Process for reducing sheeting during polymerization of alpha-olefins

Method for reducing sheet formation in the polymerization of alpha-olefins.

For the production of alpha-olefiners under polymerization.

Conventional low density polyethylene has historically been polymerized in thick-walled autoclaves or tubular reactors at pressures as high as 50,000 psi (3447.4 bar) and temperatures of 300 ° C or higher. The molecular structure of high pressure, low density polyethylene (HP-LDPE) is very complex. Changes in the arrangement of their simple structural units are essentially limitless. HP-LDPE is characterized by a complex long-chain branched molecular structure. These long chain branches have a strong effect on the melt rheology of these resins. HP-LDPEs also show a spectrum of short-chain branches, generally 1 to 6 carbon atoms in length. These short chain branches cut off crystal formation and lower the resin density.

Only recently has technology been developed to produce low density polyethylene by fluidized bed techniques at low pressures and temperatures by copolymerizing ethylene with various alpha-olefins. These low pressure LDPE (LP-LDPE) resins generally show little, if any, long chain branching and are sometimes considered linear LDPE resins. They are short-chain branched with the type and amount of comonomer used during polymerization controlling the length and frequency of the branch.

As industry professionals well know, low pressure, high! Low density polyethylenes can now be conventionally obtained by a fluidized bed process using a variety of catalyst types to produce a wide variety of low density and high density products. The appropriate range of catalysts to be used depends in part on the type of end product desired, i.e., high density, low density, extrusion grade, film grade resins, and other criteria.

The different types of catalysts that can be used to produce polyethylenes in fluidized bed reactors can be broadly classified as follows:

Type I. Silyl chromium catalysts disclosed in U.S. Patent No. 3,324,101 to Baker and Carrick and U.S. Patent No. 3,324,095 to Carrick, Karapinka and Turbet. Silyl chromate catalysts are characterized by a group of the formula:

R 0 • H - -Si - 0 - Cr - 0- 'Π R 0 wherein R is a hydrocarbon group having 1 to 14 carbon atoms. Preferred silyl chromate catalysts are bis (triarylsilyl) chromates and most preferably bis (triphenylsilyl) chromate.

This catalyst is used with a support such as silica, alumina, thorium oxide, zirconia, and the like. Other carriers such as carbon black, microcrystalline cellulose, non-sulfonated ion exchange resins and the like can also be used.

Type II. Bis (cyclopentadienyl) chromium (11) compounds disclosed in U.S. Patent 3,879,368. These bis (cyclopentadienyl) chromium (II) compounds have the formula:

<img file="FI87358C_D0001.tif" />

where R^z and R may be the same or different C1 ... C20 hydrocarbon87358 radicals and n 'and n may be the same or different integers 0 ... 5. r'- and R-hydrocarbon radicals may be saturated or unsaturated, and may contain aliphatic, alicyclic and aromatic radicals, such as methyl, ethyl, propyl, butyl, pentyl, cyclopentyl, cyclohexyl, allyl, phenyl, and naphthyl radical t.

These catalysts are supported on a support as described above.

Type III. The catalysts described in U.S. Patent 4,011,382. These catalysts contain chromium and titanium in the form of oxides and, optionally, fluorine and a support. The catalysts contain, based on the total weight of the support and chromium, titanium and fluorine, about 0.05 to 3.0, and preferably about 0.2 to 1.0,% by weight of chromium (calculated as Cr), approx. 1.5 to 9.0, and preferably about 4.0 to 7.0% by weight of titanium (calculated as Tiina), and 0.0 to about 2.5, and preferably about 0.1 to 1.0% by weight of fluorine (calculated as F).

Chromium compounds that can be used for Type III catalysts include CrCl 3 or any chromium compound that is oxidizable to CrCl 3 under the activation conditions used. At least a portion of the chromium in the supported, activated catalyst must be in the hexavalent state. Chromium compounds other than CrCl 3 that can be used are described in U.S. Patent 2,825,721 and U.S. Patent 3,622,521 and include chromium (III) acetylacetonate, chromium (III) nitrate, chromium (III) acetate, chromium ( Chloride, chromium (III) sulphate and ammonium chromate.

Titanium compounds that can be used include all those that are oxidizable under the activation conditions used as TiCl 4 and include those described in U.S. Patent No. 3,622,521 and NL Patent Application No. 72-10881.

Fluorine compounds that may be used include HF or any fluorine compound that produces HF under the activation conditions used. Fluorine compounds other than HF that can be used are described in NL Patent Application 72-10881.

Inorganic oxide materials that can be used as a support in catalyst mixtures are porous substances with a large surface area, i.e. a surface area between about 50 ... 1000 m²/ g, and an average particle size of about 20 to 200 microns. Inorganic oxides that may be used include silica, alumina, thorium oxide, zirconia, and other comparable inorganic oxides, as well as mixtures of such oxides.

Type IV. The catalysts are described in U.S. Patent Application No. 892,325, filed March 31, 1978 to FJ Karolin et al. in the name of and which is entitled Manufacture of ethylene copolymers in a fluidized bed reactor and which has been transferred to the same applicant as this application. These catalysts comprise at least one titanium compound, at least one magnesium compound, at least one electron donor compound, at least one activator compound and at least one inert support.

The structure of the titanium compound is

Ti (OR)_aX_b wherein R is a C1 ... CJ.4 aliphatic or aromatic hydrocarbon radical, or COR ', wherein r' is a C1 ... C14 aliphatic or aromatic hydrocarbon radical; X is Cl, Br or I; a is 0 or 1; b is 2 ... 4; and a + b = 3 or 4.

Titanium compounds can be used alone or in combination, and would include TiCl 3, TiCl 4, Ti (OCH 3) Cl 3, Ti (OC 9 H 5) Cl 3, Ti (OCOCH 3) Cl 3, and Ti (OCOC 6 H 5) Cl 3.

The structure of the magnesium compound is

MgX₂ wherein X is Cl, Br or I. Such magnesium compounds may be used alone or in combination and would include MgCl₂, MgBr₂ and Mgl₂. Anhydrous MgCl₂ is a preferred magnesium compound.

The Titanium compound and the magnesium compound are commonly used in a form that facilitates their dissolution in the Electron donor compound.

An Electron donor compound is an organic compound which is liquid at 25 ° C and in which the Titanium compound and the magnesium compound are partially or completely soluble. Electron donor compounds are known per se or as Lewis bases.

Electron donor compounds would include compounds such as alkyl esters of aliphatic and aromatic carboxylic acids, aliphatic ethers, cyclic ethers and aliphatic ketones.

The Catalyst may be modified with a boron halide compound having the structure

BR_CX 3-c where R is an aliphatic or aromatic hydrocarbon Radical containing from 1 to 14 carbon atoms or R 1 where r 'is also an aliphatic or aromatic hydrocarbon Radical containing from 1 to 14 carbon atoms, x' is selected from the group consisting of Cl and Br, or mixtures thereof, and c is 0 or 1, wherein R is an aliphatic or aromatic carbon6 hydrogen and O, 1 or 2, wherein R is OR^Z.

Boron halide compounds may be used alone or in combination, and would include BCl 3, ΒΒΓβ, B (C 2 H 5) Cl 2 r B (OC₂H5) C1₂, B (OC₂B₅)₂C1, B (C₆B₅) Cl2, B (OC6H₅) Cl2, B (OC₆H5) C1₂, B (C6H13) Cl₂, B (OC6Hi₃) C1₂ and B (OC6H₅)₂Cl. Boron trichloride is a particularly preferred boron compound.

The structure of the activator compound is

Al (R)_cx '_dB_e wherein x 'is Cl or ORj; R3 and R are the same or different and are C1 to C14 saturated hydrocarbon radicals, d is 0 to 1.5, e is 1 or 0, and c + d + e = 3.

Such activator compounds may be used alone or in combination.

The carriers are solids, particulate substances and would include inorganic substances such as oxides and molecular sieves of silicon and aluminum, and organic substances such as olefin polymers, e.g. polyethylene.

In general, the above-mentioned catalysts are introduced together with the polymerizable substances into a reactor with an expansion section above the straight-side section. The recycle gas enters from the bottom of the reactor and passes upwards through a gas distribution plate to a fluidized bed located in the straight side of the apparatus. The gas distribution plate works to ensure proper gas distribution and to support the resin bed when the gas flow is stopped.

The gas leaving the fluidized bed carries resin particles with it. Most of these particles come off as the gas passes through the expansion section where its velocity decreases.

The operational difficulties associated with the utilization of ka-type T ... TIT in the reactors described above have been substantially eliminated, leading to the economical and efficient production of low pressure, low or high density polyethylene resins with a wide range of applications.

Catalyst type IV has been used to satisfy certain end-use applications of ethylene resins, such as film, injection molding and centrifugal molding applications. However, attempts to prepare certain ethylene resins using type IV catalysts supported on a porous silica substrate in certain fluidized bed reactors have not been entirely satisfactory from a practical commercial point of view. This is primarily due to the formation of plates in the reactor after a short period of operation. The sheets can be considered to be characterized by a composition of fused polymeric material.

The plates vary widely in size, but are similar in most respects. They are usually about 1/4 to 1/2 inch (about 6.35 to 12.7 mm) thick and are about one to five inches (about 25.4 to 127 mm) long. a few special cases even longer. They are about 3 inches wide by more than 18 inches (about 76.2 ... 457.2 mm). The sheets have a core consisting of a fused polymer oriented longitudinally of the sheets and their surfaces covered with a granular resin fused to the core. The edges of the sheets have a hairy appearance from the filaments of the fused polymer.

After a relatively short period of time during the polymerization, plates begin to appear in the reactor, and these plates clog the product removal systems, forcing the reactor to shut down.

Accordingly, it can be seen that there is currently a need to improve polymerization techniques, which is essential for the production of polyolefin products using catalysts in fluidized bed reactors.

titanium

It is therefore an object of the present invention to provide a process for the formation or removal of sheet formation which occurs in the low pressure fluidized bed polymerization of α] faolefins in which the compounds are used as a catalyst.

titanium bases are used to substantially reduce

Another object is to provide a process for the treatment of fluidized bed reactors used in the production of polyolefins using titanium-based catalysts or other catalysts leading to a similar sheet formation phenomenon.

These and other objects will become readily apparent from the following description taken in conjunction with the accompanying drawing, which generally illustrates a typical gas phase fluidized bed polymerization process for preparing high density and low density polyolefins.

Broadly speaking, the present invention provides an improvement in a process for polymerizing alpha-olefins in a fluidized bed reactor using titanium-based catalysts or other catalysts prone to induce sheet formation during said polymerization, said improvement comprising maintaining static charge at said reactor. which would otherwise cause disc formation.

The critical static voltage level for plate formation is a complex function of resin sintering temperature, operating temperature, fluidized bed resistance, resin particle size distribution, and recycle gas composition. The static voltage can be reduced by a variety of techniques, such as treating the reactor surface to reduce static electricity generation, injecting an antistatic agent to increase the electrical conductivity of the particle surface, thereby promoting the discharge of particles; installation of appropriate equipment connected to the reactor walls designed to promote electrical discharge by creating regions of high local field strength, and neutralizing charges from the resin bed by injecting or creating ion pairs, ions, or charged particles of opposite polarity.

A particularly preferred technique generally involves treating the reactor vessel prior to polymerization by introducing a chromium-containing compound into the reaction vessel in a non-reactive atmosphere.

With particular reference to the sole figure of the drawing, a conventional fluidized bed reaction system for polymerizing alpha-olefins comprises a reactor 10 consisting of a reaction zone 12 and a rate reduction zone 14.

Reaction zone 12 comprises a bed of growing polymer particles, the polymer particles formed, and a smaller number of catalyst particles fluidized by a continuous stream of polymerizable and modifying gaseous components in the form of a make-up feed and recycle gas through the reaction zone. In order to maintain a viable fluidized bed, the rate of mass gas flow through the bed is normally maintained above the minimum flow required for fluidization and preferably n. 1.5 to 10 times G 2 and more preferably about 3 to 6 times G 2_mf. The term G_mf is used as an approved form to denote the minimum gas flow required to achieve fluidization, CY Wen and YH Yu, Mechanics of Fluidization, Chemical Engineering Progress Symposium Series, Vol. 62, pp. 100. 111 (1966).

It is highly desirable that the bed always contain particles to prevent the formation of local hot spots and to carry and distribute the particulate catalyst throughout the reaction zone. At start-up, the reactor is usually charged with a bed of particulate polymer particles before the gas flow is started. These particles may be similar in nature to or different from the polymer to be formed. When different, they are removed from the desired polymer particles formed as a first product. Sometimes the fluidized bed of the desired polymer particles is preceded by a starting bed.

The suitable fluidized bed catalyst is preferably stored in tank 16 under a gas blanket, such as nitrogen or argon, which is inert to the stored catalyst.

Fluidization is achieved at a high rate of gas recirculation to and through the bed, typically on the order of about 50 times the refill gas feed rate. The fluidized bed has the general appearance of a dense mass of viable particles in the potential free vortex flow created by the flow of gas through the bed. The pressure drop through the bed is equal to or slightly greater than the mass of the bed divided by the cross-sectional area. It thus depends on the geometry of the reactor.

The make-up gas is fed to the bed at a rate equal to the rate at which the particulate polymer product is removed. The composition of the make-up gas is determined by a gas analyzer11a 18 located above the bed. The gas analyzer determines the composition of the recycled gas and the composition of the make-up gas is adjusted accordingly to maintain a substantially constant gaseous mixture in the reaction zone.

To ensure complete fluidization, the recycle gas and, if desired, some or all of the make-up gas is returned to the reactor at a position 20 below the bed. The gas distribution plate 22 placed above the return point ensures a proper distribution of the gas and also supports the resin bed when the gas flow is stopped.

7 3 58

That part of the gas stream which does not react in the bed forms a recirculation of the exhaust gas which is removed from the polymerization zone, preferably by conveying it to a deceleration zone 14 above the bed, where the entrained particles have the opportunity to fall back into the bed.

The recycle gas is then compressed in a compressor 24 and then passed through a heat exchanger 26 where its heat of reaction is removed before it is returned to the bed. When the heat of reaction is consistently removed, there does not appear to be a noticeable temperature gradient at the top of the bed. At the bottom of the bed, in a layer of about 6 to 12 inches (about 152.4 to 304.8 mm), there is a temperature gradient between the temperature of the feed gas and the temperature of the residual bed. Thus, it has been found that the bed operates almost immediately, controlling the temperature of the recycle gas above this bottom layer of the bed zone, making it consistent with the temperature of the residual bed, thus keeping it itself at a substantially constant temperature under constant conditions. The recycle is then returned to the reactor at its base 20 and through the fluidized bed manifold 22. Compressor 24 may also be located downstream of heat exchanger 26.

Hydrogen can be used as a chain transfer agent for conventional polymerization reactions of the type discussed herein. In the case where ethylene is used as the monomer, the hydrogen / ethylene ratio used ranges from about 0 to 2.0 moles of hydrogen per mole of monomer in the gas stream.

Any gas inert to the catalyst and starting materials may also be present in the gas stream. The cocatalyst is added to the gas recycle stream upstream of its connection to the reactor , such as from a dispenser 28 through line 30.

As is well known, it is necessary to operate a fluidized bed reactor below the sintering temperature of the polymer particles.

Thus, to ensure that no sintering occurs, operating temperatures below the sintering temperatures are desirable. For the production of ethylene polymers, the operating temperature is preferably used for the production of teas, the density of which is again about 75.. . 95 e has a preferred density of about 0, 91.. . 0, 94.

n. 90.. . 100 c prod. 0, 94.. . 0, 97, when for products with

Normally, the fluidized bed reactor is operated at pressures of about 1000 psi (about 68.95 bar), and is preferably operated at pressures of about 150.. . 350 psi (ca. 10, 34.... 24, 13 bar), operating at higher pressures in these regions, favoring heat transfer because the increase in pressure increases the heat capacity per unit volume of gas.

The catalyst is injected into the bed at the same rate as its consumption at point 32 above the manifold 22. A gas that is inert to the catalyst, such as nitrogen or argon, is used to introduce the catalyst into the bed. Catalyst injection at a point above the manifold 22 is an important feature. Because the catalysts normally used are highly active, spraying on the area under the manifold can cause polymerization to begin there and occasionally cause clogging of the manifold. Spraying into a viable bed, on the other hand, helps distribute the catalyst through the bed and tends to prevent the formation of local sites of high catalyst concentration that can lead to the formation of hot spots.

For a given set of operating conditions, the fluidized bed is maintained at a substantially constant height by removing a portion of the bed as a product at the same rate as the rate of formation of the particulate polymer product. Since the rate of heat generation is directly related to the formation of the product, the temperature rise of the gas in the reactor (the difference between the temperature of the incoming gas and the outgoing gas) determines the rate of polymer formation in the gas concentration.

The particulate polymer product is preferably removed at point 34 on or near the baffle 22. The particulate polymer product is conveniently and preferably removed through successive operation of the timed valves 36 and 38 defining the pair of separation zones 40. When valve 38 is closed, valve 36 is opened to send the gas plug and product to zone 40 between it and valve 36, which is then closed. Valve 38 is then opened to allow the product to enter the external intake zone, and upon release, valve 38 is then closed to await the next art intake operation of the product.

Finally, the fluidized bed reactor is equipped with an adequate aeration system to allow aeration of the bed during start-up and erection. The reactor does not require the use of stirring means and / or wall forming means.

The reactor vessel is normally constructed of carbon steel and designed for the operating conditions described above.

To better illustrate the problems associated with the use of type IV catalysts, reference is again made to the drawing. The titanium-based catalyst (type IV) is introduced into the reactor 10 at point 32. In the normal operations, the resins specified in the normal operation, after a short period of about 36 to 72 hours, begin to form plates in the reactor 10, at the reactor wall near the reactor wall and at a distance of about half from the diameter upwards from the base of the fluidized bed. The molten resin sheets begin to appear in the separation zone 40, quickly blocking the system, causing the reactor to be shut down. More typically, plate formation begins at a production equivalent of 6 to 10 times the weight of the resin bed in the reactor.

Many possible causes were investigated in an attempt to invent and eliminate plate formation. During the study, thermocouples were installed just inside the reactor walls at a height of 14 in the direction of 1/4 and 1/2 of the reactor diameter above the gas distribution plate. In conventional operations, the surface thermocouples show the same temperatures as the fluidized bed temperature. When plate formation occurs, these thermocouples show temperature deviations up to 20 ° C above the fluidized bed temperature, thus providing a reliable indication of the occurrence of plate formation. In addition, an electrostatic voltmeter 1 inch (25.4 mm) radially from the reactor wall and 1/2 of the reactor diameter was used to measure the voltage of a 1/2 inch (12.7 mm) spherical electrode placed in a fluidized bed above the gas distribution plate. The placement was chosen because plate formation was found to begin in a ring that varies 1/4 to 3/4 of the reactor diameter in the height direction above the base of the fluidized bed. As is well known for deep fluidized bed, this corresponds to the region of the lowest mixing intensity near the wall, that is, the zero zone where the movement of the particles near the wall generally changes from upside down. Possible causes investigated include factors affecting agitation in a fluidized bed, reactor operating conditions, catalyst and resin particle size, particle size distribution, and others. A correlation was observed between plate formation and static electricity charge on the resin particles near the reactor walls. When the static voltage level of the resin particles at certain points near the reactor wall in the fluidized bed reactor is low, the reactor operates normally and no plates are formed. When the static voltage level at those locations exceeds a critical level, uncontrolled plate formation occurs and the reactor must be shut down.

Surprisingly, no significant formation of sheet formation has occurred with any resin using type IV catalysts in reactors previously used with type II catalysts or in reactors using type I catalysts. ..III catalysts.

In addition, it was found that sheet formation could be substantially reduced and in some cases completely eliminated by adjusting the static voltage in the fluidized bed at a location close to the reactor walls below the critical level of sheet formation. This critical level for sheet formation is not a fixed value, but a complex variable-dependent function, including resin sintering temperature, operating temperature, fluidized bed resistance, resin particle size distribution, and recycle gas composition.

The critical voltage level Vc for sheet formation of ethylene homopolymers and ethylene butene copolymers is primarily a function of the resin sintering temperature, reactor bed temperature, and hydrogen concentration in the recycle gas. The relationship can be expressed as follows:

Vc = -8000 - 50 Ts + 90 [H2I + 150 To where Vc = voltage in volts below which no plate formation occurs;

Ts = sintering temperature of the resin in ° C under reactor operating conditions;

To = reactor temperature in ° C; and

[H2] = mole percent hydrogen in the recycle gas.

The resin temperature of the resin under the operating conditions of the reactor is the temperature at which the settled resin bed in contact with the recycle gas of the reactor used to make the resin sintered and forms agglomerates for one minute after attempting to re-fluidize the bed after leaving the bed to ice. The sintering temperature decreases by lowering the density of the resin, increasing the melt index, and increasing the amount of dissolved monomers.

The constants in the equation were determined from data collected during reactor operation with the reactor just starting to show a tendency for plate formation through surface thermocouple temperature deviations above bed temperature, the voltage indicated by the voltage probe depicting previous changes over time due to the random nature of the fluidized bed. The critical voltage, Vc, is expressed as a time-average voltage, the voltage measurements are difficult to interpret because a static charge is generated when the plate formed due to the static charge detaches from the reactor wall. In addition, the plate formation phenomenon can start as a very local phenomenon and spread further, obscuring the interpretation of voltage readings.

Although the mechanism of plate formation is not fully understood, it is believed that the static electricity generated in the fluidized bed determines the position of the resin particles. When the charge in the particles reaches a level where the electrostatic forces that try to keep the charged particle close to the reactor wall exceed the resistance forces in the bed that try to move the particle away from the wall, the polymer layer containing polymerizable resin particles forms a non-fluidized bed near the reactor wall. Heat removal from this layer is not sufficient to remove the heat of polymerization because the non-fluidized layer near the wall has less contact with the fluidizing gas than the particles in the fluidized portion of the bed. The heat of polymerization raises the temperature of the non-fluidized bed near the reactor wall until the particles melt and sinter. At this point, other particles from the fluidized bed adhere to the sintered layer and increase in size until it becomes loose from the reactor wall. Separation of the insulation from the conductor (plate reactor wall) is known to generate more static electricity, thus accelerating the subsequent plate formation.

The technique presents several different methods by which static voltage can be reduced or eliminated. These include (1) lowering the rate of charge generation, (2) electric charge

7 3 58 increasing the discharge rate, and (3) neutralizing the electric charge. Some methods suitable for use in a fluidized bed include (1) the use of an additive to increase particle conductivity, thus providing a path for discharge, (2) the installation of grounding devices in a fluidized bed to provide additional surface area for discharging electrical charges to the ground, (3) ionizing gas or particles 4) the use of radioactive sources for the production of radiation, which generates ions to neutralize the electrostatic charges in the particles. The application of these techniques to a commercial scale, fluidized bed, polymerization reactor may not be feasible or practical. No additive used shall act as a poison to the polymerisation catalyst and shall not adversely affect the quality of the product. Therefore, water, the most widely used additive to reduce staticness in particles, cannot be used because it is a potent catalyst poison. The installation of grounding devices may, in fact, generate an additional electrical charge because the friction of the resin particles on the metal surfaces creates electrostatic charges on the resin particles. The use of ion generators and radiation sources poses serious economies of scale. Ions generated by electric charge or radiation are attracted to the reactor walls and other grounded objects and travel only a limited distance before contact with the grounded object. Thus, the ions may not travel far enough from the site of ion generation to decompose the area of ​​the bed where plate formation occurs. The formation of ions inside the fluidized bed is severely limited by the suppressive effect of the cloud of charged particles formed around the ion generator. Thus, the number of ion generation sources required can be large, which increases the complexity and hazard of radiation sources or electric charge generators in or near a pressurized hydrocarbon reactor. The study found that an effective method of treating the walls of a reactor vessel to reduce the development of static charge involves a short period of operation of the reactor, i.e., two weeks using a chromium-containing catalyst (types I ... III) in which the chromium is in the 2- or 3-state its residence time in the reactor.

Surprisingly, however, it was also found that if the walls of the reactor vessel are treated before the polymerization with a chromium-containing compound in which chromium is present in the reactor at valence 2 or 3, sheet formation during polymerization is substantially reduced and in some cases completely eliminated.

Chromium-containing compounds that are possible for use in this invention are, as previously described, those in which chromium is present in the reactor at valence 2 or 3.

By way of illustrative example only, the following compounds are shown to be suitable for this invention:

Bis (cyclopentadienyl) chromium (II) compounds of the formula:

<img file="FI87358C_D0002.tif" />

<img file="FI87358C_D0003.tif" />

where R^z and R may be the same or different C1 to C20 hydrocarbon radicals, and n 'and n may be the same or different integers 0 to 5. R^z and R hydrocarbon radicals may be saturated or unsaturated and may contain aliphatic, alicyclic and aromatic radicals such as methyl, ethyl, propyl, butyl, pentyl, cyclopentyl, cyclohexyl, allyl, phenyl and naphthyl. Other specific compounds that are suitable include chromium (III) acetylacetonate, chromium (III) nitrate, chromium (TI) or chromium (111) acetate, chromium (II) or chromium (111) chloride, chromium (II) or chromium (111) bromide, chromium) or chromium (III) fluoride, chromium (JI) or chromium (111) sulphate and polymerization catalysts prepared from chromium compounds in which the chromium is in the +2 or +3 valence state.

Due to the excellent results obtained, bis (cyclopentadienyl) chromium (chromosene) is a preferred chromium-containing compound.

In general, chromium is polymerized in such a way that the chromium-containing compound is introduced into the reactor before and can be introduced at any such way that the surfaces of the reactor walls are in contact with it.

In a preferred technique, the chromium-containing compound is dissolved in a suitable solvent and introduced into the reactor under an inert or non-reactive atmosphere. The resin bed can be used to help disperse the chromium compound through the reactor.

Suitable solvents for this purpose include, but are not limited to, benzene, toluene, isopentane, hexane, and water. The choice and use of the solvent depends on the form of the chromium-containing compound and the chosen method of application. The function of the solvent is to transport and help disperse the chromium-containing compound. Suitable inert or non-reactive gases include, but are not limited to, nitrogen, carbon dioxide, methane, ethane, and air.

The amount of chromium compound used in the process should be sufficient to achieve the desired result and can be generally determined by one skilled in the art. However, usually at least 3.5 x 10 "⁷ pounds of chromium per square foot of surface to be treated (1.7 x 10 ~ 6 kmol / m²), preferably from 1.0 x 10® to about 5 x 10 ^ moles per square foot of surface to be treated (4.9 x 10 ~ 6_{; p}ta about 2.4 x 10^-4to kmol / m²) is preferred.

The polymers to which this invention is primarily directed and which cause plate formation problems such as those mentioned above in the presence of titanium catalysts are ethylene

7 3 58 linear homopolymers or higher molar percent (_&gt; 90%) ethylene and or more Cs. . . Ce copolymers. Ca. . . Ce linear alpha-olefins having a lower molar percentage (£ 10%) of one alpha-olefin should not contain any branching at any of their carbon atoms closer than the fourth carbon atom. Preferred C3 to Ca alpha-olefins are propylene, 1-butene, 1-pentene, 1-hexene, 4-methyl-1-pentene, 1-heptene and 1-octene. This disclosure is not intended to exclude the use of this invention with an alpha-olefin homopolymer and copolymer resins in which ethylene is not a monomer.

The density of homopolymers and copolymers varies from about 0, 97.. . 0, 91. The density of the copolymer at a given melt index level is primarily controlled by Cs. . . By the amount of Ca comonomer copolymerized with ethylene. The gradual addition of larger amounts of comonomer to the copolymers results in a gradual decrease in the density of the copolymer. Each different Ca ... C_ethe amount of comonomer required to achieve the same result varies from monomer to monomer under the same reaction conditions. In the absence of comonomer, ethylene would homopolymerize.

The melt index of a homopolymer or copolymer is a reflection of its molecular weight. Polymers with relatively high molecular weights have relatively high viscosities and low melt index.

Typically, to utilize the present invention to reduce sheet formation, a reactor vessel shown in Figure 1 that is sensitive to sheet formation problems in polymerizing the materials described above using type IV catalysts is partially filled with a granular polyethylene resin purged with a non-reactive gas such as nitrogen. reactive gas through the reactor at a rate which is above the minimum fluidization rate (Gmf) of the granular polyethylene and preferably 3 to 5

Gmf. It is to be understood that the use of a resin fluidized bed in said process is a convenience and is not essential to the process. Although the non-reactive gas is recycled, a chromium-containing compound, such as chromosene, either undiluted or, preferably, dissolved in an inert solvent such as toluene, is introduced into the reactor. The concentration of the chromium-containing chemical in the inert solvent is not critical to the method but can be selected by one skilled in the art to ensure that the chromium-containing chemical is completely dissolved in the solvent. For the preferred case, a solution containing 6 to 8% by weight of chromosene in toluene is typical. Mean 4.0 x 10^-5 a mole of chromium-containing chemical is injected into the reactor for each square foot of surface to be treated (2.0 χ 10^-4 kmol / m²). The unreacted gas is recycled to bring the chromium-containing chemical into contact with the metal surfaces of the system. The treatment is performed for a sufficient time to achieve the desired result, typically from several hours to several days. In other treatments, the chemical solution can be applied to metal surfaces by painting, spraying, or other application methods familiar to those skilled in the art. After treatment, the reactor is now ready to begin polymerization in the usual manner.

Having set out the general nature of the invention, the following examples illustrate some specific embodiments of the invention. It is to be understood, however, that the present invention is not limited to the examples, as the invention may be practiced using a variety of variations.

Examples 1 to 8 were performed in the fluidized bed reactor illustrated in Figure 1. The catalyst used was a Ziegler-type, titanium-based, porous silica-supported catalyst prepared as previously described as type iv. The cocatalyst used was triethylaluminum. The products made in the examples were copolymers of ethylene and 1-butene. Hydrogen was used as a chain transfer agent to control the polymer

7 358 melt index. The reactors of Examples 1 and 2 had not been used to produce polyethylene with any catalyst except those of the type previously described as Type IV.

Example 1

The fluidized bed reactor was started under operating conditions designed to produce a film-grade low density ethylene copolymer product having a density of 0.918, a melt index of 1.0, and an adhesion temperature of 104 ° C. The reaction was started by feeding the catalyst into a reactor preloaded with a bed of granular resin similar to the product to be made. The catalyst was a mixture of 5.5 parts of titanium tetrachloride, 8.5 parts of magnesium chloride and 14 parts of tetrahydrofuran doped with 100 parts of Davison grade 952 silica, dried at 8 ° C and treated with four parts of triethylaluminum before precipitation and activated with 35 parts tri-hexylaluminum after precipitation. Prior to starting the feed, the reactor and resin bed were brought to an operating temperature of 85 ° C, cleaned of impurities by circulating nitrogen through the resin bed. The ethylene, butene and hydrogen concentrations stabilized at 53, 24 and 11%, respectively. The cocatalyst was fed at a rate of 0.3 parts of triethylaluminum per part of catalyst.

Reactor start-up was normal. After the product had been prepared for 29 hours and corresponding to 6-1 / 2 times the weight of the fluidized bed, temperature deviations of 1 ... 2 ° C above the bed temperature were observed using 1/2 of the reactor diameter located just above the reactor wall above the gas distribution plate. thermocouples. Previous experience has shown that such temperature deviations are a positive indication that resin sheets are forming in a fluidized bed. To the same extent as the bed voltage (measured using an electrostatic voltmeter connected to a 1/2-inch (12.7 mm) ha 1-shaped spherical electrode located one inch (25.4 mm) from the reactor wall 1/2 of the reactor diameter above the gas distribution plate) rose from a reading of approximately +1500 ... + 2000 volts to read more than +5000 volts and then fell back to +2000 volts over a 3-minute period. Temperature and voltage deviations continued for about 12 hours and increased in frequency and magnitude. During this period, sheets of molten polyethylene resin began to appear in the resin product. The evidence of plate formation became more severe, that is, the temperature deviations grew as high as 20 ° C above the bed temperature and remained high for long periods of time, and the voltage deviations also became more frequent. The reactor was shut down due to the extent of plate formation.

Example 2

The fluidized bed reactor used in Example 1 was started up and used to produce a linear low density ethylene copolymer suitable for extrusion or centrifugal molding having a density of 0.934, a melt index of 5 and an adhesion temperature of 118 ° C. The reaction was initiated by feeding a catalyst similar to that of Example 1 except that it was activated with 28 parts of tri-hexylaluminum to a reactor preloaded with a bed of granular resin similar to the product to be made. Prior to starting the catalyst feed, the reactor and resin bed were brought to an operating temperature of 85 ° C, and nitrogen was purified from impurities. Concentrations of ethylene (52%), butene (14%) and hydrogen (21%) were introduced into the reactor. The cocatalyst triethylaluminum was fed at 0.3 parts per part of catalyst. The reactor was operated continuously for 48 hours and during this period it produced resin corresponding to 9 times the amount of resin contained in the bed. After this 48-hour period of steady-state operation, sheets of resin melted with the normal, granular product began to appear in the reactor. During this time, the voltages measured at 1/2 of the reactor diameter above the gas distribution plate averaged +2000 volts and varied

0 ... + 10000 volts, while the surface thermocouples at the same height showed deviations&gt; 15 ° C above the bed temperature. Two hours after the first plates in the reactor product were detected, it was necessary to stop feeding the catalyst and cocatalyst to the reactor to reduce the resin production rate because the plates blocked the resin removal system. One hour later, the catalyst and cocatalyst feeds were restarted. The formation of plates continued and after two hours the catalyst and cocatalyst feed was stopped again and the reaction was terminated by injecting carbon monoxide, the voltage during this time was&gt; +12000 volts and the thermocouple deviations continued until the poison was injected. In total, the reactor operated for 53 hours and produced 10-1 / 2 resin petit volume before the reaction was stopped due to plate formation.

Example 3

The reactor of Examples 1 and 2 was treated as follows: The treatment involved charging and cleaning the bed of granular resin and drying the bed with high purity nitrogen to a water vapor concentration of less than 10 ppmv. The bed was then fluidized by circulating nitrogen. Chromosome [biscyclopentadienyl) chromium] in toluene solution was injected into the bed. 4.3 x 10 “5 pounds of chromosene was added for each square foot of steel surface in the system (2.1 χ 10^-4 kmol / m²). The bed was heated to 92 ° C and the nitrogen was recycled for 24 hours. When the treatment was complete, the bed was cooled to 40 ° C and 20 standard cubic feet (ca. 0.566 m -1) of air was injected for each pound (ca. 0.454 kg) of chromosene in the system to oxidize the chromosene before removing the resin from the reactor.

A resin bed similar to that described in Example 1 was then placed in the treated reactor. The bed was brought to 85 ° C, purified, and ethylene, butene, hydrogen, and cocatalyst stabilized at the same concentrations in Example 1.

7358 before injection of the same catalyst as in Example 1. The reactor was started under operating conditions designed to produce a low quality polyethylene copolymer product having a density of 0.918, a melt index of 1.0 and a sintering temperature of 104 ° C as in Example 1. The reactor was run for 90 hours producing an average of 3 times as much product as in Example 1. was shut down for routine inspection and maintenance. No temperature deviations were observed and no resin sheets were formed. At the end of the run, the voltage measured near the wall in the height direction 1/2 of the reactor diameter above the gas distribution plate had stabilized at about -100 volts and no major voltage deviations were observed at any time during the run.

Example 4

The reactor used in Example 3 was subsequently charged with a similar resin bed as described in Example 2. The bed was heated to 90 ° C, purified and treated with ethylene (51%), butene (13%) and hydrogen (18%) and cocatalyst (0.3 parts per parts per catalyst) stabilized before catalyst injection. The reaction started smoothly and produced a linear low density polyethylene resin having a density of 0.934, a melt index of 5 and a sintering temperature of 118 ° C.

The reactor was operated continuously for 80 hours and produced resin corresponding to 20 times the weight of the resin bed before being converted to another product grade. Thermocouples placed near the reactor wall surface 1/4 and 1/2 of the reactor diameter above the distributor showed a few short (1 minute) temperature deviations. the voltage measured near the wall in the height direction 1/2 of the reactor diameter above the gas distribution plate averaged +1200 volts and showed voltage fluctuations from 0 to +8000 volts. Some resin particles, typically 1/4 to 1 inch (6.35 to 25.4 mm), with the appearance of sintered fine particles appeared in the product removal tank and accounted for &lt;0.01 percent of the resin produced. These did not reduce the production rate of the reaction system and did not damage the quality of the resin produced.

As can be seen above, the following data correspond to the Vc formula previously expressed:

Vc = -8000 -50 (sintering temperature) +90 (hydrogen concentration) +150 (operating temperature) = -8000 -.50 (118 ° C) +90 (18%) +150 (90 ° C) = +1220 volts

Examples 5 ... 8

Four runs were performed using the reactor of Examples 1 and 2 and the method for determining the critical voltage. Different ethylene, 1-butene copolymers and / or ethylene homopolymers were used in each run, as shown in Table I.

The critical voltages, Vc, were the voltage level measured near the reactor wall (half the reactor diameter above the distributor plate), with the reactor showing signs of plate formation onset (normally small deviations of the surface thermocouple above the bed temperature). Adhesion temperatures were evaluated from experiments in which the reaction was terminated, the bed was allowed to settle for 15 minutes, and then fluidized again.

The results are shown in Table I below.

TABLE

<img file="FI87358C_D0004.tif" />

<img file="FI87358C_D0005.tif" />

<img file="FI87358C_D0006.tif" />

si *

ΙΛ

UI in φ

in «M m

«O &lt;st® r ~«

As noted in Table I for Example 5, plate formation begins to appear at more than +1000 volts. In addition, it can be seen from Table I that the critical voltage is dependent on the resin sintering temperature, operating temperature, and hydrogen concentration in the recycle gas.