Olefin Polymerization Catalyst and Method

This invention relates to a new class of polymerization catalysts, methods of making the catalysts and methods of polymerizing and interpolymerizing 1-olefins with the new catalysts. The catalysts are prepared by reacting a transition metal compound and novel magnesium silylamide compounds which have magnesium-nitrogen-silicon bonds as disclosed in commonly assigned Pullukat and Hoff U.S. Pat. No. 4,383,119 issued May 10, 1983, the disclosure of which is hereby incorporated by reference.

Catalysts of this invention are highly active and are suitable for polymerization of ethylene and other 1-olefins, particularly of 2-8 carbon atoms, and interpolymerization of these with 1-olefins of 3-20 carbon atoms, such as propylene, butene and hexene, for example, to form polymers of low and medium densities. These catalysts are suitable for particle form, solution form, and gas phase olefin polymerization since they can be added to the reactor as solids, slurries, dispersions in hydrocarbon liquids and as colloidal suspensions. These catalysts are also suitable for use in retrofitted high pressure, high temperature low density reactors.

The silylamide catalysts of this invention have an enhanced relative reactivity towards 1-alkenes in interpolymer synthesis with ethylene. Such catalysts can provide many advantages in the manufacture of ethylene interpolymers. These advantages are easily recognized by research workers in the industry. However, since only speculative guidelines can be derived from chemical theories, the discovery of such improved catalysts has been sought for a long time by empirical methods.

This will allow manufacture of low density polyethylene in low pressure units in place of conventional energy intensive high pressure units. Recently, manufacture of low density polyethylene copolymers commonly called linear low density polyethylene (LLDPE) has been accomplished. The catalyst systems that are currently being used require high comonomer ratios and thus lead to process related problems.

LLDPE products are actually interpolymers of ethylene and 1-alkenes with densities from about 0.915 to about 0.925 g/cm.sup.3. Their manufacture requires bringing together under suitable conditions ethylene, at least one 1-alkene and a catalyst. With any catalyst the amount of 1-alkene incorporated into the copolymer depends upon the molar ratio of 1-alkene to ethylene, and the particular 1-alkene involved; that is, whether it is 1-butene, or 1-hexene, etc. It is also a fact that with any known catalyst the molar ratio of 1-alkene to ethylene in the polymerization reaction greatly exceeds the molar ratio contained in the interpolymer. As a consequence it is necessary to separate the excess 1-alkene and recycle it for economical operation. Since the silylamide catalysts of this invention have higher relative reactivity towards 1-alkenes, LLDPE products can be made with a lower molar ratio of alkene to ethylene. Smaller equipment and less energy expense in recycling are required in any LLDPE process as a result.

Also in any process there is a stage in which the copolymer product is separated from the polymerization zone. At this stage, the product will contain some of the unpolymerized 1-alkene dissolved in the polymer. This unpolymerized 1-alkene must be removed. It can cause bubbling in extruders, odor problems and a fire hazard in storage. The amount of 1-alkene retained by the product is less when the concentration of 1-alkene is lower in the polymerization zone. Since the silylamide catalysts of this invention allow the desired LLDPE to be produced at lower 1-alkene concentrations, they have an advantage in this respect also.

In a slurry or particle form process, there is a practical limit to the concentration of 1-alkene in the polymerization zone. This limit is the point at which so much 1-alkene has been added that the particles of copolymer begin to dissolve resulting in reactor fouling and process failure. Since 1-hexene is a better solvent than 1-butene, the allowable concentration of 1-hexene is lower than for 1-butene. This is also the case for 1-octene. In theory, the ethylene concentration could be decreased so that the molar ratio required for any copolymer product could be reached. This approach has a limit to its usefulness since overall production rates are decreased. The silylamide catalysts of this invention allow the manufacture of 1-hexene-ethylene copolymers in the particle form process at high production rates. There is added significance to this feature in that the 1-hexene interpolymer is superior to the butene copolymer as an LLDPE film resin because the film can be made stronger.

By proper selection of the novel magnesium compounds of this invention, the electronic and steric nature of the catalyst active site can be modified. As a result, the catalyst's polymerization activity and the properties of the polymers produced can be controlled.

The alkyl magnesium silylamide compounds used in this invention are new compounds represented by the formulas: ##STR1## and ##STR2## in which R is a straight or branched chain alkyl group including alkyl groups of about 1-18 carbon atoms, aryl groups of about 6 to 14 carbon atoms, and the R groups are the same or different. The alkyl or aryl R groups are preferably unsubstituted but may have inert substituents that are nonreactive to the highly reactive R--Mg, or Mg--N bonds. These substitutents, if used, may be ether groups, tertiary amine groups, chloride and fluoride groups and combinations of these. R' is R or hydrogen; Y is R or --SiR'.sub.3, all R' are the same or different and all Y are the same or different. The preferred compounds are readily soluble in liquid alkanes such as butane, hexane and other simple liquid hydrocarbons.

The magnesium silylamide compounds I and II of this invention may be made by the methods described in Pullukat and Hoff U.S. Pat. No. 4,383,119 as follows:

One very convenient method of synthesis is to combine a silylamine which has an acidic N--H group with a mixed dialkyl magnesium compound solution, such as a solution of butyl ethyl magnesium. The silylamine N--H will displace the most reactive alkyl groups as in: ##STR3##

The following synthesis methods may also be used:

(1) A trialkyl silyl chloride is combined with a primary amine in the presence of a tertiary amine to trap HCl. To the reaction mixture a solution of dialkylmagnesium is added. Precipitated tertiary amine hydrochloride is filtered out to give a chloride and ether-free solution of alkyl magnesium silylamide. This method synthesizes a silylamine by known methods without requiring that the silylamine be separated from the reaction mixture.

(2) The following procedure using a reaction of a Grignard reagent with a lithium silylamide may also be used: ##STR4## After the addition of the lithium silylamide to the Grignard reagent in ether, a hydrocarbon solvent with a higher boiling point than the ether is added. The ether is then removed, as by distillation. The lithium chloride is then removed, as by filtration.

(3) When the ether is removed from a Grignard reagent, a solid residue remains. Extraction of this residue with a solution of silylamine provides an ether and halide free solution of alkyl magnesium silylamide.

(4) Magnesium metal, an alkyl halide, and a silylamine can be coreacted in an aromatic or aliphatic hydrocarbon to yield an alkyl magnesium silylamide.

The transition metal compound is preferably a halide, oxyhalide, alkoxyhalide, or an alkoxide from Groups IIIB, IVB, VB and VIB of the fourth and fifth periods of the periodic table and Groups VIIB and VIIIB of the fourth period. Suitable transition metal compounds include TiCl.sub.4, Ti(OR.sup.2).sub.x Cl.sub.4-x, VOCl.sub.3, VCl.sub.4, Zr(OR.sup.2).sub.x Cl.sub.4-x or mixtures of these and others where R.sup.2 are alkyl groups containing 1-10 carbon atoms or aryl groups of 6-14 carbon atoms. The R.sup.2 groups may be the same or different and x may be 0 to 4 inclusive.

The reaction between the magnesium silylamide and the transition metal is usually rapid and it is generally sufficient only to mix the components at room temperature in a hydrocarbon liquid as defined above.

A convenient way to conduct the reaction between the silylamide and the transition metal compound is in the presence of a liquid hydrocarbon solvent or diluent. After the reaction the hydrocarbon solvent can be removed to provide a catalyst in the form of a dry, free-flowing powder. However, it is not necessary to remove the solvent, and, in fact, for some processes it is desirable to suspend the catalyst in a moderately viscous oil. To promote the stability of the suspension, the catalyst may be pulverized, and agents which increase colloid stability may be added. The reaction can also be conducted in the absence of solvents in a ball mill or a ribbon blender or other suitable vessel.

Either the transition metal compound or the silylamide can be deposited on or reacted with a finely divided carrier material prior to the reaction. Suitable carriers include difficult to reduce materials such as silica, alumina, zirconia, thoria, titania, magnesia and organic polymer powders, or mixtures thereof.

Alternatively, the reaction between the silylamide and the transition metal compound can take place in the presence of such carrier and a hydrocarbon solvent. Another convenient procedure is to blend the carrier with the reaction product after the completion of the reaction.

The supported catalysts formed by these procedures may also be used as dry solids or as suspensions. If desired, the particle sizes may be reduced by grinding or ball-milling.

The reaction between silylamide and the transition metal compound may be with a Mg/transition metal atom ratio of 1.0, but the ratio may be in a range of about 0.2-100 to 1, preferably between about 0.4-30 to 1.

In some cases the activity of the reaction product is increased by the addition of an organometallic cocatalyst either prior to polymerization or simultaneously with the introduction of the catalyst to the polymerization vessel. However, some catalysts do not require cocatalysts to exhibit polymerization activity. It is preferred to use an alkylaluminum compound as a cocatalyst. The ratio of aluminum to transition metal may range from about 0-500 to 1, preferably about 1-50 to 1. Various alkylaluminum compounds can function as cocatalysts, the preferred compounds depending upon the monomers and the polymerization conditions. Typically, suitable compounds are triethylaluminum, triisobutylaluminum, diisobutylaluminum hydride, diethylaluminum chloride, ethyl aluminum dichloride, dialkylaluminum alkoxides, trioctyl aluminum, diethyl aluminum trimethylsiloxide, such as of the formula

their reaction products with small amounts of water, and complexes formed with Lewis bases like tertiary amines, ethers, and esters.

These catalysts are very effective in the polymerization of ethylene using any of the common commercial processes including solution form, particle form, and gas phase processes as well as in retrofit high pressure high temperature reactors. These processes may operate at pressures in the range of 15 psi to 50,000 psi.

Some polyethylene processes operate at about 80.degree. C. or lower while others run at temperatures of about 300.degree. C. The catalysts of this invention are intended, by appropriate changes within the scope of the disclosure, to be suitable for the entire temperature range of commercial processes, i.e. from about 25.degree. C.-300.degree. C.

These catalysts are suitable for interpolymerization in any of the above mentioned processes by conducting the polymerization with other 1-olefins, for example, propylene, 1-butene, 1-hexene, 1-octene, etc., to produce polymers of low densities. The catalysts of this invention exhibit enhanced reactivity towards monomers like 1-butene and 1-hexene so that interpolymerizations of these monomers with ethylene are more efficient.

Polymerization by the catalysts is influenced by the --SiR'.sub.3 and --Y groups and their steric requirements. As a result, varying the nature of the --SiR'.sub.3 and --Y groups alters the properties of polymers and copolymers produced with the catalysts of the invention. In specific instances, proper selection of the --SiR'.sub.3 or --Y groups or both results in enhanced relative reactivity of, for example, 1-butene or 1-hexene monomers so that copolymerization or terpolymerization of ethylene and these monomers can be conducted more efficiently.

The structure of polymers and copolymers can be controlled by proper selection of the magnesium silylamide and transition metal compounds to form the catalyst. In this way, the average molecular weight under any given conditions of polymerization and copolymerization can be regulated. As a result of such control, thermoplastic resins with better strength or processing characteristics can be produced.

The alkyl magnesium silylamides used in this invention are in many cases soluble in hydrocarbons and can be obtained without any halogen content.

This invention uses organomagnesium compounds which not only have reactive alkyl magnesium bonds but are also stable. These new compounds do not require an ether solvent as Grignard reagents do. Ethers are undesirable in many cases because the coordination of the ether diminishes the reactivity of the alkyl magnesium bond and, in some cases, modifies the catalyst activity.

The organomagnesium compounds used in this invention are superior to the dialkylmagnesiums of methyl, ethyl, n-propyl, and n-butyl in that they dissolve more easily in hydrocarbon solvents to give less viscous solutions.

The new compounds of (I) above have only one reactive alkyl magnesium bond unlike the recently discovered solutions of butyl ethyl magnesium, sec-butyl-n-butyl magnesium and similar mixtures.

No complexing agent is required for the organomagnesium compounds of this invention to induce hydrocarbon solubility.

Through the careful selection of the R' and Y groups in structures I and II above, the electronic and steric structure about the magnesium atom can be varied over a wide range and hence it can be controlled. ##STR5## In the structures listed above, Y and R' have been previously described while R.sup.4, R.sup.5, and R.sup.6 are alkyl groups of 1-14 carbon atoms or aryl groups of 6-14 carbon atoms.

Regardless of the choice of the groups R.sup.4, R.sup.5 and R.sup.6 in structure III, the ability to control the electronic and steric structure about the magnesium atom is less than that which is possible with the choice of the R' and the Y groups in structure I because the difference between --CH.sub.3, --C.sub.2 H.sub.5, and --C.sub.3 H.sub.7 is chemically less different when compared to --Si(CH.sub.3).sub.3 and --C.sub.2 H.sub.5, for example.

Hence, the novel magnesium compounds have an advantage over dialkyl magnesium compounds as catalyst components. Therefore, by using a variety of different magnesium silylamide compounds as catalyst components, catalytic sites with carefully controlled molecular architecture can be developed. Since the electronic and steric structure about the catalytic site strongly influences the polymer species produced, the careful selection of the magnesium silylamide compound can influence and hence control the molecular weight, molecular weight distribution and comonomer incorporation and distribution throughout the polymer species.

2.48 ml of titanium tetrachloride was added by syringe to a solution of 22.2 mmoles of octyl magnesium bis(trimethylsilyl)amide in about 50 ml of a mixture of hexane and heptane. The solution was contained in a three-neck flask under a purge of N.sub.2. The addition of the TiCl.sub.4 was completed in about 2 minutes. Black precipitate formed immediately which was dispersed and became a thick slurry. 90 ml of hexane was mixed in by means of the magnet bar stirrer. Within a few minutes, the entire reaction volume gelled.

The solvent was evaporated by heating the flask at 190.degree. C. A black, friable material remained which was reduced to a powder by the magnet bar.

In this catalyst preparation, the Mg/Ti atomic ratio was 1.0, and the calculated titanium content of the product was 10.0 wt.%.

A copolymerization test was conducted at 70.degree. C. in isobutane liquid with 22 wt.% 1-butene, 50 psi H.sub.2 partial pressure, and ethylene as required to maintain 350 psig. Triisobutylaluminum (TIBAL) was added to give 2.7 Al/Ti ratio.

The reactivity of the catalyst in the copolymerization was found to be 7078 g/g of catalyst per hour. The product melt index by an ASTM method was 3.9 g/10 min. and the density of the copolymer product was 0.913 g/cm.sup.3.

Similar copolymerization tests with the reaction products of dibutyl magnesium and titanium tetrachloride gave a density of 0.920 g/cm.sup.3. The lower density is an indication of larger content of short branches from the butene comonomer. Therefore, it is shown by this example that the relative reactivity of 1-butene is greater with the catalyst made with the octyl magnesium bis(trimethylsilyl)amide.

16 mmol of butyl magnesium bis(trimethylsilyl)amide dissolved in about 20 ml of heptane was mixed with 1.8 ml of TiCl.sub.4 (16 mmol). The solvent was evaporated at 110.degree. C. under flowing N.sub.2. The black residue was pulverized in the flask, with the magnet bar, then blended with 2.4 g of silica. The silica had been treated with hexamethyl disilazane and heated at 110.degree. C. prior to its use in this catalyst preparation. All of the manipulations were done so as to avoid exposure of the catalyst preparation to air and moisture.

The catalyst was tested in butene-ethylene copolymerization as described in Example 1 except that the amount of butene was 18 wt.%. The reactivity was 7928 g/g cat./hr., the melt index was 0.97, and the annealed density was 0.920 g/cm.sup.3. The bulk density of the particle form copolymer was 16 lbs/ft.sup.3.

This example shows that the catalysts of the invention can be blended with a finely divided support after the reaction product has been formed.

16 mmol of butyl magnesium bis(trimethylsilyl)amide dissolved in about 25 ml of hexane was treated with 16 mmol of TiCl.sub.4. The solvent was evaporated at 110.degree. C. under N.sub.2 and the residue was pulverized.

The catalyst was tested in ethylene-hexene copolymerization at 71.degree. C. in 500 ml of isobutane with 20 wt.% 1-hexene. TIBAL was employed as a cocatalyst. Hydrogen to give a 25 psi partial pressure, and ethylene to maintain 250 psi throughout the reaction time were added.

The product copolymer was obtained in the form of particles without signs of reactor fouling. Its annealed density was 0.919 and its melt index was 8.4.

The reactivity was 1787 g/g/cat./hr.

This example shows that an ethylene-hexene copolymer, a type of material now called linear low density polyethylene, can be made with the catalyst of this invention under particle form conditions without reactor fouling.

16 mmol of butyl magnesium bis(trimethylsilyl)amide dissolved in about 25 ml of hexane was treated with 8 mmol of TiCl.sub.4. The solvent was evaporated as in previous examples.

The catalyst was tested in butene copolymerization as described in Example 1 except that the amount of 1-butene was 15 wt.%.

The reactivity was 1295 g/g cat.hr. and the annealed density of the product was 0.925 g/cm.sup.3.

This example illustrates the use of a catalyst with a Mg/Ti atomic ratio of 2.0.

The catalyst of this example was prepared by adding 32 mmol of TiCl.sub.4 to 16 mmol of butyl magnesium bis(trimethylsilyl)amide so that the Mg/Ti atomic ratio was 0.5. The solvent was evaporated under N.sub.2 as in previous examples and then the residue was blended in the absence of solvent with 3.6 g of hexamethyl disilazane-treated silica.

The catalyst was tested in hexene copolymerization as described in Example 3, with 15 wt.% hexene.

The results of two tests are as follows:

An additional 2.4 g of the treated silica was blended into the dry catalyst preparation, and the resulting blend was tested in the same manner:

This example illustrates the preparation of a catalyst with a Mg/Ti ratio of 0.5, the dry blending of the catalyst with silica, and the use of the catalyst in hexane-ethylene copolymerization under particle form conditions.

16 mmol of butyl magnesium bis(trimethylsilyl)amide was mixed with 5.0 g of low density polyethylene powder in about 25 ml of heptane. The mixture was combined with 16 mmol of TiCl.sub.4 and the heptane was evaporated at 110.degree. C. under flowing N.sub.2.

The catalyst was tested with TIBAL cocatalyst in the copolymerization of 1-hexene and ethylene in isobutane at 68.degree. C. The amount of 1-hexene was 20 wt.%, the hydrogen partial pressure was 25 psi and ethylene was added to keep the pressure constant at 240 psig. Reactivity was 2596 g/g cat./hr. and the annealed density of the product was 0.928 g/cm.sup.3.

The melt index was 0.11 and by infrared spectroscopy the product contained 9.9 methyl groups per 1000 carbon atoms.

A similar catalyst was made with polyethylene powder except that the amount of TiCl.sub.4 was increased giving a Mg/Ti ratio of 0.5. In a hexene copolymerization test as above the reactivity was 6612 g/g cat./hr. and the density of the product was 0.918 with a melt index of 0.4 and a methyl group content of 11.0. The ratio of high load melt index to melt index was 36.

This example shows the use of polyethylene powder as a carrier, and shows a situation in which a catalyst with a Mg/Ti ratio of 0.5 is superior to a catalyst with a Mg/Ti ratio of 1.0.

A solution was prepared containing 16 mmol of the silylamide of the structure shown, ##STR6## in about 25 ml of heptane. To this solution, 16 mmol of TiCl.sub.4 was added. The solvent was evaporated at 110.degree. C. under a flow of N.sub.2, and the residue was pulverized with the magnetic stirrer.

The catalyst was tested in the particle form polymerization of ethylene at 102.degree. C. in 500 ml of isobutane. The hydrogen partial pressure was 50 psi and the total pressure of 550 psig was kept constant with ethylene as required. TIBAL solution in the amount of 9.2 mmol per gram of catalyst was added as a cocatalyst.

The reactivity was 2963 g/g cat./hr., the melt index was 0.68, and the high load melt index to melt index ratio was 27.

This example shows catalyst preparation from another magnesium silylamide of the invention.

5 g of high density polyethylene powder (a product marketed under the trademark Microthene) was combined with a solution of 16 mmol of butyl magnesium bis(trimethylsilyl)amide in 25 ml of heptane. By syringe 32 mmol of TiCl.sub.4 was added and the reaction mixture was stirred at room temperature for 30 minutes prior to solvent evaporation at 110.degree. C.

The catalyst was tested with TIBAL cocatalyst in ethylene-hexene copolymerization at 54.degree. C. with 500 ml isobutane, 20 wt.% 1-hexene, a hydrogen partial pressure of 40 psi, and ethylene sufficient to maintain a constant pressure of 205 psig.

The reactivity of the catalyst was 2756 g/g cat./hr., the melt index was 1.9, and the unannealed density of the ethylene-hexene copolymer was 0.923 g/cm.sup.3.

This example shows that the catalyst components can be blended with high density polyethylene powder.

The catalyst of Example 8 was tested in a high temperature polymerization experiment with ethylene. In this test, diethylaluminum chloride was used as the cocatalyst at a ratio of Al/Ti of 7.1.

The cocatalyst was added to 500 ml of mineral spirits (Phillips Petroleum Company product Soltrol 130) in the polymerization test vessel and the temperature was adjusted to 184.degree. C. Ethylene was added to give 450 psig pressure then 0.0875 g of the catalyst was injected quickly and the ethylene pressure was increased to 550 psig. The reaction was continued under these conditions for 15 minutes during which time the temperature increased to 233.degree. C. because of the heat of polymerization.

The reactivity was 2670 g/g cat./hr. of reaction time, or 44.5 g/g cat. per minute.

The melt index of the product polyethylene was 0.91 g/10 minutes.

This example illustrates the reactivity of the invention catalysts at high temperature and the use of diethylaluminum chloride as the cocatalyst.

A slurry was prepared containing 8 mmol of the alkyl magnesium silylamide with the structure below in 25 ml of heptane: ##STR7##

To this slurry 16 mmol of TiCl.sub.4 was added to give a mixture in which the Mg/Ti ratio was 1.0. The solvent was evaporated under nitrogen as in previous examples.

In a particle form polymerization test of ethylene with TIBAL cocatalyst the reactivity was found to be 961 g/g cat./hr.

This example shows a catalyst with another alkyl magnesium compound of the invention. In this case the silylamide compound is not readily soluble in hydrocarbon liquids at room temperature.

A catalyst was prepared from butyl magnesium bis(trimethylsilyl)amide and TiCl.sub.4 as described in Example 3. It was tested in the particle form polymerization of ethylene under the conditions of Example 7 with various organometallic cocatalysts. The results of these tests are given below.

15.95 mmol of VOCl.sub.3 was added slowly to 15.95 mmol of butyl magnesium bis(trimethylsilyl)amide dissolved in 20 ml of heptane in a N.sub.2 -purged flask. The reaction was conducted for 30 minutes then the flask was placed in an oil bath with a temperature of 100.degree. C. The volatiles were swept away by a flow of N.sub.2. A black residue remained which was readily pulverized with the magnetic stirrer.

The catalyst was tested with TIBAL cocatalyst in hexene copolymerization as described in Example 8. The reactivity was 266 g/g of vanadium per hour.

0.16 g (1.0 mM) CrCl.sub.3 was added to a flask followed by 20 ml of heptane. Next 17.9 ml of heptane solution (0.57M) of butyl-magnesium bis(trimethylsilyl)amide were added. A small amount of white precipitate formed but some of the CrCl.sub.3 still remained as a solid. The mixture was heated 100 min. at 70.degree. C. After cooling to room temperature 1.0 ml (9 mM) of TiCl.sub.4 were added. The catalyst was dried at 105.degree. C. to form a free flowing powder.

40.0 ml of a 0.25M solution of butyl magnesium bis(trimethylsilyl)amide were added to a flask. Then 1.1 ml (10 mM) of TiCl.sub.4 was added dropwise at .about.23.degree. C. A viscous suspension formed.

The catalyst was tested for ethylene polymerization as follows:

The catalyst suspension was split into two portions, one for polymerization testing with a TIBAL cocatalyst and one for testing without a cocatalyst. Polymerization testing was carried out in a glass reactor system at 55.degree. C. in heptane solvent with 35 psi ethylene pressure in the presence of 1-hexene.

In each case, the catalyst suspension and, in one case, TIBAL were syringed into the reactor after it was purged with N.sub.2 and had been charged with 500 ml of heptane. Next, 10 ml of 1-hexene was added and the ethylene (35 psi) turned on. Each copolymerization ran 30 minutes. The results are given below:

This example shows that alkyl aluminum compounds are not required for activity.

A 200 ml volume of 4-methylpentene-1, 0.2 ml of titanium tetrachloride, and 1.6 ml of a solution of butyl magnesium bis(trimethylsilyl)amide were combined in a crown cap bottle. The Ti/Mg ratio was 2/1. After 3 min., 1.0 ml of a heptane solution of triisobutylaluminum was added to make the Ti/Al ratio equal to one. A slurry of poly(4-methylpentene-1) began to form. Continuous mixing was provided by means of a magnet bar. After 63 min. of reaction time, another 0.5 ml of triisobutylaluminum solution was added to change the Ti/Al ratio to 0.67. The second addition of organoaluminum cocatalyst had no noticeable effect upon the reaction which was terminated after 74 min. with 5 ml isopropyl alcohol. The reaction solution was poured into 1 liter of methyl alcohol, and the polymer product was separated and dried.

The yield was 20 g and the conversion was 15%. The rate of conversion in this experiment was greater than literature reports for Ziegler-Natta catalyst. The polymer was pressed into films of almost perfect clarity.

The foregoing detailed description is given for clearness of understanding only, and no unnecessary limitations are to be inferred therefrom, as modifications within the scope of the invention will be obvious to those skilled in the art.