The present invention relates to a process for the gas-phase polymerization of olefins using a reactor having interconnected polymerisation zones. In particular, the invention relates to the use of a polymerization catalyst with a predefined particle size in a gas-phase polymerization reactor having interconnected polymerisation zones.

The development of olefin polymerization catalysts with high activity and selectivity, particularly of the Ziegler-Natta type and, more recently, of the metallocene type, has led to the widespread use on an industrial scale of processes in which the polymerization of olefins is carried out in a gaseous medium in the presence of a solid catalyst.

A widely used technology for gas-phase polymerization processes is the fluidized bed technology. In fluidized bed gas-phase processes, the polymer is confined in a vertical cylindrical zone (polymer bed). The reaction gases exiting the reactor are taken up by a compressor, cooled and sent back, together with make-up monomers and appropriate quantities of hydrogen, to the bottom of the polymer bed through a distribution plate. Entrainment of solid from the gas exiting the reactor is limited by an appropriate dimensioning of the upper part of the reactor (freeboard, i.e. the space between the upper bed surface and the gas exit point), where the gas velocity is reduced, and, in some designs, by the interposition of cyclones in the gases exit line. The flow rate of the circulating gaseous monomers is set so as to assure a velocity within an adequate range above the minimum fluidization velocity and below the "transport velocity". The heat of reaction is removed exclusively by cooling the circulating gas. The composition of the gas-phase controls the composition of the polymer, while the reaction kinetics is controlled by the addition of inert gases. The reactor is operated at constant pressure, normally in the range 1-4 MPa.

A significant contribution to the reliability of the fluidized bed reactor technology in the polymerization of α-olefms was made by the introduction of suitably pre-treated spheroidal catalyst of controlled dimensions and by the use of propane as a diluent component for the heat removal optimization.

A novel gas-phase polymerization process, which represents a gas-phase technology alternative to the fluidized bed reactor technology, as to the preparation of olefin polymers, is disclosed in EP-B- 1012195. The polymerization process is carried out in a gas-phase reactor having interconnected polymerization zones, where the growing polymer particles flow through a first polymerization zone (riser) under fast fluidization or transport conditions, leave said riser and enter a second polymerization zone (downcomer) through which they flow in a densifϊed form under the action of gravity, leave said downcomer and are reintroduced into the riser, thus establishing a circulation of polymer between the two polymerization zones. This polymerization process allows to obtain polymers with a broad molecular weight distribution by establishing different polymerisation conditions in the two interconnected polymerisation zones. This is achieved by introducing into the upper part of the downcomer a gas/liquid mixture, which evaporates and forms a barrier stream preventing or limiting the gases present in the riser from entering the downcomer. Accordingly, different polymerisation conditions can be maintained in the riser and in the downcomer.

The description of EP 1012195 is generic as regards the type and size of the catalyst particles to be used in the olefin polymerization: it is disclosed that suitable catalysts are those of controlled morphology, capable of giving polymers in the form of spheroidal particles having a mean dimension between 0.2 and 5 mm, preferably between 0.5 and 3 mm. It is known that when the polymerization of olefins is carried out by means of a gas-phase reactor a high-power compressor must be arranged on the recycle line to provide the continuous recycle of the reaction mixture from the top of the reactor to the bottom part of the reactor. In case of the reactor with two interconnected polymerization zones described in EP- 1012195, a high-power compressor is arranged on the recycle line in order to provide the gaseous stream with a pressure and velocity suitable to ensure fast fluidization conditions in the polymer bed present in the first polymerization zone (riser): therefore, the establishment of fast fluidisation conditions in the riser causes the recycle compressor to consume a high amount of energy, thus increasing the operating costs of the polymerization plant. Therefore, it would be desirable to decrease the power consumption required by the compressor of the recycle line, without modifying the polymer hold-up inside the first polymerization zone.

The Applicant has now found that when catalyst components of a decreased particle size are used in the above gas-phase polymerization reactor the power consumption required by the recycle compressor is remarkably reduced, thus decreasing the operating costs of the polymerization plant.

It is therefore an object of the present invention a process for the polymerisation of olefins in the presence of a solid catalyst component and a catalyst activator, the process being performed in a gas-phase reactor having interconnected polymerization zones, where the growing polymer particles flow upward through a first polymerization zone (riser) under fast fluidization conditions, leave said riser and enter a second polymerization zone (downcomer) through which they flow downward under the action of gravity, leave said downcomer and are reintroduced into the riser, thus establishing a polymer circulation between said riser and said downcomer, the process being characterized in that:

(a) said solid catalyst component has an average size P50 ranging from 20 μm to 65 μm, preferably from 40 to 60 μm;

(b) in said first polymerization zone the gas flows upward at a velocity ranging from 0.8 to 2.0 m/s.

In order to clarify the scope of present invention the average size P50 indicates a value of diameter, such that 50% of the total particles have a diameter lower than said value. The process of present invention is addressed to improve the operability of a gas-phase reactor having interconnected polymerization zones as described in EP-B- 1012195. Throughout the present description the first polymerisation zone, which comprises polymer particles flowing upwards under fast fluidisation conditions, is generally referred to as the "riser". The second polymerisation zone, which comprises polymer particles flowing downwards by gravity, is generally referred to as the "downcomer".

Fast fluidization conditions are established in the riser by feeding a gas mixture comprising one or more alpha-olefms at a velocity higher than the transport velocity of the polymer particles. The terms "transport velocity" and "fast fluidization conditions" are well known in the art; for a definition thereof, see, for example, "D. Geldart, Gas Fluidisation Technology, page 155 et seq., J. Wiley & Sons Ltd., 1986". The riser operates under fast fluidized bed conditions with gas superficial velocities higher than the average particles terminal velocities, so that the polymer particles are entrained upwards by the flow of the reacting monomers. A highly turbulent flow regime is established into the riser: this generates a good heat exchange coefficient between the single particles and the surrounding gas, and also ensures that the reaction temperature is kept reasonably constant along the reaction bed. Inside the second polymerization zone (downcomer) the polymer particles flow under the action of gravity in a densified form, so that high values of density of the solid (mass of polymer per volume of reactor) are achieved, said density of solid approaching the bulk density of the polymer. A "densified form" of the polymer implies that the ratio between the mass of polymer particles and the reactor volume is higher than 80% of the "poured bulk density" of the obtained polymer. Thus, for instance, in case of a polymer bulk density equal to 420 Kg/m , "densified conditions" of the polymer flow are satisfied if the polymer mass/reactor volume ratio is higher than 336 kg/m ³ . The "poured bulk density" of a polymer is a parameter well known to the person skilled in the art: it can be measured according to ASTM D 1895/69. In view of the above explanation, it is clear that in the downcomer the polymer flows downward in a plug flow and only small quantities of gas are entrained with the polymer particles.

The gas-phase polymerization process herewith described is not restricted to the use of any particular family of polymerization catalysts, with the proviso that the average size P50 of the catalyst particles is within the above indicated values. Any polymerization catalyst in form of a solid powder, either pre-polymerized or not, may be employed: Ziegler-Natta catalyst components, single site catalyst components and chromium-based catalyst components can be mentioned.

It is known that the morphology and average size of the polyolefϊn to be produced is strictly correlated to morphology and average size of the solid catalyst components fed to the polymerization reactor: this because the growing polymer particles replicate morphology and size of the catalyst particles. The polymerization process of the invention is based on the selection of catalyst components with an average size P50 ranging from 20 to 65 μm, preferably from 40 to 60 μm, so that the produced polyolefϊn has a particle size P50 in a range from 1000 μm to 2200 μm.

It has been found that the use of solid catalyst components of particle size significantly lower than the prior art allows to pursue a save of operating costs when establishing fast fluidization conditions inside the riser. As explained, fast fluidization conditions are established by feeding at the bottom of the riser a gas mixture at a velocity higher than the transport velocity of the polymer particles. Due to minor particle size of the polyolefin flowing upwards along the riser, the transport velocity is remarkably decreased with respect to the use of conventional catalyst particles of higher size, so that the riser may be operated at lower gas velocities: according to the invention, the upward velocity of the gas flowing in the riser ranges from 0.8 to 2.0 m/s, preferably from 1.2 to 1.8 m/s.

Accordingly, the same density of polymer in the riser (kg of polymer per m ³ of reactor) is achieved at significantly lower fluidization velocities in the riser. This means that the same polymer hold-up in the riser is obtained at lower gas velocities. This leads to a decrease of the flow rate of gas stream continuously recycled along the recycle line with a relevant energy save in the steps of compression of the gas recycle stream. The lower is the particle size of the produced polyolefϊn, the lower is the power consumption required by the recycle compressor placed on the gas recycle line: the compression stage becomes less burdensome and the size of the recycle compressor may be considerably decreased. The operative conditions of temperature and pressure in the polymerization process of the invention are those usually used in gas-phase catalytic polymerization processes. Therefore, in both riser and downcomer the temperature is generally comprised between 60 ⁰ C and 120 ⁰ C, while the pressure may range from 5 to 50 bar.

Preferred catalyst components having P50 from 20 to 65 μm exploited in the process of the invention are Ziegler-Natta catalyst components based on a titanium halide, preferably TiCU, supported on magnesium halide. Before the feeding to the polymerization said catalyst components have to be necessarily subjected to activation by contacting them with a catalyst activator, and optionally the activated catalyst particles may be successively subjected to a prepolymerization step.

The present invention will be now described in detail with reference to Figure 1 , which is only illustrative and not limitative of the scope of the present invention.

In Fig 1 Ziegler-Natta catalyst components of the claimed particle size are subjected to activation step, followed by prepolymerization in a loop reactor and successively the prepolymerized catalyst is fed to the riser of a gas-phase polymerization apparatus having two interconnected polymerization zones, as described in EP-B- 1012 195.

A Ziegler-Natta catalyst component 1 in form of a powder, an organo-aluminum compound 2 as the catalyst activator, and optionally an electron donor compound, are fed to a pre- contacting vessel 3. These components are contacted in continuous in the vessel 3 at a temperature ranging from 0 ⁰ C to 30 ⁰ C for an average residence time of 5-60 minutes.

Once activated, the catalyst particles are continuously withdrawn from the activation vessel 3 and are fed via line 4 to a prepolymerization reactor 5 to carry out the catalyst prepolymerization. The prepolymerization may be carried out in a liquid medium in whatever type of reactor: continuous stirred tank reactors (CSTR), as well as loop reactors can be used for contacting the olefin monomers with the catalyst particles. However, the prepolymerization treatment is preferably carried out in a liquid loop reactor.

The liquid medium of the prepolymerization step comprises liquid alpha-olefm monomer(s), optionally with the addition of an inert hydrocarbon solvent. Said hydrocarbon solvent can be either aromatic, such as toluene, or aliphatic, such as propane, hexane, heptane, isobutane, cyclohexane and 2,2,4-trimethylpentane. The amount of hydrocarbon solvent, if any, is lower than 40% by weight with respect to the amount of alpha-olefϊns, preferably lower than 20% by weight. Preferably the catalyst prepolymerization is carried out in the absence of inert hydrocarbon solvents.

The prepolymerization step is generally carried out at a low temperature, generally ranging from 20 to 50 ⁰ C, preferably from 25 to 40 ⁰ C. The average residence time in the prepolymerization step generally ranges from 2 to 40 minutes, preferably from 10 to 25 minutes: this parameter may be easily modified by increasing or decreasing the output of the polymeric slurry from the prepolymerizator. The polymerization degree ranges from 60 to 800 g per gram of solid catalyst component, preferably from 150 to 400 g per gram of solid catalyst component.

A slurry containing the prepolymerized catalyst particles is discharged from the loop reactor 5 and is fed via line 7 to the riser 8 of a gas-phase reactor having two interconnected polymerization zones. The gas-phase reactor comprises a riser 8 and a downcomer 9, wherein the polymer particles flow, respectively, upward under fast fluidization conditions along the direction of the arrow A and downward under the action of gravity along the direction of the arrow B. The riser 8 and the downcomer 9 are appropriately interconnected by the interconnection bends 10 and 11.

In said gas-phase reactor one or more olefin monomers are polymerized in the presence of hydrogen as the molecular weight regulator. To this aim, a gaseous mixture comprising the monomers, hydrogen and propane, as an inert diluent, is fed to the reactor through one or more lines 12, suitably placed at any point of the gas recycling system according to the knowledge of those skilled in art.

After flowing upwards in the riser 8 the polymer particles and the gaseous mixture leave the riser 8 and are conveyed to a solid/gas separation zone 13. This solid/gas separation 13 can be effected by using conventional separation means such as, for example, a centrifugal separator (cyclone) of the axial, spiral, helical or tangential type.

From the bottom of separation zone 13 the polymer particles enter the downcomer 9, while the gaseous mixture leaving the top of separation zone 13 is recycled to the riser 8 by means of the recycle line 14, equipped with means for the compression 15 and cooling 16 of the gaseous mixture.

After the compressor 15 the recycle gas is divided into two separated gaseous streams, the first one enters the connecting section 11 via line 17 and this stream favors the transfer of the polymer particles from the downcomer 9 to the riser 8. The second stream of recycle gas is fed via line 18 to the bottom of the riser 8 to establish fast fluidization conditions in this polymerization zone.

The polymer particles are discharged from the polymerization reactor via a discharge outlet 19, placed at the bottom part of the downcomer 9. According to another aspect of the invention, the gas mixture coming from the riser can be prevented from entering the downcomer by introducing a gas and/or liquid mixture of different composition through one or more introduction lines placed into the downcomer, preferably at a point close to the upper limit of the volume occupied by the densifϊed solid flowing downward along the downcomer.

The gas and/or liquid mixture of different composition to be fed into the downcomer can optionally be fed in partially or totally liquefied form. The liquefied gas mixture can also be sprinkled over the upper surface of the bed of densifϊed polymer particles; the evaporation of the liquid in the polymerisation zone will provide the required gas flow.

With reference to Figure 1, the above technical effect can be achieved by feeding a gas and/or liquid into the downcomer 9 through a line 20 placed at a suitable point of said downcomer 9, preferably in the upper part thereof. The gas and/or liquid mixture has a composition different from that of the gas mixture present in the riser 8. Said gas and/or liquid mixture partially or totally replaces the gas mixture entrained with the polymer particles entering the downcomer. The flow rate of this gas and/or liquid feed can be regulated so that a flow of gas counter-current to the flow of polymer particles is originated in the downcomer 9, particularly at the top thereof, thus acting as a barrier to the gas mixture coming from the riser 8, which is entrained among the polymer particles. It is also possible to place several feed lines 20 in the downcomer 9 at different heights, in order to better control the gas-phase composition throughout said downcomer. These additional feed lines 20 can be used to introduce gaseous or condensed monomers, optionally together with inert components. In case of adding condensed components, their evaporation in downcomer 9 contributes to remove the heat of the polymerisation reaction, thus allowing to control the temperature profile in the downcomer 9 in a reliable way. In order to control the solids recirculation between the two polymerization zones, and to provide greater resistance to backward gas flow in the section where the downcomer 9 leads into the connecting section 11, the section of the bottom of the downcomer 9 can conveniently converge into a restriction 21. Advantageously, adjustable mechanical valves can be employed, such as, for example, a throttle valve, such as a butterfly valve.

A stream of a gas, also denominated as the "dosing gas", may be fed into the lower part of the downcomer 9 by means of a line 22 placed above a suitable distance from the restriction 21. The dosing gas to be introduced through line 22 is conveniently taken from the recycle line 14, more precisely, downstream the compressor 15 and upstream the heat exchanger 16. The main function of said dosing gas is to control the solid recirculation flow from the downcomer 9 to the riser 8 through the restriction 21. The gas-phase polymerization process of the invention allows the preparation of a large number of olefin powders having an optimal particle size distribution with a low content of fines. Examples of polymers that can be obtained are: high-density poly ethylenes (HDPEs having relative densities higher than 0.940) including ethylene homopolymers and ethylene copolymers with α-olefms having 3 to 12 carbon atoms; linear poly ethylenes of low density (LLDPEs having relative densities lower than 0.940) and of very low density and ultra low density (VLDPEs and ULDPEs having relative densities lower than 0.920 down to 0.880) consisting of ethylene copolymers with one or more α-olefms having 3 to 12 carbon atoms; elastomeric terpolymers of ethylene and propylene with minor proportions of diene or elastomeric copolymers of ethylene and propylene with a content of units derived from ethylene of between about 30 and 70% by weight; isotactic polypropylene and crystalline copolymers of propylene and ethylene and/or other α-olefms having a content of units derived from propylene of more than 85% by weight; isotactic copolymers of propylene and α-olefms, such as 1-butene, with an α-olefϊn content of up to 30% by weight; impact-resistant propylene polymers obtained by sequential polymerisation of propylene and mixtures of propylene with ethylene containing up to 30% by weight of ethylene; atactic polypropylene and amorphous copolymers of propylene and ethylene and/or other α-olefms containing more than 70% by weight of units derived from propylene; polybutadiene and other polydiene rubbers.

The above mentioned bimodal polyethylene blends are particularly suitable to be subjected to injection molding for preparing shaped articles. The above mentioned polypropylene blends may be used to prepare films and fibers.

The polymerization process of the present invention can be carried out upstream or downstream other conventional polymerization technologies (either in a liquid-phase or a gas-phase) to give rise a sequential multistage polymerization process. For instance, a fluidised bed reactor can be used to prepare a first polymer component, which is successively fed to the gas-phase reactor of Fig. 1 to prepare a second and a third polymer component. Accordingly, an ethylene polymer endowed with a tri-modal molecular weight distribution can be obtained, as well as a polypropylene blend comprising three components having a different content in ethylene. As previously pointed out, the polymerization process of the invention may be performed by means of different types of solid catalyst components, with the proviso that the average size P50 of the catalyst particles fed to the reactor is within the above indicated values. Ziegler- Natta catalyst components, single site catalyst components, and chromium-based catalyst components may be employed in the present invention.

Preferred catalyst components of the invention are catalyst components comprising a magnesium halide, a titanium compound having at least a Ti-halogen bond, and optionally electron donor compounds. The magnesium halide is preferably MgCl ₂ in active form which is widely known from the patent literature as a support for Ziegler-Natta catalysts. Patents USP 4,298,718 and USP 4,495,338 were the first to describe the use of these compounds in Ziegler- Natta catalysis.

The preferred titanium compounds used in the catalyst component of the present invention are TiCU and TiCl ₃ ; furthermore, also Ti-haloalcoholates of formula Ti(OR) _n _ _y X _y can be used, where n is the valence of titanium, y is a number between 1 and n-1 X is halogen and R is a hydrocarbon radical having from 1 to 10 carbon atoms.

The preparation of the solid catalyst component having the above indicated particle size can be carried out according to several methods.

According to a preferred method, the solid catalyst component can be prepared by reacting a titanium compound of formula Ti(OR) _n _ _y X _y , where n is the valence of titanium and y is a number between 1 and n, preferably TiCU, with a magnesium chloride deriving from an adduct of suitably small particle size having formula MgCUpROH, where p is a number between 0.1 and 6, preferably from 2 to 3.5, and R is a hydrocarbon radical having 1-18 carbon atoms. The adduct can be prepared in suitable spherical form and small particle size by mixing alcohol and magnesium chloride in the presence of an inert hydrocarbon immiscible with the adduct, operating under stirring conditions at the melting temperature of the adduct (100-130 ⁰ C). Then, the emulsion is quickly quenched, thereby causing the solidification of the adduct in form of small spherical particles. A suitably small average particle size is obtained by providing to the system high energy shear stresses by way of maintaining in the mixer conditions such as to have a Reynolds (R _EM ) number 10,000 and 80,000, preferably between 30,000 and 80,000. The type of flow of a liquid inside a mixer is described by the above mentioned modified Reynolds number (ReM) which is defined by the formula Re=NL ² -d/η in which N is the number of revolutions of the stirrer per unit time, L is the characteristic length of the stirrer while d is the density of the emulsion and η is the dynamic viscosity. Due to what described above, it results that one of the methods to reduce the particle size of the adduct is that of increasing the number of revolutions of the stirrer.

According to the description of WO02/051544, particularly good results are obtained when high Reynolds numbers are maintained also during the transfer of the emulsion at the quenching stage and during the quench as well.

When providing sufficient energy to the system, it can be obtained spherical particles of the adduct that already have an average diameter sufficiently small able to generate a solid catalyst component of suitable size to obtain, upon reaction with the titanium compound a catalyst component with average particle size lower than 65 μm.

The so obtained adduct particles have average particle size determined with the method described in the characterization section below, ranging from 20 to 65 μm, preferably from

40 to 60 μm, and preferably a particle size distribution (SPAN) lower than 1.2, calculated

P90-P10 with the formula where, in a particle size distribution curve determined

P50 according to the same method, wherein P90 is the value of the diameter such that 90% of the total volume of particles have a diameter lower than that value; PlO is the value of the diameter such that 10% of the total volume of particles have a diameter lower than that value and P50 is the value of the diameter such that 50% of the total volume of particles have a diameter lower than that value.

The particle size distribution can be inherently narrowed by following the teaching of WO02/051544. However, in alternative to this method or to further narrow the SPAN, largest and/or finest fractions can be eliminated by appropriate means such as mechanical sieving and/or elutriation in a fluid stream.

The adduct particles can be directly reacted with Ti compound or it can be previously subjected to thermal controlled dealcoholation (80-130 ⁰ C) so as to obtain an adduct in which the number of moles of alcohol is generally lower than 3 preferably between 0.1 and 2.5. The reaction with the Ti compound can be carried out by suspending the adduct particles (dealcoholated or as such) in cold TiCU (generally 0 ⁰ C); the mixture is heated up to 80- 130 ⁰ C and kept at this temperature for 0.5-2 hours. The treatment with TiCU can be carried out one or more times. The electron donor compounds can be added during the treatment with TiCU. They can be added together in the same treatment with TiCU or separately in two or more treatments.

Particularly for the preparation crystalline polymers Of CH ₂ CHR olefins, where R is a Cl ClO hydrocarbon group, internal electron donor compounds can be supported on the MgCl ₂ . Typically, they can be selected among esters, ethers, amines, and ketones. In particular, the use of compounds belonging to 1,3-diethers, cyclic ethers, phthalates, benzoates, acetates and succinates is preferred.

The solid catalyst components are activated to form catalysts for the polymerization of olefins by reacting them with catalyst activators which are organo-aluminum compounds, preferably chosen among alkyl-Al compounds and in particular among the trialkyl aluminum compounds such as for example triethylaluminum, triisobutylaluminum, tri-n-butylaluminum, tri-n- hexylaluminum, tri-n-octylaluminum. It is also possible to use alkylaluminum halides, alkylaluminum hydrides or alkylaluminum sesquichlorides, such as AlEt ₂ Cl and Al ₂ EtSCIs, possibly in mixture with the above cited trialkylaluminums.

When it is desired to obtain a highly isotactic crystalline polypropylene, it is advisable to use, besides the electron-donor present in the solid catalytic component, an external electron-donor

(ED) added to the aluminium alkyl co-catalyst component or to the polymerization reactor.

Suitable external electron-donor include silanes, ethers, esters, amines, heterocyclic compounds and ketones.

A particular class of preferred external donor compounds is that of silanes of formula

Ra ⁵ Rb ⁶ Si(OR ⁷ )c, where a and b are integers from 0 to 2, c is an integer from 1 to 4 and the sum

(a+b+c) is 4; R ⁵ , R ⁶ , and R ⁷ , are alkyl, alkylen, cycloalkyl or aryl radicals with 1-18 carbon atoms optionally containing heteroatoms. Particularly preferred are the silicon compounds in which a is 1, b is 1, c is 2, at least one of R ⁵ and R ⁶ is selected from branched alkyl, cycloalkyl or aryl groups with 3-10 carbon atoms optionally containing heteroatoms and R ⁷ is a C ₁ -C ₁₀ alkyl group, in particular methyl. Examples of such preferred silicon compounds are methylcyclohexyldimethoxysilane, dicyclopentyldimethoxysilane.

Other useful solid catalyst components are Phillips catalysts, based on a chromium oxide supported on a refractory oxide, such as silica, and activated by a heat treatment. These catalysts consist of chromium (VI) trioxide, chemically fixed on silica gel. These catalysts are produced under oxidizing conditions by heating the silica gels that have been doped with chromium(III)salts (precursor or precatalyst). During this heat treatment, the chromium(III) oxidizes to chromium(VI), the chromium(VI) is fixed and the silica gel hydroxyl group is eliminated as water. When the final average size P50 of the chromium catalyst components is comprised within the range 20-65 μm, these catalyst components are suitable to be used in the process of present invention.

Other useful solid catalyst components are single-site catalysts supported on a carrier, preferably metallocene catalysts comprising:

(A) at least a transition metal compound containing at least one π bond; (B) at least a cocatalyst selected from an alumoxane or a compound able to form an alkylmetallocene cation.

A preferred class of the above compound (A) is a metallocene compound belonging to the following formula (I):

Cp(L) _q AMX _p (I) wherein

M is a transition metal belonging to group 4, 5 or to the lanthanide or actinide groups of the

Periodic Table of the Elements; preferably M is zirconium, titanium or hafnium; the substituents X, equal to or different from each other, are monoanionic sigma ligands selected from the group consisting of hydrogen, halogen, R ⁶ , OR ⁶ , OCOR ⁶ , SR ⁶ , NR ⁶ ₂ and

PR ⁶ ₂ , wherein R ⁶ is a hydrocarbon radical containing from 1 to 40 carbon atoms; preferably, the substituents X are selected from the group consisting of -Cl, -Br, -Me, -Et, -n-Bu, -sec-Bu, -

Ph, -Bz, -CH ₂ SiMe ₃ , -OEt, -OPr, -OBu, -OBz and -NMe ₂ ; p is an integer equal to the oxidation state of the metal M minus 2; q is 0 or 1 ; when q is 0 the bridge L is not present;

L is a divalent hydrocarbon moiety containing from 1 to 40 carbon atoms, optionally containing up to 5 silicon atoms, bridging Cp and A; preferably L is selected from Si(CHs) ₂ , SiPh ₂ , SiPhMe,

SiMe(SiMe ₃ ), CH ₂ , (CH ₂ ) ₂ , (CH ₂ ) ₃ or C(CH ₃ ) ₂ ;

Cp is a substituted or unsubstituted cyclopentadienyl group, optionally condensed to one or more substituted or unsubstituted, saturated, unsaturated or aromatic rings;

A has the same meaning of Cp or it is a NR ⁷ , -O, S, moiety wherein R ⁷ is a hydrocarbon radical containing from 1 to 40 carbon atoms.

The above catalyst system may be supported on an inert carrier having average size P50 in the range 20-65 μm, by depositing the compound (A), or the reaction product of the compound

(A) with the cocatalyst (B), or the cocatalyst (B) and successively the compound (A), on said inert carrier. The preferred carriers are particles of silica, alumina, magnesium halides, olefin polymers or prepolymers (i.e. polyethylenes, polypropylenes or styrene-divinylbenzene copolymers).