Solid state 13C NMR characterisation study on fourth generation Ziegler–Natta catalysts

Solid State Nuclear Magnetic Resonance 43–44 (2012) 36–41

Contents lists available at SciVerse ScienceDirect

Solid State Nuclear Magnetic Resonance journal homepage: www.elsevier.com/locate/ssnmr

Solid state 13C NMR characterisation study on fourth generation Ziegler–Natta catalysts Harri Heikkinen a,n, Tiina Liitia¨ a, Ville Virkkunen b, Timo Leinonen b, Tuulamari Helaja a, Peter Denifl b a b

VTT Technical Research Centre of Finland, Biologinkuja 7, Espoo, PO Box 1000 02044 VTT, Finland Borealis Polymers Oy, R&D, PO Box 330, FIN-06101 Porvoo, Finland

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 November 2011 Received in revised form 21 February 2012 Available online 6 March 2012

In this study, solid state 13C NMR spectroscopy was utilised to characterize and identify the metal–ester coordination in active fourth generation (phthalate) Ziegler–Natta catalysts. It is known that different donors affect the active species in ZN catalysts. However, there is still limited data available of detailed molecular information how the donors and the active species are interplaying. One of the main goals of this work was to get better insight into the interactions of donor and active species. Based on the anisotropy tensor values (d11, d22, d33) from low magic-angle spinning (MAS) 13C NMR spectra in combination with chemical shift anisotropy (CSA) calculations (daniso and Z), both the coordinative metal (Mg/Ti) and the symmetry of this interaction between metal and the internal donor in the active catalyst (MgCl2/TiCl4/electron donor) system could be identified. & 2012 Elsevier Inc. All rights reserved.

Keywords: 13C CSA ZN Coordination Donor

1. Introduction Understanding of the heterogeneous Ziegler–Natta (ZN) catalyst (MgCl2/TiCl4/electron donor) systems’ mechanisms at the molecular level plays a key role in the development of advanced catalysts for polymerisation of olefins, especially polypropylene (PP) [1]. The use of ZN catalysts in polyolefin production is still dominating the industry and novel ZN catalysts are being introduced with different compositions to optimise both the polymerisation performance and polymer properties [2]. The composition of an active ZN catalyst is generally based on MgCl2/TiCl4/electron donor (internal/external donors). The internal donor is added during preparation of the catalyst, whereas external donors are added together with cocatalysts to the polymerisation reactor. The role of a donor is to regulate the stereochemistry and molecular weight distribution (MWD) of the final product [3,4]. The ZN catalysts used are categorised into generations according to the electron donor, third (benzoate), fourth (phthalate) and fifth (diether) [5–7]. The purpose of the cocatalyst (alkylating reducing agent, e.g., AlEt3) is to activate the catalyst towards polymerisation. Since the interaction of internal donor as part of the catalyst system (MgCl2/TiCl4) plays a key role in polymerisation performance and polymer properties, better understanding on the interactions taking place in the active catalyst system provides valuable information to develop new catalyst modifications.

n

Corresponding author. Fax: þ358207227026. E-mail address: harri.heikkinen@vtt.fi (H. Heikkinen).

0926-2040/$ - see front matter & 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ssnmr.2012.02.006

Even though the molecular structures of these highly complex ZN catalyst systems have been studied extensively using spectroscopic and computational methods in addition to the X-ray diffraction (XRD) analysis, the actual coordination chemistry of the donor in the MgCl2/TiCl4 system is not well understood [8,9]. Fourier transform infrared spectroscopy (FTIR) has been used to investigate the interactions of carbonyl containing internal donors in the active catalysts [9,10]. Also, solid state 13C NMR spectroscopy has been used to elucidate the coordination chemistry of ZN catalyst systems [9,11–13]. When the NMR measurements are performed in the solid state, the physical structure of catalysts is preserved and the chemical shift anisotropy (CSA) is not averaged out by the molecular tumbling like in solutions. The chemical shift of a nucleus is thus dependent on its orientation in respect to the magnetic field, giving chemical shift tensor pattern that can be used to study the metal–donor coordination in ZN catalyst model systems, as shown by Clayden et al. [13]. More often in solid state 13C NMR spectroscopy, relatively high magicangle spinning (MAS) rates are used to remove the CSA (i.e., mimic solution state NMR) and hence acquire high resolution spectra. In a recent study of Busico et al., HR-MAS 1H spectroscopy with high spinning speed of 15 kHz was used to follow the chemisorption of donors on different MgCl2 surfaces according to the degree of coverage [11]. Resolution was shown to be further improved by two dimensional 13C–1H HETCOR experiments, and pure ethanol compounds could be identified from those coordinated to the MgCl2 [12]. Magnesium coordination sites with a given number of ligands could also be determined in the complexes. In addition, low temperature and relaxation time

H. Heikkinen et al. / Solid State Nuclear Magnetic Resonance 43–44 (2012) 36–41

2.2. ZN catalyst preparation

22

11

37

33

ppm Fig. 1. Principal components describing CSA.

measurements have also been utilised to resolve the coordinated and free compounds with different mobilities [12]. In this study, low MAS rates were used to maintain the CSA for the characterisation of metal–donor interaction in active ZN catalyst systems. The spinning sideband (ssb) pattern formed at low spinning speed is composed of three principal components (Fig. 1), [14] which can be used to characterize the CSA, and in calculating the isotropic chemical shift (diso), anisotropy value (daniso) and the asymmetry parameter (Z). The isotropic chemical shift can be calculated according to the equation diso ¼ (d11 þ d22 þ d33)/3, where the principal components follow the high frequency positive order d11 4 d22 4 d33 [14]. The anisotropy value daniso describes the largest deviation from the centre of gravity, and it can be derived from the principal components as d11  (d22 þ d33)/2. The asymmetry parameter Z indicates the shape of the spectrum based on deviation from the axially symmetric component. It can be determined as Z ¼(d33  d22)/(d11  diso), and the asymmetry parameter Z is always 0r Z r1 [15–19]. Previously, Clayden et al. [13] have also studied the interactions between model MgCl2/TiCl4 complexes and dialkylphthalates by means of 13C NMR and CSA calculations. According to the results of Clayden et al., the coordinating metal (Mg/Ti) in the catalyst complex could be identified based on the asymmetry parameter value [13]. In this paper we present results from the study of metal–donor interaction in real fourth generation ZN catalysts utilising solid state 13C NMR spectroscopy in combination with CSA calculations. Different CSA equations recommended lately were also used [14].

2. Experimental 2.1. Materials TiCl4 was used as received from Millennium Inorganic Chemicals. MgCl2 from Toho Titanium used as received for ball milling. N-Heptane from Chevron–Phillips was used after drying with 3 A˚ molecular sieves and purging with nitrogen. Phthalates, diethyl phthalate (DEP), bis-(2-ethylhexyl) phthalate (DEHP) and diisobutyl phthalate (DiBP) were obtained from Neste Chemicals and used after drying with 3 A˚ molecular sieves and purging with nitrogen. Ethyl benzoate (EBE) from Fluka (purum, 499%) was used after drying with 3 A˚ molecular sieves and purging with nitrogen. Butyl octyl magnesium (BOMAGs) as 20 wt% solution in toluene was used as recieved from Chemtura. 2-ethyl hexanol from Aldrich was used after drying with 3 A˚ molecular sieves and purging with nitrogen. Viscoplexs 1–254 from RohMax was used after purging with nitrogen. Diethylaluminium chloride (DEAC) was used as received from Aldrich. Benzoyl chloride was used as received from Aldrich. Spherical MgCl2*3EtOH support material was prepared before spray crystallisation method as described in pat. EP0614467.

2.2.1. Catalyst a In a 1 l reactor equipped with mixer, 23 g (0.1 mol) spherical support material (MgCl2*3EtOH) was added and slurried with 120 ml of heptane. The slurry was cooled down to  15 1C. Then 300 ml (2.7 mol) of cold TiCl4 ( 15 1C) was added and the temperature was slowly raised to 20 1C before 8,1 ml of bis(2ethylhexyl) phthalate (DEHP) was added. Then the temperature was increased slowly to 130 1C and kept there for 30 min. After that the solution was filtered off and another 300 ml of TiCl4 was added and mixed at 120 1C for 60 min. After that the solid was washed five times with heptane (a0 500 ml) at 90 1C and then with pentane at 30 1C. The solid was then dried with nitrogen purge. Analysis of the catalyst showed that during the synthesis DEHP transesterified completely to diethyl phthalate (DEP). 2.2.2. Catalyst b First a soluble Mg-complex was prepared by reacting butyl octyl magnesium (20 wt% solution in toluene) with 2-ethylhexanol in the molar ratio of 1 to 2.2. During the addition the reactor contents were maintained below 20 1C. The temperature of the reaction mixture was then raised to 60 1C and held at that level for 30 min with stirring, at which time reaction was complete. Benzoyl chloride (0.6 mol/mol Mg) was then added and stirring of the reaction mixture at 60 1C was continued for another 30 min. Preparation of the catalyst component was started by placing 6.5 ml titanium tetrachloride in a 50 ml glass reactor equipped with a mechanical stirrer. Mixing speed was adjusted to 300 rpm. After addition of above-mentioned Mg-complex (10.5 g), 0.6 ml of Viscoplex 1-254, and 3.3 ml of n-heptane was added. During the addition of the Mg-complex the reactor contents were maintained below 30 1C. The temperature of the reaction mixture was then raised to 90 1C over a period of 20 min and held at that level for 30 min with stirring. After settling and decanting, the solids underwent washing with 35 ml toluene (containing 0.03 ml diethyl aluminium chloride) at 90 1C for 30 min, 20 ml heptane for 20 min at 90 1C and 20 ml pentane for 10 min at 25 1C. Finally, the solids were dried at 60 1C by nitrogen purge. 2.2.3. Catalyst c A soluble Mg-complex was prepared in the same way as for Catalyst b, except 1,2-phthaloyl dichloride (0.4 mol/mol Mg) used instead of benzoyl chloride. Preparation of the catalyst component was started by placing 19.5 ml titanium tetrachloride in a 300 ml glass reactor equipped with a mechanical stirrer. Mixing speed was adjusted to 170 rpm. After addition of above-mentioned Mg-complex (32.0 g), 2.0 ml of Viscoplex 1-254, and 10.0 ml of n-heptane was added. During the addition of the Mgcomplex the reactor contents were maintained below 30 1C. The temperature of the reaction mixture was then raised to 90 1C over a period of 20 min and held at that level for 30 min with stirring. After settling and decanting, the solids underwent washing with 100 ml toluene (containing 0.11 ml diethyl aluminium chloride) at 90 1C for 30 min, 60 ml heptane for 20 min at 90 1C and 60 ml pentane for 10 min at 25 1C. Finally, the solids were dried at 60 1C by nitrogen purge. 2.2.4. Catalyst d In a 1 l reactor equipped with mixer, 23 g (0.1 mol) spherical support material (MgCl2*3EtOH) was added and slurred with 120 ml of heptane. The slurry was cooled down to  15 1C. 300 ml (2.7 mol) of cold TiCl4 (  15 1C) was then added and the temperature was slowly raised to 20 1C before 5.3 ml of diisobutyl phthalate (DiBP) was added. The temperature was then slowly

38

H. Heikkinen et al. / Solid State Nuclear Magnetic Resonance 43–44 (2012) 36–41

increased to 115 1C and kept there for 60 min. After that the solution was filtered off and another 300 ml of TiCl4 was added and mixed at 115 1C for 60 min. The solid was then washed five times with heptane (a0 500 ml) at 90 1C and then with pentane at 30 1C. After that solid was dried with nitrogen purge. 2.3. Model compound preparation 2.3.1. Model a (TiCl4*DEP) First 1.0 ml (9.0 mmol) of TiCl4 and 3 ml heptane was added in a small glass vial equipped with magnetic stirrer. Then temperature of the solution was decreased to  5 1C and then 1.8 ml (9.0 mmol) of diethyl phthalate (DEP) was added. The solid was washed with heptane and dried with nitrogen purge. 2.3.2. Model b (TiCl4*EBE) First 1.0 ml (9.0 mmol) of TiCl4 and 3 ml heptane was added in a small glass vial equipped with magnetic stirrer. Then temperature of the solution was decreased to  5 1C and then 1.3 ml (9.0 mmol) of ethylbenzoate (EBE) was added. The solid was washed with heptane and dried with nitrogen purge 2.3.3. Model c (TiCl4*DEPþDEHP (50/50)) First 1.0 ml (9.0 mmol) of TiCl4 and 3 ml heptane was added in a small glass vial equipped with magnetic stirrer. Then temperature of the solution was decreased to  5 1C and mixture of 0.9 ml (4.5 mmol) of diethyl phthalate and 1.8 ml (4.5 mmol) bis(2ethylhexyl) phthalate (DEHP) was added. The solid was washed with heptane and dried with nitrogen purge 2.3.4. Model d (MgCl2*DEP) In a dry box 1.0 g of ball milled MgCl2 (10.5 mmol), 5 ml heptane and 0.15 ml (0.74 mmol) diethyl phthalate were mixed in a mortar and pestle. After about one hour heptane was evaporated and the solid was then used as such. 2.3.5. Model e (MgCl2*EBE) In a dry box 1.0 g of ball milled MgCl2 (10.5 mmol), 5 ml heptane and 0.2 ml (1.38 mmol) ethylbenzoate (EBE) were mixed in a mortar and pestle. After about one hour heptane was evaporated and the solid was then used as such. 2.4. NMR measurements For the NMR experiments, the catalyst samples were dried at 60 1C in vacuo. All the samples were handled inertly and the samples were loaded into 6.0 mm-o.d. zirconia rotors with a push on KEL-F cap under argon atmosphere. The NMR measurements were performed with a Chemagnetics CMX 270 Infinity NMR spectrometer using nitrogen as driving and bearing gas, and a 6.0 mm double-resonance MAS NMR probe operating at 68.01 MHz. The 13C NMR spectra were acquired with variable amplitude (VA) cross polarisation (CP) magic-angle spinning (MAS) sequence with carbon background suppression [20,21]. For all the samples, 17,000 transients were accumulated using a 5 ms contact time and 3 s recycle time. The ordinary CPMAS experiments were carried out using a spinning speed between 4 and 5 kHz. For the anisotropy studies low MAS rate between 1.4 and 1.8 kHz was used to reveal the spinning side band (ssb) patterns. The chemical shifts were referenced to hexamethylbenzene (HMB) using the methyl signal (þ17.35 ppm) as an external reference. All CSA calculations were carried out according to Haeberlen convention using an online converter [22].

3. Results and discussion In this paper, we show CSA results for a series of real ZN catalysts and model compounds with varying compositions obtained with the aid of 13C low MAS NMR experiments. In the studied samples (Table 1), five two-component model compounds with donor(s) and TiCl4/MgCl2, and four active ZN catalysts composed of donor(s), TiCl4 and MgCl2 with varying amounts, have been examined. Two-component model compounds are labelled as Models a–e, and the ZN catalysts as Catalysts a–d (Table 1). Chemical structures of donor compounds used in the studied materials are shown in Fig. 2. Before CSA-analysis, VACP measurements with moderately high MAS rate (4–5 kHz) were performed. The VACP pulse sequence in combination with carbon background suppression showed to be suitable for ZN catalyst studies. Compared to normal CP sequence, VACP improves nuclei sensitivity and hence shorter acquisition times are needed. As shown in Fig. 3, the characteristic signals of the studied samples, i.e., carbonyl, aromatic and alkyl region and a-CH2 could be identified. However, no detailed information on interaction of the internal donor in the ZN catalyst (MgCl2/TiCl4/electron donor) system could be identified. Only the chemical shifts of carbonyl in model systems with Ti were somewhat higher compared to the models or catalysts with Mg due to different shielding effect, respectively. This was also seen in the chemical shifts calculated on the bases of CSA (Table 2). In addition to higher carbonyl chemical shifts in ppm very clear Lorentzian lineshapes were observed for carbonyl resonances in all the Ti containing model compounds (Fig. 4 (a) and (b)). The presence of Mg on the other hand distorts the lineshape of carbonyl signals (Fig. 4(c) and (d)). This is most likely due to the fact that interaction between Mg and the donor can take place simultaneously in different lateral cuts (100,110). [23] Two clearly different carbonyl signals were detected only in model compound d (Fig. 4(c), Table 2), and could hence indicate donor interactions between two different lateral cuts. The slow MAS rate (1.4 kHz) 13C NMR spectra of model a (TiCl4/DEP) and model d (MgCl2/DEP) systems are shown in Fig. 5, and the CSA data for the studied model compounds and ZN catalysts is given in Table 2. From the results given in Table 2, we can see that the model compounds a, b and c (donor(s):Ti in 1:1 M ratio) possess negative anisotropy values (109.6 ppm,  112.1 ppm and 95.8 ppm) and model compounds d and e (donor: Mg in 0.075:1 and 0.1:1 M ratio) positive anisotropy values (þ113.7 and þ113.5 ppm and þ 77.3 ppm). Based on the results we concluded that the sign of anisotropy daniso (þ/ ) indicates to which metal (Mg/Ti) the carbonyl is coordinating: negative value indicates Table 1 Compositions (donor molar ratio relative to Ti/Mg) of the studied ZN catalysts and model compounds. ZN Catalyst

Donora

Donor

Ti

Mg

Model a Model b Model c Model d Model e Catalyst a Catalyst b Catalyst c Catalyst d

DEP EBE DEP/DEHP DEP EBE DEP EHBE DEHP DiBP

1 1 0.5/0.5 0.075 0.1 0.05 n.mb 0.09 0.1

1 1 1 – – 0.05 n.m. 0.1 0.1

– – – 1 1 1 n.m. 1 1

a DEP ¼diethyl phthalate, EHBE¼ 2-ethylhexyl benzoate, DEHP¼ bis(2-ethylhexyl) phthalate, EBE¼ ethylbenzoate, DiBP ¼di-isobutyl phthalate. b n.m. ¼not measured.

H. Heikkinen et al. / Solid State Nuclear Magnetic Resonance 43–44 (2012) 36–41

divided into two classes: samples with Z value near 0.5 and samples with Z value near 1. Asymmetry parameter value indicates electron distribution around the carbonyl moieties and hence could give detailed information about the symmetry of the interaction taking place between the donor and the metal in the ZN catalyst system (MgCl2/TiCl4/electron donor). In general, lower asymmetry parameter value indicates more symmetrical electron distribution and larger value less symmetrical electron distribution. According to Table 2, all model compounds except compound e (MgCl2/EBE in 1:0.1 M ratio) show Z values near 0.5. This could mean that the donor in model compound e interacts simultaneously with different Mg atoms in MgCl2 lateral cuts (100,110) due to low sterical hindrance regarding the phthtalate coordination in general. Asymmetry value 0.60 in model compound b (TiCl4/EBE in 1:1 M ratio) on the other hand seems to have more coherent interaction modes, which is most likely due to the well-ordered TiCl4 crystal structure. The interpretation of asymmetry parameter values is consistent with the Lorentzian lineshape detected for model compounds with TiCl4, and more complex line shape in presence of MgCl2. A recent theoretical DFT study by Correa et al. showed that phthalates can adopt a variety of coordination modes due to the long spacing between the coordinating O atom [24]. It is also well known that activated MgCl2 contains Mg2 þ ions with coordination numbers of 4 and 5 on the (1 1 0) and (1 0 0) lateral cuts. Another DFT study by Weiss et al. [25] shows that binding of di-n-butylphthalate with activated MgCl2 takes place in different

coordination to Ti and positive to Mg. The relevance of the magnitude of anisotropy value is not known and needs further investigation to be clarified. Our results are in good agreement with the results of Clayden et al. for model compounds but are expanded over to real ZN catalysts as well [13]. The ZN catalysts a–d possess positive anisotropy values indicating coordination to Mg and do not show any evidence for the interaction of carbonyl with Ti in the catalyst systems. Furthermore, based on the asymmetry parameter value (Z) given in Table 2, both model compounds and ZN catalysts can be

Fig. 2. Chemical structures of the donor compounds.

Fig. 3.

13

39

C VACP MAS NMR spectra of Model a (left) and Model d (right). Spinning sidebands are marked with asterixes.

Table 2 Anisotropy data for the studied model compounds and ZN catalysts. Compound

Model Model Model Model

a b c d

Model e Catalyst a Catalyst b Catalyst c Catalyst d

d11

d22

d33

diso

daniso

diso experimental

(ppm7 0.3)

(ppm7 0.3)

(ppm 70.3)

(ppm7 0.3)

(ppm70.3)

(ppm7 0.2)

241.1 240.9 232.1 239.5 237.5 223.4 239.4 215.0 222.5 241.8

197.3 196.0 194.0 148.7 146.7 171.6 174.1 171.4 148.5 150.6

109.6 106.3 117.3 103.0 101.3 120.7 110.2 128.2 124.0 106.1

182.7 181.1 181.1 163.7 161.9 171.9 174.6 171.5 165.0 166.1

 109.6  112.1  95.8 þ113.7 þ113.5 þ77.3 þ97.2 þ65.2 þ86.3 þ113.5

175.3 173.6 175.2 171.6 169.6 171.5 172.8 171.4 173.6 173.4

Z (70.01)

0.60 0.60 0.60 0.60 0.60 0.99 0.99 1.00 0.43 0.59

40

H. Heikkinen et al. / Solid State Nuclear Magnetic Resonance 43–44 (2012) 36–41

Fig. 4. Carbonyl signals taken from

13

C VACP MAS NMR spectra of compounds (a) Model a, (b) Model b, (c) Model d and (d) Model e.

Fig. 5. Slow MAS rate (1.4 kHz) 13C CP NMR spectra of Model a (left) and model d (right). Principal components (d11, d22 and d33) from the carbonyl resonances are shown in the spectra. Also, the presence of amorphous MgCl2 severely broadens the spectral line widths, which is clearly seen in spectrum from model compound d.

way between different lateral cuts: on the (1 0 0) lateral cut, both carbonyls can coordinate but on the (1 1 0) cut, according to the simulation only one carbonyl can coordinate due to a mismatch. On the other hand studies by Puhakka et al. resulted in the

conclusion that different donors coordinate most likely on (1 1 0) lateral cuts thereby preventing formation of Ti species coordination on the (1 1 0) lateral cut [26]. For these reasons it would be expected that active ZN catalyst systems have more complex

H. Heikkinen et al. / Solid State Nuclear Magnetic Resonance 43–44 (2012) 36–41

interaction modes taking place simultaneously between the donor and MgCl2/TiCl4 catalyst system and hence possessing asymmetry values close to 1. In the active ZN catalyst systems studied, this is true for catalysts a and b. However, catalysts c (MgCl2/TiCl4/DEHP) and d (MgCl2/TiCl4/DiBP) showed Z values near 0.5 indicating the interactions are more alike. Clearly more complete catalyst series with controlled variation of the composition is needed to fully understand the CSA results. Also the DFT calculations in combination with other analytical techniques could be utilised in development of CSA based characterisation methods for highly complex ZN materials.

4. Conclusions Exploitation of the low MAS 13C NMR spectra in combination with CSA calculations of the carbonyl species is a promising method to reveal detailed information about the donor– transition metal interaction in active ZN catalyst materials. Methodology previously used only for model compounds could also be applied for more complex catalyst systems used in polypropylene polymerisation. Based on the anisotropy tensor values (d11, d22, d33) from low MAS 13C NMR spectra in combination of (CSA) calculations (daniso and Z), both the coordinative metal (Mg/Ti) and the symmetry of this interaction with the internal donor in the active ZN catalyst (donor/MgCl2/TiCl4) system could be identified. The sign of the anisotropy value (daniso) indicates which transition metal is coordinated to internal donor. Coordination to Ti gives negative and coordination to Mg positive values. Value of the asymmetry parameter (Z) can give information about different interactions modes. Higher asymmetry values ( 1) indicate more complex interactions, i.e., larger variety of interactions, whereas asymmetry values close to 0.5 indicate the mode of interaction taking place in a more alike manner. Further studies for more complete sample series in combination with other methods (DFT modelling, FTIR, etc.) and the actual polymerisation results is still needed to fully reveal the potential of CSA in characterisation of ZN catalysts.

41

References [1] (a) K. Ziegler, E. Holzkamp, H. Breil, H. Martin, Angew. Chem. 67 (1954) 541; (b) G. Natta, Macromol. Chem. 16 (1955) 213. ¨ [2] L.L. Bohm, Angew. Chem. Int. Ed. 42 (2003) 5010–5030. [3] J.C. Chadwick, in: G. Fink, R. Mulhaupt, H.-H. Brintzinger (Eds.), Ziegler Catalysts, Springer, Berlin, 1995, pp. 427–440. ¨ a, ¨ M. Hark ¨ onen, ¨ [4] J. Seppal L. Luciani, Macromol. Chem. 190 (1989) 2535–2550. [5] B.L. Goodal, Polypropene and other polyolefins, in: S. van der Ven (Ed.), Polymerization and Characterization, Elsevier, Amsterdam, 1990, pp. 1–133. [6] S. Parodi, R. Nocci, U. Giannini, P.C. Barbe , U. Scata, European Patent 1273595 (2003). [7] E. Albizzati, U. Giannini, G. Morini, C.A. Smith, R.C. Ziegler, in: G. Fink, R. Mulhaupt, H.-H. Brintzinger (Eds.), Ziegler Catalysts, Springer, Berlin, 1995, pp. 413–425. [8] K. Hassan, Y. Fang, B. Liu, M. Terano, Polymer 16 (2010) 3627–3635. [9] Y. Kissin, X. Liu, D.J. Pollick, N.L. Brungard, M. Chang, J. Mol. Catal. A. 287 (2008) 45–52. ¨ ¨ H. Knuuttila, P. Denifl, T. Leinonen, T. Venal ¨ ainen, ¨ [10] H.-L. Ronkk o, J. Mol. Catal. A. 278 (2007) 127–134. [11] V. Busico, M. Causa, R. Cipullo, R. Credenhino, F. Cutillo, N. Friederichs, R. Lamanna, A. Segre, V. Castelli, J. Phys. Chem. C 112 (2008) 1081–1089. [12] P. Sozzani, S. Bracco, A. Comotti, R. Simonutti, I. Camurati, J. Am. Chem. Soc. 125 (2003) 12881–12893. [13] N.J. Clayden, J.V. Jones, J. Chem. Soc. Perkins Trans. 2 (1990) 175–178. [14] J. Mason, Solid State Nucl. Magn. Reson. 2 (1993) 285. [15] U. Haeberlen, in: J.S. Waugh (Ed.), Advances in Magnetic Resonance, Academic Press, New York, 1976. [16] M. Mehring, Principles of High Resolution NMR in Solids, 2nd. edn., Springer, Berlin, 1983. [17] H.W. Spiess, in: P. Diehl, E. Fluck, R. Kosfeld (Eds.), NMR Basic Principles and Progress, vol. 15, Springer, Berlin, 1978. [18] S. Paramasivan, A. Balakrishnan, O. Dimitrenko, A. Godert, T.P. Begley, F. Jordan, T. Polenova, J. Phys. Chem. B 115 (2011) 730–736. [19] M. Mousavi, S.S.-F. Yu, D.-L. Tzou, Solid State Nucl. Magn. Reson. 36 (2009) 24–31. [20] O.B. Peersen, X. Wu, I. Kustanovich, S.O. Smith, J. Magn. Reson. 104 (1993) 334–339. [21] J.L. White, L.W. Beck, D.B. Ferguson, J.F. Haw, J. Magn. Reson. 100 (1992) 336–341. [22] http://anorganik.unituebingen.de/klaus/nmr/index.php?p=conventions/csa/ csa [26 February 2010]. [23] G. Monaco, M. Toto, G. Guerra, P. Corradini, L. Cavallo, Macromolecules 33 (2000) 8953–8962. [24] A. Correa, F. Piemontesi, G. Morini, L. Cavallo, Macromolecules 40 (2007) 9181–9189. [25] H. Weiss, M. Boero, M. Parrinello, Macromol. Symp. 173 (2001) 137–147. [26] E. Puhakka, T.T. Pakkanen, T.A. Pakkanen, Surf. Sci. 334 (1995) 289–294.