WO2012066016A1 - Process to make diblock polystyrene materials comprising syndiotactic and atactic blocks. - Google Patents

Abstract

The present invention discloses the use of a metal-based catalyst alone or in combination with a metallic compound of formula M2(R')n, acting as chain transfer agent, to initiate the stereospecific chain-growth polymerisation of styrene and prepare subsequently sPS-b-aPS stereoblock polystyrenic materials, possibly rubber-toughened, and useful for applications requiring a high thermal and/or chemical resistance.

Description

PROCESS TO MAKE DIBLOCK POLYSTYRENE MATERIALS COMPRISING SYNDIOTACTIC AND ATACTIC BLOCKS-

FIELD OF THE INVENTION.

The present invention concerns the use of a metal-based catalyst alone or in combination with a metallic compound of formula M²(R')_n, acting as chain transfer agent, to initiate the stereospecific chain-growth polymerisation of styrene and eventually prepare stereoblock polystyrenic materials by subsequent radical polymerisation of styrene, possibly in the presence of elastomers.

BRIEF DESCRIPTION OF THE PRIOR ART.

Syndiotactic polystyrene (sPS) has been first described by Ishihara et al. (Ishihara, N.; Seimiya, T.; Kuramoto, M.; Uoi, in M. Macromolecules 1986, 19, 2464; JP-A- 62187708; EP-A-210615). Its unique mechanical properties make it valuable for industrial purposes and specialty materials. However, despite the remarkable heat and chemical resistance properties induced by its crystalline phase, sPS is, just as aPS, inherently brittle, which is an acute issue since it has to compete with inherently tough engineering plastics such as polyamides. For many applications, sPS toughness appears insufficient and its impact-modification is often highly desirable. The incorporation of glass fibres, often used to reinforce sPS, remains a palliative solution as they only enhance stress energy dissipation when the strain applied is perpendicular to their orientation direction. Such fillers are far less effective than rubbery particles as impact modifiers.

Comparatively, aPS impact modification has been practiced for a long time, either by blending aPS with multi-block rubbery copolymers, mainly of the styrene-butadiene- styrene (SBS) type, or by merely polymerising styrene by a radical pathway in the presence of polybutadiene (PB) rubber leading to high-impact polystyrene, commonly referred to as HiPS. During the conventional HiPS manufacturing process, polybutadiene rubber chains are partly grafted in situ by polystyrene chain. The resulting PB-g-PS emulsifies the rubbery particles formed within the aPS continuous phase. HiPS are consequently composite thermoplastics made of soft particles well-dispersed and anchored within a stiff polystyrene matrix and exhibiting excellent elongation and impact properties.

Dienic rubber grafting is on the contrary very difficult to achieve in the catalytic metallocene processes required for manufacturing syndiotactic polystyrene. This limitation has led to many attempts in the past to increase sPS toughness by blending sPS with rubbers. The most commonly used impact modifiers reported for sPS in EP-A-318793 are specialty block copolymers such as high-molecular weight hydrogenated SBS block copolymers, such as for example SEBS, Kraton™ G 1651 from Kraton Performance polymers Inc., possibly combined with olefin rubbers, especially EPR as disclosed in EP-A-324398, or core-shell impact modifiers as disclosed in EP-A-755972. Ethylene-styrene random interpolymers having a styrene content as high as 70% made from the Insite™ technology as disclosed in US-A- 6063872, combined with ethylene-octene copolymers, have also been reported as efficient impact-modifying systems for sPS. However, sPS impact-modification through compounding with specialty rubbers resulted in significant extra-costs compared to direct-reaction processes, not to mention the price of the impact modifiers. sPS and aPS are miscible in the amorphous state, i.e. mostly in the molten state, as long as the sPS fraction has not extensively developed crystallinity. Overcoming the weak properties of sPS such as high brittleness, and of HiPS such as low thermal and chemical resistance, by blending them together, has always been considered with great interest. The mechanical properties, especially the Izod impact and elongation at break, of HiPS can be maintained even after blending with 20% of sPS, while the resistance of the resulting material to aggressive chemicals (fat oils, solvents, hydrocarbons... ) can be dramatically improved, as demonstrated by, e.g., T. Takebe et al. in Syndiotactic Polystyrene, Jurgen Schellenberg Ed. (2010). Idemitsu Kosan Co., Ltd. thus commercialised sPS/HiPS blends under the Xarec™ E tradename. This interesting combination of properties, that could make sPS/HiPS blends as performing as some engineering plastics, can however only be reached if the sPS part can develop enough crystallinity. Unfortunately, mixing aPS with sPS results in a sharp reduction of the crystallisation rate of the latter, due mostly to an impairment of the growth of the a crystal form (cf. F-C. Chiu et al., Polymer 43 (2002) 4879-4886). This major drawback dictates constraining processing conditions with high mould temperatures and very long cycle times for injection-moulded sPS/HiPS articles. If such processing conditions are not applied, sPS and HiPS melt-blending results most likely in fully amorphous rubber-toughened PS, and the mere solvent- induced crystallisation of the sPS fraction of the blend exposed to aggressive solvents is often insufficient for improving significantly the chemical resistance of the blend.

It is thus highly desirable to find efficient and inexpensive pathways for sPS impact modification. Till now, all the direct-reaction processes reported for making impact- modified sPS consisted mostly of reactor blending approaches, such as for example sPS production in the presence of polybutadiene rubber without in situ grafting reactions as described in EP 440,014, or involved non-commercial graft precursors such as ethylene-propylene-p-(3-butenylstyrene) macromonomers as described in EP 559, 108.

It is the objective of the present invention to make both and at the same time possible:

(i) The production of sPS-Jb-aPS stereoblock copolymers acting efficiently as crystallisation accelerators for sPS/HiPS blends;

(ii) The production of impact-modified semi-crystalline polystyrene, namely rubber-toughened sPS-Jb-aPS, with in situ grafting of the toughening rubber phase in a direct-reaction process.

A major advantage of the present invention is that it makes possible the versatile production of impact-resistant polystyrene of very high thermal and chemical resistance.

To the best of our knowledge, the literature does not describe any sPS-Jb-aPS block copolymers either made by chain-growth polymerisation in presence of chain transfer agents (CTAs), or by any other means. Synthesis of sPS under homogenous conditions relies usually on hemi- or post- metallocene titanium (IV/III) complexes activated with MAO or B(C6F₅)₃ or [CPh)][B(C6F₅)₄]. Suitable Ti complexes were selected, for instance, from:

Unsubstituted or substituted cyclopentadienyl complexes of titanium were considered as the most active catalysts for the syndiospecific polymerisation of styrene. They were extensively studied in experimental and theoretical works summarised, for example, by Rodrigues et al. (Rodrigues, A.-S.; Kirillov, E.; Carpentier, J.-F in. Coord. Chem. Rev. 2008, 252, 21 15) or by Schellenberg (Schellenberg, J. in Prog. Polym. Sci. 2009, 34, 688-718), for determining that:

(i) the polymerisation occurred in all cases by a poly-insertion mechanism,

(ii) the regiochemistry of styrene insertion was secondary in both the initiation and termination steps,

(iii) the addition to the monomer double bond was cis,

(iv) the stereoselectivity of the insertion step was controlled by the chirality of the growing chain end,

(v) the active catalyst was a Ti(lll) species derived from the Ti(IV) precursor.

Among these precursors, pentamethylcyclopentadienyl-titanium-tribenzyl (Cp^*TiBn₃) is known to produce almost pure sPS, giving better polymer yield and control of the average molecular weight in comparison to the corresponding trimethyl-titanium complex, as described for example by Grassi et al (Grassi, A.; Lamberti, C; Zambelli, A.; Mingozzi, I in. Macromolecules 1997, 30, 1884-1889).

Living syndiospecific homo- or co-polymerisation of styrene or substituted styrenic monomers have been reported by Kawabe et al. (Kawabe, M.; Murata, M.; Soga, K. in Macromol. Rapid. Commun. 1999, 20, 569-572; Kawabe, M.; Murata, M. in /. polym. Sci: Part A: Polym. Chem. 2001 , 39, 3692-3706; Kawabe, M.; Murata, M. in Macromol. Chem. Phys. 2002, 203, 24-30) using the catalyst system Cp^*TiMe₃/B(C₆F5)3/AI(Oct)3.

In 2004, Carpentier et al. (Kirillov, E.; Lehmann, C. W.; Razavi, A.; Carpentier, J.-F. in J. Am. Chem. Soc. 2004, 126, 12240; Rodrigues, A.-S.; Kirillov, E.; Vuillemin, B.; Razavi, A.; Carpentier, J.-F. in J. Mol. Cat. A: Chem. 2007, 273, 87; Perrin, L; Sarazin, Y.; Kirillov, E.; Carpentier, J.-F.; Maron, L. in Chem. Eur. J. 2009, 15, 3773; Carpentier, J.-F.; Kirillov, E.; Razavi, A. in Eur. Pat. Appl. 04/290847 (31/03/2004); I 1/O05 051 369; WO 05 095 470; Eur. Pat. 2005/1005) described the syndiotactic homo-polymerisation of styrene mediated with a series of discrete ansa-metallocene ally I complexes of early lanthanides acting as mono-component catalysts, therefore not requiring the addition of activator and/or co-catalyst, and yielding virtually pure syndiotactic polystyrenes having more than 99% rrrrr with good control over the polymerisation.

On the other hand, it is known from Pelletier et al. (Pelletier, J. F.; Mortreux, A.; Olonde, X.; Bujadoux, K. Angew. Chem. Int. Ed. 1996, 35, 1854) or from Bogaert et al. (Bogaert, S.; Chenal, T.; Mortreux, A.; Nowogrocki, G.; Lehmann, C. W.; Carpentier, J.-F. Organometallics 2001 , 20, 199) that dialkylmagnesium compounds (MgR₂) acted as chain transfer agents (CTA) during the polymerisation of ethylene in the presence of neodymium catalysts. In this case, the transfer reactions were beneficial for the polymerisation, as they stabilised the growing polymer chains to reversible dormant species, and tended to narrow the PDI of the polymers obtained. More, the use of CTA enhanced greatly the catalyst efficiency and productivity, as the catalyst loading (vs monomer) could be substantially reduced by adjusting the amount of CTA. The latter amount of CTA allowed also controlling the molecular weight of the polymer chains. Based on this "chain growth polymerisation" principle, Rodrigues et al. (Rodrigues, A.-S.; Kirillov, E.; Vuillemin, B.; Razavi, A.; Carpentier, J.-F. J. Mol. Cat. A: Chem. 2007, 273, 87) synthesised some highly syndiotactic oligostyrenes using an ansa-chloroneodymocene catalyst precursor combined with diallyl- or dialkyl-magnesium compounds that acted as catalyst activator and chain transfer agents, as seen in Scheme 1 showing a simplified mechanism for the chain- growth polymerisation of styrene mediated by such binary neodymocene- dialkylmagnesium systems.

Scheme 1

Examples of transfer reactions with dialkylmagnesium compounds leading to copolymers remain scarce. In 2005, Boisson et al. (Thuilliez, J.; Monteil, V.; Spitz, R.; Boisson C. in Angew. Chem. Int. Ed. 2005, 44, 2593) reported the statistic copolymerisation of ethylene with butadiene using a neodymocene / butyloctylmagnesium system. More recently, Mortreux, Visseaux et al. (P. Zinck, A. Valente, A. Mortreux, M. Visseaux in Polymer 2007, 48, 4609) reported the block- copolymerisation of syndiotactic enriched oligostyrene (macro-initiator) with isoprene using a system based on a bis(borohydride) group 3 metallocene and n- butylethylmagnesium, as illustrated in Scheme 2 showing the sequential copolymerisation of styrene and isoprene mediated by a scandium catalyst in the presence of butylethylmagneisum as chain transfer agent.

Scheme 2 On the other hand, processes involving the in situ functionalisation of sPS chains and the subsequent transitioning to radical polymerisation have been studied. In the literature, examples of sPS-graff-aPS copolymers have been reported by Sen (Lui, S.; Sen, A. Macromolecules 2000, 33, 5106-51 10), by Entezami (Abbasian, M.; Rahmani, S.; Mohammadi, R.; Entezami, A. A. J. Appl. Polym. Sci. 2007, 104, 61 1 - 619), by Wang (Gao, Y.; Li, H. ; Wang, X. Int. Int. 2007, 56, 976-983), and by Endo et al. (Senoo, K.; Endo, K.; Tosaka, M.; Murakami, S.; Kohjiya, S. Macromolecules 2001 , 34, 1267-1273; Endo, K.; Senoo, K. Polymer 1999, 40, 5977-5980)

In the first three examples, sPS-graff-aPS copolymers were synthesized by a three- step process consisting of:

1 . synthesis of regular syndiotactic polystyrene using CpTiCl3/MAO as catalyst;

2. functionalisation of this sPS by random insertion, in the polymer framework, of bromine atoms and a-phenyl-X-acetyl groups (X = CI or Br), respectively;

3. growth of aPS segments from these halide functionalities via atom transfer radical polymerisation (ATRP).

Instead, the strategy reported by Endo involved a two-step process with the preliminary preparation of an atactic polystyrene macromer bearing a styryl end- group and its syndiospecific copolymerisation with styrene using CpTiCl3/MAO as catalyst.

Due to the inherent nonreactive nature of sPS, the presence of functional groups is necessary to obtain new polymeric materials. As discussed above, the two conventional pathways for functionalisation of polyolefins are: 1 ) post-polymerization modification and 2) direct catalytic introduction of functional groups. In general, the fist method avoids the issues of catalyst functional-group tolerance and catalyst poisoning, while the second method involves incorporation of chain-transfer agents into olefin polymerisation process. The chain transfer reaction in the coordination olefin polymerisation was usefully employed and confirmed to be a good strategy to achieve end-functionalised polyolefins as reported by Amin et al. (Amin, S. B.; Marks, T. Angew. Chem. Int. Ed. 2008, 47, 2006-2025). In 1999, Chung et al. (Xu, G.; Chung, T. C. Macromolecules 1999, 32, 8689-8692) reported the synthesis of borane-terminated sPS prepared by using 9- borabicyclo[3.3.1 ]-nonane (9-BBN) as CTA during titanocene-catalysed styrene polymerisation. The borane terminal groups were then used either to generate hydroxyl end-groups (sPS-OH), or to generate alkoxyl radicals (sPS-0°) in order to eventually initiate the radical polymerisation of methyl methacrylate and obtain sPs- Jb-PMMA materials.

Functionalisation of syndiotactic polystyrene, using both aforementioned strategies, was recently reviewed by Zink et al. (Zink, P.; Bonnet, F.; Mortreux, A.; Visseaux, M. Prog. Polym. Sci. 2009, 34, 362-392).

Almost at the same time, Grassi et al. (Pastorino, R.; Capacchione, C; Ferro, R.; Milione, S.; Grassi, A. Macromolecules 2009, 42, 2480-2487) reported the preparation of syndiotactic polystyrene macromers end-capped with a bromine atoms, using stereospecific polymerisation of styrene with a CpTiR₃-B(C6F₅)₃ catalyst and sequential reaction of living Ti-sPS chains with N-bromosuccinimide (NBS).

Styrene is a very versatile monomer, which can be polymerised either by radical, anionic, cationic or coordination-insertion mechanism. Radical pathways proved to be quite useful. In particular, the past few years have witnessed the rapid growth in development and understanding of new controlled radical polymerisation (CRP) methods. One of the most efficient CRP method is nitroxide-mediated polymerisation (NMP) that requires stable-free nitroxide radicals or alkoxyamines as disclosed by Hawker et al. (Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. Rev. 2001 , 101, 3661 and references therein); or atom transfer radical polymerisation (ATRP) involving alkyl halides, metallic salts and adequate ligands as studied by Wanga et al. (Wanga, J.; Matyjaszewsky, K. J. Am. Chem. Soc, 1995, 117, 5614- 5615); or the reversible addition-fragmentation chain transfer polymerisation (RAFT) using dithiocarbonyl derivatives as disclosed by Rizzardo et al. (Rizzardo, E.; Chiefari, J.; Mayadunne, R.; Moad, G.; Thang, S. Macromol. Symp. 2001, 174, 209- 21 1 ); or iodine transfer polymerisation (ITP) and reversible iodine transfer polymerization (RITP) using alkyl iodides as disclosed by David et al. (David, G.; Boyer, C; Tonnar, J.; Ameduri, B.; Lacroix-Desmazes, P.; Boutevin, B. Chem. Rev. 2006, 106, 3936-3962). All these CRPs methods allow to achieve a very good control of the polymerisation, giving polymers with narrow molecular weight distribution and tailored with reactive (end-)groups that can be easily converted into other functional groups or can be used to initiate again a new polymerisation with different monomers in order to eventually produce block copolymers. Despite these advantages, the main problem of these polymerisations is the lack of stereocontrol and, in all cases, the production of atactic polymers only.

Also, it is well known that the bulk free radical polymerisation of vinyl monomers, especially styrene, can be a living process in the presence of a small amount of control agents such as TEMPO (2,2,6,6-tetramethylpiperidine-N-oxyl). In 1994, Hawker et al. (Hawker, C. J. in J. Am. Chem. Soc. 1994, 116, 1 1314; Hawker, C. J.; Hedrick, J. L. in Macromolecules 1995, 28, 2993) or Rizzardo (Rizzardo, E. in Chem. Aust. 1978, 54, 32) described the controlled polymerisation of vinyl monomers mediated by TEMPO and polystyrenes with accurately controlled molecular weights and polydispersity indices as low as 1 .1 (around 1 .1 ). Since then, numerous initiators based on nitroxides have been synthesized and applied successfully to the so-called pseudo-living nitroxide-mediated polymerization (NMP) of vinyl monomers as reviewed in Hawker, C. J. ; Bosman, A. W.; Harth, E. in Chem. Rev. 2001 , 101, 3661 and in references cited therein. This is summarised in Scheme 3 wherein the upper line represents typical monomers and the bottom line represents initiators used for nitroxide-mediated polymerisation (NMP).

Scheme 3 It is also known that alkyl-lithium compounds are efficient reagents for neutralising an excess of TEMPO, yielding the corresponding diamagnetic alkylated alkoxy-amines. The reaction has been extended to other alkyl metals, including Grignard reagents, and was found to be very efficient with dialkyl-samarium compounds as disclosed in Whitesides and Newirth (Whitesides, G. M.; Newirth, T. L. in J. Org. Chem. 1975, 40, 3443) or in Nagashima and Curran (Nagashima, T.; Curran, D. P. in Synlett 1996, 4, 330).

Recently the end-capping of di(polyethylenyl)magnesium (Mg(PE)₂) chains with TEMPO, resulting from the catalytic chain-growth polymerisation of ethylene, has been achieved by D'Agosto et al. (Lopez, R. G.; Boisson, C ; D'Agosto, F. ; Spitz, R.; Boisson, F.; Bertin, D.; Tordo, P. in Macromolecules 2004, 37, 3540), taking advantage of the radical-like reactivity of dialkylmagnesium compounds as illustrated in Scheme 4. The functionalisation efficiency was, however, moderate, notably due to the formation of side-products similar to those encountered during the disproportionation of unstable radicals.

Scheme 4.

Among the various controlled radical polymerisation routes of vinyl monomers, atom transfer radical polymerisation (ATRP) has gained rapidly increasing interest due to excellent control of the chain-growth and narrow molecular weight distributions. This kind of polymerisation, reported by Matyjaszewski et al. (Wanga, J.; Matyjaszewsky, K. in J. Am. Chem. Soc, 1995, 117, 5614-5615), relies on the reversible activation of a dormant alkyl (benzyl) halide through halogen abstraction by a transition metal complex in addition to the classical free-radical polymerisation scheme. ATRP of styrene has been widely used, for example by Matyjaszewsky et al. (Matyjaszewsky, K.; Xia, J. Chem. Rev. 2001 , 101, 2921 -2990) to produce copolymers with different composition (gradient/statistical or block copolymers), or different topology (graft- copolymers or star polymers).

LIST OF FIGURES.

Figure 1 represents the MALDI-ToF-MS spectrum of the reaction product of Mg(sPS)₂ and TEMPO.

Figure 2 represents the SEC trace of the material recovered via nitroxide-mediated polymerisation (NMP) of styrene with sPS-TEMPO using the first embodiment of the invention, respectively for the starting block having a number average molecular weight M_n of 3 620 g/mol as indicated and for the final diblock copolymer having blocks of 4 900 g/mol and 35 780 g/mol as indicated.

Figure 3 represents the SEC trace of the material recovered via nitroxide- mediated polymerisation (NMP) of styrene with sPS-TEMPO using the first embodiment of the invention, respectively for the starting block having M_n of 4 630 g/mol as indicated and for the final diblock copolymer having blocks of 3 840 g/mol and 90 740 g/mol as indicated.

Figure 4 represents the SEC trace of the material recovered via thermal polymerisation of styrene with Mg(sPS)2 using the second embodiment of the invention respectively for the starting block having M_n of 4 630 g/mol as indicated and for the final copolymer with M_n of 44 470 g/mol as indicated.

Figure 5 represents the SEC trace of the material recovered via thermal polymerisation of styrene with Mg(sPS)2 using the second embodiment of the invention respectively for the starting block M_n of 8 810 g/mol as indicated and for the final copolymer with M_n of 32 990 g/mol as indicated.

Figure 6 represents the SEC trace of the material recovered via catalytic polymerisation of styrene with Mg(aPS)2 using the third embodiment of the invention respectively for the starting block M_n of 5 020 g/mol as indicated and for the final copolymer with M_n of 12 030 g/mol as indicated.

Figure 7 represents the SEC trace of the material recovered via catalytic polymerisation of styrene with Mg(aPS)2 using the third embodiment of the invention respectively for the starting block M_n of 10 560 g/mol as indicated and for the final copolymer with M_n of 10 970 g/mol as indicated.

Figure 8 represents the DSC curve of a sPS-b-aPS diblock copolymer.

Figure 9 represents the conversion rate expressed in % as function of polymerisation time expressed in minutes for styrene polymerisation catalysed by Cp^*Ti(CH₂Ph)3/B(C₆F5)3/AI(Oct)3 in ratio 1/1/1 at a temperature of 0°C using 65 equivalents of styrene (Table 4, lines 10-14) for conversion represented by■ and for number-average molecular weight M_n represented by o.

Figure 10 represents the conversion rate expressed in % as function of polymerisation time expressed in minutes for styrene polymerisation catalysed by Cp^*Ti(CH₂Ph)3/B(C₆F5)3/AI(Oct)3 in ratio 1/1/1 at a temperature of 0°C using 325 equivalents of styrene (Table 4, lines 10-14) for conversion represented by■ and for number-average molecular weight M_n represented by o.

Figure 1 1 represents the ¹H NMR spectrum (C2D2CI4, 353 K) of sPS-end capped bromine atoms.

Figure 12 represents high temperature Gel Permeation Chromatography (GPC) chromatograms of the copolymers; the dashed line represents the starting sPS-Br block and the solid line represents the final copolymer.

Figure 13 represents the ¹H NMR spectrum (C₂D₂CI₄, 353 K) of sPS-b-aPS copolymer

Figure 14 represents a plot of melting enthalpy ΔΗ in terms of % of syndiotactic polystyrene present either in sPS/aPS blends, represented by squares or in sPS-b- aPS diblock copolymers represented by circles.

Figure 15 represents a plot of crystallisation enthalpy ΔΗ in terms of % of syndiotactic polystyrene present either in sPS/aPS blends, represented by squares or in sPS-b-aPS diblock copolymers represented by circles.

Figure 16 represents a plot of melting enthalpy ΔΗ (squares) and of melting temperature (circles) in terms of % of syndiotactic polystyrene present in sPS-b-aPS diblock copolymers. Figure 17 represents the DSC curve for a aPS/sPS blend of composition 90/10 (blend 6 in table 7).

Figure 18 represents the DSC curve for a aPS/sPS blend of composition 90/10 additivated with 15 wt%, based on the total weight of the mix, of sPS-b-aPS block copolymer (blend 1 1 in table 7).

Figure 19 represents a plot of the reciprocal value t_p ^"1 of the crystallisation peak time (t_p) as a function of crystallisation temperature Tc for syndiotactic polystyrene, represented by squares, and for a sPS-b-aPS copolymer with similar amount of aPS and sPS represented by circles.

Figure 20 represents a plot of the reciprocal value t_p ^"1 of the crystallisation peak time (t_p) as a function of crystallisation temperature Tc for syndiotactic polystyrene, represented by squares (■ ), for a aPS/sPS blend represented by triangles ( A), for a sPS-b-aPS block copolymer with similar amount of aPS and sPS blocks represented by circles ( · ), and for a blend aPS/sPS additivated with 15 wt%, based on the total weight of the mix, of sPS-aPS copolymer represented by rhombuses (♦) at a crystallisation temperature Tc of 244 °C.

Figure 21 represents an Avrami plot for syndiotactic polystyrene, represented by squares, and for a sPS-b-aPS copolymer with similar amount of aPS and sPS blocks represented by circles at a crystallisation temperature Tc of 245 °C.

Figure 22 represents an Avrami plot for syndiotactic polystyrene, represented by squares (■), for a aPS/sPS blend represented by triangles ( A), for a sPS-b-aPS block copolymer with similar amount of aPS and sPS blocks represented by circles ( · ), and for a blend aPS/sPS additivated with 15 wt%, based on the total weight of the mix, of sPS-aPS copolymer represented by rhombuses (♦) at a crystallisation temperature Tc of 244 °C.

Figure 23 represents the ¹H NMR spectrum (300 MHz, C₂D₂CI₄, 333K) of the crude sPS-Jb-aPS/PBD produced in Run 1 of Table 10. SUMMARY OF THE INVENTION.

It is an objective of the present invention to disclose techniques to access sPS-Jb- aPS block copolymers and rubber-toughened sPS-Jb-aPS.

It is also an objective of the present invention to disclose a direct-reaction and one- pot process yielding sPS-Jb-aPS, or impact-modified sPS-Jb-aPS.

It is another objective of the present invention to produce syndiospecific di(polystyryl) metallic species in a controlled manner.

It is yet another objective of the present invention to use such syndiospecific di(polystyryl) metallic species directly as macro-initiators for the free radical polymerisation of styrene.

It is a further objective of the present invention to end-cap syndiospecific di(polystyryl) metallic species with a nitroxide radical for the subsequent nitroxide- mediated polymerisation (NMP) of styrene.

It is yet a further objective of the present invention to use sPS-Jb-aPS copolymers as crystallisation accelerators in mixtures comprising atactic and syndiotactic polystyrenes.

A purpose of the present invention is the versatile production of semi-crystalline impact-modified polystyrene dedicated to applications requiring a high thermal and chemical resistance.

Accordingly, the foregoing objectives are realised as described in the independent claims. Preferred embodiments are disclosed in the dependent claims.

DETAILED DESCRIPTION OF THE INVENTION.

The present invention discloses a method for preparing polystyrene materials containing a syndiotactic polystyrene block linked to an atactic polystyrene block, namely sPS-Jb-aPS block copolymers, wherein a metal-based catalyst component of formula [L_nXx]M¹R_n, alone or in combination with a compound M²(R')_n' acting as a chain transfer agent (CTA), initiates the stereospecific chain-growth polymerisation of styrene, and wherein M¹ is a metal selected from Group 3-5 of the Periodic Table, LnXx is a monanionic or a dianionic ligand selected from cyclopentadienyl-type ligands and related compounds or a phenolate or an amido-type ligand, all of these ligands possibly bearing additional donor groups, M² is an element selected from Group 1 to 13 of the Periodic Table, R is hydrogen or an alkyl or allyl or benzyl group having up to 12 carbon atoms, R' is hydrogen or an alkyl or allyl or benzyl group having up to 12 carbon atoms, n and n' are an integers ranging from 1 to 4 depending on the nature and oxidation state of the metal M¹ and M², respectively, and of the nature of the L_nX_x ligand.

The different possibilities of the present method are schematically represented in Scheme 5, wherein Δ means heat.

[LnXx]M R Ph Phi

a) Chain growth polymerisation M²(R)n ^{► M}^

(cat) b) chemical fu notion alisation c) Synthesis of block copolymer

Scheme 5 As illustrated in Scheme 5, the M²(R)_n' compound, which acts as a chain-tranfer agent (CTA), leads to the formation of M²(sPS)_n' species. This intermediate product can be used as such, but it can optionally be end-capped with (a) a nitroxide radical thereby allowing the nitroxide-mediated polymerisation (NMP) of styrene, or (b) with dioxygen in the case of boryl-sPS species leading to boryl-O-O-sPS species, allowing radical polymerisation upon simple heating (Δ), , or (c) with halogens thereby allowing either atom transfer radical polymerisation (ATRP) or reversible iodine transfer polymerisation (RITP) of styrene.

LnXx is preferably selected from cyclopentadienyl-type ligands of general formula C5R"5 where R" are equal or different and selected from hydrogen, alkyl, aryl, trialkylsilyl or hetero-functionalized substituents, and all related Cp-type ligands such as indenyl and fluorenyl derivatives, substituted or not. L can also be a non- cyclopentadienyl ligand and selected from regular ligands used in post-metallocenes derivatives, for instance imino-phenolate derivatives, amido derivatives and all combinations derived from such phenolate and amido derivatives, with possible donor functionalities including imino, alkoxy, amino... . groups.

M¹ is preferably selected from Nd, Y, Sc, Ti, Zr

M² is preferably selected from Mg, Zn, Al, B

R is preferably selected from methyl or higher alkyl groups such as ethyl, butyl, hexyl and octyl, benzyl, allyl (C3H5) or allyl groups substituted at the 1 and/or 3 positions

R' is preferably selected from methyl or higher alkyl groups such as ethyl, butyl, hexyl and octyl, or benzyl groups, or allyl (C3H5) groups, or allyl groups substituted at the 1 and/or 3 positions

Among the preferred metal-based catalysts, one can cite {CpCMe₂Flu}Nd(allyl) (where Cp = C5H4 and Flu = 9-fluorenyl), or (C5Me₅)Ti(Bn)₃ activated by a stoichiometric amount of a Lewis acid such as B(C6F₅)₃ or [Ph₃C][B(C6F₅)4] or [HNMe₂Ph][B(CeF₅)4], or [{C5Me4SiMe3}Sc(CH2SiMe3)]⁺[SiMe₃CH2B(C6F5)3]^"

In a first embodiment according to the present invention, the method for preparing the aPS-b-sPS copolymer comprises the steps of: a) growing a first syndiotactic sPS block in the presence of a metal-based catalyst system [L_nXx]M¹ R_n, and a M²(R')_n' chain transfer agent to generate a M²(sPS)n' product;

b) end-capping the sPS chains of the latter product with a suitable agent such as a nitroxide radical and converting to two sPS-nitroxide blocks;

c) growing an atactic aPS block from the latter functionality and forming an sPS- jb-aPS polymer.

This embodiment can be schematically represented in Scheme 6 wherein M¹ is Nd M² is Mg and the functional group capped on sPS chain ends is a nitroxide introduced via TEMPO.

Scheme 6

Polymerisations are carried out at temperatures ranging from 20 °C to 150 °C, more preferably from 60 °C to 130°C. Functionalisation reactions are carried out at temperatures ranging from 20 °C to 150 °C, more preferably from 60 °c to 130 °C. In a second embodiment according to the present invention, the method for preparing the aPS-b-sPS copolymer can be carried out in a one-pot process and comprises the steps of: a) growing a first syndiotactic sPS block in the presence of a metal-based catalyst system [L_nXx]M¹ R_n, and a M²(R')_n' chain-transfer agent to generate a M²(sPS)n' product;

b) deactivating the catalyst;

c) chain-growing a second atactic aPS block by a thermal, non catalytic process to prepare a M²(aPS-£>-sPS)_n product and forming a aPS-Jb-sPS copolymer upon hydrolysing/quenching of the previous M²(aPS-£>-sPS)_n product.

It can be schematically represented in Scheme 7 wherein M¹ is Nd and M² is Mg.

(cat)

Scheme 7

The selective deactivation of the Nd catalyst, required for growing aPS segments onto Mg(sPS)2, and thus for producing aPS-Jb-sPS stereoblock copolymers, is carried out with ethers, nitriles, simple amines or other species coordinating to lanthanides.

This deactivation step is necessary to stop the production of sPS that otherwise takes over that of aPS. Both polymerisation steps are carried out at temperatures ranging from 20 °C to 160 °C, more preferably from 80 °C to 150 °C.

In a third embodiment according to the present invention, the method for preparing the aPS-Jb-sPS copolymer can be carried out in one-pot process and comprises the steps of: a) chain growing a first atactic aPS block by a thermal process in the presence of a M²(R')n' chain transfer agent to generate a M²(aPS)_n' product;

b) chain-growing a second sPS block in the presence of a metal-based catalyst system [L_nXx]M¹ R_n, and the M²(aPS) product of step a) to generate M²(sPS-Jb- aPS)n' product and forming a aPS-Jb-sPS block upon hydrolysing/quenching of the previous M²(aPS-Jb-sPS)_n product.

It can be schematically represented in Scheme 8 wherein M¹ is Nd and M² is Mg.

(cat)

Scheme 8

Both polymerisation steps are carried out at temperatures ranging from 20 °C to 160 °C, more preferably from 80 °C to 150 °C.

In a fourth embodiment according to the present invention, the method for preparing the aPS-Jb-sPS block copolymer comprises the steps of: a) growing a first sPS block in the presence of a metal-based catalyst system [L_nXx]M¹ Rn, and in situ end-capping said sPS block with a halogen atom; b) growing an aPS block via atom transfer radical polymerisation (ATRP) or reversible iodine transfer polymerisation (RITP) and eventually forming sPS-Jb- aPS copolymers.

It can be schematically represented in Scheme 9 wherein the halogen is bromine and M¹ is Ti, N-bromosuccinimide (NBS) is the bromination agent, and CuBr and pentamethyldiethylenetriamine (PMDETA) are the components necessary to achieve the ATRP step.

anisole, 1 30°C

Scheme 9

The stereospecific polymerisation step is carried out at temperatures ranging from -20 °C to 160 °C, more preferably from 0 to 100 °C, in an hydrocarbon solvent such as heptanes, toluene and xylenes. The ATRP polymerisation step is carried out at temperatures ranging from 60 °C to 180 °C, more preferably from 100 °C to 150 °C and in an organic solvent, preferably an ether such as anisole.

As can be understood from the methods used in the present invention, the syndiospecific and atactic blocks are grown sequentially by different methods. The length of the sPS is selected when catalysed with a metal-based catalyst system and that of the aPS block is selected during the radicalar graft of styrene onto the sPS block. The present methods thus produce sPS-b-aPS block copolymers that can be accurately tailored.

Optionally, the same approaches can be used using a rubber in styrene solution as starting material, preferably a polybutadiene in styrene solution. This approach leads to impact-modified semi-crystalline polystyrene composites. In such process, the metallocene polymerisation of the rubber-in-styrene solution produces first a sPS and rubber in styrene solution. The subsequent radical polymerisation of the remaining styrene is used advantageously to generate in situ some polystyrene- grafted rubber chains that stabilise the final rubber in polystyrene composites which contains crystalline domains resulting from the partial crystallisation of the syndiotactic PS produced in the first step. This versatile and inexpensive process leads consequently to impact-modified semi-crystalline styrenic composites that can compare with polyamide or ABS in terms of impact resistance, ductility and stress- cracking resistance.

Impact-resistant semi-crystalline polystyrene can thus be obtained by using sPS-Jb- aPS block copolymer as a crystallisation accelerator in the preparation of sPS/HiPS blends.

This impact-resistant semi-crystalline polystyrene can advantageously be used in applications requiring a high heat and/or chemical resistance, such as the manufacturing of refrigerator liners, electrical & electronic appliances or automotive parts.

The sPS-b-aPS block copolymers of the present invention are characterised by a faster crystallisation rate than that of homopolymers of styrene. It is further observed that:

• The more sPS in the stereoblock, the faster the crystallisation.

• At a given sPS/aPS ratio, the diblock copolymers crystallise faster than blends for comparable Mn values.

• The addition of from 5 to 15 wt%, based on the weight of the total mixture, of stereoblock copolymers sPS-Jb-aPS to aPS/sPS blend increases the crystallisation rate of the mixture with respect to an aPS/sPS blend having the same percentage of sPS as the mixture. aPS-Jb-sPS block copolymers can thus be used as accelerators in the polymerisation of styrene or in sPS/aPS blends, wherein the crystallisation rate increases with increasing amount of sPS in the block copolymer. EXAMPLES.

Materials.

HPLC grade heptanes, toluene and m-xylene were purchased from VWR. Dry diethyl ether was purchased from Aldrich. These solvents were distilled under argon from a sodium mirror prior to use.

Technical grade dichloromethane (DCM), chloroform, acetone, acetonitrile, methylbutylketone (MEK) and methanol were purchased from VWR and used to precipitate, wash and extract the synthesised polystyrenes and their corresponding copolymers.

The initiator {Me₂C(Cp)(Flu)}Nd(1 ,3-(SiMe₃)2-C₃H₃) was prepared with all steps performed under argon accoding to the method disclosed by Rodrigues et al. (Rodrigues, A. S.; Kirillov, E.; Roisnel, T.; Razavi, A.; Vuillemin, B.; Carpentier, J.-F. Angew. Chem. Int. Ed. 2007, 46, 7240). The {Me₂C(Cp)(Flu)}H₂ pro-ligand (1.94 g, 7.12 mmol, 1 equivalent(s)), provided by Total Petrochemicals, was dissolved in diethyl ether (10 ml_), cooled down at 0 °C prior addition of n-BuLi (8.9 ml_ of a 1 .6 M solution in hexanes, 14.2 mmol, 2 equivalent(s)). The mixture was stirred vigorously overnight at room temperature and NdCl3(THF)_{3 5} (2.81 g, 7.12 mmol, 1 equivalent(s)) was added via a bent pipe as a light blue powder. The reaction mixture was stirred for 1 day at room temperature, turning red, and diethyl ether was replaced by toluene (10 ml_) prior addition of 1 ,3-bis(trimethylsilyl)allyl potassium (1 .60 g, 7.13 mmol, 1 .01 equivalent(s)). After another day of stirring at room temperature, the solution was filtered, the solids were washed with pentane (2 x 10 ml_) and the deep red solution was evaporated in vacuo to afford a sticky solid. It was then triturated twice with heptanes to afford a crimson brittle powder after extensive drying (Yield: 3.77 g, 74%).

TEMPO (2,2,6,6-tetramethylpiperidine-N-oxyl) was purchased from Aldrich and sublimated under static vacuum at 40 °C prior to use. Styrene (99.5%) was supplied by Aldrich, dried over CaH₂ overnight, distilled by heating at a temperature of 50 °C under dynamic vacuum at 10^"2 Bar and stored at a temperature of 4 °C away from light under argon.

Pentamethylcyclopentadienyltitanium(IV)tribenzyl (Cp^*TiBn₃) was synthesised according to the method described by Mena et al. (Mena, M.; Royo, P.; Serrano, R. Organometallics 1989, 8, 476-482).

Trispentafluorophenylborane B(C6F₅)₃ was synthesised according to the method disclosed by Massey et al. (Massey, A.G.; Park, A. J. J. Organomet. Chem. 1964, 2, 245-250).

Diethylzinc (ZnEt₂, 1 .0 M solution in hexane), dibutylmagnesium (MgBu₂, 1 .0 M solution in heptane), triethylborane (BEt₃, neat), trimethylaluminum (AIMe₃, 2.0 M solution in hexane), triisobutylaluminum (AI(iBu)₃, neat) were purchased from Aldrich and used as received. Trisoctylaluminum (AIOct₃, 0.3 M solution in heptane) was purcheased from Acros and used as received. The molarity of these reagents was periodically checked by titration following reported procedures, for instance in the case of MgBu₂ with a stock solution of sec-butanol in toluene in the presence of phenantroline as disclosed in watson et al. (Watson, S. C; Eastham, J. F. J. Organomet. Chem. 1967, 9, 165). N-bromosuccinimmide (NBS), CuBr, 1 , 1 ,4,7,7- pentamethyldiethylenetriamine (PMDETA) were purchased from Aldrich, while 1 , 1 ,2,2-tetrachloroethane-d₂ was purchased from Acros.

The typical procedures used in all steps of the four embodiments are broadly disclosed in the following paragraphs. Detailed experimental procedures will then be described.

Catalytic polymerisation of styrene in the presence of MgR?.

In a glove-box, 13.3 mg of the neodymium catalyst, (22 pmol, 1 equivalent(s)) were placed into a Schlenk flask. All subsequent operations were performed on a Schlenk line, using standard techniques. The required amounts of solvent (25 ml_) and styrene (25.4 ml_, 0.222 mol, 10 000 equivalent(s) with respect to Nd) were added by syringe and the resulting reddish solution was heated for 30 s before the addition of transfer agent (MgBu₂) (220 pmol, 10 equivalent(s) with respect to Nd) via a canula. The reaction mixture was heated during the appropriate period of time at steady temperature leading to a dark orange solution. In order to analyse the sPS blocks produced, the polymerisation was quenched by addition of a dichloromethane solution containing 5% of acidified methanol (1 M/HCI). Polystyrenes were fully dissolved in dichloromethane (or chloroform), precipitated and washed with methanol to be dried out at 70 °C under vacuum.

End-capping of Mg(sPS)? with TEMPO.

The temperature of a first Schlenk flask containing the propagating species Mg(sPS)2 (1 .57 mL of a solution obtained as described above and being not quenched with acidified methanol, 13.7 mmol, 1 equivalent(s)), in toluene or m- xylene (1 .5 mL), was placed into an oil bath at 85 °C and kept under argon on a Schlenk-line. In the glove box, TEMPO (188 mg, 1 .23 mmol, 4.5 equivalent(s) With respect to Mg(sPS)2) was placed into a second Schlenk flask and transferred further via a bent pipe to the first Schlenk flask. The reaction mixture was stirred for 3 h, cooled down to room temperature. For analytical purposes, the mixture was quenched by addition of a dichloromethane solution containing 5% of acidified methanol (1 M/HCI). Polystyrene-TEMPO materials were fully dissolved in dichloromethane (or chloroform), precipitated and washed with methanol to be dried out at 70 °C under vacuum.

Copolvmerisation of sPS-TEMPO with styrene via NMP and formation of sPS-b-aPS according to the first embodiment of the invention.

All subsequent operations were performed on a Schlenk line, using standard techniques. In a typical experiment, the air stable macro-initiator syncf/o-polystyryl- TEMPO (sPS-TEMPO) (500 mg, 230 pmol, 1 equivalent(s)) was placed into a Schlenk flask under argon. The required amounts of m-xylene (13 mL) and styrene (13 mL, 0.134 mol, 500 equivalent(s) with respect to sPS-TEMPO) were added by syringe and the resulting solution was heated up overnight at 135 °C. The polymerisation was stopped by cooling down the reaction medium to room temperature. Copolymers were fully dissolved in dichloromethane (or chloroform), precipitated and washed with methanol to be dried out at 70 °C under vacuum.

Thermal polymerisation of styrene in the presence of MgBu? and formation of Mg(aPS)? All operations were performed under argon on a Schlenk line using standard techniques. In a typical experiment, the required amounts of m-xylene (5 ml_), styrene (4.6 ml_, 40 mmol, 200 equivalent(s)) and MgBu₂ (200 μΙ_, 200 μηποΙ, 1 equivalent(s)) were added by syringe and the resulting solution was heated up overnight at 145 °C. The polymerisation was stopped by cooling down the reaction media to room temperature and, for analytical purposes, the Mg(aPS)2 chains were converted to two distinct aPS chains by addition of a solution of dichloromethane containing 5% of acidified methanol (1 M/HCI). Polystyrenes were fully dissolved in dichloromethane precipitated and washed with methanol to be dried out at 70 °C under vacuum.

Thermal copolymerisation of Mg(sPS)? with styrene and formation of sPS-b-aPS according to the second embodiment of the invention.

All operations were performed on a Schlenk line, using standard techniques. In a typical experiment, a desired amount of styrene (20 ml_, 174 mmol, 500 equivalent(s)) was added by syringe into a Schlenk flask containing the propagating species Mg(sPS2) (10 ml_ of the solution prepared as described above, 87 mmol, 1 equivalent(s)) in m-xylene (30 ml_). The reaction mixture was stirred overnight at 145 °C then quenched at room temperature by addition of a dichloromethane solution containing 5% of acidified methanol (1 M/HCI). Polystyrenes were fully dissolved in dichloromethane (or chloroform), precipitated and washed with methanol to be dried out at 70 °C under vacuum.

Chain-growth copolymerisation of a aPS block with Mg(sPS)? and formation of aPS- b-sPS according to the third embodiment of the invention.

All operations were performed on a Schlenk line, using standard techniques. In a typical experiment, the temperature of a first Schlenk flask containing the propagating species Mg(aPS2) (9.2 ml_ of a solution prepared as described above, 80 mmol, 2 500 equivalent(s)) in m-xylene (20 ml_) was set at 50 °C. A desired amount of neodymium initiator (15.9 mg, 27 pmol, 1 equivalent(s)) and styrene (200 μΙ_, 200 μηιοΙ, 8 equivalent(s)) were added, leading to a clear orange solution. The reaction mixture was stirred overnight at the desired temperature. For analytical purposes, the mixture was quenched to room temperature by addition of a dichloromethane solution containing 5% of acidified methanol (1 M/HCI). Polystyrenes were fully dissolved in dichloromethane (or chloroform), precipitated and washed with methanol to be dried out at 70 °C under vacuum.

Polymerisation of styrene using Cp^*TiBn₃/B(CeF5)₃ catalyst system in combination with chain transfer agents selected from ZnEt?, MgBu?, BEt₃, AIMe₃, AliBu, or AIOct₃.

All manipulations of air and/or water-sensitive compounds were performed under a argon atmosphere using standard Schlenk techniques or glove-box. In a typical polymerisation procedure, a 100 mL flask, equipped with a magnetic bar, was charged sequentially with toluene (30 mL), styrene (5.00 mL) and chain transfer agent (0.88 mmol). The resulting solution was thermostated at 27 °C and the polymerisation was started by injection of the catalytic solution (dark red solution), prepared by adding a colourless solution of B(C6F₅)₃ (23 mg, 44 pmol in 2 mL of toluene) to a red solution of Cp^*TiBn₃ (20 mg, 44 pmol in 3 mL of toluene). After 12 min, the mixture was poured into acidified methanol. When present, the polymers were recovered by filtration and dried at 45 °C in a vacuum oven.

Polymerisation of styrene, in situ functionalisation with bromine atoms and formation of sPS-Br according to the first step of the fourth embodiment of the invention.

All manipulations of air and/or water-sensitive compounds were performed under an argon atmosphere using standard Schlenk techniques or glove-box. In a typical polymerisation procedure, a 100 mL flask, equipped with a magnetic bar, was charged sequentially with toluene (33 mL), styrene (1.00 mL) and AI(Oct)₃ (132 pmol). After equilibration of the resulting clear solution at the required polymerisation temperature, the reaction was started by injecting the catalytic solution (dark red solution), prepared by adding a colourless solution of B(C6F₅)₃ (69 mg, 132 pmol in 2 mL of toluene) to a red solution of Cp^*TiBn₃ (60 mg, 132 pmol in 3 mL of toluene). The polymerisation was terminated after the prescribed time by adding the brominating agent NBS (N-bromosuccinimide, 6.6 mmol) and keeping the mixture under stirring for 1 h at room temperature. The polymer was coagulated in acidified methanol, recovered by filtration, washed with large excess of acetonitrile, and dried in vacuo at 45 °C. ATRP of styrene and formation of sPS-b-aPS according to the second step of the fourth embodiment of the invention.

All operations were performed on a Schlenk line, using standard techniques. In a typical experiment, a 25 mL flask equipped with a magnetic stir bar was charged sequentially with brominated sPS (0.200 g, 1 1.6 pmol), CuBr (35 pmol), anisole (10 mL), pentamethyldiethylenetriamine (PMDETA) (35 pmol) and styrene (1.00 mL). In order to remove the oxygen, 3 freeze-pump-thaw cycles were performed and finally the flask was filled with argon. The resulting degassed suspension was stirred for 10 min. at room temperature and then was placed in an oil bath at 130 °C for the requested polymerisation time. The reaction was terminated by pouring the clear homogeneous solution into a large excess of acidic methanol. The copolymer was recovered by filtration, washed with fresh methanol, and dried under vacuum at 45°C.

Extraction of a PS containing a syndiotactic block.

Approximately 150 mL of boiling MEK was used to extract ca. 5 g of (co)polymers placed in a Soxhlet cartridge. All soluble material was recovered by evaporating the MEK, and both soluble and not soluble polymers were dried out at 70 °C under vacuum to be weighted.

Extraction of atactic fraction from sPS-b-aPS copolymers.

In a typical experiment, 25 mL of acetone were added in a vial containing ca. 0.3 g of copolymer. The resulting suspension was stirred at room temperature overnight. The insoluble white solid was recovered by filtration, while the soluble portion was obtained by evaporating the acetone. Both soluble and non soluble fractions were dried out at 45 °C under vacuum to be weighted and analysed by NMR spectroscopy

Polymer characterisation Molecular weights (M_n and M_w) and polydispersities (M_w/M_n) of polystyrene and copolymers were determined by high temperature gel permeation chromatography (GPC) using Waters GPC-V2000 Rl detector equipped with a PL GEL mixed-B column. The measurements were recorded at 150 °C at 1 mL/min using 1 ,2,4- trichlorobenzene as solvent stabilised with ditertbutyl-methyl-4-phenol.

Melting points (T_m) of the polystyrene and copolymers were measured by differential scanning calorimetry (DSC) using a DSC 131 Setaram instrument in argon flow with a heating and cooling rate of 10 "C.min^"1 in the range 30 °C to + 300 °C. Melting temperatures were reported for the second heating cycle.

NMR spectra of polymers and copolymers were recorded on a Bruker AM-500 spectrometer in 1 , 1 ,2,2-tetrachloroethane-d₂ at several temperature 353K, 333K, 323K and reported relative to tetramethylsilane.

Example 1 .

Polymerisation of styrene with (Me?CfCp)(Flu)Nd(1 ,3-(SiMea)?-CaHa))/ MgBu? and formation of Mg(sPS)?

A first series of experiments was undertaken to assess the efficiency of the {Me₂C(Cp)(Flu)Nd(1 ,3-(SiMe3)2-C₃H3)} / MgBu₂ system to synthesise sPS blocks in chain growth polymerisation and yielding Mg(sPS)₂ products. Different solvents and temperatures were tested in order to reach the highest possible molecular weight M_n.

5 000 equivalent(s) of styrene with 15 equivalents of MgBu₂ and 1 equivalent of Nd- initiator (5 000/1/15) were polymerised quantitatively, at 100 °C, in toluene, m- xylene, heptanes or bulk. These polymerisations were quite controlled with an observed number average molecular weight M_n comprised between 15 600 and 19 000 g.mol^"1, close to the theoretical value of 17 300 g.mol^"1, and a polydispersity index (PDI) value around 1 .80, wherein PDI is defined as the ratio M_w/M_n of the weight average molecular weight Mw to the number average molecular weight Mn. By opposition to heptanes or bulk conditions, aromatics solvents allowed polymerisations with soluble sPS and mechanical stirring (homogenous conditions).

10 000 equivalent(s) of styrene with 15 equivalent(s) of MgBu₂ and 1 equivalent(s) of Nd catalyst (10 000/1/15) were homogeneously polymerised quantitatively in aromatic solvents at 100 °C. Syndiotactic polystyrenes exhibited a M_n around 17 500 g.mol^"1, lower than the theoretical value of 34 600 g.mol^"1, with some PDI values ranging from 1 .70 to 2.45. It is worth mentioning that these molecular weights were obtained from RT SEC and might represent only a fraction of soluble sPS from the whole sample. High temperature SEC analysis should give some M_n values closer to 34 600 g.mol^"1.

Higher loadings of styrene, with more than 10 000 equivalent(s) of monomer with respect to Nd-initiator, led to controlled polymerisation, as long as the ratio styrene to MgBu₂ remained smaller than 350. This value of 350 seems to be the maximum ratio allowing efficient transfer with magnesium. Polymerisations carried out with 50 000 equivalent(s) of styrene often failed due to the extreme sensitivity of the initiator.

Polymerisations proceeded faster at high temperature and transfer to magnesium was strongly enhanced. The initiator and MgBu₂ were stable up to a temperature of about 120 °C.

In summary, these experiments established that:

^■ The system {Me₂C(Cp)(Flu)Nd(1 ,3-(SiMe₃)2-C₃H₃)} / MgBu₂ was very efficient to synthesise syndiotactic polystyrene blocks grafted onto magnesium, i.e., Mg(sPS)₂.

^■ The reaction was tolerant to aromatic solvents such as toluene or xylenes, allowing homogenous polymerisation at temperature of 100 °C and above.

^■ Polystyrenes with M_n up to 20 000 g.mol^"1 were characterised by room temperature (RT) SEC and polymerisations remained controlled as long as the concentration of MgBu₂ reached a minimum value necessary to keep the transfer efficient.

The results are displayed in Table 1 . TABLE 1.

Example 2

End-capping of Mg(sPS)? with TEMPO, according to the first embodiment of the invention.

The efficiency of TEMPO to cap some Mg(sPS)2 chains was tested in a series of examples. The reaction conditions for the capping of polyethylene were carried out as described by D'Agosto et al (Lopez, R. G.; Boisson, C; D'Agosto, F.; Spitz, R.; Boisson, F.; Bertin, D.; Tordo, P. in Macromolecules 2004, 37, 3540) They are based on the reactivity of Grignard and dialkylmagnesium compounds towards radicals. An aliquot of growing chains, having a low molecular weight M_n to ensure their full solubility in THF, was analysed by SEC prior to and after addition of TEMPO. The reaction of three batches of Mg(sPS)2 chains having respectively M_n = 2 100 g.mol^"1 and PDI = 1 .91 , M„ = 6 690 g.mol^"1 and PDI = 1 .57, and M_n = 10 140 g.mol^"1 and PDI = 2.04 led to three new polymers with M_n = 2 200 g.mol^"1 and PDI = 1 .71 , Mn = 6 690 g.mol^"1 and PDI = 1 .58, and M_n = 1 1 070 g.mol^"1 and PDI = 2.04. SEC traces did not show other PS distribution. This ruled out side-reactions, such as coupling between PS growing chains, during the capping process.

A MALDI-ToF-MS analysis of the polymers after reaction with TEMPO revealed two distributions as seen in Figure 1 . One of these distributions fitted the expected capped sPS-TEMPO macromolecules while the second fitted the sPS-H chains. The latter chains could arise from hydrolysis of Mg(sPS)2 during functionalisation with TEMPO and/or from fragmentation of sPS-TEMPO during the MS analysis. This analysis revealed the presence of the targeted sPS-TEMPO materials but was not an indication of the efficiency of the functionalisation reaction.

These experiments thus established that the capping of Mg(sPS)2 with TEMPO occurred without undesired coupling reaction with growing chains, but the efficiency of this functionalisation step could not be assessed by MALDI-ToF-MS techniques.

Example 3.

Synthesis of sPS-b-aPS via NMP of styrene onto sPS-TEMPO according to the first embodiment of the invention.

A series of experiments was carried out to assess the efficiency of sPS-TEMPO products to initiate the growth of an atactic polystyrene block (aPS) and form a sPS- jb-aPS copolymer. Reaction conditions described by Hawker et al. for living NMP of styrene with TEMPO were used. The copolymerisations were monitored by SEC analysis.

The NMP of 100 equivalent(s) of styrene with 1 equivalent(s) of sPS-TEMPO (M_n = 3 620 g.mol^"1 and PDI = 1 .34) proceeded quantitatively at a temperature of 130 °C and yielded two populations of polystyrene respectively with M_n = 4 900 g.mol^"1, PDI = 1 .41 (49%) and with M_n = 35 780 g.mol^"1, PDI = 1 .29 (51 %). The first polymer distribution had its M_n close to that of the starting sPS-TEMPO and likely arose from some sPS-H chains formed during the synthesis of PS-TEMPO. The second polymer had its M_n way above the expected M_n at 14 000 g.mol^"1. This can be explained by the fact that there is less macro-initiator than expected, due to the presence of inactive sPS-H chains. Both distributions overlapped on the SEC, but NMP was controlled and the (co)polymer sPS-Jb-aPS exhibited a narrow M_n distribution characterised by a low PDI as seen in Figure 2.

The NMP of 500 equivalent(s) of styrene with 1 equivalent(s) of sPS-TEMPO (M_n = 2 200 g.mol^"1; PDI = 1 .71 ) yielded also two distributions of polystyrene respectively with M_n = 3 840 g.mol^"1, PDI = 1 .77 (22%) and with M_n = 90 740 g.mol^"1, PDI = 1 .92 (78%) likely for the same reasons as described hereabove. It is represented in Figure 3.

These experiments thus established that the NMP of sPS-TEMPO with styrene was controlled and led smoothly to sPS-Jb-aPS copolymers and that the sPS-H chains issued from the synthesis of sPS-TEMPO were not active in NMP and led to a blend of homo-sPS and sPS-Jb-aPS.

Example 4

Thermal polymerisation of styrene with MgBu? and formation of Mg(aPS)?

A series of experiments was carried out to monitor the thermal polymerisation of styrene in the presence of MgBu₂. Different styrene/MgBu₂ ratios were tested in order to increase the molecular weight M_n of the aPS blocks in Mg(aPS)2 product.

200 equivalent(s) of styrene with 1 equivalent(s) of MgBu₂ were polymerised with a yield of 89% after a period of time of 21 h. The aPS blocks exhibited a M_n of about 19 730 g.mol^"1, somewhat larger than the theoretical value of 14 300 g.mol^"1. They had a polydispersity index equal to 2.61 .

Polymerisations with higher loadings of styrene of up to 800 equivalent(s) with respect to MgBu₂, afforded Mg(aPS)₂ blocks with PDIs ranging from 2.50 to 3.00 and molecular weights of aPS blocks higher than expected. MgBu₂ did not initiate the polymerisation in these experiments, but acted as chain transfer agent (CTA), narrowing the PDI below 3 for this type of free radical polymerisation. Under these conditions, the polymerisations were quite sluggish and the yields decreased with decreasing quantities of styrene for the same reaction time as seen in Table 2.

TABLE 2.

These experiments established that aPS blocks with M_n of up to 60 000 g.mol^"1 were easily produced by slow thermal polymerisation of styrene in the presence of MgBu₂ acting as chain transfer agent (CTA) and preventing extreme broadening of the PDIs. The Mn values reached were higher than expected, reflecting moderate transfer efficiency.

Example 5

Thermal polymerisation of styrene with Mg(sPS)? according to the second

embodiment according to the present invention.

A series of experiments was carried out to assess the efficiency of thermal polymerisation of styrene in the presence of Mg(sPS)2 to form a sPS-Jb-aPS copolymer, after neutralisation of the {Me₂C(Cp)(Flu)Nd(1 ,3-(SiMe₃)2-C₃H₃)} catalyst. Different amounts of styrene and Mg(sPS)2 with variable chain sizes were chosen to change the composition of the resulting di-block copolymers.

100 equivalent(s) of styrene and 1 equivalent(s) of Mg(sPS)2 (M_n = 4 630 g.mol^"1 and PDI = 1 .81 ) were polymerised quantitatively at a temperature of 145 °C and yielded a sPS-Jb-aPS copolymer having M_n = 44 470 g.mol^"1 and PDI = 3.07. The chromatogram trace showed a clean takeover of the radical polymerisation over the chain-growth polymerisation along with a small amount of terminated chains at 4 630 g.mol^"1 arising from Mg(sPS)2 as can be seen in Figure 4.

500 equivalent(s) of styrene and 1 equivalent(s) of Mg(sPS)2 (M_n = 8 810 g.mol^"1 and PDI=1 .70) were polymerised quantitatively at a temperature of 145 °C and yielded a sPS-Jb-aPS copolymer having M_n = 32 990 g.mol^"1 and PDI = 2.87. The chromatogram trace also showed a clean increase of the weight with minor sign of terminated chains as seen in Figure 5.

These experiments established that sPS-Jb-aPS copolymers could be synthesised in high yields with minor termination side-reactions by thermal polymerisation of styrene in the presence of Mg(sPS)2, after deactivation of the Nd, and that these experimental results could be reproduced.

The selective deactivation of the Nd catalyst, required for growing aPS segments onto Mg(sPS)2, and eventually producing aPS-Jb-sPS stereoblock copolymers, was possible using ethers, nitriles, simple amines and other species coordinating to lanthanides.

Example 6

Catalytic chain-growth copolymerisation of styrene with Mg(aPS)? according to the third embodiment according to the present invention.

A series of experiments was carried out to assess the efficiency of the chain-growth polymerisation of styrene in the presence of {Me₂C(Cp)(Flu)Nd(1 ,3-(SiMe₃)2-C3l-l3)} as catalyst and Mg(aPS)2, to form eventually a stereoblock copolymer aPS-Jb-sPS. Different amounts of styrene and Mg(aPS)2 with variable chain sizes were chosen to change the composition of the block-copolymers.

20 180 equivalent(s) of styrene, were polymerised quantitatively (M_n = 5 020 g.mol^"1 and PDI = 2.04) in presence of 81 equivalent(s) of Mg(aPS)2 and 1 equivalent(s) of neodymium catalyst at 100 °C and yielded, after hydrolysis, a aPS-Jb-sPS copolymer with Mn = 12 030 g.mol^"1 and PDI = 2.04. The chromatogram trace showed a clean takeover of the chain-growth polymerisation over the radical polymerisation; no terminated chains at 5 020 g.mol^"1 from Mg(aPS)2 were detected as seen in Figure 6. 2 500 equivalent(s) of styrene (M_n = 10 560 g.mol^"1 and PDI = 1 .73) were polymerised quantitatively in presence of 17 equivalent(s) of Mg(aPS)2 and 1 equivalent(s) of neodymium catalyst at 100 °C and yielded, after hydrolysis, a aPS-Jb- sPS copolymer with M_n = 10 970 g.mol^"1 and PDI = 1 .68. The chromatogram trace showed a clean takeover of the chain-growth polymerisation over the radical polymerization as seen in Figure 7. It must be noted that the copolymer was poorly soluble in THF at room temperature, as expected from the presence of large sPS blocks.

These experiments established that aPS-Jb-sPS copolymers could be synthesised in high yields by catalytic polymerisation of styrene in the presence of Mg(aPS)2. No evidence of significant termination side-reactions was observed and the experimental results were successfully reproduced.

Example 7.

Extraction and DSC analysis of the copolymers.

The extraction in boiling MEK of different sPS-Jb-aPS and aPS-Jb-sPS copolymers led to soluble fractions ranging from 7 to 42 wt-%, whereas the extraction of homo-sPS gave roughly 5 wt-% of soluble material. The aPS block of the sPs-Jb-aPS copolymers likely increased the solubility in MEK of the diblock materials, as compared to homo-sPS polymers. Moreover, the percentages (wt-%) of soluble PS were low compared to the quantities of styrene required for a second atactic block and were consistent with the formation of block copolymers instead of mixtures of two homopolymers.

DSC analyses of the copolymers confirmed the presence of crystalline sPS blocks with melting temperatures ranging from 253 °C to 265 °C as seen in Figure 8.

Example 8.

Chain growth polymerisation of styrene with Cp^*TiBn₃/B(CeFs)3 system and several chain transfer agents (CTA)

A first series of experiments was undertaken to search a suitable chain transfer agent to be used in combination with the Cp^*TiBn₃/B(C6F₅)₃ catalyst, in order to produce sPS blocks via chain growth polymerisation. Different CTAs and different titanium/CTA ratios were evaluated. Diethylzinc, dibutylmagnesium and triethylborane turned out to be poisons for the catalytic system and, in all cases, the solution immediately changed colour once the dark red catalytic solution was added to the solution containing the CTA. It changed from red to deep yellow, green and clear yellow, respectively. On the contrary, some AIR₃ compounds proved to be compatible with the Cp^*TiBn₃/B(C6F₅)₃ catalyst system. Therefore, several polymerisations were performed by changing the Al/Ti ratio.

Among the aluminium compounds investigated, trimethylaluminium, used in ratio 20: 1 with respect to titanium, resulted in depressing the polymerisation activity by almost 90%. With 50 equivalent(s) of AIMe₃ the catalytic system became inactive in the polymerisation of styrene. With 20 equivalent(s) of AI(Oct)₃ (Oct = n-octyl), the polymerisation of styrene proceeded more efficiently and an increase of the conversion of 34%, as well as an increase of the number of polystyrene chains per titanium moiety up to 1 .9 were observed. This compares advantageously with the results of 29% conversion and 0.5 polystyrene chains per titanium moiety obtained in the presence of other CTAs. With 100 equivalent(s) of AI(Oct)₃ with respect to Cp^*TiBn₃/B(C6F₅)₃, the polymerisation of styrene was completely inhibited.

In the case of AI(iBu)₃, the catalytic system Cp^*TiBn₃/B(C6F₅)₃ remained active for up to 200 equivalent(s) of Al compound. A progressive decrease of both the conversion and molecular weight was observed by increasing the Al/Ti ratio. The polydispersity index remained unchanged between 2.2 and 2.3, as well as the number of polystyrene chains per titanium moiety. This observation was more likely related to the scavenger ability of the AI(iBu)₃ than to the real reversible transfer reaction. The transfer efficiency of the Al compound, if any, was very low.

The results are displayed in Table 3 TABLE 3.

The polymerisations were carried out as follows: 44 μΓηοΙ of Ti catalyst; 44 μΓηοΙ of boron cocatalyst; 35 mL of toluene; 5 mL of styrene; at a temperature of 27 °C ; during a period of time of 12 minutes.

These experiments established that diethylzinc, dibutylmagnesium and triethylborane were not suitable chain transfer agents in the polymerisation of styrene using Cp*TiBn₃/B(C6F₅)₃ system; that some AIR₃ compounds were compatible with the Cp*TiBn₃/B(C6F₅)₃ catalyst system. The maximum loading of AIR₃, before complete inhibition of the catalytic system, depended upon the nature of the alkyl group. According to the results obtained with AliBu₃, the transfer efficiency seemed quite low. Example 9.

Synthesis of sPS-Br according to the fourth embodiment of the present invention.

A series of experiments was carried out to assess the efficiency of NBS as a brominating agent and to improve the percentage of sPS-Br chain already reported by Grassi et al. (Pastorino, R.; Capacchione, C; Ferro, R.; Milione, S.; Grassi, A. Macromolecules 2009, 42, 2480-2487). Significant results are summarised in Table 4. In all experiments, the styrene concentration in the feed, and polymerisation time were properly calibrated in order to obtain low molecular weight polymers suitable for the NMR structural characterisation of the chain end groups. Different conditions such as the amount of catalyst, the Al/Ti ratio, and the polymerisation temperature and time, were investigated. In all cases, the number of polystyrene chains per titanium moiety was lower than 1 , confirming the difficulty to control the number of active species in the polymerisation. Despite this difficulty, it has been found that at a temperature of 0 °C, using relatively large amounts of catalyst (132 pmol), in the presence of 1 equivalent(s) of tris(octyl)aluminium, nearly 100% of the sPS chains were functionalised (i.e., brominated). In other cases, keeping the temperature at 0 °C, but either reducing the amount of catalyst to 88 pmol or using AI(Oct)₃/Ti ratios varying between 0 and 20, the percentage of sPS-Br was of about 80%.

Additional experiments were carried out increasing the initial styrene-to-Ti ratio from 65 to 325. They resulted in production of sPS-Br with higher molecular weight and percentage of functionalisation ranging between 85-97%. The system was pseudo- living, as was observed with lower Styrene/Ti ratio of 65. Figures 9 and 10 represent the relationship between conversion (■) and number-average molecular weight M_n (o) as function of polymerisation time in styrene polymeriation using respectively 65 equivalents and 325 equivalents of styrene. The molecular weights were very different, ca. 15000 g/mol for 65 equivalents of styrene vs ca. 80,000 g/mol for 325 equivalents of styrene. In both series of experiments, however, the experimental M_n values reached a plateau for conversions ranging between 40 and 50%.

All samples were characterised by NMR spectroscopy. The ¹H NMR signals of the bromine functionalised chain-end groups were attributed according to literature and the percentage of brominated chains was determined by the integral ratio of the signals of the brominated chain over the sum of the unsaturated end-groups and brominated end-groups as seen in Figure 1 1 . The M_n values of the polymers were also determined by the integral ratio of the methine in the chain over the sum of all chain end-groups. The M_n values calculated by ¹H NMR were in good agreement with the M_n values obtained by HT-GPC analysis; on the other hand, there was a systematic discrepancy between the theoretical M_n values, determined from the styrene-to-Ti ratio and conversion values, and the M_n values obtained from either NMR or GPC measurements.

TABLE 4.

Ti [Ti]/[AI]/[St T t Yield sPSBr M_n,theo M_n,NMR _n,_Gpc M_w/M_n PS

(μιηοΙ) ] (°c) (min) (g) (%) (g/mol) (g/mol) (g/mol) chains/Ti

88 1/1/100 27 5 0.38 51 4400 5 500 7 200 1.9 0.6

132 1/1/65 27 5 0.60 88 4 500 8 000 6 100 1.9 0.6

88 1/0/100 0 30 0.28 83 3 100 21 700 25 200 1.7 0.15

88 1/1/100 0 30 0.32 82 3 600 17 100 16 900 1.8 0.2

88 1/2/100 0 30 0.47 81 5 300 13 200 nd nd nd

88 1/5/100 0 30 0.58 78 6 600 10 000 12 700 2.2 0.7

88 1/10/100 0 30 0.40 79 4 600 9 700 nd nd nd

88 1/20/100 0 30 0.20 78 2 300 8 200 7 700 2.6 0.3

132 1/0/65 0 30 0.24 80 1 700 25 600 nd nd nd

132 1/1/65 0 7 0.22 93 1500 9 500 8 900 1.8 0.1

132 1/1/65 0 15 0.30 96 2 300 11 200 11 100 1.7 0.2

132 1/1/65 0 30 0.61 98 4 600 17 300 13 600 2.0 0.3

132 1/1/65 0 30 0.72 96 5 500 17 000 14 900 1.9 0.3

132 1/1/65 0 45 0.91 98 6 700 12 500 13 000 1.9 0.5

132 1/1/327 0 120 3.38 85 25 500 53 600 79 400 1.7 0.5

132 1/1/325 15 0.6 97 4500 35000 52300 1.7 0.1

132 1/1/325 30 1.2 94 9000 44400 58200 1.7 0.1 132 1/1/325 60 1.8 87 13600 43000 61500 2.0 0.2

132 1/1/325 90 2.7 86 20400 38100 77500 1.8 0.3

132 1/1/325 120 3.4 85 25500 53600 79400 1.7 0.5

132 1/1/325 150 4.3 89 32500 69000 84000 1.9 0.5

The polymerisations were carried out with 1 equivalent of B(C₆F₅)₃; 38 mL of toluene; 1 mL of styrene. Ά in this table stands for AI(Oct)₃.

These experiments established that NBS is an efficient brominating agent. The present conditions allowed the preparation of nearly perfectly functionalised syndiotactic polystyrene. The efficiency of this functionalisation step was assessed by NMR spectroscopy.

Example 10.

Synthesis of sPS-b-aPS via ATRP of styrene onto sPS-Br.

A series of experiments was carried out to assess the efficiency of sPS-Br products as macroinitiators for growing an atactic polystyrene block (aPS) via ATRP and for forming a sPS-Jb-aPS copolymer. Reaction conditions for sPS-graft-aPS were as described by Sen et al. (Lui, S.; Sen, A. Macromolecules 2000, 33, 5106-51 10). The CuBr and PMDETA were always used in equimolar ratio. Different loadings of styrene and different polymerisation times were used in order to prepare materials with different compositions. Unlike the usual ATRP processes in which the organic halide is soluble in the reaction medium, the brominated sPS were poorly soluble. Nevertheless, as reported in Table 5, the reactions produced diblock sPS-Jb-aPS copolymers with different compositions as displayed in Table 6. Thanks to the "living" nature of the ATRP, a set of experiments were performed starting from different portions of the same sPS-Br sample by simply prolonging the reaction time, thereby producing copolymers with different lengths of the atactic block.

The copolymers were fully characterised by HT-GPC, NMR spectroscopy and by DSC. Formation of diblock sPS-Jb-aPS copolymers were confirmed by GPC analysis, by the shift of the trace toward higher molecular weight values after the ATRP as seen in Figure 12.

A very good agreement was observed between the theoretical M_n value for the aPS block and the M_n value determined from GPC for the same block, calculated by subtracting the M_n value of the sPS-block from the M_n of the diblock copolymer. These observations confirmed the living character of the ATRP.

TABLE 5.

Ex Yield_st sPSBr [Br]/[Cu]/ t Yield_f conv_ATRp Mn.theo Mn.GPC M_w/M_n

(g) (%) [PMDETA]/[St] (h) (g) (%) (g/mol) (g/mol)

1 0.300 51 1/2/2/300 6 0.440 31 9 600 nd nd

2 0.250 82 1/6/6/2 740 16 0.366 1 1 31 300 nd nd

3 0.100 83 1/6/6/2 740 20 0.215 1 1 31 300 53 800 2.3

4 0.400 88 1/6/6/300 20 1 .490 61 19 000 25 400 3.2

5 0.265 80 1/6/6/450 22 0.750 54 25 300 nd nd

6 0.175 81 1/6/6/400 22 0.245 15 6 200 nd nd

7 0.200 77 1/6/6/600 22 0.513 35 21 800 nd nd

8 0.223 80 1/6/6/700 22 0.299 8.4 6 100 nd nd

9 0.400 78 1/6/6/300 22 0.920 58 18 100 25 500 2.4

10 0.100 78 1/6/6/400 22 0.271 38 15 800 nd nd

1 1 0.350 79 1/6/6/400 22 0.640 32 13 300 nd nd

12 0.150 79 1/1/1/400 4.5 0.174 5.3 2 200 7 800 2.6

13 0.200 98 1/3/3/600 65 0.600 44 27 400 30 200 2.3

14 0.210 98 1/3/3/1 400 144 1 .238 57 82 900 39 100 2.3

15 0.200 96 1/3/3/600 65 0.630 47 29 300 33 000 2.3

16 0.200 96 1/3/3/600 24 0.374 19 1 1 800 23 600 2.0

17 0.200 96 1/3/3/600 4 0.247 5 3 100 17 700 1.8

18 0.210 98 1/3/3/1000 96 1 .150 58 60.000 51 000 2.4

19 0.500 85 1/5/5/1800 106 1 .260 42 79 700 nd nd The polymerisations were carried out with 5 to 20 ml_ of anisole and at a temperature of 130 °C.

The compositions of the synthesised sPS-Jb-aPS are displayed in Table 6. TABLE 6.

The diblock copolymers sPS-Jb-aPS were characterised by ¹H and ¹³C NMR spectroscopy. For instance, Figure 13 shows the ¹H NMR spectrum of a copolymer where the signals of the aPS and sPS blocks are clearly visible. In all cases, no signals of the starting sPS-Br were detected, indicating that all the functionalised chains initiated the ATRP of styrene. Some experiments were performed in order to study the chain end-groups of the atactic block. As expected for the ATRP mechanism, the bromine atoms were found at the end of the atactic block, and the signals were attributed according to the literature for related atactic polystyrene obtained via ATRP, for example in Liu and Sen (Lui, S.; Sen, A. Macromolecules 2000, 33, 5106-51 10) or in Chen et al. (Chen, J.; Cui, K.; Zhag, S.; Xie, P.; Zhao, Q.; Huang, J.; Shi, L; Li, G.; Ma, Z. Macromol. Rapid. Commun. 2009, 30, 532-538).

The sPS-Br samples and their corresponding sPS-Jb-aPS copolymers were characterised by DSC and all data are reported for the second heating cycle. In almost all cases, even in the presence of large fractions of aPS block, a melting temperature was observed around 270-272 °C, which is characteristic of sPS. In some cases, a lower melting temperature ranging between 264 °C and 266 °C was observed. With respect to the melting temperatures of the correponding functionalised homopolymers sPS-Br, a maximun decrease of ca. 5 °C in the melting temperature was observed for the block copolymers sPS-Jb-aPS.

Interestingly, the melting and crystallisation enthalpies were found to be both related to the percentage of aPS in the diblock copolymers. The graphs presented in Figures 14 and 15 show respectively the good fit of the AH(melting) and AH(crystallisation) values for the synthesised copolymers with the calibration made for blends of aPS and sPS in variable ratio. The latter blends were prepared with laboratory synthesised aPS and sPS samples having similar molecular weigths of about M_n= 50 000 g/mol and similar molecular weight distributions ranging between 1 .3 and 2.0.

It is known that every factor that disturbs the organisation of crystalline domains causes a decrease in the enthalpy. For the blends as well as for the diblock copolymers, the fraction of atactic material resulted in a decrease of the melting enthalpy with an almost perfectly linear trend.

Additional studies were conducted on copolymers having the same length for the sPS block and variable length for the aPS block. The melting enthalpy of sPS segment decreased with an increase of the molecular weight of the atactic block. The melting temperature remained between 269.4 °C and 270.9 °C as seen in Figure 16.

Thermal properties of the aPS-Jb-sPS diblock copolymers have also been carried out and compared with blends of aPS and sPS homopolymers having comparable percentage of aPS and sPS components and with blends of aPS and sPS homopolymers additivated with aPS-Jb-sPS diblock copolymers. For example, the last two diblock copolymers of Table 4 have been characterised using different crystallisation flow rates.

All blends were prepared using classical solution mixture procedure. The proper amount of sPS, aPS and optionally sPS-Jb-aPS were dissolved at a temperature of 130 °C in 1 ,2,4-trichlorobenzene (TCB) to form 3 wt% solutions. The resulting solutions were then casted onto a stainless dish warmed at a temperature of 200 °C to rapidly evaporate the TCB solvent. Finally, the blends were kept in a vacuum oven at a temperature of 150 °C for 24 h before being characterised. For comparative purposes, sPS and aPS homopolymers, and the diblock copolymers underwent the same treatment. Differential scanning calorimetry (DSC) analyses were performed on a Setaram DSC 131 apparatus at a heating rate of 5, 10 and 20 °C/min, under a continuous flow of helium, using aluminium capsules. Melting temperatures were reported for the second heating cycle.

The results are displayed in Table 7, as well as the results obtained under the same crystallisation flow rates for several blends of aPS and sPS homopolymers additivated or not with aPS-b-sPS diblock copolymers. DSC curve of aPS/sPS blend of composition 90/10 (blend 6 in Table 7) is represented in Figure 17 whereas the DSC curve of the same blend additivated with 15 wt%, based on the total weight of the mix, of sPS-b-aPS copolymer (blend 1 1 in Table 7) is represented in Figure 18.

TABLE 7.

DSC flow rate 10°C/min 5°C/min 10°C/min 20°C/min

Samples T_m AH_m Tc Tc Tc AHc Tg

[°C] [j/g] [°C] [j/g] [°C] [j/g] [°C] [j/g] [°C] aPS-b-sPS

55/45 aPS/sPS 269.1 8.11 242.6 -6.3 238.2 -6.5 233.5 7.9 106.1 aPS-b-sPS

80/20 aPS/sPS 268.8 2.29 237.7 -2.7 231.8 -2.9 223.4 -2.85 104.3

Pure sPS

0/100 aPS/sPS 271.9 13.33 245.9 -12.9 241.2 -12.4 234.8 -12.8

Blend 7

25/75 aPS/sPS 271.6 11.49 243.2 -10.7 238.2 -10.2 231.8 -10.4 107.0 Blend 4

50/50 aPS/sPS 269.9 7.5 239.2 -7.4 234.1 -7.6 225.8 -8.0 102.5

Blend 5

75/25 aPS/sPS 270.4 4.76 236.2 -4.5 228.3 -4.7 213.7 -4.1 96

Blend 9

80/20 aPS/sPS 270.1 3.2 233.4 -2.9 223.5 -2.6 212.6 -1.1 105.5

Blend 10

85/15 aPS/sPS 269.5 2.3 229.8 -2.2 212.6 -2.1 200.0 -0.3 106.0

Blend 6

90/10 aPS/sPS 269.4 0.34 n.o. n.o. n.o. n.o. n.o. n.o. n.o.

Blend 8 107.7

50/50 aPS/sPS 270.4 8.7 241.1 -7.1 235.4 -7.1 228.1 -6.8 150.7 15% sPS-fa-aPS^a

Blend 1 107.3

80/20 aPS/sPS 270.9 4.8 236.0 -4.7 229.2 -3.7 224.8 -4.3 153.8 15% sPS-fa-aPS

Blend 14

90/10 aPS/sPS 268.9 1.7 n.o. n.o. n.o. n.o. n.o. n.o. 104.7 5% sPS-fa-aPS

Blend 11 104.9

90/10 aPS/sPS 270.0 2.5 235.2 -2.8 228.3 -1.9 213.4 -1.9 153.6 15% sPS-fa-aPS a: the sPS-Jb-aPS diblock copolymer added to the blend was the diblock copolymer of entry 1 in this table.

It is well known that the crystallisation temperature is strictly dependent upon the crystallisation rate: reducing the flow rates results in an increase of the crystallisation temperature T_c of blends, especially with low molecular weight aPS. Experiments performed by changing the crystallisation flow rates respectively to 5°C/min, 10°C/min and 20°C/min, were performed either on the blends or on the diblock copolymers and results reported in table 7 are in line with this trend, even with high molecular weight aPS.

In addition, T_c values were found to be related to the percentage of aPS in the blends: the more aPS in the blends, the lower the crystallisation temperature, as seen in Figure 19. This trend is even more obvious at high flow rates (20 °C/min).

The diblock copolymers showed a similar behavior but in all cases, for a given aPS/sPS ratio, the crystallisation of the diblock copolymer proceeded faster (i.e., at a higher T_c value) than for the corresponding sPS/aPS blend. That can be seen by comparing lines 1 and 5 and lines 2 and 7 of Table 7. The effect of sPS-Jb-aPS stereoblock materials on the crystallysation behavior of aPS/sPS blends was tested by adding from 5 to 15% of stereoblock copolymer to aPS/sPS blends. The results are reported in Table 7. Interestingly, the addition of 15% of copolymer affected positively the crystallysation behaviour of the blends. Crystallisation proceeded faster when the stereoblock materail was added. This effect was amplified in the blends with high contents (90%) of aPS. It was further observed that the beneficial effect of the stereoblock copolymers on the crystallisation behavior blends increased with increasing crystallisation flow rate. isothermal crystallisation studies were also performed in order to better understand the crystallisation behaviour of this copolymer, used either alone or as additive in the aPS/sPS blends. For comparative purposes the same studies were also conducted on pure sPS and on aPS/sPS blends.

The samples were held for 5 minutes at a temperature of 300 °C to erase the thermal history, then they were quickly cooled at a rate of 40 °C/min to various preset temperatures ranging between 225 °C and 260 °C. For each temperature, a crystallisation peak time (t_p) was defined as the time when the crystallisation peak appeared. It is generally accepted that t_p ^"1 is proportional to the overall crystallisation rate as disclosed for example by Chui and Peng (Chui, F-C, Peng, C-C, Polymer 2002, 4879-4886). Figure 19 represents the reciprocal value of t_p (t_p ^"1) versus crystallisation temperature (T_c) for pure sPS and for a copolymer. The crystallisation rate decreased with increasing T_c for each sample. It was observed that a copolymer, comprising up to 50 % of aPS crystallises faster than pure sPS at the same reciprocal value of t_p. Figure 20 represents the reciprocal value of t_p (t_p ^"1) versus crystallisation temperature (T_c) for pure sPS and for a diblock copolymer, for a regular blend aPS/sPS and for a blend modified with 15wt%, based on the total weight of the mix, of diblock copolymer. At given T_c, the pure sPS crystallizes the fastest (i.e., the highest t_p ^"1 value), followed by copolymer and aPS/sPS blend modified with sPS-Jb-aPS (i.e. similar t_p ^"1 values) and by aPS/sPS blend which showed the lowest t_p ^"1 value. As observed above the new diblock copolymer alone or used as additive in the aPS/sPS blends crystallised faster than regular aPS/sPS blend having the same relative amounts of sPS and sPS as the diblock copolymer. To quantitatively describe the macroscopic evolution of the crystallinity during primary crystallisation under quiescent isothermal condition, a number of macrokinetic models have been proposed over the past 60 years. One of the most largely used is the Avrami model (Avrami, M J. Chem. Phys. 1939, 7, 1 103 ; Avrami, M J. Chem. Phys. 1940, 8, 212; Avrami, M J. Chem. Phys. 1941 , 9, 177). The relative crystallinity as a function of time, V_c(t), is related to the crystallization time t by equation

V_c = 1 - expt(-Ktⁿ) wherein K is the rate constant, containing the nucleation and the growth parameter in an isothermal crystallisation process, and n is the Avrami exponent whose value depends on the nucleation mechanism and on the nature of crystal growth.

Using the Avrami equation described hereabove, parameters n and K were calculated by plotting Log(-ln(1 -V_c)) vs. Log (t) and evaluating the slope and intercept of the best fitting line. For consistency, the slope (n) and intercept (Log K) values were taken from the portion 0.05< V_c < 0.50 of the Avrami plot.

The crystallisation data at a temperature of 245 °C, represented by the Avrami plots are displayed in Figure 21 . A similar value of n was obtained for sPS homopolymer and for copolymer sPB-Jb-aPS, suggesting that the crystallisation mechanism was the same in both cases. Concerning the value of constant K, it was also observed that the copolymer apparently crystallised faster than pure sPS

The Avrami plots for pure sPS, block copolymers, blends and for blends modified with block copolymer at a temperature of 244 °C are reported in Figure 22. The experimental values of crystallisation rate (K) and of Avrami exponent (n) are reported in the Table 8. The n values were very similar for all samples, suggesting a tridimensional crystals growth. The k values for the crystallisation rate were very different thereby confirming the trends observed hereabove for crystallisation rate, i.e.: pure sPS >> sPS-Jb-aPS block copolymer ~ aPS/sPS blend modified with sPS-Jb- aPS > aPS/sPS blend. TABLE 8.

In summary, these experiments established that:

• The more sPS in the stereoblock, the faster the crystallisation.

• At a given sPS/aPS ratio, the diblock copolymers crystallise faster than blends for comparable Mn values.

• Stereoblock copolymers sPS-Jb-aPS affect positively the crystallisation

behavior of aPS/sPS blends. aPS-Jb-sPS block copolymers can thus be used as accelerators in the polymerisation of styrene or in sPS/aPS blends, wherein the crystallisation rate increases with increasing amount of sPS in the block copolymer.

In general for sPS, crystallisation from the melt strongly depends upon the experimental conditions, such as for example the maximum temperature (T_max) at which the sample is heated, the residence time of the melt at that temperature (t_max), the cooling rate and upon the crystalline form of the starting material as discussed by De Rosa et al. (De Rosa, C; Riuz de Ballenteros, O.; Di Gennaro, M.; Auriemma, F. Polymer, 2003, 44, 1861 -1870). Guerra et al. (Guerra, G.; Vitagliano, V.; De Rosa, C; Petraccone, V.; Corradini, P. Macromolecules 1990, 23, 1539) reported that for moderate cooling rates ranging between 5 and 20 °C/min, the crystalline form present in sPS sample was essentially related to the thermal history of the melt. The thermal history can be erased by heating the sample at a temperature of at least 50°C above the melting temperature (T_max) and by keeping it for a period of time ranging between 2 and 5 min (t_max) at T_max. Additional experiments were performed on the di-block copolymers and on the blends holding the samples for 5 minutes at Tmax of 335 °C before changing the crystallisation flow rates respectively to 5 °C/min, 10 °C/min, 20 °C/min. Results are reported in Table 9. The di-block copolymers showed very negligible changes in the crystallisation temperatures when compared to those reported in the Table 7 wherein the samples were heated to T_max= 300 °C. For the blends, the same treatments brought significantly lower Tc values. It was also observed that copolymers crystallised faster than the corresponding blends. This trend was amplified at high flow rates, for example at 20 °C/min.

TABLE 9.

Example 1 1

The syndiospecific polymerisation of a polybutadiene rubber solution in styrene was carried out using the Cp^*TiBn₃/B(C₆F₅)₃ catalyst system. It demonstrated that the presence of the unsaturated polydienic elastomer did not interfere with styrene polymerisation rate.

All manipulations of air and/or water-sensitive compounds were performed under a argon atmosphere using standard Schlenk techniques or using a glove-box. The commercial polybutadiene rubber (M_n = 247 000 g.mol^"1, PDI = 2.0; 1 ,4-c/s = 34.5%, 1 ,4-frans = 51 %, 1 ,2- = 14.5%) was purified from antioxidant by dissolution in chloroform and reprecipitation into methanol. This two-step procedure was repeated once and the resulting product was washed twice with dry toluene and finally dried in vacuum prior to use. In a typical polymerisation procedure, a 250 ml_ glass flask, equipped with a magnetic stirring bar, was charged sequentially with toluene (50 mL), purified polybutadiene (2.0 g) and styrene (10.00 mL). The resulting solution was thermostated in a water bath set at 27 °C and the polymerisation was started by injection of the dark red catalytic solution, preliminarily prepared by adding a colourless solution of B(C6F₅)₃ (46 mg, 88 pmol in 2 mL of toluene) to a red solution of Cp^*TiBn₃ (40 mg, 88 pmol in 3 mL of toluene). Polymerisation was carried out with one equivalent of B(C6F₅)₃ in 50 mL of dry toluene for 30 minutes at a temperature of 27 °C. The mixture was thenpoured into acidified methanol. The polymers were recovered by filtration and dried at 45 °C in a vacuum oven; the yield of 9.5 g was similar to that observed when starting from neat styrene in the absence of any polybutadiene rubber.

Example 12.

Production of rubber-toughened semi-crystalline PS was carried out by ATRP of styrene in the presence of a commercial polybutadiene rubber using sPS-Br as a macroinitiator.

All operations were performed on a Schlenk line, using standard techniques. The commercial polybutadiene rubber (M_n = 247 000 g.mol^"1, PDI = 2.0; 1 ,4-c/s = 34.5%, 1 ,4-frans = 51 %, 1 ,2- = 14.5%) was purified from antioxidant by dissolution in chloroform and reprecipitation into methanol. This two-step procedure was repeated once and the resulting product was washed twice with dry toluene and finally dried in vacuum prior to use. In a typical experiment, a 25 mL glass flask equipped with a magnetic stirring bar was charged sequentially with sPS-Br (0.200 g, 1 1.6 pmol), purified polybutadiene (50 mg), CuBr (35 pmol), anisole (15 mL), pentamethyldiethylenetriamine (PMDETA) (35 pmol) and styrene (1 .50 mL). 3 freeze-pump-thaw cycles were performed to de-oxygenate the reaction medium, and finally the flask was filled with argon. The resulting degassed suspension was stirred for 10 min. at room temperature and then was placed in an oil bath at 130 °C for the requested polymerisation time. The reaction was terminated by pouring the clear homogeneous solution into a large excess of acidic methanol. The copolymer was recovered by filtration, washed with fresh methanol, and dried under vacuum at 45

°C. The results are displayed in Table 10.

Table 10

Run Starting sPS-Br sPS- PDB Time Final Mn_Cop Mn_sPS-Br Mn_PB

Br/Cu/PMDETA/St

[g] [%] [h] [g] [xio³] [xio³] [xio³]

Ϊ 0.200 98 1/3/3/1000 0.050 65 Ϊ07 447 Ϊ3 247.2

2 0.500 86 1/5/5/1800 0.100 105 1.85 82.6 79.4 247.2

3 0.210 96 1/3/3/400 0.050 72 0.57 nd 15.0 247.2

The spectrum of the crude material as obtained from polymerisation is shown in

Figure 23, wherein the signals of all components sPS, aPS and PBD are clearly visible, demonstrating that a rubber-toughened semi-crystalline polystyrene material was effectively prepared. The polybutadiene rubber was exposed to radicals during styrene radical polymerisation, thereby undergoing radical grafting reactions, leading to the in situ production of PS-g-aPS emulsifying grafted macromolecules, exactly as observed in a conventional HiPS manufacturing process.

Claims

CLAIMS.

1 . A method for preparing polystyrene materials containing a syndiotactic polystyrene block linked to an atactic polystyrene block, namely sPS-Jb-aPS block copolymers, wherein a metal-based catalyst component of formula [L_nXx]M¹R_n, alone or in combination with a compound M²(R')_n' acting as a chain transfer agent (CTA), initiates the stereospecific chain-growth polymerisation of styrene, and wherein M¹ is a metal selected from Group 3-5 of the Periodic Table, L_nX_x is a monanionic or a dianionic ligand selected from cyclopentadienyl-type ligands and related compounds or a phenolate or an amido-type ligand, all of these ligands optionally bearing additional donor groups, M² is an element selected from Group 1 to 13 of the Periodic Table, R is hydrogen or an alkyl or allyl or benzyl group having up to 12 carbon atoms, R' is hydrogen or an alkyl or allyl or benzyl group having up to 12 carbon atoms, n and n' are an integers ranging from 1 to 4 depending on the nature and oxidation state of the metal M¹ and M², respectively, and of the nature of the LnXx ligand.

2. The method of claim 1 wherein the aPS-b-sPS copolymer is prepared by the steps of: a) growing a first syndiotactic sPS block in the presence of the metal-based catalyst system [L_nXx]M¹R_n, and the M²(R')_n' chain transfer agent to generate a M²(sPS)_n' product; b) end-capping the sPS chains of the latter product with a suitable agent such as a nitroxide radical and converting to two sPS-nitroxide blocks; c) growing an atactic aPS block from the latter functionality and forming an sPS-Jb-aPS polymer.

3. The method of claim 1 wherein the aPS-b-sPS copolymer is prepared by the steps of: a) growing a first syndiotactic sPS block in the presence of the metal-based catalyst system [L_nXx]M¹R_n and the M²(R')_n' chain transfer agent to generate a M²(sPS)_n' product; and deactivating the polymerisation catalyst by adding a strong donort ligand; b) deactivating the catalyst; c) chain-growing a second atactic aPS block by a thermal, non catalytic process to prepare a M²(aPS-Jb-sPS)_n product and forming a aPS-Jb-sPS stereoblock copolymer upon hydrolysing/quenching of the previous M²(aPS-Jb-sPS)n product.

The method of claim 1 wherein the aPS-Jb-sPS copolymer is prepared by the steps of: a) chain growing a first atactic aPS block by a thermal process in the presence of the M²(R')_n' chain transfer agent to generate a M²(aPS)_n' product; b) chain-growing a second sPS block in the presence of the metal-based catalyst system [L_nXx]M¹R_n, and the M²(aPS)_n' product of step a) to generate a M²(sPS-Jb-aPS)_n' product and forming a aPS-Jb-sPS block upon hydrolysing/quenching of the previous M²(aPS-Jb-sPS)_n' product.

The method of claim 1 wherein the aPS-Jb-sPS stereoblock copolymer is prepared by the steps of : a) growing a first sPS block in the presence of the metal-based catalyst system [L_nXx]M¹ R_n, and in situ end-capping said sPS block with a halogen atom;

b) growing an aPS block via atom transfer radical polymerisation (ATRP) or reversible iodine transfer polymerisation (RITP) and forming sPS-Jb-aPS copolymers.

The method of any one of the preceding claims wherein L_nX_x is selected from cyclopentadienyl-type ligands of general formula CsR"5 where R" are equal or different and selected from hydrogen, alkyl, aryl, trialkylsilyl or hetero- functionalised substituents, and all related Cp-type ligands such as indenyl and fluorenyl derivatives, substituted or not or L is a non-cyclopentadienyl ligand, as used in post-metallocenes derivatives, and selected from imino- phenolate derivatives, amido derivatives and all combinations thereof, with possible donor functionalities including imino, alkoxy or amino Groups.

7. The method of any one of the preceding claims wherein M¹ is selected from Nd, Y, Sc, Ti or Zr.

8. The method of any one of the preceding claims wherein M² is selected from Mg, Zn, Al or B.

9. The method of any one of the preceding claims wherein R is selected from methyl or higher alkyl groups such as ethyl, butyl, hexyl and octyl, benzyl, allyl (C3H5) or allyl groups substituted at the 1 and/or 3 positions.

10. The method of any one of the preceding claims wherein R' is preferably selected from methyl or higher alkyl groups such as ethyl, butyl, hexyl and octyl, benzyl, allyl (C3H5) or allyl groups substituted at the 1 and/or 3 positions

1 1 . The method of any one of the preceding claims wherein the metal-based catalysts is selected from {CpCMe₂Flu}Nd(allyl), or (C5Me₅)Ti(Bn)₃ activated by a stoichiometric amount of a Lewis acid such as B(C6F₅)₃ or [Ph₃C]⁺[B(C₆F5)4] or [HNMe₂Ph]⁺[B(C₆F₅)₄] , or [{C5Me₄SiMe3}Sc(CH2Silv1e3)]⁺ [SiMe₃CH₂B(CeF5)3] .

12. The method of any one of the preceding claims wherein the starting material is a rubber in styrene solution, preferably a polybutadiene in styrene solution.

13. sPS-Jb-aPS diblock copolymers obtained by the process of the present invention, characterised in that the lengths of the sPS and aPS blocks are tailored by the polymerisation conditions.

14. Use of sPS-Jb-aPS block copolymer as crystallisation accelerator in sPS/aPS blends, wherein the crystallisation rate increases with increasing amount of sPS in the block copolymer.

15. Impact-resistant semi-crystalline polystyrene obtainable by the process of claim 12.