- Email: [email protected]
Structure-stereospecificity relationships of propylene polymerization using substituted ansa-silylene(fluorenyl)(amido) titanium complexes
Journal of Organometallic Chemistry 804 (2016) 95e100
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry journal homepage: www.elsevier.com/locate/jorganchem
Structure-stereospecificity relationships of propylene polymerization using substituted ansa-silylene(fluorenyl)(amido) titanium complexes Ryo Tanaka a, Chie Yanase a, Zhengguo Cai b, Yuushou Nakayama a, Takeshi Shiono a, * a Department of Applied Chemistry, Graduate School of Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima, Hiroshima, 739-8527, Japan b State Key Lab of Chemical Fibers & Polymer Materials, College of Material Science & Engineering, Donghua University, Shanghai 201620, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 July 2015 Received in revised form 3 December 2015 Accepted 21 December 2015 Available online 24 December 2015
ansa-Fluorenylamidotitanium complexes bearing various substituents on the nitrogen and fluorene (2ad) were synthesized. The structures of the complexes were characterized by 1H and 13C NMR, and X-ray crystal analyses were performed for complexes 2a, 2b and 2d. The coordination mode of the fluorenyl group to the metal center was changed from h3 to h1 when a bulky group was introduced on the nitrogen or 2,3-position of the fluorenyl ring. Syndiotactic-specificity of the catalyst for the propylene polymerization was reduced when bulky group was introduced on the nitrogen. Least-square fitting analysis of the steric pentad distributions revealed that the stereodefect was mainly formed by the chain migration without monomer insertion, which is accelerated by the h1-coordination of the fluorenyl group. © 2015 Elsevier B.V. All rights reserved.
Keywords: Titanium complex Propylene polymerization Stereospecificity
1. Introduction Constrained geometry catalysts (CGCs), a series of group 4 metal complexes having an ansa-monocyclopentadienylamido ligand, are widely accepted as catalysts of olefin polymerization with various copolymerization ability and stereospecificity since the first example was introduced from industry [1e4]. Among them, fluorenylamido-ligated titanium or zirconium complexes [5e21] are known as catalysts for the copolymerization of ethylene with a-olefins [9,10], a-olefins with styrene [11,12], and norbornene with a-olefins [13e15]. These high copolymerization abilities are ascribed to a h1- or h3-coordination of the fluorenyl group to the metal center, compared with a h5-coordination of the cyclopentadienyl group. These catalysts conduct syndiotactic (syn)specific polymerization of a-olefins because of their high regioselectivity and Cs-symmetry [16e21]. However, the qualitative analysis of the relationship between the structure and the stereospecificity has not been performed well. We previously reported that the introduction of tBu groups on the 3,6-position of the fluorenyl ligand (1b) enhanced the synspecificity of a-olefin polymerization compared with the original complex (1a) [19]. Razavi et al. also reported that the dichloro
* Corresponding author. E-mail address: [email protected] (T. Shiono). http://dx.doi.org/10.1016/j.jorganchem.2015.12.028 0022-328X/© 2015 Elsevier B.V. All rights reserved.
derivative of 1b showed high syn-specificity for propylene polymerization [16,17]. However, the introduction of alkyl groups to the 2,3,6,7-position (1c) reduced the syn-specificity [21], as opposed to the corresponding zirconium complex which showed very high syn-specificity [18]. With these expertise of stereospecificitystructure relationships, we were motivated to investigate how the amido and fluorenyl groups affect the propylene polymerization ability of the fluorenylamidotitanium complex. Therefore, we designed new fluorenylamidotitanium complexes 2a-2d with bulky N-alkyl group and/or 2,3-substituted fluorenyl ligand (Fig. 1). Herein, we investigated the relationships between propylene polymerization behavior and the structure of these complexes. 2. Experimental section 2.1. General All manipulations were performed under nitrogen gas using standard Schlenk techniques. MMAO was gratefully donated by Tosoh-Finechem Co. Propylene was purchased from Takachiho Chemicals Co. and purified by passing it through dry and deoxygenation columns (DC-A4 and GC-RX, Nikka Seiko Co.). Dry toluene and Et2O were purchased from Kanto Chemical Co., Inc. and distilled over sodium metal before use. Hexane was dried over 4A molecular sieves overnight and degassed under nitrogen flow before use. Other chemicals were used as purchased. 9-
96
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
(Chlorodimethylsilyl)fluorene, 1,1,4,4-tetramethyl-2,2,3,3tetrahydrobenzo[b]fluorene and titanium complex 1a were synthesized according to the literature [20,22]. NMR spectra were recorded by a Varian system500 spectrometer calibrated with residual non-deuterated solvent (1H: d ¼ 7.26 ppm (CHCl3), 7.15 ppm (C6D5H)) or solvent (13C: d ¼ 128.1 ppm (C6D6)). High temperature NMR measurement of the polymer was performed with a JEOLLA500 spectrometer at 130 C using 1,1,2,2-tetrachloroethane-d2 as a solvent and for the calibration (13C: 74.7 ppm). Molecular weight of the polymer was determined by a Polymer Laboratories PL-GPC210 chromatograph at 140 C using o-dichlorobenzene as an eluent. The calibration was performed by polystyrene standard. Polymer concentration in injecting solution was 1 mg/mL and injection volume was 0.2 mL. X-ray diffraction measurements were performed on a Rigaku R-AXIS RAPID system with Mo Ka radiation (l ¼ 0.71069 Å, 2qmax ¼ 55 , crystal-to-detector distance ¼ 110 mm). The intensities were corrected for Lorentz and polarization effects. The structures were solved using a combination of direct methods (SHELXS-86) [23] and Fourier synthesis (DIRDIF94) [24], and refined by least-squares (SHELXL-97) [25] on Fo [2]. The non-hydrogen atoms were refined anisotropically and hydrogen atoms were placed using riding models. Least square fitting for the nine stereopentad distributions of the obtained polypropylene (PP) was performed by Microsoft Excel 2010 Solver program. 2.2. Synthesis of titanium complexes 2.2.1. Synthesis of adamantyl-substituted fluorenylamino ligand (3a) n BuLi (3.9 mL, 1.6 M solution in hexane, 6.4 mmol) was added dropwise to a solution of 1-adamantylamine (0.88 g, 5.8 mmol) in Et2O (40 mL) at 0 C. The mixture was stirred for 3 h at room temperature and the resulting solution was added dropwise to a solution of 9-(chlorodimethylsilyl)fluorene (1.5 g, 5.9 mmol) in Et2O (30 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane and the evaporation of the solvent gave yellowish oil of ligand 3a (2.0 g, 91%). 1H NMR (C6D6, 500 MHz): d ¼ 7.82 (m, 2H), 7.69e7.71 (m, 2H), 7.31 (m,4H), 3.82 (s, 1H), 1.91 (s, 3H), 1.47e1.57 (m, 12H), 0.57 (s, 1H), 0.05 (s, 6H). 2.2.2. Synthesis of 10-chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4tetrahydrobenzo[b]fluorene To a solution of 1,1,4,4-tetramethyl-2,2,3,3-tetrahydrobenzo[b] fluorene (2.2 g, 7.9 mmol) in Et2O (25 mL) was added dropwise n BuLi (5.3 mL, 1.6 M solution in hexane, 8.7 mmol) at 0 C. After stirring for 4 h at room temperature, the supernatant liquid was removed and the remaining lithium salt was suspended in hexane (25 mL). The suspension was added to a solution of dichlorodimethylsilane (4.7 mL, 40 mmol) in hexane at 78 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The solid was extracted with hexane and evapolation of the solvent gave yellowish oil of 10chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4-tetrahydrobenzo[b]fluorene (2.7 g, 91%). 1H NMR (CDCl3, 500 MHz): d ¼ 7.82 (d, 1H, J ¼ 10 Hz), 7.76 (s, 1H), 7.62 (d, 1H, J ¼ 10 Hz), 7.57 (s, 1H), 7.36 (dd, 1H, J ¼ 10 Hz), 7.27 (m, 1H), 4.00 (s, 1H), 1.74 (s, 4H), 1.33e1.37 (m, 12H), 0.20 (s, 3H), 0.12 (s, 3H). 2.2.3. Synthesis of adamantyl-substituted tetrahydrobenzo[b] fluorenylamino ligand (3b) n BuLi (3.4 mL, 1.6 M solution in hexane, 5.5 mmol) was added dropwise to a solution of cumylamine (0.76 g, 5.0 mmol) in Et2O
(40 mL) at 0 C. The mixture was stirred for 3 h at room temperature and the solution was added dropwise to a solution of 10chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4-tetrahydrobenzo[b]fluorene (1.6 g, 4.2 mmol) in Et2O (30 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane and the evaporation of the solvent gave yellowish oil of ligand 3b (1.8 g, 87%). 1H NMR (C6D6, 500 MHz): d ¼ 7.98 (s, 1H), 7.85 (d, 1H, J ¼ 7 Hz), 7.75 (s, 1H), 7.71 (d, 1H, J ¼ 9 Hz), 7.32 (m, 2H), 3.79 (s, 1H), 1.90 (s, 3H), 1,72 (m, 4H), 1.24e1.56 (m, 24H) 0.46 (s, 1H), 0.09 (s, 3H), 0.03 (s, 3H). 2.2.4. Synthesis of cumyl-substituted fluorenylamino ligand (3c) n BuLi (3.8 mL, 1.6 M solution in hexane, 6.0 mmol) was added dropwise to a solution of cumylamine (0.86 mL, 6.0 mmol) in Et2O (20 mL) at 0 C. The mixture was stirred for 4 h at room temperature and the resulting solution was added dropwise to a solution of 9-(chlorodimethylsilyl)fluorene (1.5 g, 6.0 mmol) in Et2O (20 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane (20 mL x3) and the evaporation of the solvent gave yellowish oil of ligand 3c (2.1 g, 98%). 1H NMR (C6D6, 500 MHz): d ¼ 7.65 (dd, 2H, J ¼ 8 Hz, 1 Hz), 7.40 (dd, 2H, J ¼ 8 Hz, 1 Hz), 7.17-6.89 (brm, 9H), 3.57 (s, 1H), 1.09 (s, 6H), 0.56 (brs, 1H), 0.28 (s, 6H). 2.2.5. Synthesis of cumyl-substituted tetrahydrobenzo[b] fluorenylamino ligand (3d) n BuLi (7.1 mL, 1.6 M solution in hexane, 12 mmol) was added dropwise to a solution of cumylamine (1.6 mL, 11 mmol) in Et2O (40 mL) at 0 C. The mixture was stirred for 3 h at room temperature and the solution was added dropwise to a solution of 10chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4-tetrahydrobenzo[b]fluorene (4.2 g, 11 mmol) in Et2O (40 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane and the evaporation of the solvent gave yellowish oil of ligand 3d (5.0 g, 97%). 1H NMR (C6D6, 500 MHz): d ¼ 7.96 (s, 1H), 7.83 (d, 1H, J ¼ 9 Hz), 7,65 (s, 1H), 7.57 (d, 1H, J ¼ 9 Hz), 7.45-7.07 (m, 7H), 3.71 (s, 1H), 1.72 (s, 4H), 1.24e1.42 (m, 18H) 0.75 (s, 1H), 0.04 (s, 3H), 0.07 (s, 3H). 2.2.6. Synthesis of fluorenyl-adamantylamido titanium complex (2a) To a solution of 3a (2.0 g, 5.3 mmol) in Et2O (30 mL) was slowly added excess MeLi (22 mL, 1.14 M solution in Et2O, 25 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.57 mL, 5.3 mmol) in hexane (40 mL) at room temperature, and the resulting red solution was stirred for 1 h. After the solvent was removed, the residue was extracted with hexane (100 mL) and MeMgBr (4.2 mL, 3.0 M solution in Et2O, 13 mmol) was added to the solution. After the resulting brown suspension was stirred for 1 h at room temperature, the solvent was removed and the residue was extracted with hexane (120 mL). The solution was concentrated to approximately 20 mL and cooled overnight at 30 C to give 2a as orange crystals (0.87 g, 33%). 1H NMR (C6D6, 500 MHz): d ¼ 7.84 (d, 2H, J ¼ 9 Hz), 7.70 (d, 2H, J ¼ 9 Hz), 7.24 (dd, 2H, J ¼ 9 Hz), 7.13 (dd, 2H, J ¼ 9 Hz), 2.00 (s, 6H), 1.97 (s, 3H), 1.54 (s, 6H), 0.72 (s, 6H), 0.00 (s, 6H), 13C NMR (C6D6, 125 MHz): d ¼ 135.1, 129.3, 124.9, 123.8, 82.1, 60.2, 56.9, 47.8, 36.6, 30.8, 6.4 (2 peaks in the aromatic region should be overlapped with solvent peaks). Anal. Calc. for C27H35NSiTi: C, 72.14; H, 7.85; N, 3.12. Found: C, 71.73; H, 7.45; N, 3.28.
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
2.2.7. Synthesis of tetrahydrobenzo[b]fluorenyl-adamantylamido titanium complex (2b) To a solution of 3b (1.8 g, 3.7 mmol) in Et2O (30 mL) was slowly added excess MeLi (15 mL, 1.14 M solution in Et2O, 17 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.40 mL, 3.7 mmol) in hexane (40 mL) at room temperature, and the resulting red solution was stirred for 1 h. After the solvent was removed, the residue was extracted with hexane (80 mL) and MeMgBr (3.0 mL, 3.0 M solution in Et2O, 9.0 mmol) was added to the solution. After the resulting black suspension was stirred for 1 h at room temperature, the solvent was removed and the residue was extracted with hexane (120 mL). The solution was concentrated to approximately 15 mL and cooled overnight at 30 C to give 2b as red crystals (0.59 g, 29%). 1H NMR (C6D6, 500 MHz): d ¼ 8.08 (s, 1H), 7.86 (d, 1H, J ¼ 7 Hz), 7.82 (s, 1H), 7.73 (d, 1H, J ¼ 7 Hz), 7.24 (dd, 1H, J ¼ 7 Hz, 7 Hz), 7.15 (br, 1H), 2.03 (s, 6H), 1.96 (s, 3H), 1.65 (s, 4H), 1.54 (br, 6H), 1.41 (s, 3H), 1.37 (s, 6H), 1.28 (s, 3H), 0.83 (s, 3H), 0.75 (s, 3H), 0.06 (s, 3H), 0.05 (s, 3H), 13C NMR (C6D6, 125 MHz): d ¼ 146.3, 144.7, 135.5, 133.6, 129.5, 127.9, 125.1, 124.7, 123.6, 120.7, 81.3, 60.0, 56.0, 55.6, 47.8, 36.6, 35.5, 35.3, 35.2, 35.0, 33.7, 33.0, 32.8, 32.6, 30.8, 6.4, 6.4 (2 peaks in the aromatic region should be overlapped with solvent peaks). Anal. Calc. for C35H49NSiTi: C, 75.10; H, 8.82; N, 2.50. Found: C, 75.17; H, 9.10; N, 2.66. 2.2.8. Synthesis of fluorenyl-cumylamido titanium complex (2c) To a solution of 3c (2.1 g, 5.9 mmol) in Et2O (30 mL) was slowly added excess MeLi (22 mL, 1.14 M solution in Et2O, 25 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.63 mL, 5.9 mmol) in hexane (20 mL) at room temperature, and the resulting red solution was stirred for 1 h at room temperature. After the solvent was removed, the residue was extracted with Et2O (30 mL x2) and MeMgBr (4.9 mL, 3.0 M solution in THF, 15 mmol) was added to the solution. After the resulting brown suspension was stirred for 3 h at room temperature, the solvent was removed and the residue was extracted with hexane (120 mL). The solution was concentrated to approximately 20 mL and cooled overnight at 30 C to give 2c as orange crystals (0.89 g, 34%). 1H NMR (C6D6, 500 MHz): d ¼ 7.77 (d, 2H, J ¼ 8 Hz), 7.61 (d, 2H, J ¼ 8 Hz), 7.32 (d, 2H, J ¼ 7 Hz), 7.21 (t, 2H, J ¼ 7 Hz), 7.11 (m, 4H), 7.04 (dd, 1H, J ¼ 8 Hz, 7 Hz), 1.80 (s, 6H), 0.28 (s, 6H), 0.01 (s, 6H), 13C NMR (C6D6, 125 MHz): d ¼ 150.2, 135.0, 129.1, 128.2, 127.9, 127.9, 126.8, 126.4, 125.0, 123.8, 83.0, 62.8, 57.8, 33.8, 4.8. Anal. Calc. for C26H31NSiTi: C, 72.04; H, 7.21; N, 3.23. Found: C, 72.22; H, 7.56; N, 3.27. 2.2.9. Synthesis of tetrahydrobenzo[b]fluorenyl-cumylamido titanium complex (2d) To a solution of 3d (2.6 g, 5.6 mmol) in Et2O (30 mL) was slowly added excess MeLi (22 mL, 1.14 M solution in Et2O, 25 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.62 mL, 5.7 mmol) in hexane (40 mL) at room temperature, and the resulting red solution was stirred for 1 h. After the solvent was removed, the residue was extracted with hexane (100 mL) and MeMgBr (4.7 mL, 3.0 M solution in Et2O, 14 mmol) was added to the solution. After the resulting brown suspension was stirred for 1 h at room temperature, the solvent was removed and the residue was extracted with hexane (70 mL). The solution was concentrated and cooled overnight at 30 C to give 2d as red crystals (0.65 g, 21%). 1 H NMR (C6D6, 500 MHz): d ¼ 8.06 (s, 1H), 7.82 (d, 1H, J ¼ 8 Hz), 7.78 (s, 1H), 7.68 (d, 1H, J ¼ 8 Hz), 7.39 (d, 2H, J ¼ 7 Hz), 7.22 (t, 1H, J ¼ 7 Hz), 7.13e7.19 (m, 3H), 7.08 (dd, 1H, J ¼ 8 Hz), 1.89 (s, 3H), 1.82 (s, 3H), 1.66 (s, 4H), 1.40 (s, 3H), 1.37 (s, 3H), 1.34 (s, 6H), 0.36 (s, 6H), 0.09 (s, 3H), 0.03 (s, 3H), 13C NMR (C6D6, 125 MHz): d ¼ 150.5,
97
146.5, 144.8, 135.5, 133.7, 129.4, 126.8, 126.5, 125.1, 124.8, 123.7, 120.8, 82.3, 62.7, 56,7. 56.5, 35.5, 35.3, 35.2, 35.0, 34.3, 33.7, 33.5, 33.0, 32.8, 32.6, 5.1, 4.6 (4 peaks in the aromatic region should be overlapped with solvent peaks). Anal. Calc. for C34H45NSiTi: C, 75.11; H, 8.34; N, 2.58. Found: C, 74.79; H, 8.73; N, 2.48. 2.3. General polymerization procedure of propylene To a 100-mL 2-necked flask, MMAO (6.5 wt% Al in hexane, 1.8 mL, 4.0 mmol) and heptane (27.2 mL) were charged under nitrogen and stirred for 30 min at 0 C. N2 in headspace of the flask was evacuated and the solvent was saturated with an atomospheric pressure of propylene. The polymerization was started under the flow of propylene by the addition of a solution of titanium catalyst (20 mmol) in heptane (1.0 mL). After the polymerization was conducted for 10 min under an atomospheric pressure of propylene, the reaction mixture was poured into 200 mL of MeOH containing 4 mL of concentrated HCl and the precipitated polymer was collected by filtration. The obtained polymer was dried for 6 h under vacuum at 60 C to be a constant weight. 3. Results and discussions Fluorenylamido titanium complex 2a-2d were synthesized according to the synthetic method of the corresponding t-butylamidotitanium complex 1a (Scheme 1) [20]. The ligands 3a-3d were synthesized from fluorene or 2,3-alkyl-substituted fluorene by chlorodimethylsilylation at the 9-position of the fluorenyl ring followed by amination. Successive treatment with excess methyllithium, an equimolar amount of TiCl4 and 2 equivalent of MeMgBr gave the corresponding dimethyltitanium complexes in moderate yields. Single crystal of 2a, 2b and 2d suitable for the X-ray analysis was obtained and their ORTEP drawings are shown at Fig. 2. The bond lengths between the titanium center and the fluorenyl carbon C(1), C(2), and C(5) were 2.23e2.43 Å, respectively, indicating that the fluorenyl group is coordinating to the titanium center with h3- to h1-form, which was the same tendency as 1a (Table 1). Miller et al. evaluated the hapticity of the fluorenyl ligands of some zirconium complexes by comparing the CeC bond lengths of the fivemembered ring [8,11]. They used the value (aþb-c-d)/2 as a parameter of h1-form tendency, where a, b, c and d are the bond lengths of C(1)-C(2), C(1)-C(5), C(2)-C(3), and C(4)-C(5), respectively. The (aþb-c-d)/2 value of 1c, 2b and 2d were 0.033e0.048 Å, which were much larger than those of the other h5-fluorenyl complexes including previously reported zirconium complexes (0.014e0.028 Å). These results indicated that the alkylation of 2,3position made the fluorenyl group to coordinate in h1-form. We applied Tolman cone angle [26] to estimate the steric effect of the amino group (Table 2). These values indicate that tBu and 1adamantyl groups have similar steric bulkiness (65e68 ) and cumyl (a, a-dimethylbenzyl) group is more sterically hindered (82 ). Next, propylene polymerization using titanium complexes 2a2d was performed (Table 3, run 4e7). The complexes were activated by MMAO and the polymerization was conducted with a semi-batch reaction under 1 atm of propylene at 0 C in heptane. The previously reported results using 1a-1c are also shown for the comparison (run 1e3). Polymer with high molecular weight (~30,000) was obtained by each complex. All the polypropylene obtained by 2a-2d was amorphous that showed no melting point in the DSC analysis. 13C NMR analysis of the obtained PPs indicated no 2,1-inserted, namely, regioirregular propylene unit. Among Cs-symmetric complexes 1a-1c, 2a and 2c, only 2c gave atactic PP, while the other complexes produced syn-PP. The stereospecificities of C1-symmetric complexes 2b and 2d resemble
98
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
Fig. 1. Structures of fluorenylamidotitanium complexes 1a-c, 2a-2d.
Scheme 1. Synthesis of ansa-dimethylsilylene(fluorenyl) (cumylamido)titanium complexes 2a-2d.
Fig. 2. Structure of fluorenylamidotitanium complexes 2a, 2b and 2d.
Table 1 Selected bond lengths (Å) of 1a-1c, 2a, 2b and 2d. Parameters
1a17
1b12
1c13
2a
2b
2d
TieC(1) TieC(2) TieC(3) TieC(4) TieC(5) C(1)-C(2) (a) C(1)-C(5) (b) C(2)-C(3) (c) C(4)-C(5) (d) (a þ b e c e d)/2
2.366(3) 2.415(3) 2.572(3) 2.573(3) 2.418(3) 1.448(2) 1.455(3) 1.425(2) 1.426(2) 0.026
2.247(3) 2.422(3) 2.619(3) 2.609(3) 2.401(3) 1.462(4) 1.447(4) 1.424(4) 1.436(4) 0.025
2.229(4) 2.424(4) 2.620(1) 2.589(1) 2.412(3) 1.461(7) 1.464(6) 1.411(7) 1.419(8) 0.048
2.241(4) 2.399(5) 2.624(5) 2.643(4) 2.433(4) 1.441(6) 1.456(5) 1.436(4) 1.433(5) 0.014
2.226(4) 2.387(4) 2.585(5) 2.585(7) 2.391(6) 1.464(5) 1.454(10) 1.423(9) 1.419(6) 0.038
2.248(3) 2.401(3) 2.608(4) 2.606(3) 2.422(3) 1.445(5) 1.458(5) 1.428(5) 1.410(5) 0.033
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100 Table 2 Tolman's cone angle of amido groups of complexes 1a-1c, 2a, 2b and 2b.
q ( ) 67 65 67 68 68 82
N-alkyl Bu t Bu t Bu Ad Ad Me2PhC t
1a 1b 1c 2a 2b 2d
Table 3 Polymerization of propylene with Ti complex 1a-2d activated by MMAO.
Run
Cat.
Time (min)
Aa
Mnb (104)
PDIb
rrc (%)
mrc (%)
mmc (%)
1 2d 3e 4 5 6 7
1a 1b 1c 2a 2b 2c 2d
15 7 0.5 30 10 3 3
160 920 11,000 100 500 1500 1600
3.3 17 20 6.0 4.0 3.3 3.2
1.7 1.5 2.9 1.6 1.8 1.7 1.9
73 93 61 71 66 33 36
21 5 29 23 26 48 41
6 2 10 5 8 19 23
a b c d e
with those of the corresponding Cs-symmetric complex 2a and 2c. These results indicated that the introduction of sterically demanded cumyl group reduce the stereospecificity, whereas alkyl group on 2,3-position of the fluorenyl group has little influence on the stereospecificity. To reveal the origin of the different stereospecificity, we applied Busico's chain propagation model described with the probability of enantiofacial preference (s) and chain migration probability without monomer insertion (P) [27,28]. They previously reported that the stereopentad distributions of PP obtained by 3,6disubstituted fluorenylamidotitanium complex closely fit to this statistical model [17]. Our experimental result using 1a also Table 4 Stereopentad distribution of the obtained PPs (%) from 1a and 2c. 1a
2c
mmmm mmmr rmmr mmrr mmrm þ rmrr rmrm rrrr mrrr mrrm
1 1 4 9 8 4 59 13 1 93.2 4.4 1.03
4 9 6 12 23 13 10 16 7 95.0 42.2 0.82
a
Least square error of nine distributions from the fitting model.
Fluorenylamido-ligated titanium complexes 2a-2d were synthesized and characterized. The syn-specificity of propylene polymerization was reduced when bulky group was introduced to amide or 2,3-position of the fluorenyl ring. The least-square fitting analysis of the stereosequence of 2c assuming the enantiomorphicsite control and chain migration model indicated that the reduction of syn-specificity was mainly ascribed to the chain migration without monomer insertion. The promotion of chain migration is probably because of the sterically hindered environment around the metal center. Acknowledgements We are grateful to Tosoh Finechem Co. for generous donation of chemicals. We also gratefully acknowledge the Natural Science Center for Basic Research and Development (N-BARD), Hiroshima University, for the high temperature NMR measurement. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jorganchem.2015.12.028. References
Cat.
P RMSa
supported this model. The amount of isolated meso-sterodefect increased when polar solvent such as chlorobenzene was used, which indicated that the rate of chain migration became faster than that of insertion in the polar solvent [29]. s and P values were calculated from stereopentad distributions of the polymer obtained by 1a and 2c (Table 4). Both of the distributions were within 2% error to the calculated s and P values. The resulting s values (93e95%) were not so different between 1a and 2c, but the P value of 2c (42.2%) was much larger than that of the 1a (4.4%). Therefore, lower stereospecificity of 2c than 1a is because of more frequent chain migration without monomer insertion. This is probably because fluorenyl group is forced to coordinate to the metal center with h1-form by the sterically hindered cumyl group during the polymerization, which resulted in the acceleration of chain migration. Considering the difference between complex 1a and 2c, the bulky cumyl group on 2d should also facilitate frequent chain migration without monomer insertion. Under the frequent chain migration condition, the preference of propagation from one conformer over the other conformer should give isotactic polymer. However, the obtained polymer from 2d was atactic, which has almost the same stereotriad distribution as that by 2c. This result shows that the propagation pathways from two distinct coordination sites in 2d were not energetically different enough to give isotactoid. 4. Conclusion
Activity as kg mol Ti1 h1. Determined by GPC calibrated with polystyrene standard. Stereotriad distributions determined by13C NMR. Data taken from ref. [12] using dried MMAO. Data taken from ref. [13] using 2.0 mmol of dried MMAO.
s
99
[1] J.C. Stevens, F.J. Timmers, D.R. Wilson, G.F. Schmidt, P.N. Nickias, R.K. Rosen, G.W. Knight, S.-Y. Lai, EP0416815, 1991. [2] J.A.M. Canich, US5264405, 1993. [3] A.L. McKnight, R.M. Waymouth, Chem. Rev. 98 (2010) 2587. [4] H. Braunschweig, F.M. Breitling, Coord. Chem. Rev. 250 (2006) 2691. [5] J. Okuda, F.J. Schattenmann, S. Wocadlo, W. Massa, Organometallics 14 (1995) 789. [6] H.V.R. Dias, Z. Wang, S.G. Bott, J. Organomet. Chem. 508 (1996) 91. [7] H.V.R. Dias, Z. Wang, J. Organomet. Chem. 539 (1997) 77. [8] L.J. Irwin, J.H. Reibenspies, S.A. Miller, Polyhedron 24 (2005) 1314. [9] J. Chai, K.A. Abboud, S.A. Miller, Dalton Trans. 42 (2013) 9139. [10] Y. Nakayama, Y. Sogo, Z. Cai, T. Shiono, J. Polym. Sci. Part A Polym. Chem. 51 (2013) 1223. [11] G. Xu, Macromolecules 31 (1998) 2395. [12] E. Kirillov, A. Razavi, J.-F. Carpentier, J. Mol. Catal. Part A Chem. 249 (2006)
100
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
230. [13] T. Shiono, M. Sugimoto, T. Hasan, Z. Cai, T. Ikeda, Macromolecules 41 (2008) 8292. [14] T. Shiono, M. Sugimoto, T. Hasan, Z. Cai, Macromol. Chem. Phys. 214 (2013) 2239. [15] R. Tanaka, T. Ikeda, Y. Nakayama, T. Shiono, Polymer 56 (2015) 218. [16] A. Razavi, U. Thewalt, J. Organomet. Chem. 621 (2001) 267. [17] V. Busico, R. Cipullo, F. Cutillo, G. Talarico, A. Razavi, Macromol. Chem. Phys. 204 (2003) 1269. [18] L.J. Irwin, S.A. Miller, J. Am. Chem. Soc. 127 (2005) 9972. [19] Z. Cai, T. Ikeda, M. Akita, T. Shiono, Macromolecules 38 (2005) 8135. [20] K. Nishii, H. Hagihara, T. Ikeda, M. Akita, T. Shiono, J. Organomet. Chem. 691 (2006) 193. [21] T. Shiono, R. Harada, Z. Cai, Y. Nakayama, Top. Catal. 52 (2009) 675. [22] S.A. Miller, J.E. Bercaw, Organometallics 23 (2004) 1777.
[23] G.M. Sheldrick, SHELXS-86: Program for Crystal Structure Determination, €ttingen, Go €ttingen, Germany, 1986. University of Go [24] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, S. Garcia-Granda, R.O. Gould, J.M.M. Smits, C. Smykalla, The DIRDIF Program System, Technical Report of the Crystallography Laboratory, University of Nijmegen, Nijmegen, The Netherland, 1992. [25] G.M. Sheldrick, SHELXL-97: Program for Crystal Structure Refinement, Uni€ ttingen, Go €ttingen, Germany, 1997. versity of Go [26] C.A. Tolman, J. Am. Chem. Soc. 96 (1974) 53. [27] V. Busico, R. Cipullo, P. Corradini, L. Landriani, M. Vacatello, A.L. Segre, Macromolecules 28 (1995) 1887. [28] V. Busico, R. Cipullo, Prog. Polym. Sci. 26 (2001) 443. [29] K. Nishii, T. Matsumae, E. O'Dare, T. Shiono, T. Ikeda, Macromol. Chem. Phys. 205 (2004) 363.
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry journal homepage: www.elsevier.com/locate/jorganchem
Structure-stereospecificity relationships of propylene polymerization using substituted ansa-silylene(fluorenyl)(amido) titanium complexes Ryo Tanaka a, Chie Yanase a, Zhengguo Cai b, Yuushou Nakayama a, Takeshi Shiono a, * a Department of Applied Chemistry, Graduate School of Engineering, Hiroshima University, 1-4-1 Kagamiyama, Higashi-Hiroshima, Hiroshima, 739-8527, Japan b State Key Lab of Chemical Fibers & Polymer Materials, College of Material Science & Engineering, Donghua University, Shanghai 201620, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 July 2015 Received in revised form 3 December 2015 Accepted 21 December 2015 Available online 24 December 2015
ansa-Fluorenylamidotitanium complexes bearing various substituents on the nitrogen and fluorene (2ad) were synthesized. The structures of the complexes were characterized by 1H and 13C NMR, and X-ray crystal analyses were performed for complexes 2a, 2b and 2d. The coordination mode of the fluorenyl group to the metal center was changed from h3 to h1 when a bulky group was introduced on the nitrogen or 2,3-position of the fluorenyl ring. Syndiotactic-specificity of the catalyst for the propylene polymerization was reduced when bulky group was introduced on the nitrogen. Least-square fitting analysis of the steric pentad distributions revealed that the stereodefect was mainly formed by the chain migration without monomer insertion, which is accelerated by the h1-coordination of the fluorenyl group. © 2015 Elsevier B.V. All rights reserved.
Keywords: Titanium complex Propylene polymerization Stereospecificity
1. Introduction Constrained geometry catalysts (CGCs), a series of group 4 metal complexes having an ansa-monocyclopentadienylamido ligand, are widely accepted as catalysts of olefin polymerization with various copolymerization ability and stereospecificity since the first example was introduced from industry [1e4]. Among them, fluorenylamido-ligated titanium or zirconium complexes [5e21] are known as catalysts for the copolymerization of ethylene with a-olefins [9,10], a-olefins with styrene [11,12], and norbornene with a-olefins [13e15]. These high copolymerization abilities are ascribed to a h1- or h3-coordination of the fluorenyl group to the metal center, compared with a h5-coordination of the cyclopentadienyl group. These catalysts conduct syndiotactic (syn)specific polymerization of a-olefins because of their high regioselectivity and Cs-symmetry [16e21]. However, the qualitative analysis of the relationship between the structure and the stereospecificity has not been performed well. We previously reported that the introduction of tBu groups on the 3,6-position of the fluorenyl ligand (1b) enhanced the synspecificity of a-olefin polymerization compared with the original complex (1a) [19]. Razavi et al. also reported that the dichloro
* Corresponding author. E-mail address: [email protected] (T. Shiono). http://dx.doi.org/10.1016/j.jorganchem.2015.12.028 0022-328X/© 2015 Elsevier B.V. All rights reserved.
derivative of 1b showed high syn-specificity for propylene polymerization [16,17]. However, the introduction of alkyl groups to the 2,3,6,7-position (1c) reduced the syn-specificity [21], as opposed to the corresponding zirconium complex which showed very high syn-specificity [18]. With these expertise of stereospecificitystructure relationships, we were motivated to investigate how the amido and fluorenyl groups affect the propylene polymerization ability of the fluorenylamidotitanium complex. Therefore, we designed new fluorenylamidotitanium complexes 2a-2d with bulky N-alkyl group and/or 2,3-substituted fluorenyl ligand (Fig. 1). Herein, we investigated the relationships between propylene polymerization behavior and the structure of these complexes. 2. Experimental section 2.1. General All manipulations were performed under nitrogen gas using standard Schlenk techniques. MMAO was gratefully donated by Tosoh-Finechem Co. Propylene was purchased from Takachiho Chemicals Co. and purified by passing it through dry and deoxygenation columns (DC-A4 and GC-RX, Nikka Seiko Co.). Dry toluene and Et2O were purchased from Kanto Chemical Co., Inc. and distilled over sodium metal before use. Hexane was dried over 4A molecular sieves overnight and degassed under nitrogen flow before use. Other chemicals were used as purchased. 9-
96
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
(Chlorodimethylsilyl)fluorene, 1,1,4,4-tetramethyl-2,2,3,3tetrahydrobenzo[b]fluorene and titanium complex 1a were synthesized according to the literature [20,22]. NMR spectra were recorded by a Varian system500 spectrometer calibrated with residual non-deuterated solvent (1H: d ¼ 7.26 ppm (CHCl3), 7.15 ppm (C6D5H)) or solvent (13C: d ¼ 128.1 ppm (C6D6)). High temperature NMR measurement of the polymer was performed with a JEOLLA500 spectrometer at 130 C using 1,1,2,2-tetrachloroethane-d2 as a solvent and for the calibration (13C: 74.7 ppm). Molecular weight of the polymer was determined by a Polymer Laboratories PL-GPC210 chromatograph at 140 C using o-dichlorobenzene as an eluent. The calibration was performed by polystyrene standard. Polymer concentration in injecting solution was 1 mg/mL and injection volume was 0.2 mL. X-ray diffraction measurements were performed on a Rigaku R-AXIS RAPID system with Mo Ka radiation (l ¼ 0.71069 Å, 2qmax ¼ 55 , crystal-to-detector distance ¼ 110 mm). The intensities were corrected for Lorentz and polarization effects. The structures were solved using a combination of direct methods (SHELXS-86) [23] and Fourier synthesis (DIRDIF94) [24], and refined by least-squares (SHELXL-97) [25] on Fo [2]. The non-hydrogen atoms were refined anisotropically and hydrogen atoms were placed using riding models. Least square fitting for the nine stereopentad distributions of the obtained polypropylene (PP) was performed by Microsoft Excel 2010 Solver program. 2.2. Synthesis of titanium complexes 2.2.1. Synthesis of adamantyl-substituted fluorenylamino ligand (3a) n BuLi (3.9 mL, 1.6 M solution in hexane, 6.4 mmol) was added dropwise to a solution of 1-adamantylamine (0.88 g, 5.8 mmol) in Et2O (40 mL) at 0 C. The mixture was stirred for 3 h at room temperature and the resulting solution was added dropwise to a solution of 9-(chlorodimethylsilyl)fluorene (1.5 g, 5.9 mmol) in Et2O (30 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane and the evaporation of the solvent gave yellowish oil of ligand 3a (2.0 g, 91%). 1H NMR (C6D6, 500 MHz): d ¼ 7.82 (m, 2H), 7.69e7.71 (m, 2H), 7.31 (m,4H), 3.82 (s, 1H), 1.91 (s, 3H), 1.47e1.57 (m, 12H), 0.57 (s, 1H), 0.05 (s, 6H). 2.2.2. Synthesis of 10-chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4tetrahydrobenzo[b]fluorene To a solution of 1,1,4,4-tetramethyl-2,2,3,3-tetrahydrobenzo[b] fluorene (2.2 g, 7.9 mmol) in Et2O (25 mL) was added dropwise n BuLi (5.3 mL, 1.6 M solution in hexane, 8.7 mmol) at 0 C. After stirring for 4 h at room temperature, the supernatant liquid was removed and the remaining lithium salt was suspended in hexane (25 mL). The suspension was added to a solution of dichlorodimethylsilane (4.7 mL, 40 mmol) in hexane at 78 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The solid was extracted with hexane and evapolation of the solvent gave yellowish oil of 10chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4-tetrahydrobenzo[b]fluorene (2.7 g, 91%). 1H NMR (CDCl3, 500 MHz): d ¼ 7.82 (d, 1H, J ¼ 10 Hz), 7.76 (s, 1H), 7.62 (d, 1H, J ¼ 10 Hz), 7.57 (s, 1H), 7.36 (dd, 1H, J ¼ 10 Hz), 7.27 (m, 1H), 4.00 (s, 1H), 1.74 (s, 4H), 1.33e1.37 (m, 12H), 0.20 (s, 3H), 0.12 (s, 3H). 2.2.3. Synthesis of adamantyl-substituted tetrahydrobenzo[b] fluorenylamino ligand (3b) n BuLi (3.4 mL, 1.6 M solution in hexane, 5.5 mmol) was added dropwise to a solution of cumylamine (0.76 g, 5.0 mmol) in Et2O
(40 mL) at 0 C. The mixture was stirred for 3 h at room temperature and the solution was added dropwise to a solution of 10chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4-tetrahydrobenzo[b]fluorene (1.6 g, 4.2 mmol) in Et2O (30 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane and the evaporation of the solvent gave yellowish oil of ligand 3b (1.8 g, 87%). 1H NMR (C6D6, 500 MHz): d ¼ 7.98 (s, 1H), 7.85 (d, 1H, J ¼ 7 Hz), 7.75 (s, 1H), 7.71 (d, 1H, J ¼ 9 Hz), 7.32 (m, 2H), 3.79 (s, 1H), 1.90 (s, 3H), 1,72 (m, 4H), 1.24e1.56 (m, 24H) 0.46 (s, 1H), 0.09 (s, 3H), 0.03 (s, 3H). 2.2.4. Synthesis of cumyl-substituted fluorenylamino ligand (3c) n BuLi (3.8 mL, 1.6 M solution in hexane, 6.0 mmol) was added dropwise to a solution of cumylamine (0.86 mL, 6.0 mmol) in Et2O (20 mL) at 0 C. The mixture was stirred for 4 h at room temperature and the resulting solution was added dropwise to a solution of 9-(chlorodimethylsilyl)fluorene (1.5 g, 6.0 mmol) in Et2O (20 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane (20 mL x3) and the evaporation of the solvent gave yellowish oil of ligand 3c (2.1 g, 98%). 1H NMR (C6D6, 500 MHz): d ¼ 7.65 (dd, 2H, J ¼ 8 Hz, 1 Hz), 7.40 (dd, 2H, J ¼ 8 Hz, 1 Hz), 7.17-6.89 (brm, 9H), 3.57 (s, 1H), 1.09 (s, 6H), 0.56 (brs, 1H), 0.28 (s, 6H). 2.2.5. Synthesis of cumyl-substituted tetrahydrobenzo[b] fluorenylamino ligand (3d) n BuLi (7.1 mL, 1.6 M solution in hexane, 12 mmol) was added dropwise to a solution of cumylamine (1.6 mL, 11 mmol) in Et2O (40 mL) at 0 C. The mixture was stirred for 3 h at room temperature and the solution was added dropwise to a solution of 10chlorodimethyl-1,1,4,4-tetramethyl-1,2,3,4-tetrahydrobenzo[b]fluorene (4.2 g, 11 mmol) in Et2O (40 mL) at 0 C. The resulting mixture was stirred overnight at room temperature and the solvent was removed in vacuo. The remaining solid was extracted with hexane and the evaporation of the solvent gave yellowish oil of ligand 3d (5.0 g, 97%). 1H NMR (C6D6, 500 MHz): d ¼ 7.96 (s, 1H), 7.83 (d, 1H, J ¼ 9 Hz), 7,65 (s, 1H), 7.57 (d, 1H, J ¼ 9 Hz), 7.45-7.07 (m, 7H), 3.71 (s, 1H), 1.72 (s, 4H), 1.24e1.42 (m, 18H) 0.75 (s, 1H), 0.04 (s, 3H), 0.07 (s, 3H). 2.2.6. Synthesis of fluorenyl-adamantylamido titanium complex (2a) To a solution of 3a (2.0 g, 5.3 mmol) in Et2O (30 mL) was slowly added excess MeLi (22 mL, 1.14 M solution in Et2O, 25 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.57 mL, 5.3 mmol) in hexane (40 mL) at room temperature, and the resulting red solution was stirred for 1 h. After the solvent was removed, the residue was extracted with hexane (100 mL) and MeMgBr (4.2 mL, 3.0 M solution in Et2O, 13 mmol) was added to the solution. After the resulting brown suspension was stirred for 1 h at room temperature, the solvent was removed and the residue was extracted with hexane (120 mL). The solution was concentrated to approximately 20 mL and cooled overnight at 30 C to give 2a as orange crystals (0.87 g, 33%). 1H NMR (C6D6, 500 MHz): d ¼ 7.84 (d, 2H, J ¼ 9 Hz), 7.70 (d, 2H, J ¼ 9 Hz), 7.24 (dd, 2H, J ¼ 9 Hz), 7.13 (dd, 2H, J ¼ 9 Hz), 2.00 (s, 6H), 1.97 (s, 3H), 1.54 (s, 6H), 0.72 (s, 6H), 0.00 (s, 6H), 13C NMR (C6D6, 125 MHz): d ¼ 135.1, 129.3, 124.9, 123.8, 82.1, 60.2, 56.9, 47.8, 36.6, 30.8, 6.4 (2 peaks in the aromatic region should be overlapped with solvent peaks). Anal. Calc. for C27H35NSiTi: C, 72.14; H, 7.85; N, 3.12. Found: C, 71.73; H, 7.45; N, 3.28.
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
2.2.7. Synthesis of tetrahydrobenzo[b]fluorenyl-adamantylamido titanium complex (2b) To a solution of 3b (1.8 g, 3.7 mmol) in Et2O (30 mL) was slowly added excess MeLi (15 mL, 1.14 M solution in Et2O, 17 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.40 mL, 3.7 mmol) in hexane (40 mL) at room temperature, and the resulting red solution was stirred for 1 h. After the solvent was removed, the residue was extracted with hexane (80 mL) and MeMgBr (3.0 mL, 3.0 M solution in Et2O, 9.0 mmol) was added to the solution. After the resulting black suspension was stirred for 1 h at room temperature, the solvent was removed and the residue was extracted with hexane (120 mL). The solution was concentrated to approximately 15 mL and cooled overnight at 30 C to give 2b as red crystals (0.59 g, 29%). 1H NMR (C6D6, 500 MHz): d ¼ 8.08 (s, 1H), 7.86 (d, 1H, J ¼ 7 Hz), 7.82 (s, 1H), 7.73 (d, 1H, J ¼ 7 Hz), 7.24 (dd, 1H, J ¼ 7 Hz, 7 Hz), 7.15 (br, 1H), 2.03 (s, 6H), 1.96 (s, 3H), 1.65 (s, 4H), 1.54 (br, 6H), 1.41 (s, 3H), 1.37 (s, 6H), 1.28 (s, 3H), 0.83 (s, 3H), 0.75 (s, 3H), 0.06 (s, 3H), 0.05 (s, 3H), 13C NMR (C6D6, 125 MHz): d ¼ 146.3, 144.7, 135.5, 133.6, 129.5, 127.9, 125.1, 124.7, 123.6, 120.7, 81.3, 60.0, 56.0, 55.6, 47.8, 36.6, 35.5, 35.3, 35.2, 35.0, 33.7, 33.0, 32.8, 32.6, 30.8, 6.4, 6.4 (2 peaks in the aromatic region should be overlapped with solvent peaks). Anal. Calc. for C35H49NSiTi: C, 75.10; H, 8.82; N, 2.50. Found: C, 75.17; H, 9.10; N, 2.66. 2.2.8. Synthesis of fluorenyl-cumylamido titanium complex (2c) To a solution of 3c (2.1 g, 5.9 mmol) in Et2O (30 mL) was slowly added excess MeLi (22 mL, 1.14 M solution in Et2O, 25 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.63 mL, 5.9 mmol) in hexane (20 mL) at room temperature, and the resulting red solution was stirred for 1 h at room temperature. After the solvent was removed, the residue was extracted with Et2O (30 mL x2) and MeMgBr (4.9 mL, 3.0 M solution in THF, 15 mmol) was added to the solution. After the resulting brown suspension was stirred for 3 h at room temperature, the solvent was removed and the residue was extracted with hexane (120 mL). The solution was concentrated to approximately 20 mL and cooled overnight at 30 C to give 2c as orange crystals (0.89 g, 34%). 1H NMR (C6D6, 500 MHz): d ¼ 7.77 (d, 2H, J ¼ 8 Hz), 7.61 (d, 2H, J ¼ 8 Hz), 7.32 (d, 2H, J ¼ 7 Hz), 7.21 (t, 2H, J ¼ 7 Hz), 7.11 (m, 4H), 7.04 (dd, 1H, J ¼ 8 Hz, 7 Hz), 1.80 (s, 6H), 0.28 (s, 6H), 0.01 (s, 6H), 13C NMR (C6D6, 125 MHz): d ¼ 150.2, 135.0, 129.1, 128.2, 127.9, 127.9, 126.8, 126.4, 125.0, 123.8, 83.0, 62.8, 57.8, 33.8, 4.8. Anal. Calc. for C26H31NSiTi: C, 72.04; H, 7.21; N, 3.23. Found: C, 72.22; H, 7.56; N, 3.27. 2.2.9. Synthesis of tetrahydrobenzo[b]fluorenyl-cumylamido titanium complex (2d) To a solution of 3d (2.6 g, 5.6 mmol) in Et2O (30 mL) was slowly added excess MeLi (22 mL, 1.14 M solution in Et2O, 25 mmol) at room temperature and the mixture was stirred for 4 h. The solution of dilithium species was added to a solution of TiCl4 (0.62 mL, 5.7 mmol) in hexane (40 mL) at room temperature, and the resulting red solution was stirred for 1 h. After the solvent was removed, the residue was extracted with hexane (100 mL) and MeMgBr (4.7 mL, 3.0 M solution in Et2O, 14 mmol) was added to the solution. After the resulting brown suspension was stirred for 1 h at room temperature, the solvent was removed and the residue was extracted with hexane (70 mL). The solution was concentrated and cooled overnight at 30 C to give 2d as red crystals (0.65 g, 21%). 1 H NMR (C6D6, 500 MHz): d ¼ 8.06 (s, 1H), 7.82 (d, 1H, J ¼ 8 Hz), 7.78 (s, 1H), 7.68 (d, 1H, J ¼ 8 Hz), 7.39 (d, 2H, J ¼ 7 Hz), 7.22 (t, 1H, J ¼ 7 Hz), 7.13e7.19 (m, 3H), 7.08 (dd, 1H, J ¼ 8 Hz), 1.89 (s, 3H), 1.82 (s, 3H), 1.66 (s, 4H), 1.40 (s, 3H), 1.37 (s, 3H), 1.34 (s, 6H), 0.36 (s, 6H), 0.09 (s, 3H), 0.03 (s, 3H), 13C NMR (C6D6, 125 MHz): d ¼ 150.5,
97
146.5, 144.8, 135.5, 133.7, 129.4, 126.8, 126.5, 125.1, 124.8, 123.7, 120.8, 82.3, 62.7, 56,7. 56.5, 35.5, 35.3, 35.2, 35.0, 34.3, 33.7, 33.5, 33.0, 32.8, 32.6, 5.1, 4.6 (4 peaks in the aromatic region should be overlapped with solvent peaks). Anal. Calc. for C34H45NSiTi: C, 75.11; H, 8.34; N, 2.58. Found: C, 74.79; H, 8.73; N, 2.48. 2.3. General polymerization procedure of propylene To a 100-mL 2-necked flask, MMAO (6.5 wt% Al in hexane, 1.8 mL, 4.0 mmol) and heptane (27.2 mL) were charged under nitrogen and stirred for 30 min at 0 C. N2 in headspace of the flask was evacuated and the solvent was saturated with an atomospheric pressure of propylene. The polymerization was started under the flow of propylene by the addition of a solution of titanium catalyst (20 mmol) in heptane (1.0 mL). After the polymerization was conducted for 10 min under an atomospheric pressure of propylene, the reaction mixture was poured into 200 mL of MeOH containing 4 mL of concentrated HCl and the precipitated polymer was collected by filtration. The obtained polymer was dried for 6 h under vacuum at 60 C to be a constant weight. 3. Results and discussions Fluorenylamido titanium complex 2a-2d were synthesized according to the synthetic method of the corresponding t-butylamidotitanium complex 1a (Scheme 1) [20]. The ligands 3a-3d were synthesized from fluorene or 2,3-alkyl-substituted fluorene by chlorodimethylsilylation at the 9-position of the fluorenyl ring followed by amination. Successive treatment with excess methyllithium, an equimolar amount of TiCl4 and 2 equivalent of MeMgBr gave the corresponding dimethyltitanium complexes in moderate yields. Single crystal of 2a, 2b and 2d suitable for the X-ray analysis was obtained and their ORTEP drawings are shown at Fig. 2. The bond lengths between the titanium center and the fluorenyl carbon C(1), C(2), and C(5) were 2.23e2.43 Å, respectively, indicating that the fluorenyl group is coordinating to the titanium center with h3- to h1-form, which was the same tendency as 1a (Table 1). Miller et al. evaluated the hapticity of the fluorenyl ligands of some zirconium complexes by comparing the CeC bond lengths of the fivemembered ring [8,11]. They used the value (aþb-c-d)/2 as a parameter of h1-form tendency, where a, b, c and d are the bond lengths of C(1)-C(2), C(1)-C(5), C(2)-C(3), and C(4)-C(5), respectively. The (aþb-c-d)/2 value of 1c, 2b and 2d were 0.033e0.048 Å, which were much larger than those of the other h5-fluorenyl complexes including previously reported zirconium complexes (0.014e0.028 Å). These results indicated that the alkylation of 2,3position made the fluorenyl group to coordinate in h1-form. We applied Tolman cone angle [26] to estimate the steric effect of the amino group (Table 2). These values indicate that tBu and 1adamantyl groups have similar steric bulkiness (65e68 ) and cumyl (a, a-dimethylbenzyl) group is more sterically hindered (82 ). Next, propylene polymerization using titanium complexes 2a2d was performed (Table 3, run 4e7). The complexes were activated by MMAO and the polymerization was conducted with a semi-batch reaction under 1 atm of propylene at 0 C in heptane. The previously reported results using 1a-1c are also shown for the comparison (run 1e3). Polymer with high molecular weight (~30,000) was obtained by each complex. All the polypropylene obtained by 2a-2d was amorphous that showed no melting point in the DSC analysis. 13C NMR analysis of the obtained PPs indicated no 2,1-inserted, namely, regioirregular propylene unit. Among Cs-symmetric complexes 1a-1c, 2a and 2c, only 2c gave atactic PP, while the other complexes produced syn-PP. The stereospecificities of C1-symmetric complexes 2b and 2d resemble
98
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
Fig. 1. Structures of fluorenylamidotitanium complexes 1a-c, 2a-2d.
Scheme 1. Synthesis of ansa-dimethylsilylene(fluorenyl) (cumylamido)titanium complexes 2a-2d.
Fig. 2. Structure of fluorenylamidotitanium complexes 2a, 2b and 2d.
Table 1 Selected bond lengths (Å) of 1a-1c, 2a, 2b and 2d. Parameters
1a17
1b12
1c13
2a
2b
2d
TieC(1) TieC(2) TieC(3) TieC(4) TieC(5) C(1)-C(2) (a) C(1)-C(5) (b) C(2)-C(3) (c) C(4)-C(5) (d) (a þ b e c e d)/2
2.366(3) 2.415(3) 2.572(3) 2.573(3) 2.418(3) 1.448(2) 1.455(3) 1.425(2) 1.426(2) 0.026
2.247(3) 2.422(3) 2.619(3) 2.609(3) 2.401(3) 1.462(4) 1.447(4) 1.424(4) 1.436(4) 0.025
2.229(4) 2.424(4) 2.620(1) 2.589(1) 2.412(3) 1.461(7) 1.464(6) 1.411(7) 1.419(8) 0.048
2.241(4) 2.399(5) 2.624(5) 2.643(4) 2.433(4) 1.441(6) 1.456(5) 1.436(4) 1.433(5) 0.014
2.226(4) 2.387(4) 2.585(5) 2.585(7) 2.391(6) 1.464(5) 1.454(10) 1.423(9) 1.419(6) 0.038
2.248(3) 2.401(3) 2.608(4) 2.606(3) 2.422(3) 1.445(5) 1.458(5) 1.428(5) 1.410(5) 0.033
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100 Table 2 Tolman's cone angle of amido groups of complexes 1a-1c, 2a, 2b and 2b.
q ( ) 67 65 67 68 68 82
N-alkyl Bu t Bu t Bu Ad Ad Me2PhC t
1a 1b 1c 2a 2b 2d
Table 3 Polymerization of propylene with Ti complex 1a-2d activated by MMAO.
Run
Cat.
Time (min)
Aa
Mnb (104)
PDIb
rrc (%)
mrc (%)
mmc (%)
1 2d 3e 4 5 6 7
1a 1b 1c 2a 2b 2c 2d
15 7 0.5 30 10 3 3
160 920 11,000 100 500 1500 1600
3.3 17 20 6.0 4.0 3.3 3.2
1.7 1.5 2.9 1.6 1.8 1.7 1.9
73 93 61 71 66 33 36
21 5 29 23 26 48 41
6 2 10 5 8 19 23
a b c d e
with those of the corresponding Cs-symmetric complex 2a and 2c. These results indicated that the introduction of sterically demanded cumyl group reduce the stereospecificity, whereas alkyl group on 2,3-position of the fluorenyl group has little influence on the stereospecificity. To reveal the origin of the different stereospecificity, we applied Busico's chain propagation model described with the probability of enantiofacial preference (s) and chain migration probability without monomer insertion (P) [27,28]. They previously reported that the stereopentad distributions of PP obtained by 3,6disubstituted fluorenylamidotitanium complex closely fit to this statistical model [17]. Our experimental result using 1a also Table 4 Stereopentad distribution of the obtained PPs (%) from 1a and 2c. 1a
2c
mmmm mmmr rmmr mmrr mmrm þ rmrr rmrm rrrr mrrr mrrm
1 1 4 9 8 4 59 13 1 93.2 4.4 1.03
4 9 6 12 23 13 10 16 7 95.0 42.2 0.82
a
Least square error of nine distributions from the fitting model.
Fluorenylamido-ligated titanium complexes 2a-2d were synthesized and characterized. The syn-specificity of propylene polymerization was reduced when bulky group was introduced to amide or 2,3-position of the fluorenyl ring. The least-square fitting analysis of the stereosequence of 2c assuming the enantiomorphicsite control and chain migration model indicated that the reduction of syn-specificity was mainly ascribed to the chain migration without monomer insertion. The promotion of chain migration is probably because of the sterically hindered environment around the metal center. Acknowledgements We are grateful to Tosoh Finechem Co. for generous donation of chemicals. We also gratefully acknowledge the Natural Science Center for Basic Research and Development (N-BARD), Hiroshima University, for the high temperature NMR measurement. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jorganchem.2015.12.028. References
Cat.
P RMSa
supported this model. The amount of isolated meso-sterodefect increased when polar solvent such as chlorobenzene was used, which indicated that the rate of chain migration became faster than that of insertion in the polar solvent [29]. s and P values were calculated from stereopentad distributions of the polymer obtained by 1a and 2c (Table 4). Both of the distributions were within 2% error to the calculated s and P values. The resulting s values (93e95%) were not so different between 1a and 2c, but the P value of 2c (42.2%) was much larger than that of the 1a (4.4%). Therefore, lower stereospecificity of 2c than 1a is because of more frequent chain migration without monomer insertion. This is probably because fluorenyl group is forced to coordinate to the metal center with h1-form by the sterically hindered cumyl group during the polymerization, which resulted in the acceleration of chain migration. Considering the difference between complex 1a and 2c, the bulky cumyl group on 2d should also facilitate frequent chain migration without monomer insertion. Under the frequent chain migration condition, the preference of propagation from one conformer over the other conformer should give isotactic polymer. However, the obtained polymer from 2d was atactic, which has almost the same stereotriad distribution as that by 2c. This result shows that the propagation pathways from two distinct coordination sites in 2d were not energetically different enough to give isotactoid. 4. Conclusion
Activity as kg mol Ti1 h1. Determined by GPC calibrated with polystyrene standard. Stereotriad distributions determined by13C NMR. Data taken from ref. [12] using dried MMAO. Data taken from ref. [13] using 2.0 mmol of dried MMAO.
s
99
[1] J.C. Stevens, F.J. Timmers, D.R. Wilson, G.F. Schmidt, P.N. Nickias, R.K. Rosen, G.W. Knight, S.-Y. Lai, EP0416815, 1991. [2] J.A.M. Canich, US5264405, 1993. [3] A.L. McKnight, R.M. Waymouth, Chem. Rev. 98 (2010) 2587. [4] H. Braunschweig, F.M. Breitling, Coord. Chem. Rev. 250 (2006) 2691. [5] J. Okuda, F.J. Schattenmann, S. Wocadlo, W. Massa, Organometallics 14 (1995) 789. [6] H.V.R. Dias, Z. Wang, S.G. Bott, J. Organomet. Chem. 508 (1996) 91. [7] H.V.R. Dias, Z. Wang, J. Organomet. Chem. 539 (1997) 77. [8] L.J. Irwin, J.H. Reibenspies, S.A. Miller, Polyhedron 24 (2005) 1314. [9] J. Chai, K.A. Abboud, S.A. Miller, Dalton Trans. 42 (2013) 9139. [10] Y. Nakayama, Y. Sogo, Z. Cai, T. Shiono, J. Polym. Sci. Part A Polym. Chem. 51 (2013) 1223. [11] G. Xu, Macromolecules 31 (1998) 2395. [12] E. Kirillov, A. Razavi, J.-F. Carpentier, J. Mol. Catal. Part A Chem. 249 (2006)
100
R. Tanaka et al. / Journal of Organometallic Chemistry 804 (2016) 95e100
230. [13] T. Shiono, M. Sugimoto, T. Hasan, Z. Cai, T. Ikeda, Macromolecules 41 (2008) 8292. [14] T. Shiono, M. Sugimoto, T. Hasan, Z. Cai, Macromol. Chem. Phys. 214 (2013) 2239. [15] R. Tanaka, T. Ikeda, Y. Nakayama, T. Shiono, Polymer 56 (2015) 218. [16] A. Razavi, U. Thewalt, J. Organomet. Chem. 621 (2001) 267. [17] V. Busico, R. Cipullo, F. Cutillo, G. Talarico, A. Razavi, Macromol. Chem. Phys. 204 (2003) 1269. [18] L.J. Irwin, S.A. Miller, J. Am. Chem. Soc. 127 (2005) 9972. [19] Z. Cai, T. Ikeda, M. Akita, T. Shiono, Macromolecules 38 (2005) 8135. [20] K. Nishii, H. Hagihara, T. Ikeda, M. Akita, T. Shiono, J. Organomet. Chem. 691 (2006) 193. [21] T. Shiono, R. Harada, Z. Cai, Y. Nakayama, Top. Catal. 52 (2009) 675. [22] S.A. Miller, J.E. Bercaw, Organometallics 23 (2004) 1777.
[23] G.M. Sheldrick, SHELXS-86: Program for Crystal Structure Determination, €ttingen, Go €ttingen, Germany, 1986. University of Go [24] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, S. Garcia-Granda, R.O. Gould, J.M.M. Smits, C. Smykalla, The DIRDIF Program System, Technical Report of the Crystallography Laboratory, University of Nijmegen, Nijmegen, The Netherland, 1992. [25] G.M. Sheldrick, SHELXL-97: Program for Crystal Structure Refinement, Uni€ ttingen, Go €ttingen, Germany, 1997. versity of Go [26] C.A. Tolman, J. Am. Chem. Soc. 96 (1974) 53. [27] V. Busico, R. Cipullo, P. Corradini, L. Landriani, M. Vacatello, A.L. Segre, Macromolecules 28 (1995) 1887. [28] V. Busico, R. Cipullo, Prog. Polym. Sci. 26 (2001) 443. [29] K. Nishii, T. Matsumae, E. O'Dare, T. Shiono, T. Ikeda, Macromol. Chem. Phys. 205 (2004) 363.