Group 4 salicyloxazolines are potent polymerization ...

2 downloads 0 Views 138KB Size Report
Of the post-metallocene olefin polymerization catalysts, those based on salicylaldiminato (e.g. I) complexes of group 4 transition metals have shown perhaps the ...
COMMUNICATION

Dalton

Robyn K. J. Bott,a Max Hammond,b Peter N. Horton,c Simon J. Lancaster,a Manfred Bochmann*a and Peter Scott*b a Wolfson Materials and Catalysis Centre, School of Chemical Sciences and Pharmacy, University of East Anglia, Norwich, UK NR4 7TJ. E-mail: [email protected] b University of Warwick, Coventry, UK CV4 7AL. E-mail: [email protected]; Fax: +44 24 7657 2710; Tel: +44 24 7652 3238 c EPSRC National Crystallography Service, Department of Chemistry, University of Southampton, Southampton, UK SO17 1BJ

www.rsc.org/dalton

Group 4 salicyloxazolines are potent polymerization catalysts†

Received 12th July 2005, Accepted 27th September 2005 First published as an Advance Article on the web 13th October 2005

Octahedral titanium and zirconium complexes based on salicyloxazoline ligands with sterically demanding orthosubstituents provide a new family of extremely active ethene polymerization catalysts [up to 108 g PE (mol bar h)−1 ] which are in some cases “single site”. Of the post-metallocene olefin polymerization catalysts, those based on salicylaldiminato (e.g. I) complexes of group 4 transition metals have shown perhaps the greatest promise.1 Some examples display very high activity, and efforts have been made to improve catalyst longevity, especially at the high temperatures favoured industrially.2 Catalyst deactivation may occur, for example, via the alkylation of the C=N bonds. We have previously observed that 1,2-migratory insertion to the imine functionality by metal-bound alkyl species occurs in related chelate complexes, and hindering this pathway dramatically increases catalyst stability.3 The imine functionality has also been shown to be directly reducible by aluminium compounds present in the co-catalyst mixtures.4

ligands such as I massively outperform the analogous complexes of II in this area.6,7 The introduction of bulky substituents frequently results in improved catalytic activity in non-metallocene systems.8 With this in mind, our two laboratories independently began investigating a series of Group 4 salicyloxazoline complexes in which the ligands of type HLn carry substituents R1 in the orthoposition of the phenol ring. The proligands HL1 –HL6 were synthesised from salicylic acids in fair to good yields (ESI†),9 Reaction of the Na salts of these ligands with MCl4 or MCl4 (THF)2 gave the precatalysts Ln 2 MCl2 (M = Ti, Zr) (Scheme 1), while protonolysis of M(CH2 Ph)4 with HLn afforded the benzyl complexes Ln 2 Zr(CH2 Ph)2 (n = 3, 5, 6).

DOI: 10.1039/b509807f

Scheme 1

Complexes based on salicyloxazolinato ligands such as II have several potential advantages over the salicylaldiminates. The catalysts are not susceptible to migratory insertion or reduction at the imine functionality, and in addition, replacement of the aromatic amine group in I by a substituent derived from an aminoalcohol will eliminate this source of toxicity, and permitting the use of the resulting polymers in a wider range of applications. However, zirconium and hafnium complexes based on II have previously been shown to give only very low activity for ethene polymerization,5 and it is perhaps for this reason that oxazoline complexes have scarcely been used in alkene polymerization catalysis. Notably also, nickel complexes of † Electronic supplementary information (ESI) available: Synthetic procedures and characterising data for all compounds, crystal data for [L2 2 ZrCl2 ] and [L4 2 ZrCl2 ]. See DOI: 10.1039/b509807f This journal is

©

Reagents: (i) NaH, THF, then MCl4 ; (ii) M(CH2 Ph)4 .

The 1 H NMR spectra of these complexes indicate that the major isomers present were the C 2 symmetric cis-trans-cis species (as shown in Scheme 1).10 Smaller amounts of a C 1 symmetric cis-cis-cis species were occasionally observed.† The molecular structures of L4 2 ZrCl2 (Fig. 1) and L2 2 ZrCl2 determined by Xray crystallography confirmed the geometry of the C 2 symmetric isomers.‡ The complexes with less sterically demanding phenoxy units gave poor ethene polymerization activity5 on activation with MAO (e.g. entries 1–4, Table 1) but the bulkier complexes were potent catalysts. This is especially marked for the zirconium systems which show productivities and peak activities of the order 107 –108 g PE (mol Zr)−1 h−1 bar−1 (entries 7, 8, 12, 13). The productivity of the system is also strongly influenced by the bulk of the substituent on the oxazoline ring. For example, measured productivity/activity for L6 ZrCl2 is three orders of magnitude higher than for L5 ZrCl2 (entries 10, 12); out of necessity the former was run with a much lower catalyst loading. A similar comparison is possible for the titanium complexes (entries 9, 11). We expect that both the above steric effects arise in the need to protect the phenolate O atom from attack by electrophilic Al in the system, and subsequent transfer of ligand to that metal to

The Royal Society of Chemistry 2005

Dalton Trans., 2005, 3611–3613

3611

Table 1 Polymerization of ethene using [MLn 2 Cl2 ]/MAO

Entry 1 2 3 4 5 6 7 8 9 10 11 12 13

Complex 2

L 2 TiCl2 L2 2 TiCl2 L3 2 TiCl2 b L3 2 ZrCl2 b L4 2 TiCl2 a L4 2 TiCl2 L4 2 ZrCl2 a L4 ZrCl2 a L5 2 TiCl2 c L5 2 ZrCl2 c L6 2 TiCl2 c L6 2 ZrCl2 c L6 2 ZrCl2 a

Cat./lmol 10 50 18 14 10 50 0.1 0.5 17 2.8 1.4 0.3 1.0

Equiv. MAO 1000 1000 1000 1000 1000 1000 10 000 1000 500 3000 6000 30 000 1000

C2 /bar 1 7 1.2 1.2 1 7 1 7 1.2 1.2 1.2 1.2 1.2

T/◦ C 22 25 22 22 22 25 22 25 24 22 24 26 24

t/min 60 60 60 60 10 10 1 1 60 60 60 20 1

Yield/g

Productivity/g PE (mol bar h)−1

Peak activity/g PE (mol bar h)−1

Mn

Mw

0.072 2.55 0.117 0.038 0.838 2.26 0.462 6.88 0.190 0.147 0.410 5.172 0.860

7.2 × 10 7.3 × 103 5.4 × 103 2.2 × 103 5.0 × 105 1.9 × 105 2.8 × 108 1.2 × 107 9.3 × 103 43 × 103 2.4 × 105 4.3 × 107 5.2 × 107

— — — — — — — — 22.6 × 103 86.7 × 103 4.5 × 105 5.9 × 107 —

228 000 550 000 2000 2300 940 000 250 000 1100 8300 7300 760 104 000 1900 —

544 000d 1 080 000e 348 000f 75 200f 1 700 000d 610 000g 2000 22 000 106 000f 77 000 179 000h 38 100 —

3

a

Conditions: solvent toluene (50 ml). b Conditions: solvent toluene (175 ml). c Conditions: solvent toluene (100 ml), peak activity was measured directly using gas burette system. Concentration of MAO was kept constant between runs, resulting in varying Al/metal ratios. d Multimodal molecular weight distribution observed. Highest weight distinct peak reported here. e Polydispersity 2.0. f Broad multimodal molecular weight distribution observed. g Polydispersity 2.5. h Polydispersity 1.7.

Fig. 1 Molecular structure of [L4 2 ZrCl2 ]. Selected interatomic distances ˚ ) and angles (◦ ): N1–Zr 2.3217(16), O(1)–Zr 1.9965(13), Cl(1)–Zr (A 2.4239(6), C(11)–N(1) 1.306(2), C(11)–O(2) 1.335(2), C(13)–N(1) 1.505(2); O(1)–Zr–O(1i ) 172.81(7), O(1)–Zr–N(1) 97.13(6), O(1)–Zr– N(1i ) 77.59(5), N(1)–Zr–N(1i ) 87.15(8), N(1)–Zr–Cl(1i ) 167.11(4), O(1)– Zr–Cl(1) 90.41(4), N)1)–Zr–Cl(1) 89.72(4), Cl(1i )–Zr–Cl(1) 95.97(3).

give less active catalysts.8 In the structure of L4 2 ZrCl2 (Fig. 1), O(1) is shielded by the tert-butyl group and C(15) of the second ligand. Size exclusion chromatography of most of the polymers produced with titanium catalysts indicated bimodal or multimodal molecular weight distributions. We attribute this behaviour to the presence of several active species in the reaction mixture. It seems likely that the presence of structural isomers in the system is responsible at least in part for this behaviour.11 At 7 bar ethene pressure, some catalysts based on the smaller metal Ti showed single site behaviour (entries 2 and 6), as did the highly active catalyst L6 2 TiCl2 at 1.2 bar (entry 11). Remarkably, while titanium catalysts produced polyethylene with very high molecular weight, the analogous zirconium complexes gave low to medium molecular weight products, though with substantially higher activity. Evidently in the less congested Zr system the relative rate of chain transfer is significantly enhanced. Tests with L4 2 TiCl2 /MAO showed that at ambient temperatures this catalyst system is very long-lived, i.e. the polymer mass increased linearly over the whole duration of the experiment (2 h) (Fig. 2). Ethene uptake measurements using a gas burette system confirmed this. Also the peak activity measured for L6 2 ZrCl2 using this apparatus correlates very well with the productivity measured over a 1 min experiment for the same catalyst (Table 1; entries 12, 13). In contrast, gas uptake at higher temperatures indicated in all cases that the catalyst system 3612

Dalton Trans., 2005, 3611–3613

Fig. 2 Time dependence of PE mass and M n produced with L2 2 TiCl2 /MAO (Al/Ti = 1000, toluene 50 ml, 22 ◦ C).

was thermally unstable. For example, a plot of the activity of L6 2 ZrCl2 /MAO at 50 ◦ C is shown in Fig. 3. It is worth noting that after 5 min the ethene uptake is still ca. 4 × 106 g (mol bar h)−1 , i.e. “highly active”.12

Fig. 3

Gas consumption for L6 2 ZrCl2 /MAO at 50 ◦ C.

Although the salicyloxazoline ligands used here are highly resistant to decomposition by 1,2-migratory insertion or intermolecular reduction by Al compounds, there remained the possibility that they could be subject to acid-promoted opening of the oxazoline ring. However, exposure of a sample of L6 2 ZrCl2 to excess of MAO, followed by careful hydrolysis led to isolation

of the free proligand HL6 in good yield. On storing a mixture of L6 2 ZrCl2 and MAO (Al/Zr = 45, [Zr] = 0.004 M) for 2 days at room temperature, L6 2 ZrCl2 was completely converted to L6 AlMe2 ·MAO. The latter was assigned by comparison of its 1 H NMR spectrum with that for L6 AlMe2 formed in the system L6 2 ZrCl2 –AlMe3 –MAO (Zr : Al : AlMAO = 1 : 85 : 3.5; MAO was added as a catalyst for the ligand exchange process). NMR shifts of L6 AlMe2 differ slightly from those of L6 AlMe2 ·MAO. There are striking similarities between this new and highly potent salicyloxazoline catalyst system and that based on salicylaldimines I. The crystallographically characterised precatalysts are of the same basic structure, and NMR spectra of both systems indicate the presence of a mixture of structural isomers.4a The most productive catalysts in both series contain bulky aryloxides and have a similar steric demand at the “imine” unit. The trend in productivity with respect to the metal used (Zr > Ti)8 and the properties of the polymers produced are also familiar.1e As we have shown, the oxazoline ligand does not decompose under catalytic conditions, and the main catalyst decomposition pathway is ligand transfer to Al in both series.13 There is a growing body of evidence in favour of this scenario, such as the comparatively lower activities of monosalicylaldimine complexes.14 This is a particularly surprising conclusion given that the catalysts in question are presumed to be the alkyl cations, e.g. [(I)2 Zr–R]+ (R = growing polymer chain) which as a result of the charge might be expected to be particularly susceptible to (nucleophilic) reduction at imine. This work was supported by the Engineering and Physical Sciences Research Council, BP Chemicals plc and the Institute of Applied Catalysis. We thank Dr K. Bryliakov (UEA) for assistance with NMR investigations.

Notes and references ‡ Nonius KappaCCD area detector (φ scans and x scans to fill asymmetric unit sphere). Crystal data for [L2 2 ZrCl2 ]: C24 H28 Cl2 N2 O4 Zr; ˚ ; orthorhombic, space M = 570.60; T = 120(2) K; k = 0.71073 A ˚; group P21 21 21 ; a = 11.4174(18), b = 13.9809(9), c = 15.0376(17) A ˚ 3 ; Z = 4; Dc = 1.579 Mg m−3 ; l = 0.715 mm−1 . V = 2400.4(5) A F(000) = 1168. Reflections collected 26703, independent reflections 5479 (Rint = 0.0502); data/restraints/parameters 5479/0/305; goodness-offit on F 2 = 1.032. Final R indices [F 2 > 2r(F 2 )]: R1 = 0.0286, wR2 = 0.0618; R indices (all data): R1 = 0.0368, wR2 = 0.0648. Crystal data for [L4 2 ZrCl2 ]: C30 H40 Cl2 N2 O4 Zr, M = 654.76; T = 120(2) K; k = ˚ ; monoclinic, space group C2/c; a = 15.947(3), b = 8.3503(17), 0.71073 A

˚ 3 ; Z = 4; Dc = 1.431 Mg ˚ , b = 90.41(3)◦ ; V = 3038.6(11) A c = 22.819(5) A m−3 ; l = 0.575 mm−1 ; F(000) = 1360. Reflections collected 15884; independent reflections 3478 (Rint = 0.0383); data/restraints/parameters 3478/0/183; goodness-of-fit on F 2 = 1.041. Final R indices [F 2 > 2r(F 2 )]: R1 = 0.0303, wR2 = 0.0698; R indices (all data): R1 = 0.0400, wR2 = 0.0739. CCDC reference numbers 268638 and 268639. For crystallographic data in CIF or other electronic format see DOI: 10.1039/b509807f 1 (a) S. Matsui and T. Fujita, Catal. Today, 2001, 66, 63; (b) G. W. Coates, J. Chem. Soc., Dalton Trans., 2002, 467; (c) H. Makio, N. Kashiwa and T. Fujita, Adv. Synth. Catal., 2002, 344, 477; (d) Y. Suzuki, H. Terao and T. Fujita, Bull. Chem. Soc. Jpn., 2003, 76, 1493; (e) M. Mitani, J. Saito, S. I. Ishii, Y. Nakayama, H. Makio, N. Matsukawa, S. Matsui, J. I. Mohri, R. Furuyama, H. Terao, H. Bando, H. Tanaka and T. Fujita, Chem. Record, 2004, 4, 137. 2 N. Matsukawa, S. Matsui, M. Mitani, J. Saito, K. Tsuru, N. Kashiwa and T. Fujita, J. Mol. Catal. A: Chem., 2001, 169, 99. 3 (a) P. D. Knight, A. J. Clarke, B. S. Kimberley, R. A. Jackson and P. Scott, Chem. Commun., 2002, 352; (b) M. Sanz, T. Cuenca, M. Galakhov, A. Grassi, R. K. J. Bott, D. L. Hughes, S. J. Lancaster and M. Bochmann, Organometallics, 2004, 23, 5324. 4 (a) S. Matsui, M. Mitani, J. Saito, Y. Tohi, H. Makio, N. Matsukawa, Y. Takagi, K. Tsuru, M. Nitabaru, T. Nakano, H. Tanaka, N. Kashiwa and T. Fujita, J. Am. Chem. Soc., 2001, 123, 6847; (b) J. Saito, M. Mitani, S. Matsui, Y. Tohi, H. Makio, T. Nakano, H. Tanaka, N. Kashiwa and T. Fujita, Macromol. Chem. Phys., 2002, 203, 59. 5 P. G. Cozzi, E. Gallo, C. Floriani, A. Chiesi-Villa and C. Rizzoli, Organometallics, 1995, 14, 4994. 6 T. R. Younkin, E. F. Connor, J. I. Henderson, S. K. Friedrich, R. H. Grubbs and D. A. Bansleben, Science, 2000, 287, 460. 7 W. Zhao, Y. Qian, J. Huang and J. Duan, J. Organomet. Chem., 2004, 689, 2614. 8 Y. Suzuki, H. Terao and T. Fujita, Bull. Chem. Soc. Jpn., 2003, 76, 1493, and references therein. ¨ 9 (a) H. Vorbruggen and K. Krolikiewicz, Tetrahedron, 1993, 49, 9353; (b) H. Vorbruggen and K. Krolikiewicz, Tetrahedron Lett., 1981, 22, 4471. 10 P. D. Knight and P. Scott, Coord. Chem. Rev., 2003, 242, 125. 11 Y. Tohi, H. Makio, S. Matsui, M. Onda and T. Fujita, Macromolecules, 2003, 36, 523. 12 G. J. P. Britovsek, V. C. Gibson and D. F. Wass, Angew. Chem., Int. Ed., 1999, 38, 428. 13 H. Makio and T. Fujita, Macromol. Symp., 2004, 213, 221. 14 (a) D. A. Pennington, D. L. Hughes, M. Bochmann and S. J. Lancaster, Dalton Trans., 2003, 3480; (b) D. A. Pennington, W. Clegg, S. J. Coles, R. W. Harrington, M. B. Hursthouse, D. L. Hughes, M. E. Light, M. Schormann, M. Bochmann and S. J. Lancaster, Dalton Trans., 2005, 561.

Dalton Trans., 2005, 3611–3613

3613