WO2025219537A1 - Process for propylene polymerization with optimized prepolymerization conditions - Google Patents

Abstract

The present invention relates to a process for the preparation of a propylene polymer, the process comprising the steps: a) prepolymerizing propylene in the presence of a metallocene catalyst in a first reactor, yielding a prepolymer (A), wherein the prepolymerization is conducted at defined conditions, b) transferring the prepolymer (A) to a second reactor, and c) polymerizing propylene and optionally one or more of C₂ and C₄ to C₁₀ alpha olefin comonomers in the presence of the prepolymer (A) in the second reactor, yielding a propylene polymer (B). The present invention also relates to a propylene polymer obtainable or obtained by the process.

Description

Process for propylene polymerization with optimized prepolymerization conditions The present invention relates to a process for the preparation of a propylene polymer in the presence of a metallocene catalyst at defined conditions in the prepolymerization step. Background Propylene polymers are widely used for different applications such as packaging applications or in the field of automotive. For their numerous applications, a good cost/performance ratio of the propylene polymers is important. Costs are inter alia dependent on the productivity of a process, i.e., the amount of polymer produced per amount of catalyst used. Thus, it is a constant need to improve the productivity of propylene polymerization processes. Costs can also be reduced by improving the process conditions in terms of the wear-off and maintenance of the reactor equipment and good operability of the polymerization process. The general performance of a polymer is connected with its physical properties. These properties are adjustable by the utilization of particular catalysts, reaction conditions and the kind and content of comonomers. However, it is difficult to accurately predict polymer properties and intense research is needed in order to establish the intended properties. In particular, in the field of propylene polymers prepared by metallocene catalysts, there is still a high need for improvement of polymer properties. It has been surprisingly found that the above-mentioned objects can be achieved by the process of the present invention. Summary of the invention The present invention relates to a process for the preparation of a propylene polymer, the process comprising the steps: a) prepolymerizing propylene in the presence of a metallocene catalyst in a first reactor, yielding a prepolymer (A), wherein the prepolymerization is conducted at the following conditions: (i) a temperature in the range of from 22 to 27 °C, (ii) a pressure in the range of from 3.5 to 6.5 MPa (gauge), (iii) a residence time in the range of from 15 to 35 min, (iv) an H2/C3 feed ratio in the range of from 0.02 to 0.25 mol/kmol, and (v) a feed of an antistatic agent in an amount in the range of from 0.5 to 10.0 ppm by weight of the propylene feed, b) transferring the prepolymer (A) to a second reactor, and c) polymerizing propylene and optionally one or more of C2 and C4 to C10 alpha olefin comonomers in the presence of the prepolymer (A) in the second reactor, yielding a propylene polymer (B). The present invention is based on the surprising finding that selective optimization of the process conditions in the prepolymerization step: temperature, pressure, residence time, H₂/C₃ feed ratio and the content of the antistatic agent, an improved process of preparing a propylene polymer can be provided. This process is characterized by an excellent operability during all polymerization stages, and thus an optimum reactor balance in the multistage process was reached. No instability during the polymerization process was observed such as plugging or fouling of the reactor. These improved reaction conditions have positive impact on the maintenance of the reactors. Further, the process is characterized by excellent productivity and use of the catalyst, i.e., high polymer amounts are obtained by relatively low amounts of catalyst used. This was particularly reached for a process comprising a prepolymerization step and one or two polymerization stage(s). Thus, the polymerization costs may be significantly reduced by the process of the present invention. The present invention also relates to a propylene polymer obtainable or obtained by the process. The final polymer is characterized by an excellent morphology. It has excellent average particle size with very low contents of agglomerates. The content of gel inclusions in the final polymer is very low. This improved morphology of the polymer facilitates its operability and enables a broader field of applications at lower costs. Detailed description of the invention The present invention relates to a process for the preparation of a propylene polymer. A “propylene polymer” as used herein denotes a propylene homopolymer and a propylene copolymer. Accordingly, the propylene polymer obtainable or obtained by the process of the present invention may be a propylene homopolymer or a propylene copolymer. Preferably, the propylene polymer obtainable or obtained by the process of the present invention is a propylene homopolymer. A “propylene homopolymer” as used herein denotes a propylene polymer comprising, based on the total weight of the propylene polymer, at least 99.0 wt.- %, preferably at least 99.2 wt.-% and more preferably at least 99.5 wt.-%, of units derived from propylene. Preferably, the propylene homopolymer comprises or substantially consists of only units derived from propylene, i.e., units derived from other monomers (such as ethylene and non-propylene alpha-olefin units) are below 1.0 wt.-%, more preferably below 0.5 wt.-%, based on the total weight of the propylene polymer. A “propylene copolymer” as used herein denotes a propylene polymer comprising, based on the total weight of the propylene polymer, at least 50.0 wt.- %, preferably at least 70.0 wt.-% and more preferably at least 90.0 wt.-%, of units derived from propylene, and additionally units derived from non-propylene comonomers. The non-propylene comonomers typically comprise at least one of ethylene and C₄ to C₁₀ alpha olefin comonomers. Preferably, the non-propylene comonomers are ethylene and/or C₄ (e.g., n-butene) and/or C₆ (e.g., n-hexene) comonomers, more preferably ethylene comonomers. The content of comonomer units in the propylene polymer can be determined by ¹³C{¹H} NMR spectroscopy as described herein. Catalyst The propylene polymer is prepared in the presence of a metallocene catalyst, preferably in the presence of at least one metallocene catalyst. Metallocene catalysts are well known in the art. A metallocene catalyst typically comprises a metallocene/activator reaction product impregnated in a porous support at maximum internal pore volume. The metallocene complex comprises a ligand which is typically bridged, a transition metal of group IVa to VIa, and an organoaluminum compound. The catalytic metal compound is typically a metal halide. The metallocene catalyst used according to the present invention is preferably a supported metallocene catalyst. Any suitable supported metallocene catalyst for the preparation of propylene polymers may be used. It is preferred that the metallocene catalyst comprises a metallocene complex, a co-catalyst system comprising a boron-containing co-catalyst and/or aluminoxane co-catalyst, and a support, preferably a support comprising or consisting of silica. Examples of suitable metallocene compounds are given, among others, in EP 629631, EP 629632, WO 00/26266, WO 02/002576, WO 02/002575, WO 99/12943, WO 98/40331, EP 776913, EP 1074557, WO 99/42497, EP 2402353, EP 2729479 and EP 2746289. The metallocene complex is ideally an organometallic compound (C) which comprises a transition metal (M) of Group 3 to 10 of the Periodic Table (IUPAC 2007) or of an actinide or lanthanide. The term “an organometallic compound (C)” denotes any metallocene compound of a transition metal which bears at least one organic (coordination) ligand and exhibits the catalytic activity alone or together with a cocatalyst. The transition metal compounds are well known in the art and are particularly compounds of metals from Group 3 to 10, e.g., Group 3 to 7, or 3 to 6, such as Group 4 to 6 of the Periodic Table, (IUPAC 2007), as well as lanthanides or actinides. In an embodiment the organometallic compound (C) has the following formula (I): (L)mRnMXq (I) wherein “M” is a transition metal (M) transition metal (M) of Group 3 to 10 of the Periodic Table (IUPAC 2007); each “X” is independently selected from monoanionic ligands, such as a σ-ligand; each “L” is independently selected from organic ligands which coordinate to the transition metal “M”; “R” is a bridging group linking said organic ligands (L); “m” is 1, 2 or 3, preferably 2; “n” is 0, 1 or 2, preferably 1; “q” is 1, 2 or 3, preferably 2; and m+q is equal to the valency of the transition metal (M). “M” is preferably selected from the group consisting of zirconium (Zr), hafnium (Hf) and titanium (Ti), more preferably selected from the group consisting of zirconium (Zr) and hafnium (Hf). In a preferred embodiment, each organic ligand (L) is independently selected from (a) a substituted or unsubstituted cyclopentadienyl or a bi- or multicyclic derivative of a cyclopentadienyl which optionally bear further substituents and/or one or more hetero ring atoms from a Group 13 to 16 of the Periodic Table (IUPAC); or (b) an acyclic η¹- to η ⁴- or η ⁶-ligand composed of atoms from Groups 13 to 16 of the Periodic Table, and in which the open chain ligand may be fused with one or two, preferably two, aromatic or non-aromatic rings and/or bear further substituents; or (c) a cyclic η ¹- to η ⁴- or η ⁶-, mono-, bi- or multidentate ligand composed of unsubstituted or substituted mono-, bi- or multicyclic ring systems selected from aromatic or non-aromatic or partially saturated ring systems, such ring systems containing optionally one or more heteroatoms selected from Groups 15 and 16 of the Periodic Table. Organometallic compounds (C), preferably used according to the present invention, have at least one organic ligand (L) belonging to the group (a) above. Such organometallic compounds are called metallocenes. More preferably, at least one of the organic ligands (L), preferably two organic ligands (L), is (are) selected from the group consisting of cyclopentadienyl, indenyl, tetrahydroindenyl and fluorenyl, which can be independently substituted or unsubstituted. Further, in case the organic ligands (L) are substituted, it is preferred that at least one organic ligand (L), preferably two organic ligands (L), comprise one or more substituents independently selected from C1 to C20 hydrocarbyl and silyl groups, which optionally contain one or more heteroatoms selected from Groups 14 to 16 of the Periodic Table and/or are optionally substituted by halogen atom(s), The term C1 to C20 hydrocarbyl group, whenever used herein, includes C1 to C20 alkyl, C₂ to C₂₀ alkenyl, C₂ to C₂₀ alkynyl, C₃ to C₂₀ cycloalkyl, C₃ to C₂₀ cycloalkenyl, C₆ to C₂₀ aryl, C₇ to C₂₀ alkylaryl or C₇ to C₂₀ arylalkyl groups or mixtures of these groups such as cycloalkyl substituted by alkyl. Further, two substituents, which can be same or different, attached to adjacent C-atoms of a ring of the ligands (L) can also taken together form a further mono- or multicyclic ring fused to the ring. Preferred hydrocarbyl groups are independently selected from linear or branched C₁ to C₁₀ alkyl groups, optionally containing one or more heteroatoms of Groups 14 to 16 of the Periodic Table, such as O, N or S, and substituted or unsubstituted C6 to C20 aryl groups. Linear or branched C₁ to C₁₀ alkyl groups, optionally containing one or more heteroatoms of Groups 14 to 16, are more preferably selected from methyl, ethyl, propyl, isopropyl, tertbutyl, isobutyl, C5-6 cycloalkyl, OR and SR, wherein R is a C₁ to C₁₀ alkyl group, C₆ to C₂₀ aryl groups are more preferably phenyl groups, optionally substituted with one or two C1 to C10 alkyl groups as defined above. A “σ-ligand” or “sigma-ligand” denotes a group bonded to the transition metal (M) via a sigma bond. Further, the ligands “X” are preferably independently selected from the group consisting of hydrogen, halogen, C1 to C20 alkyl, C1 to C20 alkoxy, C2 to C20 alkenyl, C₂ to C₂₀ alkynyl, C₃ to C₁₂ cycloalkyl, C₆ to C₂₀ aryl, C₆ to C₂₀ aryloxy, C₇ to C₂₀ arylalkyl, C₇ to C₂₀ arylalkenyl, -SR”, -PR”₃, -SiR”₃, -OSiR”₃ and -NR”₂, wherein each R” is independently selected from hydrogen, C1 to C20 alkyl, C2 to C20 alkenyl, C2 to C20 alkynyl, C3 to C12 cycloalkyl and C6 to C20 aryl. More preferably, “X” ligands are selected from halogen, C₁ to C₆ alkyl, C₅ to C₆ cycloalkyl, C1 to C6 alkoxy, phenyl and benzyl groups. The bridging group “R” may be a divalent bridge, preferably selected from –R’2C-, –R’₂C-CR’₂-, –R’₂Si-, -R’₂Si-Si R’₂-, -R’₂Ge-, wherein each R’ is independently a hydrogen atom, C1 to C20 alkyl, C2 to C10 cycloalkyl, tri(C1-C20-alkyl)silyl, C6- C20- aryl, C7- C20 arylalkyl and C7- C20-alkylaryl group. More preferably, the bridging group “R” is a divalent bridge selected from –R’₂C-, –R’₂Si-, wherein each R’ is independently selected from a hydrogen atom, C₁ to C20 alkyl, C2 to C10 cycloalkyl, C6- C20-aryl, C7- C20 arylalkyl and C7- C20-alkylaryl group. Another subgroup of the organometallic compounds (C) of formula (I) is known as non-metallocenes, wherein the transition metal (M), preferably a Group 4 to 6 transition metal, suitably Ti, Zr or Hf, has a coordination ligand other than a cyclopentadienyl ligand. The term “non-metallocene” denotes herein compounds, which bear no cyclopentadienyl ligands or fused derivatives thereof, but one or more non- cyclopentadienyl η-, or σ-, mono-, bi- or multidentate ligand(s). Such ligands can be chosen e.g. from the groups (b) and (c) as defined above and described e.g. in WO 01/70395, WO 97/10248, WO 99/41290, and WO 99/10353), and further in V. C. Gibson et al., in Angew. Chem. Int. Ed., engl., vol 38, 1999, pp 428447, the disclosures of which are incorporated herein by reference. However, the organometallic compound (C) used according to the present invention is preferably a metallocene as defined above. Metallocenes are described in numerous patents. In the following just a few examples are listed; EP 260130, WO 97/28170, WO 98/46616, WO 98/49208, WO 98/040331, WO 99/12981, WO 99/19335, WO 98/56831, WO 00/34341, WO 00/148034, EP 423101, EP 537130, WO 2002/02576, WO 2005/105863, WO 2006097497, WO 2007/116034, WO 2007/107448, WO 2009/027075, WO 2009/054832, WO 2012/001052 and EP 2532687, the disclosures of which are incorporated herein by reference. Further, metallocenes are described widely in academic and scientific articles. In a preferred embodiment the organometallic compound (C) has the following formula (Ia): (L)2RnMX2 (Ia) wherein “M” is Zr or Hf; each “X” is a σ-ligand; each “L” is an optionally substituted cyclopentadienyl, indenyl or tetrahydroindenyl; “R” is SiMe2 bridging group linking said organic ligands (L); “n” is 0 or 1, preferably 1. The metallocene catalyst complexes used in accordance with the present invention are preferably asymmetrical. Asymmetrical means simply that the two ligands forming the metallocene are different, that is, each ligand bears a set of substituents that are chemically different. The metallocene catalyst complexes used in accordance with the present invention are typically chiral, racemic bridged bisindenyl C1-symmetric metallocenes in their anti-configuration. Although such complexes are formally C₁-symmetric, the complexes ideally retain a pseudo-C₂-symmetry since they maintain C2-symmetry in close proximity of the metal center although not at the ligand periphery. By nature of their chemistry both anti and syn enantiomer pairs (in case of C₁-symmetric complexes) are formed during the synthesis of the complexes. For the purpose of this invention, racemic-anti means that the two indenyl ligands are oriented in opposite directions with respect to the cyclopentadienyl-metal-cyclopentadienyl plane, while racemic-syn means that the two indenyl ligands are oriented in the same direction with respect to the cyclopentadienyl-metal-cyclopentadienyl plane, as shown in the scheme below.

Racemic Anti Racemic Syn Formula (I), and any sub formulae, are intended to cover both syn- and anti- configurations. Preferred metallocene catalyst complexes are in the anti- configuration. The metallocene catalyst complexes used according to the present invention are generally employed as the racemic-anti isomers. Ideally, at least 95 mol-%, such as at least 98 mol-%, especially at least 99 mol-%, of the metallocene catalyst complex is in the racemic-anti isomeric form. More preferably, the metallocene catalyst is of formula (II) (as described in WO 2019/179959 A1): Formula (II) wherein “Mt” is Hf or Zr; each “X” is a sigma-ligand; each “R^1” are the same or different and are independently a CH2-R⁷ group, with R⁷ being H or linear or branched C1-6-alkyl group, C3-8 cycloalkyl group or C6-10 aryl group, each “R^2” is independently a –CH=, -CY=, –CH2-, -CHY- or -CY2- group, wherein Y is a C1-10 hydrocarbyl group and wherein n is 2-6, each “R³” and “R⁴” independently are the same or different and are independently a hydrogen, a linear or branched C1-C6-alkyl group, an OY group, a C7-20 arylalkyl, C7-20 alkylaryl group and C6-20 aryl group, whereby at least one R³ per phenyl group and at least one R⁴ is not hydrogen, and optionally two adjacent R³ or R⁴ groups can be part of a ring including the phenyl carbons to which they are bonded, “R⁵” is a linear or branched C₁-C₆-alkyl group, C_7-20 arylalkyl, C_7-20 alkylaryl group or C₆-C₂₀-aryl group, “R⁶” is a C(R⁸)3 group, with R⁸ being a linear or branched C1-C6 alkyl group, and each “R” is independently a C1-C20-hydrocarbyl. It is preferred that “Mt” is Zr. Preferably, each “X” is independently a hydrogen atom, a halogen atom, C1-6 alkoxy group or an R’ group, where R’ is a C1-6 alkyl, phenyl or benzyl group. Most preferably, X is chlorine, benzyl or a methyl group. Preferably, both X groups are the same. The most preferred options are two chlorides, two methyl or two benzyl groups, especially two chlorides. Each “R” is independently a C₁-C₂₀-hydrocarbyl, such as C₆-C₂₀-aryl, C₇-C₂₀- arylalkyl or C₇-C₂₀-alkylaryl. The term C_1-20 hydrocarbyl group also includes C_1- 20 alkyl, C2-20 alkenyl, C2-20 alkynyl, C3-20 cycloalkyl, C3-20 cycloalkenyl, C6-20 aryl groups, C7-20 alkylaryl groups or C7-20 arylalkyl groups and mixtures of these groups such as cycloalkyl substituted by alkyl. Unless otherwise stated, preferred C1-20 hydrocarbyl groups are C1-20 alkyl, C4-20 cycloalkyl, C5-20 cycloalkyl-alkyl groups, C7-20 alkylaryl groups, C7-20 arylalkyl groups or C6-20 aryl groups. Preferably, both R groups are the same. It is preferred that R is a C₁-C₁₀- hydrocarbyl or C6-C10-aryl group, such as methyl, ethyl, propyl, isopropyl, tertbutyl, isobutyl, C5-6-cycloalkyl, cyclohexylmethyl, phenyl or benzyl, more preferably both R are a C1-C6-alkyl, C3-8 cycloalkyl or C6-aryl group, such as a C₁-C₄-alkyl, C_5-6 cycloalkyl or C₆-aryl group and most preferably both R are methyl or one is methyl and the other cyclohexyl. Most preferably, the bridge is -Si(CH3)2-. Each “R¹” are the same or different and are independently a CH₂-R⁷ group, with R⁷ being H or linear or branched C1-6-alkyl group, like methyl, ethyl, n-propyl, i- propyl, n-butyl, i-butyl, sec-butyl and tert.-butyl, C3-8 cycloalkyl group (e.g. cyclohexyl) or C_6-10 aryl group (preferably phenyl). Preferably, both R¹ groups are the same and are a CH₂-R⁷ group, with R⁷ being H or linear or branched C1-C4-alkyl group, more preferably, both R¹ groups are the same and are a CH₂-R⁷ group, with R⁷ being H or a linear or branched C₁- C₃-alkyl group. Most preferably, both R¹ are methyl. Each “R²” is independently a –CH=– -CY=, –CH2-, -CHY- or -CY2- group, wherein Y is a C1-10 hydrocarbyl group, preferably a C1-4 hydrocarbyl group and where n is 2-6, preferably 3-4. Each “R³” and “R⁴” are independently the same or different and are independently hydrogen, a linear or branched C1-C6-alkyl group, an OY group or a C_7-20 arylalkyl, C_7-20 alkylaryl group or C_6-20 aryl group, preferably hydrogen, a linear or branched C₁-C₆-alkyl group or C_6-20 aryl groups, and optionally two adjacent R³ or R⁴ groups can be part of a ring including the phenyl carbons to which they are bonded. More preferably, R³ and R⁴ are hydrogen or a linear or branched C₁-C₄ alkyl group or an OY-group, wherein Y is a is a C_1-4 hydrocarbyl group. Even more preferably, each R³ and R⁴ are independently hydrogen, methyl, ethyl, isopropyl, tert-butyl or methoxy, especially hydrogen, methyl or tert-butyl_, wherein at least one R³ per phenyl group and at least one R⁴ is not hydrogen. Thus, preferably one or two R³ per phenyl group are not hydrogen, more preferably on both phenyl groups the R³ groups are the same, like 3´,5´-di-methyl or 4´- tert-butyl for both phenyl groups. For the indenyl moiety preferably one or two R⁴ on the phenyl group are not hydrogen, more preferably two R⁴ are not hydrogen and most preferably these two R⁴ are the same like 3´,5´-di-methyl or 3´,5´-di-tert-butyl_. R⁵ is a linear or branched C1-C6-alkyl group such as methyl, ethyl, n-propyl, i- propyl, n-butyl, i-butyl, sec-butyl and tert-butyl, C7-20 arylalkyl, C7-20 alkylaryl group or C₆-C₂₀ aryl group. R⁵ is a preferably a linear or branched C₁-C₆ alkyl group or C6-20 aryl group, more preferably a linear C1-C4 alkyl group, even more preferably a C1-C2 alkyl group and most preferably methyl. R⁶ is a C(R⁸)₃ group, with R⁸ being a linear or branched C₁-C₆ alkyl group. Each R is independently a C₁-C₂₀-hydrocarbyl, C₆-C₂₀-aryl, C₇-C₂₀-arylalkyl or C7-C20-alkylaryl. Preferably each R⁸ are the same or different with R⁸ being a linear or branched C₁-C₄-alkyl group, more preferably with R⁸ being the same and being a C₁-C₂-alkyl group. Most preferably, all R⁸ groups are methyl. In a further preferred embodiment, the organometallic compound (C) has the following formula (III) (as described in WO 2019/179959 A1): Formula (III) wherein “Mt” is Zr or Hf, preferably Zr; each “R³” and “R⁴” are independently the same or different and are independently hydrogen or a linear or branched C₁-C₆-alkyl group, whereby at least on R³ per phenyl group and at least one R⁴ is not hydrogen. Specific metallocene catalyst complexes include: rac-anti-dimethylsilanediyl[2- methyl-4,8-bis-(4’-tert-butylphenyl)-1,5,6,7-tetrahydro-s-indacen-1-yl][2-methyl- 4-(3’,5’-dimethyl-phenyl)-5-methoxy-6-tert-butylinden-1-yl] zirconium dichloride; rac-anti-dimethylsilanediyl[2-methyl-4,8-bis-(3’,5’-dimethylphenyl)-1,5,6,7- tetrahydro-s-indacen-1-yl][2-methyl-4-(3’,5’-dimethylphenyl)-5-methoxy-6-tert- butylinden-1-yl]zirconium dichloride; rac-anti-dimethylsilanediyl[2-methyl-4,8-bis-(3’,5’-dimethylphenyl)-1,5,6,7- tetrahydro-s-indacen-1-yl][2-methyl-4-(3’,5’-ditert-butyl-phenyl)-5-methoxy-6- tert-butylinden-1-yl] zirconium dichloride or their corresponding zirconium dimethyl analogues. MC-1 MC-2 MC-3 The ligands required to form the metallocene catalysts used according to the present invention can be synthesized by any process and the skilled organic chemist would be able to devise various synthetic protocols for the manufacture of the necessary ligand materials. WO 2007/116034, for example, discloses the necessary chemistry. Synthetic protocols can also generally be found in WO 2002/02576, WO 2011/135004, WO 2012/084961, WO 2012/001052, WO 2011/076780 and WO 2015/158790. To form an active catalytic species, typically a cocatalyst is also employed as is well known in the art. According to the present invention, a cocatalyst system comprising a boron containing cocatalyst and/or an aluminoxane cocatalyst may be used in combination with the above defined metallocene catalyst. The aluminoxane cocatalyst can be one of formula (IV): where “n” is usually from 6 to 20 and “R” is as defined below. Aluminoxanes are formed on partial hydrolysis of organoaluminum compounds, for example those of the formula AlR₃, AlR₂Y and Al₂R₃Y₃ where “R” can be, for example, C1-C10 alkyl, preferably C1-C5 alkyl, or C3-C10 cycloalkyl, C7-C12 arylalkyl or alkylaryl and/or phenyl or naphthyl, and where Y can be hydrogen, halogen, preferably chlorine or bromine, or C₁-C₁₀ alkoxy, preferably methoxy or ethoxy. The resulting oxygen-containing aluminoxanes are not in general pure compounds but mixtures of oligomers of the formula (III). The preferred aluminoxane is methylaluminoxane (MAO). Since the aluminoxanes used according to the present invention as cocatalysts are not, owing to their mode of preparation, pure compounds, the molarity of aluminoxane solutions hereinafter is based on their aluminum content. According to the present invention, also a boron containing cocatalyst can be used instead of the aluminoxane cocatalyst or the aluminoxane cocatalyst can be used in combination with a boron containing cocatalyst. It will be appreciated by the person skilled in the art that where boron based cocatalysts are employed, it is normal to pre-alkylate the complex by reaction thereof with an aluminum alkyl compound, such as TIBA. This procedure is well known and any suitable aluminum alkyl, e.g. Al(C₁-C₆ alkyl)₃ can be used. Preferred aluminum alkyl compounds are triethylaluminium, tri- isobutylaluminum, tri-isohexylaluminum, tri-n-octylaluminum and tri- isooctylaluminum. Alternatively, when a borate cocatalyst is used, the metallocene complex is in its alkylated version, that is for example a dimethyl or dibenzyl metallocene complex can be used. Boron based cocatalysts of interest include those of formula (V) BY3 (V) wherein “Y” is the same or different and is a hydrogen atom, an alkyl group of from 1 to about carbon atoms, an aryl group of from 6 to about 15 carbon atoms, alkylaryl, arylalkyl, haloalkyl or haloaryl each having from 1 to 10 carbon atoms in the alkyl radical and from 6-20 carbon atoms in the aryl radical or fluorine, chlorine, bromine or iodine. Preferred examples for Y are methyl, propyl, isopropyl, isobutyl or trifluoromethyl, unsaturated groups such as aryl or haloaryl like phenyl, tolyl, benzyl groups, p-fluorophenyl, 3,5- difluorophenyl, pentachlorophenyl, pentafluorophenyl, 3,4,5-trifluorophenyl and 3,5- di(trifluoromethyl)phenyl. Preferred options are trifluoroborane, triphenylborane, tris(4-fluorophenyl)borane, tris(3,5-difluorophenyl)borane, tris(4- fluoromethylphenyl)borane, tris(2,4,6-trifluorophenyl)borane, tris(penta- fluorophenyl)borane, tris(tolyl)borane, tris(3,5-dimethyl-phenyl)borane, tris(3,5- difluorophenyl)borane and/or tris (3,4,5-trifluorophenyl)borane. Particular preference is given to tris(pentafluorophenyl)borane. However, it is preferred that borates are used, i.e., compounds containing a borate 3⁺ ion. Such ionic cocatalysts preferably contain a non-coordinating anion such as tetrakis(pentafluorophenyl)borate and tetraphenylborate. Suitable counterions are protonated amine or aniline derivatives such as methylammonium, anilinium, dimethylammonium, diethylammonium, N- methylanilinium, diphenylammonium, N,N-dimethylanilinium, trimethylammonium, triethylammonium, tri-n- butylammonium, methyldiphenylammonium, pyridinium, p-bromo-N,N- dimethylanilinium or p-nitro-N,N-dimethylanilinium. Preferred ionic compounds which can be used according to the present invention include: triethylammoniumtetra(phenyl)borate, tributylammoniumtetra(phenyl)borate, trimethylammoniumtetra(tolyl)borate, tributylammoniumtetra(tolyl)borate, tributylammoniumtetra(pentafluorophenyl)borate, tripropylammoniumtetra- (dimethylphenyl)borate, tributylammoniumtetra(trifluoromethylphenyl)borate, tributylammoniumtetra(4-fluorophenyl)borate, N,N-dimethylcyclohexyl- ammoniumtetrakis(pentafluorophenyl)borate, N,N-dimethylbenzylammonium- tetrakis(pentafluorophenyl)borate, N,N-dimethylaniliniumtetra(phenyl)borate, N,N-diethylaniliniumtetra(phenyl)borate, N,N-dimethylaniliniumtetrakis(penta- fluorophenyl)borate, N,N-di(propyl)ammoniumtetrakis(pentafluorophenyl)borate, di(cyclohexyl)ammoniumtetrakist(pentafluorophenyl)borate, triphenyl- phosphoniumtetrakis(phenyl)borate, triethylphosphoniumtetrakis(phenyl)borate, diphenylphosphoniumtetrakis(phenyl)borate, tri(methylphenyl)phosphonium- tetrakis(phenyl)borate, tri(dimethylphenyl)phosphoniumtetrakis(phenyl)borate, triphenylcarbeniumtetrakis(pentafluorophenyl)borate, and/or ferrocenium- tetrakis(pentafluorophenyl)borate. Preference is given to triphenylcarbeniumtetrakis(pentafluorophenyl) borate, N,N-dimethylcyclohexylammoniumtetrakis(pentafluorophenyl)borate and N,N- dimethylbenzylammoniumtetrakis(pentafluorophenyl)borate. It has been surprisingly found that certain boron cocatalysts are especially preferred. Preferred borates for use in the present invention therefore comprise the trityl ion. Thus, the use of N,N-dimethylammonium-tetrakispentafluorophenylborate and Ph3CB(PhF5)4 and analogues therefore are especially preferred. According to the present invention, the preferred cocatalysts are aluminoxanes, more preferably methylaluminoxanes, combinations of aluminoxanes with Al- alkyls, boron or borate cocatalysts, and combination of aluminoxanes with boron- based cocatalysts. Suitable amounts of cocatalyst are well known to the person skilled in the art. The molar ratio of boron to the metal ion of the metallocene may be in the range of from 0.5:1 to 10:1 mol/mol, preferably from 1:1 to 10:1 mol/mol, especially 1:1 to 5:1 mol/mol. The molar ratio of Al in the aluminoxane to the metal ion of the metallocene may be in the range of from 1:1 to 2000:1 mol/mol, preferably from 10:1 to 1000:1 mol/mol, and more preferably from 50:1 to 900:1 mol/mol, and most preferably from 600:1 to 800:1 mol/mol. The metallocene catalyst used in the polymerization process of the present invention may be in supported form. The preferable support used comprises, preferably consists of, silica. The person skilled in the art is aware of the procedures required to support a metallocene catalyst. Especially preferably, the support is a porous material so that the complex may be loaded into the pores of the support, e.g., using a process analogous to those described in WO 94/14856, WO 95/12622 and WO 2006/097497. The average particle size of the support can be typically in the range of from 10 to 100 μm. However, it has turned out that certain advantages can be obtained if the support has an average particle size in the range of from 15 to 80 μm, preferably from 18 to 50 μm. The particle size distribution of the support is described in the following. The support preferably has a D₅₀ in the range of from 10 to 80 µm, preferably from 18 to 50 µm. Furthermore, the support preferably has a D10 in the range of from 5 to 30 µm and a D₉₀ in the range of from 30 and 90 µm. Preferably, the support has a SPAN value in the range of from 0.1 to 1.1, preferably from 0.3 to 1.0. The average particle size of the metallocene catalyst is preferably of from 20 to 50 µm, more preferably from 25 to 45 µm, and most preferably from 30 to 40 μm. The particle size distribution of the metallocene catalyst is described in the following. The metallocene catalyst preferably has a D50 in the range of from 30 to 80 µm, preferably from 32 to 50 µm and most preferably from 34 to 40 µm. Furthermore, the metallocene catalyst preferably has a D₁₀ of at most 29 µm, more preferably in the range of from 15 to 29 µm, more preferably from 20 to 28 µm, and most preferably from 25 to 27 µm. The metallocene catalyst preferably has a D90 of at least 45 µm, more preferably in the range of from 45 to 70 µm and most preferably from 40 to 60 µm. The average pore size of the support can be in the range of from 10 to 100 nm, preferably from 20 to 50 nm and the pore volume in the range of from 1 to 3 ml/g, preferably from 1.5 to 2.5 ml/g. BET surface area of silica support materials are determined according to ASTM D3663 and porosity parameters based on BJH according to ASTM D4641. Examples of suitable support materials are, for instance, ES757 produced and marketed by PQ Corporation, Sylopol 948 produced and marketed by Grace or SUNSPERA DM-L-303 silica produced by AGC Si-Tech Co. Supports can be optionally calcined prior to the use in catalyst preparation in order to reach optimal silanol group content. All or part of the preparation steps can be done in a continuous manner. The formed catalyst preferably has good stability/kinetics in terms of longevity of reaction, high activity and the catalysts enable low ash contents. Process The process according to the present invention is conducted in the presence of a metallocenes catalyst, which is preferably a supported metallocene catalyst. Supported metallocene catalysts, such as silica supported catalysts, may show very complex polymerization behavior and the polymerization process can be subdivided into several phases. During the first minutes of polymerization, the catalyst activity can reach high values resulting in an uncontrollable fragmentation process which in turn can lead to decrease of catalyst activity due to increased external mass and heat transport phenomena. More particularly, the exothermic heat generated due to the polymerization reaction cannot be properly dissipated, thus, leading to local particle overheating (i.e., the temperature difference between the surface of the growing polymer particles and the bulk temperature attains high values). Therefore, the polymer that is produced on the surface of the growing polymer particle becomes sticky, thus leading to increased risk of particle agglomeration, with concomitant effects on process performance and reactor operability. A polymerization kinetic as described above requires a new design of the pre- polymerization process with respect to temperature, monomer concentration and residence time. In the initial phase (first activity peak) the temperature and monomer concentration must be low to avoid overheating of the formed polymer and to avoid formation of agglomerates. In the second phase, the monomer concentration and the temperature are typically higher to accelerate the catalyst fragmentation process. According to the present invention, the polymerization is carried out in the presence of hydrogen in the prepolymerization step. Hydrogen is typically employed to help control polymer properties, such as polymer molecular weight. Sometimes, hydrogen is not added to the polymerization process (e.g., in step a) or c)). The skilled artisan will appreciate, however, that hydrogen may be generated during the polymerization process. Thus, if hydrogen is present in the polymerization reaction, it may originate from hydrogen which has been added as a reactant and/or hydrogen produced as a side product during polymerization. According to the present invention, a H2 feed is used at least in the prepolymerization step a) and preferably also in the polymerization step c). According to the present invention, step a) of the process is a prepolymerization step. The purpose of the prepolymerization is to polymerize a small amount of polymer onto the catalyst at a low temperature and/or a low monomer concentration. By prepolymerization it is possible to improve the performance of the catalyst in slurry and/or to modify the properties of the final polymer. The prepolymerization step is typically conducted as slurry polymerization. Use of a prepolymerization step generally provides the advantage of minimizing leaching of catalyst components. In step c) of the process of the present invention, a lower catalyst activity helps in controlling the polymerization reaction more accurately, allowing an easier fine tuning the process parameters and hence operating closer to the ideal conditions. Otherwise, for a high catalyst activity, the reaction is more difficult to control and the flexibility in adjusting the operating conditions is reduced. Preferably, step a) is conducted as slurry polymerization, more preferably as bulk polymerization. “Bulk polymerization” denotes a polymerization process wherein the polymerization is conducted in a liquid monomer essentially in the absence of an inert diluent. However, as is known in the art, the monomers used in commercial production are never pure but always contain aliphatic hydrocarbons as impurities. For instance, the propylene monomer may contain up to 5 % of propane as an impurity. As propylene is consumed in the reaction and also recycled from the reaction effluent back to the polymerization, the inert components tend to accumulate, and thus the reaction medium may comprise up to 40 wt.-% of other compounds than the monomer. It is to be understood, however, that such a polymerization process is still within the meaning of “bulk polymerization”, as defined above. The slurry polymerization, preferably the bulk polymerization, may be conducted in any known reactor used for slurry polymerization. Such reactors include a continuous stirred tank reactor and a loop reactor. It is especially preferred to conduct the prepolymerization in a loop reactor. Preferably, the first reactor is a loop reactor. In such reactors, the slurry is circulated with a high velocity along a closed pipe by using a circulation pump. Loop reactors are generally known in the art and examples are given, for instance, in US-A-4582816, US-A-3405109, US-A-3324093, EP-A-479186 and US-A-5391654. It is thus preferred to conduct the prepolymerization as a slurry polymerization in a loop reactor. The amount of monomer is typically such that from 0.1 to 1000 g of monomer per one gram of solid catalyst component is polymerized in the prepolymerization step. As the person skilled in the art knows, the catalyst particles recovered from a continuous prepolymerization reactor do not all contain the same amount of prepolymer. Instead, each particle has its own characteristic amount, which depends on the residence time of that particle in the prepolymerization reactor. As some particles remain in the reactor for a relatively long time and some for a relatively short time, then also the amount of prepolymer on different particles is different and some individual particles may contain an amount of prepolymer which is outside the above limits. However, the average amount of prepolymer on the catalyst typically is within the limits specified above. In step a) of the process, a prepolymer (A) is produced in a first reactor in the presence of propylene at specifically defined prepolymerization conditions, namely: (i) a temperature in the range of from 22 to 27 °C, preferably from 23 to 26 °C, more preferably from 24 to 25 °C, (ii) a pressure in the range of from 3.5 to 6.5 MPa (gauge), preferably from 4.5 to 6.5 MPa (gauge), more preferably from 5.0 to 6.0 MPa (gauge), (iii) a residence time in the range of from 15 to 35 min, preferably from 20 to 35 min, more preferably from 25 to 30 min, (iv) an H₂/C₃ feed ratio in the range of from 0.02 to 0.25 mol/kmol, preferably from 0.02 to 0.20 mol/kmol, more preferably from 0.05 to 0.20 mol/kmol, and (v) a feed of an antistatic agent in an amount in the range of from 0.5 to 10.0 ppm, preferably from 2.5 to 6.5 ppm, more preferably from 3.5 to 5.5 ppm, by weight of the propylene feed. It has been surprisingly found that a combination of the process conditions (i) to (v) as defined above in the prepolymerization step enables an optimum operability of a propylene polymerization process, wherein process disturbance by plugging and fouling of the reactor could be avoided. A temperature of ca.20 °C – as often used in prepolymerization steps – did not enable to conduct a process without disturbance. An increase of temperature to at least 22 °C allowed for better process operability. The process can be further improved by further enhancing the temperature. The increase of the prepolymerization temperature has further positive effect on the prepolymerization degree as depicted in Figure 4. Moreover, in order to avoid disturbance during the polymerization process, the presence of an antistatic agent in the prepolymerization step is required. However, as the increasing content of the antistatic agent has a negative impact on the prepolymerization degree (as shown in Figure 3), an antistatic agent feed in an amount in the range of from 0.5 to 10.0 ppm by weight of the propylene feed has proven particularly suitable to reach the best balance of properties. It is to be understood that in step a), propylene is provided as a feed. The antistatic agent is also provided as a feed, and its amount is based on the amount (by weight) of the propylene. Further improvement can be reached by using an antistatic agent feed in an amount in the range of from 2.5 to 6.5 ppm, such as 3.5 to 5.5 ppm, by weight of the propylene feed. Generally, any antistatic agent known in the art for propylene polymerization processes may be used. Typically, an antistatic agent may be selected from the group consisting of glycerol monostearate (GMS), hydrogenated tallow fatty acids, blends of GMS and hydrogenated tallow fatty acids, polypyrrole, carbon nanotubes, carbon black, carbon fiber, graphite fiber, fatty acid alkanolamide, blends of GMS and fatty acid dialkanolamide, anionic hydrocarbyl sulfonate, N,N-bis(2- hydroxyethyl)alkoxypropylbetaine, lauric diethanol amide, alkyl-bis(2- hydroxyethyl)amine, quaternary ammonium compound, polyetheresteramide, tertiary amine, blends of GMS and tertiary amine, stearyldiethanolamine, alkyl phosphate, ethoxylated secondary alcohols, glycerol distearate, blends of GMS and glycerol distearate, sodium alkyl sulfonate, neutralized alcohol phosphate, polyethylene glycol monolaurate, sodium alkyl sulfonate, sorbitan monolaurate, sorbitan monooleate, ethoxylated sorbitan monolaurate, oleochemical derivatives, polyethylene glycol, polypropylene glycol, polyoxyethylene sorbitan monooleate, polyoxyethylene sorbitan monolaurate, polyethylene oxide, blends of polyethylene oxide and sorbitan monolaurate, blends of polyethylene oxide and sorbitan monooleate, polyether block amides, and any combinations thereof. Preferably, an antistatic agent is selected from the group consisting of glycerol monostearate, sorbitan monolaurate, sorbitan monooleate, polyethylene glycol, polypropylene glycol, and combinations thereof, more preferably the antistatic agent is sorbitan monooleate. It has been found important to use hydrogen in the prepolymerization process, in particular in order to obtain satisfying polymer morphology. An H2/C3 feed ratio in the range of from 0.02 to 0.25 mol/kmol has been proven useful to reduce the content of polymer agglomerates in the final propylene polymer. By lowering the H2/C3 feed ratio, the content of polymer agglomerates can be further reduced. For example, a minimum of agglomerates can be reached at an H₂/C₃ feed ratio in the range of from 0.02 to 0.10 mol/kmol. Generally, employment of hydrogen in the prepolymerization step may be further used to control the molecular weight of the prepolymer. Finally, it has been found that optimum process conditions can be reached when a pressure in the range of from 3.5 to 6.5 MPa (gauge) and a residence time in the range of from 15 to 35 min are used in the prepolymerization step. As it is well known in the art the average residence time τ can be calculated from equation (1) below: τ _{ൌ ^ೃ} ொ_^ equation (1) wherein V_R is the volume of the reaction space (in case of a loop reactor, the volume of the reactor; in case of the fluidized bed reactor, the volume of the fluidized bed) Q_o is the volumetric flow rate of the product stream (including the polymer product and the fluid reaction mixture). Thus, the selection of conditions (i) to (v) as defined above in the prepolymerization step enables an optimum operability of a propylene polymerization process. In some embodiments, step a) of the process is conducted at the following prepolymerization conditions: (i) a temperature in the range of from 23 to 26 °C, (ii) a pressure in the range of from 4.5 to 6.5 MPa (gauge), (iii) a residence time in the range of from 20 to 35 min, (iv) an H2/C3 feed ratio in the range of from 0.02 to 0.20 mol/kmol, and (v) a feed of an antistatic agent in an amount in the range of from 2.5 to 6.5 ppm by weight of the propylene feed. In some embodiments, step a) of the process is conducted at the following prepolymerization conditions: (i) a temperature in the range of from 24 to 25 °C, (ii) a pressure in the range of from 5.0 to 6.0 MPa (gauge), (iii) a residence time in the range of from 25 to 30 min, (iv) an H2/C3 feed ratio in the range of from 0.05 to 0.20 mol/kmol, and (v) a feed of an antistatic agent in an amount in the range of from 3.5 to 5.5 ppm by weight of the propylene feed. In one particularly optimized embodiment of the process, step a) of the process is conducted at the following prepolymerization conditions: (i) a temperature at about 25 °C, (ii) a pressure in the range of from 5.0 to 6.0 MPa (gauge), (iii) a residence time in the range of from 25 to 30 min, (iv) an H₂/C₃ feed ratio in the range of from 0.05 to 0.10 mol/kmol, and (v) a feed of an antistatic agent in an amount in the range of from 3.5 to 5.5 ppm by weight of the propylene feed. The process according to the present invention shows good results in terms of production rate and productivity in all stages of the process. This, in turn, has advantages in cost efficiency of the process. Typically, the production rate for the first reactor is in the range of from 0.5 to 1.0 kg PP/h, and/or the productivity/prepolymerization degree is in the range of from 130 to 250 g PP/g cat, preferably from 150 to 200 g PP/g cat. (PP=polypropylene, cat= catalyst) The prepolymerization degree in the first reactor is calculated by dividing the production rate in the first reactor by the catalyst feed to the first reactor. The catalyst feed in the first reactor is typically in the range of from 2.0 to 6.0 g cat/h, preferably from 3.0 to 5.0 g cat/h. Preferably, in step a) the ratio of a feed of metallocene catalyst to the feed of propylene is in the range of from 0.02 to 0.15 g/kg, more preferably from 0.05 to 0.12 g/kg. The content of the prepolymer (A) produced in step a) typically is in the range of from 1.0 to 10.0 wt.-%, preferably from 1.0 to 7.0 wt.-%, based on the total weight of the final propylene polymer (preferably being propylene polymer (B) or (C)). Depending on the number of reaction stages used in the process, in some embodiments, the content of the prepolymer (A) is in the range of from 1.0 to 5.0 wt.-%, preferably from 1.5 to 4.0 wt.-% and more preferably from 2.0 to 3.0 wt.-%, based on the total weight of the final propylene polymer (e.g., for a process with three polymerization stages, incl. the prepolymerization stage, such as propylene polymer (C)). In other embodiments, the content of the prepolymer (A) is in the range of from 2.0 to 10.0 wt.-%, preferably from 2.5 to 8.0 wt.-% and more preferably from 3.0 to 6.0 wt.-%, based on the total weight of the final propylene polymer (e.g., for a process with two polymerization stages, incl. the prepolymerization stage, such as propylene polymer (B)). It is also preferred that in step a) only propylene is used as the monomer for prepolymerization, i.e., that no additional comonomers are added. It has been found that the presence of other comonomers may have negative impact on the stability of the process rection. It is further preferred that no external co-catalyst is added. Generally, a co- catalyst may be used as a scavenger to eliminate catalyst poison. However, external addition of co-catalyst may lead to leaching of catalyst components from the catalyst and fouling of reactor walls. The process of the present invention does not require external addition of a co-catalyst for good polymerization results. In step b) of the process, the prepolymer (A) obtained in step a) is transferred to a second reactor, preferably directly transferred to a second reactor. Preferably, the prepolymer is transferred to the second reactor in the form of a slurry. The slurry will generally comprise the prepolymer, unreacted monomer and the metallocene catalyst. The slurry may be withdrawn from the first reactor either continuously or intermittently. A preferred way of intermittent withdrawal is the use of settling legs where slurry is allowed to concentrate before withdrawing a batch of the concentrated slurry from the reactor. The use of settling legs is disclosed, among others, in US-A-3374211, US-A-3242150 and EP-A-1310295. Continuous withdrawal is disclosed, among others, in EP 891990, EP 1415999, EP 1591460 and WO 2007/025640. The continuous withdrawal is advantageously combined with a suitable concentration method, as disclosed in EP 1310295 and EP 1591460. It is preferred to withdraw the slurry from the first reactor continuously. Preferably, the prepolymer (A) withdrawn from the first reactor is directly transferred to the second reactor to produce the propylene polymer (B) in step c). In the sense of the present invention, “directly” means that the slurry is introduced from the first reactor into the second reactor without any separation step in-between. Similar to step a), step c) in the second reactor is preferably conducted as a slurry polymerization, the slurry polymerization preferably being a bulk polymerization. Preferably, the second reactor is a loop reactor. Preferably, the temperature in the second reactor (preferably a loop reactor) in step c) is in the range of from 55 to 100 °C, more preferably from 60 to 90 °C and most preferably from 65 to 80 °C. Generally, in step c), the pressure in the second reactor (preferably a loop reactor) may be in the range of from 0.1 to 10.0 MPa (gauge). Preferably, the pressure is in the range of from 3.5 to 6.5 MPa (gauge) and more preferably from 5.0 to 6.0 MPa (gauge). Generally, the average residence time in the second reactor (preferably a loop reactor) in step c) may be in the range of from 5 to 120 min. Preferably, the average residence time is in the range of from 15 to 45 min and more preferably from 20 to 40 min. Preferably, the metallocene catalyst used in step a) is present in the second reactor during the polymerization in step c). This is accomplished by transferring, preferably via the slurry, the metallocene catalyst used in step a) into the second reactor. If needed, fresh metallocene catalyst may be added into the second reactor in step c), however, this is not preferred. In step c) of the process, propylene polymer (B) is produced in the presence of propylene and optionally one or more of C₂ and C₄ to C₁₀ alpha olefin comonomers. Accordingly, in step c) of the process, either a propylene homopolymer or a propylene copolymer with one or more of C2 and C4 to C10 alpha olefin comonomers is prepared. In some preferred embodiments of the process, step c) comprises polymerizing only propylene (i.e., in the absence of other comonomers) in the presence of the prepolymer (A) in the second reactor, yielding a propylene polymer (B). The propylene polymer (B) in these embodiments is a propylene homopolymer. In some embodiments of the process, step c) comprises polymerizing propylene and one or more of C2 and C4 to C10 alpha olefin comonomers in the presence of the prepolymer (A) in the second reactor, yielding a propylene polymer (B). The propylene polymer (B) in these embodiments is a propylene copolymer, comprising one or more of C2 and C4 to C10 alpha olefin comonomers. Preferably, the comonomers are ethylene and/or C₄ (e.g., n-butene) and/or C₆ (e.g., n- hexene) comonomers, more preferably ethylene comonomers. Hydrogen is typically introduced into the polymerization stage in step c) for controlling the melt flow rate (e.g., MFR2) of the propylene polymer. The amount of hydrogen needed to reach the desired melt flow rate (e.g., MFR₂) depends on the catalyst used and the polymerization conditions, as will be appreciated by the skilled artisan. In one embodiment, the melt flow rate MFR₂ of the propylene polymer (B), determined according to ISO 1133 at 2.16 kg and 230 °C, is in the range of from 1 to 50 g/10 min, preferably from 5 to 20 g/10 min, such as 5 to 12 g/10 min. It has been surprisingly found that the morphology of the propylene polymer can be further improved by controlling the hydrogen feed. An H₂/C₃ feed ratio in the range of from 0.20 to 0.40 mol/kmol in the second reactor has been proven particularly suitable. In some embodiments, the production rate for the second reactor is in the range of from 20 to 50.0 kg PP/h, and/or the productivity is in the range of from 2 to 50 kg PP/g cat, preferably from 4 to 20 kg PP/g cat, such as from 5 to 10 kg PP/g cat. (PP=polypropylene, cat= catalyst) The production rate is suitably controlled with the catalyst feed rate. It is also possible to influence the production rate by suitable selection of the monomer concentration. The desired monomer concentration can then be achieved by suitably adjusting the propylene feed rate. In the second reactor in step c), a propylene polymer (B) is produced. The propylene polymer (B) may represent the final propylene polymer produced by the process according of the present invention or may be a fraction of the final propylene polymer, in case additional reaction stages are used. Accordingly, in some embodiments, the propylene polymer (B) represents the final propylene polymer produced by the process according of the present invention. In these embodiments, the propylene polymer (B) may be either used directly or may undergo further preparation steps such as compounding with standard polymer additives as described below and/or extrusion, pulverization, pelletization etc. In some other embodiments, the process according to the present invention comprises the further steps: d) transferring the propylene polymer (B) to a third reactor, and e) polymerizing propylene and optionally one or more of C₂ and C₄ to C₁₀ alpha olefin comonomers in the presence of the propylene polymer (B) in the third reactor, yielding a propylene polymer (C). Accordingly, in these embodiments, in step d) of the process, the propylene polymer (B) obtained in step c) is transferred to a third reactor (preferably directly transferred), wherein an additional propylene polymer fraction is produced, yielding a propylene polymer (C). In step e) of the process, propylene polymer (C) is produced in the presence of propylene and optionally one or more of C2 and C4 to C10 alpha olefin comonomers. Accordingly, in step e) of the process, either a propylene homopolymer fraction or a propylene copolymer fraction with one or more of C₂ and C4 to C10 alpha olefin comonomers is prepared. In some embodiments of the process, step e) comprises polymerizing only propylene (i.e., in the absence of other comonomers) in the presence of the propylene polymer (B) in the third reactor, yielding a propylene polymer (C). The propylene polymer fraction prepared in the third reactor of these embodiments is a propylene homopolymer fraction. Based on the nature of the propylene polymer (B), the final propylene polymer (C) is either a propylene homopolymer or a propylene copolymer. In some embodiments of the process, step e) comprises polymerizing propylene and one or more of C2 and C4 to C10 alpha olefin comonomers in the presence of the propylene polymer (B) in the third reactor, yielding a propylene polymer (C). The propylene polymer (C) in these embodiments is a propylene copolymer, comprising one or more of C2 and C4 to C10 alpha olefin comonomers. Preferably, the comonomers are ethylene and/or C₄ (e.g., n-butene) and/or C₆ (e.g., n- hexene) comonomers, more preferably ethylene comonomers. According to the present invention, in the second reactor, a propylene homopolymer or a propylene copolymer may be produced as propylene polymer (B). Similarly, in the third reactor, a propylene homopolymer or a propylene copolymer may be produced, wherein propylene polymer (C) is a propylene homopolymer or a propylene copolymer. In some embodiments, in both the second and third reactor, a propylene copolymer is produced, wherein the propylene polymer (C) is a propylene copolymer. In some embodiments, in the second reactor, a propylene copolymer is produced and in the third reactor, a propylene homopolymer is produced, wherein the propylene polymer (C) is a propylene copolymer. In some embodiments, in the second reactor, a propylene homopolymer is produced and in the third reactor, a propylene copolymer is produced, wherein the propylene polymer (C) is a propylene copolymer. In some preferred embodiments, in both the second and third reactor, a propylene homopolymer is produced, wherein the propylene polymer (C) is a propylene homopolymer. The copolymers are as defined above and preferably comprise ethylene and/or C₄ (e.g., n-butene) and/or C6 (e.g., n-hexene) comonomers, more preferably ethylene comonomers. Preferably, the third reactor in step d) is a gas phase reactor, more preferably a fluidized bed gas phase reactor. Generally, any suitable gas phase reactor known in the art may be used. For gas phase reactors, in addition to the monomer (i.e., propylene) feed and the hydrogen feed (if needed), generally a non-reactive gas such as nitrogen or low boiling point hydrocarbons (such as propane) is fed to the reactor. Preferably, the temperature in the third reactor (preferably a gas phase reactor) in step e) is in the range of from 55 to 100 °C, more preferably from 60 to 90 °C and most preferably from 70 to 85 °C. Generally, in step e), the pressure in the third reactor (preferably a gas phase reactor) may be in the range of from 0.1 to 5.0 MPa (gauge). Preferably, the pressure is in the range of from 1.0 to 4.0 MPa (gauge) and more preferably from 2.0 to 3.5 MPa (gauge). Generally, the average residence time in the third reactor (preferably a gas phase reactor) in step e) may be in the range of from 1 to 10 h. Preferably, the average residence time is in the range of from 1 to 5 h and more preferably from 2 to 4 h. Total productivity of the second and third reactor may be even higher than that of the second reactor alone. In some embodiments, the productivity is in the range of from 5.0 to 100.0 kg PP/g cat, preferably from 8.0 to 20.0 kg PP/g cat, such as from 10.5 to 18.0 kg PP/g cat and/or the total production rate for the second and third reactor is in the range of from 46 to 70.0 kg PP/h. (PP=polypropylene, cat= catalyst) In one embodiment, the melt flow rate MFR2 of the propylene polymer (C), determined according to ISO 1133 at 2.16 kg and 230 °C, is in the range of from 1 to 50 g/10 min, preferably from 5 to 20 g/10 min, such as from 5 to 13 g/10 min. By the process according to the present invention, an excellent balance between the polymerization stages can be reached, as depicted in Figure 2. A polymer split between the second and third reactor of 40:60 to 60:40 is considered important to obtain a good reactor balance for processes employing three reactors, incl. the prepolymerization reactor. A good reactor balance facilitates stable reactions with a good production rate. In the third reactor in step e), a propylene polymer (C) is produced. The propylene polymer (C) may represent the final propylene polymer produced by the process according of the present invention or may be a fraction of the final propylene polymer, in case still additional reaction stages are used. Accordingly, in some embodiments, the propylene polymer (C) represents the final propylene polymer produced by the process according of the present invention. In these embodiments, the propylene polymer (C) may be either used directly or may undergo further preparation steps such as compounding with standard polymer additives as described below and/or extrusion, pulverization, pelletization etc. In some other embodiments, the process according to the present invention comprises the further steps: f) transferring the propylene polymer (C) to a fourth reactor, g) polymerizing propylene and optionally one or more of C₂ and C₄ to C₁₀ alpha olefin comonomers in the presence of the propylene polymer (C) in the fourth reactor, yielding a propylene polymer (D). In these embodiments, the propylene polymer (D) may represent the final propylene polymer produced by the process according of the present invention or may be a fraction of the final propylene polymer, in case still additional reaction stages are used. For example, further polymerization in a fifth reactor and optionally sixth reactor may be carried out. Any of the fourth, fifth and sixth reactor is preferably a gas phase reactor and the polymerization is preferably conducted at the conditions generally described for the third reactor above. Preferably, the final polymer is obtained as propylene polymer (B) or propylene polymer (C) after the second or third reactor, respectively. Preferably, the final propylene polymer is a propylene homopolymer. A suitable process is the above-identified slurry-gas phase process, such as developed by Borealis and known as the Borstar® technology. In this respect, reference is made to the EP applications EP 887379A1 and EP 517868A1. Propylene polymer The process according to the present invention produces a propylene polymer. Accordingly, the present invention also relates to a propylene polymer obtainable or obtained by the process. The propylene polymer may contain standard polymer additives. These polymer additives typically form less than 5.0 wt.%, such as less than 2.0 wt.% of the polymer material. Additives, such as antioxidants, phosphites, cling additives, pigments, colorants, fillers, antistatic agent, processing aids, clarifiers and the like may thus be added during the polymerization process. These additives are well known in the industry and their use will be familiar to the artisan. Any additives which are present may be added as an isolated raw material or in a mixture with a carrier polymer, i.e., in so-called master batch. The propylene polymer is characterized by excellent morphology. The content of agglomerated particles (e.g., particles with sizes of more than 2.0 mm) is reduced. This improved morphology of the polymer facilitates its operability and enables a broader field of applications at lower costs. In preferred embodiments, the content of particles having a particle size of more than 2.0 mm and less than or equal to 4.0 mm is less than 5 wt.-%, preferably less than 3 wt.-%, and/or the content of particles having a particle size of more than 4.0 mm is less than 1 wt.-%, based on the total weight of all particles of the propylene polymer and determined as described herein below. Further, the content of gel inclusions (indicated as gel index) in the propylene polymer is also reduced. In preferred embodiments, the gel index is less than 10, preferably less than 8, determined as described herein below. Preferably, the propylene polymer is a propylene homopolymer. Figures The present invention is further illustrated by the figures, wherein Figure 1 shows the content of particles of > 2 mm in the propylene polymer obtained after the GPR step, Figure 2 shows the polymer split between the loop reactor (second reactor) and the GPR (third reactor), Figure 3 shows the correlation between the prepolymerization degree and the content of the antistatic agent in the prepolymerization step, and Figure 4 shows the correlation between the prepolymerization degree and the temperature in the prepolymerization step. Experimental part Measurement methods The measurement methods apply to the parameters mentioned in the detailed description of the invention above and the examples below. a) Melt Flow Rate The melt flow rate (MFR) is determined according to ISO 1133 and is indicated in g/10 min. The MFR is an indication of the melt viscosity of the polymer. The MFR is determined at 230 °C for PP. The load under which the melt flow rate is determined is usually indicated as a subscript, for instance MFR2 is measured under 2.16 kg load (condition D). b) Particle size and particle size distribution The particle size distribution of the catalyst and catalyst support was determined using laser diffraction measurements by Coulter LS 200. The particle size and particle size distribution is a measure for the size of the particles. The D-values (D₁₀ (or d10), D₅₀ (or d50) and D₉₀ (or d90)) represent the intercepts for 10%, 50% and 90% of the cumulative mass of sample. The D-values can be thought of as the diameter of the sphere which divides the sample’s mass into a specified percentage when the particles are arranged on an ascending mass basis. For example, the D₁₀ is the diameter at which 10% of the sample's mass is comprised of particles with a diameter less than this value. The D50 is the diameter of the particle where 50% of a sample's mass is smaller than and 50% of a sample's mass is larger than this value. The D₉₀ is the diameter at which 90% of the sample's mass is comprised of particles with a diameter less than this value. The D50 value is also called median particle size. From laser diffraction measurements according to ISO 13320 the volumetric D-values are obtained, based on the volume distribution. The distribution width or span of the particle size distribution is calculated from the D-values D₁₀, D₅₀ and D₉₀ according to equation (3): Span = (D90-D10)/D50 equation (3) Particle size determination of the polymer was carried out via digital image analysis by Camsizer P4 from the Company Retsch Technology GmbH. The measuring principle is a dynamic image analysis according to ISO 13322- 2. c) Differential scanning calorimetry (DSC) Differential scanning calorimetry (DSC) analysis, melting temperature (T_m) and melt enthalpy (Hm), crystallization temperature (Tc), and heat of crystallization (Hc, Hcr) are measured with a TA Instrument Q200 d^ifferential _{scanning calorimetry (DSC) on 5 to 7 mg samples. DSC is run} according to ISO 11357 / part 3 /method C2 in a heat / cool / heat cycle with a scan rate of 10 °C/min in the temperature range of -30 to +225 °C. Crystallization temperature (Tc) and heat of crystallization (H_c) are d^{etermined from the cooling step, while melting temperature (T} _m ^{) and melt} enthalpy (Hm) are determined from the second heating step. Throughout the patent the term Tc or (Tcr) is understood as Peak temperature of crystallization as determined by DSC at a cooling rate of 10 K/min (i.e.0.16 K/sec). d) Quantification of microstructure by NMR spectroscopy Quantitative nuclear-magnetic resonance (NMR) spectroscopy was used to quantify the comonomer content of the polymers, comonomer dyad sequence distribution and sequence order parameter quantification. Quantitative ¹³C{¹H} NMR spectra were recorded in the solution-state using a Bruker Avance III 400 NMR spectrometer operating at 400.15 and 100.62 MHz for ¹H and ¹³C respectively. All spectra were recorded using a ¹³C optimized 10 mm extended temperature probehead at 125 °C using nitrogen gas for all pneumatics. Approximately 200 mg of material was dissolved in 3 ml of 7,2-tetrachloroethane-d/2 (TCE-d/2) along with chromium-(lll)- acetylacetonate (Cr(acac)3) resulting in a 65 mM solution of relaxation agent in solvent (Singh, G., Kothari, A., Gupta, V., Polymer Testing 285 (2009), 475). To ensure a homogenous solution, after initial sample preparation in a heat block, the NMR tube was further heated in a rotatory oven for at least 1 hour. Upon insertion into the magnet the tube was spun at 10 Hz. This setup was chosen primarily for the high resolution and quantitatively needed for accurate ethylene content quantification. Standard single-pulse excitation was employed without NOE, using an optimized tip angle, 1 s recycle delay and a bi-level WALTZ16 decoupling scheme (Zhou, Z., Kuemmerle, R., Qiu, X., Redwine, D., Cong, R., Taha, A., Baugh, D. Winniford, B., J. Mag. Reson.187 (2007) 225; Busico, V., Carbonniere, P., Cipullo, R., Pellecchia, R., Severn, J., Talarico, G., Macromol. Rapid Commun. 2007, 28, 1128). A total of 6144 (6k) transients were acquired per spectra. Comonomer content quantification of poly(propylene-co-ethylene) polymers Quantitative ¹³C{¹H} NMR spectra were processed, integrated and relevant quantitative properties determined from the integrals using proprietary computer programs. All chemical shifts were indirectly referenced to the central methylene group of the ethylene block (EEE) at 30.00 ppm using the chemical shift of the solvent. This approach allowed comparable referencing even when this structural unit was not present. Characteristic signals corresponding to the incorporation of ethylene were observed (Cheng, H. N., Macromolecules 17 (1984), 1950) and the comonomer fractions calculated as the fraction of ethylene and propylene in the polymer with respect to all monomer in the polymer: The comonomer fraction was quantified using the method of Wang et. al. (Wang, W-J., Zhu, S., Macromolecules 33 (2000), 1157) through integration of multiple signals across the whole spectral region in the ¹³C{¹H} spectra. This method was chosen for its robust nature and ability to account for the presence of regio-defects when needed. The mole percent comonomer incorporation was calculated from the mole fraction: E [mol%] = 100* fE The weight percent comonomer incorporation was calculated from the mole fraction: E [wt.-%] = 100* (fE* 28.06) / ( (fE* 28.06) + ((1-fE)* 42.08)) Comonomer dyad sequences determination Comonomer sequence distribution was quantified at the dyad level using the characteristic signals corresponding to the incorporation of ethylene into propylene-ethylene homopolymers (Cheng, H. N., Macromolecules 17 (1984), 1950). Integrals of respective sites were taken individually, the regions of integration described in the article of Wang et. al. were not applied for dyad sequence quantification. It should be noted that due to overlapping of the signals of Tβδ and Sγγ, the compensation equations were applied for integration range of these signals using the sites Sβδ and Sγδ: Sγγ = (I(Sβδ) - l(Sγδ))/2 Tβδ = I(Tβδ + Sγγ) - (l(Sβδ) - l(Sγδ))/2 With characteristic signals corresponding to regio defects observed (Resconi, L, Cavallo, L, Fait, A., Piemontesi, F., Chem. Rev. 2000, 100, 1253; Cheng, H. N., Macromolecules 17 (1984), 1950; Wang, W-J., Zhu, S., Macromolecules 33 (2000), 1157) the correction for the influence of the regio defects on comonomer contents was required. In case of 2,1-erythro mis-insertions presence the signal from ninth carbon (S2ie 9) of this microstructure element (Resconi, L., Cavallo, L., Fait, A., Piemontesi, F., Chem. Rev. 2000, 100, 1253) was chosen for compensation. In case of 2,1 regio irregular propene units in structure with one successive ethylene units presence, the signal from Tγγ (Cheng, H. N., Macromolecules 17 (1984), 1950; Wang, W-J., Zhu, S., Macromolecules 33 (2000), 1157) was chosen for compensation. The constitutive equations were: EP = 2* Tδδ + Tβδ + 2* Tγγ = 2*I(Tδδ) + I(Tβγ + Sγγ) - (l(Sβδ) - l(Sγδ))/2 + 2*I(Tγγ) EE = Sγγ + Sγδ + (Sδδ - Sγδ/2)/2 = 0.5*I(Sβγ) + 0.5*I(Sδδ) + 0.25*l(Sγδ) PP = Tβδ/2 + Tββb + 3*S_21e9 + 2* Tγγ = 0.5* (I(Tβδd + Sγγ) - (I(Sβδ) - l(Sγδ))/2) + I(Tββ) + 2*I(Tγγ) + 3*I(S_21e9) Note that for simplicity the two indistinguishable reversible PE and EP dyads are termed EP i.e. EP = PE + EP. The mole fraction of each dyad was determined through normalisation to the sum of all dyads. XX = PP + EP + EE fPP = PP/XX fEP = EP/XX fEE = EE/XX Sequence order parameter description and quantification Sequence order parameter, χ as it is defined by Koenig (Koenig92: Spectroscopy of Polymers, Lack. L Koenig. American Chemical Society, Washington, DC 1992) (or “Koenig B-value” as it is named in WO 2010/078479 A1), yields information about whether the distribution of the structures is random, i.e. can be described by Bernoullian statistics, and whether it tends towards an alternating or block distribution. This parameter can be determined by the formula: B_Koenig = fEP/(2* fE* fP) (IV) e) Xylene cold soluble fraction The xylene cold soluble fraction (XS) was determined according to ISO 16152 at 25 °C. f) Bulk density The bulk density was determined on the polymer powder according to ISO 60:1977 at 23 °C using a 100 cm³ cylinder. g) Gel index The gel content was measured via gel count with a gel counting apparatus consisting of a measuring extruder, ME 25 / 5200 V1, 25*25D, with five temperature conditioning zones adjusted to a temperature profile of 170/180/190/190/190 °C, an adapter and a slit die (with an opening of 0.5 * 150 mm). Attached to this were a chill roll unit (with a diameter of 13 cm with a temperature set of 50 °C), a line camera (CCD 4096 pixel for dynamic digital processing of grey tone images) and a winding unit. For the gel count measurements, the materials were extruded at a screw speed of 30 rounds per minute, a drawing speed of 3-3.5 m/min and a chill roll temperature of 50 °C to make thin cast films with a thickness of 70 μm and a width of approximately 110 mm. The resolution of the camera is 25 μm x 25 μm on the film. The camera works in transmission mode with a constant grey value (auto.set. margin level = 170). The system is able to decide between 256 grey values from black = 0 to white = 256. For detecting gels, a sensitivity level dark of 25% is used. For each material the average number of gel dots on a film surface area of 10 m² was inspected by the line camera. The line camera was set to differentiate the gel dot size according to the following: Gel size (the size of the longest dimension of a gel) Size class 1: 100 to 299 µm Size class 2: 300 µm to 599 µm Size class 3: 600 μm to 999 μm Size class 4: above 1000 μm The gel counts for the gels of the different size classes were measured and are given as counts per m². They represent the gel content of the respective size classes. The total gel content is the sum of these gel contents. For the determination of the gel index, the counts in the respective size classes were multiplied with a particular weigh factor as given below. The sum of the counts of each size class multiplied with the weigh factor represents the gel index (GI). Size class 1: 100 to 299 µm weight factor: 0.1 Size class 2: 300 µm to 599 µm weight factor: 1.0 Size class 3: 600 μm to 999 μm weight factor: 5.0 Size class 4: above 1000 μm weight factor: 10.0 Example: Size class 1: 17 counts x weight factor: 0.1 = 1.7 Size class 2: 5 counts x weight factor: 1.0 = 5.0 Size class 3: 2 counts x weight factor: 5.0 = 10.0 Size class 4: 0 counts x weight factor: 10.0 = 0.0 gel index (GI) = sum = 16.7 Materials The following catalyst was used in the processes according to the comparative and inventive examples as described in Table 1. A metallocene catalyst as described in WO 2019/179959 A1 was used. A steel reactor equipped with a mechanical stirrer and a filter net was flushed with nitrogen and the reactor temperature was set to 20 °C. Next silica grade DM-L-303 from AGC Si-Tech Co, pre-calcined at 600 °C (10 kg) was added from a feeding drum followed by careful pressuring and depressurizing with nitrogen using manual valves. Then toluene (43.5 kg) was added. The mixture was stirred for 30 min. Next 30 wt.% solution of MAO in toluene (17.5 kg) from Lanxess was added via feed line on the top of the reactor within 140 min. The reaction mixture was then heated up to 90 °C and stirred at 90 °C for additional two hours. The slurry was allowed to settle, and the mother liquor was filtered off. The catalyst was washed twice with toluene (43.5 kg) at 90 °C, following by settling and filtration. Finally, MAO treated SiO₂ was dried at 60 °C under nitrogen flow for 2 hours and then for 14 hours under vacuum (~0.5 barg) with stirring. MAO treated support was collected as a free-flowing white powder found to contain 15.0% Al by weight. 30 wt.% MAO in toluene (2 kg) was added into a steel nitrogen blanked reactor via a burette at 20 °C. Toluene (12.8 kg) was then added under stirring. 129 g of the metallocene was added from a metal cylinder followed by flushing with 1 kg toluene. The mixture was stirred for 60 minutes at 20 °C. Trityl tetrakis(pentafluorophenyl) borate (127.2 g) was then added from a metal cylinder followed by a flush with 1 kg of toluene. The mixture was stirred for 1 h at room temperature. The resulting solution was added to a stirred cake of MAO- silica support prepared as described above over 2 hours. The cake was stirred for 30 minutes and then allowed to stay without stirring for 30 minutes, followed by drying under N2 flow at 60 °C for 2 h and additionally for 15 h under vacuum (~0.5 barg) under stirring. Examples The following examples were carried out in a pilot plant, comprising a reactor sequence consisting of a prepolymerization reactor, a loop reactor and a gas phase reactor. The antistatic agent in the prepolymerization step was sorbitan monooleate (SPAN 80, Sigma-Aldrich). The polymer powder obtained after the GPR reactor was dried with purge bin type plug flow product drier. The average residence time was about 2 h and the temperature was 680 °C The polymer powder was compounded with the following additives: pentaerythrityl-tetrakis(3-(3’,5’-di-tert. butyl-4- hydroxyphenyl)-propionate (Irganox 1010 FF, BASF): 0.05 wt.-%; tris (2,4-di-t- butylphenyl) phosphite (Irgafos 168 FF, BASF): 0.05 wt.-%; and calcium stearate (Ceasit FI, Baerlocher): 0.04 wt.-%, extruded by a ZSK 70 extruder and pelletized to obtain pellets with sizes of 3 to 5 mm. Process conditions and properties of the polymers are depicted in Table 1a (for the comparative examples) and Table 1b (for the inventive examples). Table 1a Example CE1 CE2 CE3 CE4 Prepolymerization reactor (unstable) (unstable) (unstable) Temp. (°C) 20 20 30 25 Press. (kPa) 5300 5349 5500 5400 Catalyst feed (g/h) 4.0 4.6 3.9 7.6 C3 feed (kg/h) 54.6 54.6 55 54 Feed H₂/C₃ ratio (mol/kmol) 0.31 0.31 0.20 0.20 H2 feed (g/h) 0.8 0.8 0.4 0.4 Antistatic feed (wt.-ppm/h)* 0 2.25 4.0 20 Residence time (h) 0.45 0.45 0.42 0.45 Production rate (kg/h) 0.9 0.8 1.2 0 Prepolymerization degree (g P_{P/g cat)} ^{200 174 350 4} Polymer Split (%) 2.6 2.6 3.0 0 Loop reactor (plugging) (plugging) (unstable) Temp. (°C) 75 75 75 75 Press. (kPa) 5350 5289 5400 5300 C3 feed (kg/h) 165 165 154 155 Feed H₂/C₃ ratio (mol/kmol) 0.26 0.26 0.28 0.28 Production rate (kg/h) 30.3 30.3 32 20 Polymer Split (%) 64.8 64.8 67 40 Polymer residence time (h) 0.3 0.3 0.4 0.4 MFR₂ (g/10 min) 6.8 6.8 7.0 8.0 XS (%) 1.6 1.6 1.6 1.6 APS (mm) 0.82 0.82 0.90 0.60 Bulk Density (kg/m³) 461 461 440 466 Productivity (kg PP/g cat) 7.6 6.6 GPR reactor (unstable) (unstable) (unstable) Temp. (°C) 80 80 80 80 Press. (kPa) 2474 2474 2500 2500 C3 feed (kg/h) 215 215 205 205 H₂/C₃ ratio (mol/kmol) 2.93 2.93 2.9 2.9 Polymer residence time (h) 3.06 3.06 3.0 3.0 Polymer Split (%) 32.6 32.6 30 60 MFR₂ (g/10 min) 6.1 6.1 7.0 9.0 XS (%) 1.2 1.2 1.5 1.5 Bulk Density (kg/m³) 465 465 430 470 Loop + GPR Production rate (_kg/h) ^{45.0 45.0 41 38} Loop + GPR Productivity (kg P_{P/g cat)} ^{11.0 9.8 10.5 5.0} Material after GPR, Powder No data No data No data a_vailable available available MFR2 (g/10 min) 6.0 XS (%) 1.3 Bulk Density (kg/m³) 475 PS < 0.106 mm (wt.-%) 0.03 PS > 0.106 mm ≤ 0.250 mm (_wt.-%) ^0.44 PS > 0.250 mm ≤ 0.355 mm (_wt.-%) ^1.95 PS > 0.355 mm ≤ 0.850 mm (_wt.-%) ^65.59 PS > 0.850 mm ≤ 2.0 mm (_wt.-%) ^21.29 PS > 2.0 mm ≤ 4.0 mm (wt.- %₎ ^9.61 PS > 4.0 mm (wt.-%) 1.09 APS (mm) 1.037 Material after GPR, Pellet No data No data No data a_vailable available available MFR₂ (g/10 min) 5.8 Tm (°C) 150.3 Tcr (°C) 114.0 Gel index PP (OCS) 11.9 * antistatic feed is relative to the C₃ feed Table 1b Example IE1 IE2 IE3 Prepolymerization reactor Temp. (°C) 24 25 25 Press. (kPa) 5317 5331 5358 Catalyst feed (g/h) 3.7 4.5 4.8 C₃ feed (kg/h) 51.1 51.7 52.2 Feed H2/C3 ratio (mol/kmol) 0.19 0.16 0.07 H₂ feed (g/h) 0.4 0.4 0.2 Antistatic feed (wt.-ppm/h)* 4.5 4.5 4.5 Residence time (h) 0.48 0.47 0.47 Production rate (kg/h) 0.7 0.7 0.8 Prepolymerization degree (g _{PP/g cat)} ^{191 157 167} Polymer Split (%) 2.4 2.5 2.2 Loop reactor Temp. (°C) 75 75 75 Press. (kPa) 5276 5279 5283 C₃ feed (kg/h) 152 152 152 Feed H2/C3 ratio (mol/kmol) 0.30 0.28 0.26 Production rate (kg/h) 28.8 28.6 36.1 Polymer Split (%) 48.6 56.5 54.3 Polymer residence time (h) 0.4 0.4 0.5 MFR2 (g/10 min) 10.3 17.3 7.4 XS (%) 1.6 1.5 0.9 APS (mm) 0.70 0.74 0.73 Bulk Density (kg/m³) 458 470 452 Productivity (kg PP/g cat) 7.8 6.4 7.5 GPR reactor Temp. (°C) 80 80 80 Press. (kPa) 2600 2600 2600 C₃ feed (kg/h) 203 203 196 H2/C3 ratio (mol/kmol) 2.71 2.75 3.10 Polymer residence time (h) 3.10 3.20 3.00 Polymer Split (%) 49.0 41.0 43.5 MFR₂ (g/10 min) 7.3 11.8 7.3 XS (%) 1.6 1.6 0.8 Bulk Density (kg/m³) 453 459 459 Loop + GPR Production rate _(kg/h) ^{56.4 48.5 63.8} Loop + GPR Productivity (kg _{PP/g cat)} ^{15.4 10.9 13.4} Material after GPR, Powder MFR2 (g/10 min) 6.2 9.6 8.1 XS (%) 1.5 1.6 0.9 Bulk Density (kg/m³) 477 470 478 PS < 0.106 mm (wt.-%) 0.03 0.04 0.02 PS > 0.106 mm ≤ 0.250 mm _(wt.-%) ^{0.83 1.63 0.65} PS > 0.250 mm ≤ 0.355 mm (_wt.-%) ^{2.88 3.56 2.56} PS > 0.355 mm ≤ 0.850 mm (_wt.-%) ^{73.79 72.10 76.50} PS > 0.850 mm ≤ 2.0 mm (_wt.-%) ^{19.61 20.84 19.99} PS > 2.0 mm ≤ 4.0 mm (wt.- %₎ ^{1.95 1.54 0.24} PS > 4.0 mm (wt.-%) 0.91 0.28 0.04 APS (mm) 0.829 0.803 0.764 Material after GPR, Pellet MFR₂ (g/10 min) n.m. 8.6 7.7 Tm (°C) n.m. 148.9 149.2 Tcr (°C) n.m. 113.4 113.2 Gel index PP (OCS) n.m. 6.6 5.8 n.m. = not measured * antistatic feed is relative to the C3 feed As depicted in Tables 1a and 1b, higher productivity is reached for the inventive examples when compared to the Comparative Example CE2. The Comparative Examples CE1, CE3 and CE4 could not be collected for measurements due to process disturbance. In the absence of an antistatic agent and at a temperature of 20 °C in the prepolymerization step, the reactor could not be stably run for the Comparative Example CE1. The same was true for an antistatic agent feed of 20 wt-ppm/h (Comparative Example CE4) and for a temperature of 30 °C (Comparative Example CE3). Disturbance was also observed for Comparative Example CE2, however, at a reduced level such that the polymer could be prepared and collected. The content of agglomerated particles (PS > 2.0 mm) and the gel index are significantly reduced in the examples according to the invention. No significant change in melting point and crystallization temperature was observed. Correlation between the content of the antistatic agent feed and the prepolymerization degree is depicted in Table 2. (other conditions: T = 20 °C, P = 5300 kPa, catalyst feed = 4.0 g/h, C3 feed = 53 kg/h, feed H2/C3 ratio = 0.2 mol/kmol, H2 feed = 0.4 g/h, residence time = 0.35 h, production rate = 0.75 kg/h) Table 2 Antistatic feed (wt.-ppm) Prepolymerization degree (g PP/g cat) 0 200 2 170 4 155 6 120 8 80 The prepolymerization degree linearly decreases with the content of the antistatic agent (as depicted in Figure 3). Correlation between the temperature during prepolymerization and the prepolymerization degree is depicted in Table 3. (other conditions: antistatic feed = 4 wt.-ppm, P = 5300 kPa, catalyst feed = 4 g/h, C₃ feed = 53 kg/h, feed H₂/C₃ ratio = 0.2 mol/kmol, H2 feed = 0.4 g/h, residence time = 0.35 h, production rate = 0.75 kg/h) Table 3 Temperature (°C) Prepolymerization degree (g PP/g cat) 20 170 23 220 25 250 27 260 28 310 The prepolymerization degree linearly increases with the temperature of the prepolymerization step (as depicted in Figure 4).

Claims

Claims 1. A process for the preparation of a propylene polymer, the process comprising the steps: a) prepolymerizing propylene in the presence of a metallocene catalyst in a first reactor, yielding a prepolymer (A), wherein the prepolymerization is conducted at the following conditions: (i) a temperature in the range of from 22 to 27 °C, (ii) a pressure in the range of from 3.5 to 6.5 MPa (gauge), (iii) a residence time in the range of from 15 to 35 min, (iv) an H₂/C₃ feed ratio in the range of from 0.02 to 0.25 mol/kmol, and (v) a feed of an antistatic agent in an amount in the range of from 0.5 to 10.0 ppm by weight of the propylene feed, b) transferring the prepolymer (A) to a second reactor, and c) polymerizing propylene and optionally one or more of C2 and C4 to C10 alpha olefin comonomers in the presence of the prepolymer (A) in the second reactor, yielding a propylene polymer (B). 2. The process according to claim 1, further comprising d) transferring the propylene polymer (B) to a third reactor, and e) polymerizing propylene and optionally one or more of C₂ and C₄ to C₁₀ alpha olefin comonomers in the presence of the propylene polymer (B) in the third reactor, yielding a propylene polymer (C). 3. The process according to any one of the preceding claims, wherein the content of the prepolymer (A) in the propylene polymer (B) is in the range of from 1.0 to 10.0 wt.-%, based on the total weight of the propylene polymer (B), and, if the propylene polymer (C) is present, the content of the prepolymer (A) in the propylene polymer (C) is in the range of from 1.0 to 10.0 wt.-%, based on the total weight of the propylene polymer (C). 4. The process according to any one of the preceding claims, wherein the propylene polymer (B) is a propylene homopolymer, and, if the propylene polymer (C) is present, the propylene polymer (C) is a propylene homopolymer. 5. The process according to any one of the preceding claims, wherein the feed of an antistatic agent in the first reactor in step a) is in an amount in the range of from 2.5 to 6.5 ppm, preferably 3.5 to 5.5 ppm, by weight of the propylene feed. 6. The process according to any one of the preceding claims, wherein step a) is conducted in slurry. 7. The process according to any one of the preceding claims, wherein the first reactor is a loop reactor. 8. The process according to any one of the preceding claims, wherein the second rector is a loop reactor and/or the third reactor, if present, is a gas phase reactor. 9. The process according to any one of the preceding claims, wherein prepolymerization in step a) is conducted at the following conditions: (i) a temperature in the range of from 23 to 26 °C, preferably from 24 to 25 °C, (ii) a pressure in the range of from 4.5 to 6.5 MPa (gauge), preferably from 5.0 to 6.0 MPa (gauge), (iii) a residence time in the range of from 20 to 35 min, preferably from 25 to 30 min, (iv) an H₂/C₃ feed ratio in the range of from 0.02 to 0.20 mol/kmol, preferably from 0.05 to 0.20 mol/kmol, and (v) a feed of an antistatic agent in an amount in the range of from 2.5 to 6.5 ppm, preferably from 3.5 to 5.5 ppm, by weight of the propylene feed. 10. The process according to any one of the preceding claims, wherein the antistatic agent is selected from the group consisting of glycerol monostearate (GMS), hydrogenated tallow fatty acids, blends of GMS and hydrogenated tallow fatty acids, polypyrrole, carbon nanotubes, carbon black, carbon fiber, graphite fiber, fatty acid alkanolamide, blends of GMS and fatty acid dialkanolamide, anionic hydrocarbyl sulfonate, N,N-bis(2- hydroxyethyl)alkoxypropylbetaine, lauric diethanol amide, alkyl-bis(2- hydroxyethyl)amine, quaternary ammonium compound, polyetheresteramide, tertiary amine, blends of GMS and tertiary amine, stearyldiethanolamine, alkyl phosphate, ethoxylated secondary alcohols, glycerol distearate, blends of GMS and glycerol distearate, sodium alkyl sulfonate, neutralized alcohol phosphate, polyethylene glycol monolaurate, sodium alkyl sulfonate, sorbitan monolaurate, sorbitan monooleate, ethoxylated sorbitan monolaurate, oleochemical derivatives, polyethylene glycol, polypropylene glycol, polyoxyethylene sorbitan monooleate, polyoxyethylene sorbitan monolaurate, polyethylene oxide, blends of polyethylene oxide and sorbitan monolaurate, blends of polyethylene oxide and sorbitan monooleate, polyether block amides, and any combinations thereof. 11. The process according to any one of the preceding claims, wherein in step c) and step e), if present, the temperature in the second and optional third reactor is in the range of from 60 to 90 °C. 12. The process according to any one of the preceding claims, wherein in step a) the ratio of a feed of metallocene catalyst to the feed of propylene is in the range of from 0.02 to 0.15 g/kg, preferably from 0.05 to 0.12 g/kg. 13. The process according to any one of the preceding claims, wherein the metallocene catalyst comprises a metallocene complex and a support, wherein the support comprises silica. 14. The process according to claim 13, wherein the metallocene complex is an organometallic compound (C), the organometallic compound (C) has the following formula (I): (L)mRnMXq (I) wherein “M” is a transition metal (M) of Group 3 to 10 of the Periodic Table (IUPAC 2007); each “X” is independently selected from monoanionic ligands, such as a σ- ligand; each “L” is independently selected from organic ligands which coordinate to the transition metal “M”; “R” is a bridging group linking said organic ligands (L); “m” is 1, 2 or 3, preferably 2; “n” is 0, 1 or 2, preferably 1; “q” is 1, 2 or 3, preferably 2; and m+q is equal to the valency of the transition metal (M). 15. A propylene polymer obtainable or obtained by the process according to any one of the preceding claims.