be suitable for some of the required products or reactor conditions. In this case, the mixture of catalyst and cocatalyst can be diluted further with one or more additional solvents, referred to as diluents. These diluents may be the same as the solvent(s) originally used for the catalyst and cocatalyst, or may be completely different. The actual concentration of catalyst and cocatalyst in the liquid droplets can then be varied over a wide range of conditions, obviating the need for catalyst and cocatalyst solutions of varying concentrations. In practice, the feedrates of catalyst and cocatalyst are adjusted to maintain the target production rates, the optimum catalyst/cocatalyst economics, and the desired resin properties. The amount of solvent and diluent added to the catalyst and cocatalyst can be changed without affecting the production rate. This allows for droplet size control while maintaining a stable process. The use of two or more solvents or diluents having different evaporation rates has the additional advantage of allowing one to control better the rate of solvent evaporation from the liquid droplets. This can be used to moderate the tendency of the droplets either to overheat or to wet and adhere.

The diluent can be combined directly with and introduced into the polymerization reactor with the liquid feedstream of catalyst, cocatalyst, and solvent(s). Alternatively, the diluent can be fed to the polymerization reactor separately from the liquid feedstream of catalyst, cocatalyst, and solvent(s), for example as an induced condensing agent. Or diluent can be both added to the liquid feedstream of catalyst, cocatalyst, and solvent(s) and fed separately to the reactor. When diluent is separately added to the reactor, the relative rate of solvent evaporation from the liquid droplets can be controlled by adjusting the amount of diluent vapor in the cycle gas. When the level of diluent is increased in the cycle gas, the evaporation rate of solvent in the liquid droplets decreases, which leads to increased polyolefin particle size under some conditions. The amount of solvent in the liquid droplets can then be decreased by, for example, feeding less solvent with the catalyst, and the liquid droplet size can be decreased by feeding more inert carrier gas with the unsupported catalyst. The liquid droplets can form new, stable polymer particles more readily under the conditions of increased dew point or condensation in the cycle gas.

. In a preferred embodiment, the average particle size of a polyolefin produced using an unsupported catalyst in a gas phase, fluidized bed polymerization reactor is controlled by adjusting at least one of: a) the size of the liquid droplets containing the catalyst introduced into the polymerization reactor, and b) the concentration of catalyst composition in the liquid droplets, while introducing said liquid droplets into a particle lean zone in the reactor in the manner described in copending U.S. Application Serial No. 08/659,764 filed June 6, 1996. A particle lean zone can be established in the reactor by feeding the unsupported catalyst in any manner such that the liquid droplets containing the catalyst do not immediately contact a substantial portion of the resin particles of the fluidized bed. Generally, the particle density in the particle lean zone is at least 10 times lower than that in the fluidized bed. In the time period elapsing from when the liquid droplets leave the nozzle until they contact the particles in the bed, new polymer particles are formed. The time between a liquid droplet being introduced into the reactor and its contacting the particles in the bed ranges from about 0.01 seconds to 60 seconds, preferably about 0.01 to 30 seconds, and most preferably about 0.01 seconds to 5 seconds.

A particle lean zone may be a section of the reactor that normally does not contain the fluidized bed, such as the disengaging section, the gas recirculation system, or the area below the distributor plate. The particle lean zone may also be created by deflecting resin away from the unsupported catalyst spray with a stream of gas. In a preferred embodiment, the liquid feedstream containing unsupported catalyst, accompanied by an inert carrier gas, is surrounded by at least one other gas which serves to move or deflect resin particles of the bed out of the path of the liquid droplets of unsupported catalyst as they enter the fluidization zone and away from the area of catalyst entry, thereby providing a particle lean zone. In a particularly preferred embodiment, the unsupported catalyst in the carrier gas is surrounded by at least two other gases, the first gas serving primarily to deflect resin particles of the bed out of the path of the liquid droplets and the second gas primarily to prevent the injection tube or nozzle tip from becoming clogged.

The average particle size of the polyolefin produced ranges from about 0.01 to 0.06, preferably about 0.015 to 0.04, most preferably about 0.02 to 0.03 inches. In a fluidized bed reactor, low particle size can result in high fines levels, which present handling difficulties due to possible dust-cloud formation and explosion. Also, small particles can more readily be entrained from the fluidized bed, leading to fouling in the cycle gas piping and equipment, such as the cooler or distributor plate. Small particles are also more susceptible to electrostatic forces, leading to sheeting or wall fouling. Large average particle size, on the other hand, can often lead to bed segregation and other fluidization problems, as well as a decrease in the resin bulk density.

Control of polyolefin average particle size according to the invention can improve reactor operability by avoiding these problems. For example, during reactor start-up or during transitions, or for normal production of some products with unfavorable electrostatic tendencies, the average particle size of the polyolefin produced can be increased to reduce the possibility of reactor sheeting or other ill effects of static, perhaps to a 0.04 to 0.06 inch range. When the static risk decreases the particle size can be adjusted back to a middle range. Likewise, the average particle size of the polyolefin can also be increased when producing products that tend to foul the cycle gas line. Fewer particles are entrained and fouling is reduced.

When transitioning the reactor from one polyolefin grade to another, the amount of offgrade polyolefin produced is proportional to the bed weight of polymer in the reactor. Prior to a transition, the average particle size of the polyolefin can be increased, causing the bulk density to decrease. For constant bed level transitions, the bed weight concomitantly decreases to compensate for the reduced bulk density. This requires fewer pounds of transition grade resin to be made. The increased average particle size also facilitates low bed level transitions so that offgrade production can be decreased.

The following examples further illustrate the invention.

EXAMPLES

The catalyst used for Examples 1 through 4 was a Zr-based metallocene at a 2 wt % solution in n-hexane. The solution was used as made for Examples 1 and 2, but was diluted with 1-hexene for Examples 3 and 4 to 1.33 wt-% catalyst with 32.9% 1-hexene and 65.8% hexane.

Catalyst was mixed in line with MMAO 3A (modified methyl alumoxane) as received from Akzo Nobel at 7.1 wt % Al. Additional dilution was performed by adding isopentane to the mixture before introducing it to the reactor. Catalyst and MMAO feedrates were adjusted to provide a final Al:Zr molar ratio between 330 and 340. The reactor was 2.4 m in diameter and was operated with a bed height of 11.6 m and a superficial gas velocity of approximately 0.6 m/s. Total reactor pressure was 1960 kPa. ATMER-163, marketed by ICI, was added as necessary to the reactor to control the buildup of electrostatic charge.

The catalyst atomization devices used in all examples were located at the end of a 1/4" (0.635 cm) OD stainless steel tube, and they could be removed from the reactor during operation. This tube passed through a 3/4-inch (1.9 cm) schedule-40 pipe. A stream of 1000 to 1180 kg/hr of ethylene monomer at a temperature between 85 and 95°C was fed through the annular space between the ^-inch tube and the 3/4- inch pipe. This monomer stream is referred to as a nozzle cleaning gas. The 3/4-inch pipe was located in the center of a six-inch pipe (15.2 cm), through which was fed between 22,700 and 29,500 kg/hr of cycle gas, known as particle deflecting gas. The six-inch pipe extended 53 cm into the reactor, the 3/4-inch pipe extended 61 cm into the reactor, and the spray nozzle extended 66 cm into the reactor, at a location 2.4 m above the distributor plate.

EXAMPLES 1 and 2 A seed bed was charged to the reactor and it was dried to 9 ppm water. It was pressurized to 790 kPa of nitrogen and then 22.7 kg/hr of 10 wt % TEAL in isopentane were fed to the reactor and allowed to circulate for 1 hour. The conditions listed in the Table below were established in the reactor. Catalyst was fed through a perpendicular spray nozzle, located within the stream of 22,700 kg/hr of cycle gas, as described above. Catalyst and MMAO were mixed for 15 to 30 seconds. The reactor was started with a nitrogen carrier rate of 27.2 kg/hr. This caused the APS of the resin to drop rapidly from 0.66 to 0.356. cm, which was not acceptable for good operation. The nitrogen carrier rate was then decreased to 14.3 lbs/hr and the APS increased to 0.533 cm, where it remained stable, and desirable.

This shows that by proper control of the inert carrier gas rate, the APS can be controlled. In other cases, if the APS is too large, the carrier gas rate can be increased to lower the APS back to an acceptable range.

EXAMPLES 3 and 4

These examples show that the APS can also be controlled by adjusting the amount of diluent added to the catalyst and cocatalyst mixture.

The reactor was operating with the perpendicular spray nozzle and with the 1.33 wt % catalyst in the mixed hexane/hexene solvent, which was mixed with the MMAO for 19 minutes before being diluted with isopentane and fed to the reactor. Several hundred pounds per hour of isopentane were additionally fed to the reactor at a separate location to induce condensing mode operation. The amount of isopentane in the cycle gas was increased to about 5 mole percent.

The reactor was operated with the conditions listed in Example 3, and the APS was stable at 0.483 mm. The isopentane feedrate then was decreased for Example 4, as shown in the Table, and the APS increased to 0.610 mm, demonstrating that the APS can be controlled while in condensing mode, and that it can be controlled with the amount of isopentane diluent added to the catalyst mixture. It also was demonstrated throughout the course of a 10 day run with the perpendicular spray nozzle that the APS could be kept in a narrow range by adjusting the nitrogen and isopentane carrier rates independently, or in concert.

TABLE

Example 2 3 4 5

Catalyst feedrate 0.34 0.77 0.64 0.64 (kg/hr)

MMAO feedrate 1.7 3.8 2.0 2.0 (kg/hr)

Isopentane 6.8 4.7 6.6 4.2 feedrate (kg/hr)

Nitrogen feedrate 27.2 14.3 13.6 13.6 (kg/hr)

Reactor 76 76 76 76 temperature (°C)

Ethylene partial 1310 1310 1530 1530 pressure (kPa)

Molar C&IC ratio 0.0255 0.0258 0.0246 0.0235

Resin density 0.917 0.918 0.918 0.918 (g/cm³)

Average particle size (mm) Initial 0.660 0.356 0.457 0.483 Final 0.356 0.533 0.483 0.610

Bulk density 378 362 366 348

Morphology S Spphheerrees Spheres and Spheres and Spheres and and small small small small clusters clusters clusters clusters

Claims

We claim:

1. A process for producing a polyolefin in a gas phase polymerization reactor, which comprises:

(i) introducing monomer into the polymerization reactor;

2. The process of claim 1, wherein the liquid droplets of unsupported catalyst are introduced into the polymerization reactor in an inert carrier gas, and the size of the liquid droplets is adjusted by changing the flow rate of the inert carrier gas into the polymerization reactor.

3. The process of claim 1, wherein the liquid droplets comprise a solution of the unsupported catalyst in a solvent, and the concentration of catalyst in the liquid droplets is adjusted by changing the amount of solvent in the solution.

4. The process of claim 3, further comprising feeding a diluent to the polymerization reactor separately from the liquid droplets containing the unsupported catalyst, such that the rate of evaporation of the solvent from the liquid droplets in the polymerization reactor is controlled by the feed rate of the diluent.

5. The process of claim 1, wherein the catalyst is selected from the group consisting of Ziegler-Natta catalysts, metallocene catalysts, rare earth metal catalysts, and mixtures thereof.

6. The process of claim 5, wherein the catalyst is a metallocene catalyst having a formula selected from the group consisting of:

mixtures thereof wherein:

M is Zr or Hf;

L is a substituted or unsubstituted, ╧Ç-bonded ligand; each Q can be the same or different and is independently selected from the group consisting of -0-, -NR³-, -CR³2- and -S-;

Y is either C or S;

Z is selected from the group consisting of -OR³, -NR³2, - CR³3, -SR³, -SiR³3, -PR³2 and -H, with the proviso that when Q is - NR³- then Z is selected from the group consisting of -OR³, -NR³2, -

SR³, -SiR³3, -PR³2 and -H; n is 1 or 2;

A is a univalent anionic group when n is 2 or A is a divalent anionic group when n is 1;

R³ can be the same or different and is independently a group containing carbon, silicon, nitrogen, oxygen, and/or phosphorus and one or more R³ groups may optionally be attached to the L substituent;

T is a bridging group selected from the group consisting of an alkylene or arylene group containing from 1 to 10 carbon atoms optionally substituted with carbon or heteroatoms, germanium, silicone and alkyl phosphine; and m is 2 to 7.

7. The process of claim 6, wherein the metallocene catalyst is selected from the group consisting of indenyl zirconium tris(diethylcarbamate), indenyl zirconium tris(trimethylacetate), and (methylcyclopentadienyl) zirconium tris(trimethylacetate).

8. The process of claim 5, wherein the catalyst is a rare earth metal catalyst selected from the group consisting of neodymium carboxylates, neodymium alcoholates, neodymium acetylacetonates, neodymium halides, neodymium allyl derivatives, and mixtures thereof.

9. The process of claim 1 wherein the polyolefin produced is selected from the group consisting of ethylene homopolymers, ethylene copolymers and terpolymers employing C3-C12 alpha olefins, propylene homopolymers, propylene copolymers employing C4-C12 alpha olefins, polybutadiene, ethylene propylene rubbers, and ethylene propylene diene rubbers.

10. A method of controlling the average particle size of a polyolefin produced by contacting in a gas phase polymerization reactor monomer with liquid droplets containing an unsupported catalyst, which comprises adjusting at least one of: a) the size of the droplets, and b) the concentration of catalyst in the droplets.

11. The method of claim 10, wherein the liquid droplets of unsupported catalyst are introduced into the polymerization reactor in an inert carrier gas, and the size of the liquid droplets is adjusted by changing the flow rate of the inert carrier gas into the polymerization reactor.

12. The method of claim 10, wherein the liquid droplets comprise a solution of the unsupported catalyst in a solvent, and the concentration of catalyst in the liquid droplets is adjusted by changing the amount of solvent in the solution.

13. The method of claim 10, further comprising feeding a diluent to the polymerization reactor separately from the liquid droplets containing the unsupported catalyst, such that the rate of evaporation of the diluent from the liquid droplets in the polymerization reactor is controlled by the feed rate of the diluent.

14. The method of claim 10, wherein the catalyst is selected from the group consisting of Ziegler-Natta catalysts, metallocene catalysts, rare earth metal catalysts, and mixtures thereof.

15. The method of claim 14, wherein the catalyst is a metallocene catalyst having a formula selected from the group consisting of:

mixtures thereof wherein:

M is Zr or Hf; L is a substituted or unsubstituted, ╧Ç-bonded ligand; each Q can be the same or different and is independently selected from the group consisting of -0-, -NR³-, -CR³2- and -S-;

Y is either C or S;

SR³, -SiR³3, -PR³2 and -H; n is 1 or 2;

16. The method of claim 15, wherein the metallocene catalyst is selected from the group consisting of indenyl zirconium tris(diethylcarbamate), indenyl zirconium tris(trimethylacetate), and (methylcyclopentadienyl) zirconium tris(trimethylacetate).

17. The method of claim 14, wherein the catalyst is a rare earth metal catalyst selected from the group consisting of neodymium carboxylates, neodymium alcoholates, neodymium acetylacetonates, neodymium halides, neodymium allyl derivatives, and mixtures thereof.

18. The method of claim 10 wherein the polyolefin produced is selected from the group consisting of ethylene homopolymers, ethylene copolymers and terpolymers employing one or more C3-C12 alpha olefins, propylene homopolymers, propylene copolymers employing one or more C4-C12 alpha olefins, polybutadiene, ethylene propylene rubbers, and ethylene propylene diene rubbers.