Preparation of “Constrained Geometry” Titanium Complexes of [1,2]Azasilinane Framework for Ethylene/1-Octene Copolymerization

The Me2Si-bridged ansa-Cp/amido half-metallocene, [Me2Si(η5-Me4C5)(NtBu)]TiCl2, termed a “constrained-geometry catalyst (CGC)”, is a representative homogeneous Ziegler catalyst. CGC derivatives with the [1,2]azasilinane framework, in which the amide alkyl substituent is joined by the Si-bridge, were prepared, and the catalytic performances of these species was studied. Me4C5HSi(Me)(CH2CH=CH2)-NH(C(R)(R’)CH=CH2) (R, R’ = H or methyl; Me4C5H = tetramethylcyclopentadienyl) was susceptible to ring closure metathesis (RCM) when treated with Schrock’s Mo-catalyst to afford -Si(Me4C5H)(Me)CH2CH=CHC(R)(R’)NH- containing a six-membered ring framework. Using the precursors and the products of RCM, various CGC derivatives, i.e., [-Si(η5-Me4C5)(Me)CH2CH=CHC(R)(H)N-]TiMe2 (13, R = H; 15, R = Me), [-Si(η5-Me4C5)(Me)CH2CH2CH2CH2N]TiMe2 (14), [(η5-Me4C5)Si(Me)(CH2CH=CH2)NCH2CH=CH2]TiMe2 (16), [(η5-Me4C5)Si (Me)(CH=CH2)NCH2CH=CH2]TiMe2 (17), and [(η5-Me4C5)Si(Me)(CH2CH3)NCH2CH2CH3]TiMe2 (18), were prepared. The catalytic activity of the newly prepared complexes was lower than that of CGC when activated with [Ph3C][B(C6F5)4]/iBu3Al. However, the catalytic activity of these species was improved by using tetrabutylaluminoxane ([iBu2Al]2O) instead of iBu3Al and the activity of 14/[Ph3C][B(C6F5)4]/[iBu2Al]2O was comparable to that of CGC/[Ph3C][B(C6F5)4]/iBu3Al (4.7 and 5.0 × 106 g/mol-Ti, respectively). Advantageously, the newly prepared complexes produced higher molecular weight poly(ethylene-co-1-octene)s than CGC.


Synthesis and Characterization
The present strategy for construction of the [1,2]azasilinane framework involves the ring closure metathesis (RCM) of the two olefin units attached to the nitrogen and silicon atoms (Scheme 1). The precursor (3, Me4C5H-Si(Me)(CH2CH=CH2)-NH(CH2CH=CH2)) for RCM was easily prepared according to the scheme applied in the construction of the prototype CGC ligand by using commercially available chemicals, tetramethylcyclopentadiene (Me4C5H2), allyl(dichloro)methylsilane ((CH2=CHCH2)(Me)SiCl2), and allylamine. Initial RCM trials of 3 using Ru-catalysts (Grubbs 1st and 2nd generation catalysts and Hoveyda-Grubbs catalyst) for the synthesis of 6 were unsuccessful [40]. No ethylene gas was detected in the reaction monitored by 1 H-NMR. Instead, migration of the double bond in the allylamine unit was observed. In contrast, the Mo-catalyst (Schrock catalyst, 2 mol %) cleanly converted 3 to the desired cyclic compound 6 [41][42][43][44]. When a closed-NMR tube with C6D6 was used, the signal of generated ethylene was observed at 5.24 ppm and the vinyl signals that were observed in the spectrum of 3 at 4.8-5.2 ppm and 5.7-5.9 ppm completely disappeared, while two new =CH signals were observed at 5.8 and 5.5 ppm (Figure 1). The RCM product 6 could be separated from the Mo-catalyst by vacuum distillation at 60 °C. The derivatives bearing methyl substituents at the α-position of the amino group (7 and 8) were also successfully prepared by RCM of the corresponding precursors 4 and 5, which were prepared using 1-methylallylamine or 1,1-dimethylallylamine instead of allylamine. Hydrogenation of 6 using a Pd/C catalyst afforded 9 bearing the desired [1,2]azasilinane framework; the double bonds in the tetramethylcyclopentadiene units remained intact during the hydrogenation process. The characteristic vinyl signals in the 1 H-NMR spectrum of 6 completely disappeared after hydrogenation. However, hydrogenation of the derivatives 7 and 8 bearing methyl substituent(s) at the α-position of the amino group was unsuccessful; the substrates remained intact under the identical hydrogenation conditions using the Pd/C catalyst. The five-membered analogue 11 was also successfully prepared via RCM of the vinylsilane derivative using the same Mo catalyst. Hydrogenation of 11 using the Pd/C catalyst was also successful to afford 12 bearing the [1,2]azasilolidine framework (Scheme 1).

Chart 1.
Ortho-phenylene-bridged half-metallocene complexes and the complexes targeted in this work.

Synthesis and Characterization
The present strategy for construction of the [1,2]azasilinane framework involves the ring closure metathesis (RCM) of the two olefin units attached to the nitrogen and silicon atoms (Scheme 1). The precursor (3, Me 4 C 5 H-Si(Me)(CH 2 CH=CH 2 )-NH(CH 2 CH=CH 2 )) for RCM was easily prepared according to the scheme applied in the construction of the prototype CGC ligand by using commercially available chemicals, tetramethylcyclopentadiene (Me 4 C 5 H 2 ), allyl(dichloro)methylsilane ((CH 2 =CHCH 2 )(Me)SiCl 2 ), and allylamine. Initial RCM trials of 3 using Ru-catalysts (Grubbs 1st and 2nd generation catalysts and Hoveyda-Grubbs catalyst) for the synthesis of 6 were unsuccessful [40]. No ethylene gas was detected in the reaction monitored by 1 H-NMR. Instead, migration of the double bond in the allylamine unit was observed. In contrast, the Mo-catalyst (Schrock catalyst, 2 mol %) cleanly converted 3 to the desired cyclic compound 6 [41][42][43][44]. When a closed-NMR tube with C 6 D 6 was used, the signal of generated ethylene was observed at 5.24 ppm and the vinyl signals that were observed in the spectrum of 3 at 4.8-5.2 ppm and 5.7-5.9 ppm completely disappeared, while two new =CH signals were observed at 5.8 and 5.5 ppm (Figure 1). The RCM product 6 could be separated from the Mo-catalyst by vacuum distillation at 60 • C. The derivatives bearing methyl substituents at the α-position of the amino group (7 and 8) were also successfully prepared by RCM of the corresponding precursors 4 and 5, which were prepared using 1-methylallylamine or 1,1-dimethylallylamine instead of allylamine. Hydrogenation of 6 using a Pd/C catalyst afforded 9 bearing the desired [1,2]azasilinane framework; the double bonds in the tetramethylcyclopentadiene units remained intact during the hydrogenation process. The characteristic vinyl signals in the 1 H-NMR spectrum of 6 completely disappeared after hydrogenation. However, hydrogenation of the derivatives 7 and 8 bearing methyl substituent(s) at the α-position of the amino group was unsuccessful; the substrates remained intact under the identical hydrogenation conditions using the Pd/C catalyst.
The five-membered analogue 11 was also successfully prepared via RCM of the vinylsilane derivative using the same Mo catalyst. Hydrogenation of 11 using the Pd/C catalyst was also successful to afford 12 bearing the [1,2]azasilolidine framework (Scheme 1).  The metalation of 6 using the methods introduced by Resconi et al. (treatment of 4 equiv. of MeLi in diethyl ether then with TiCl4) was unsuccessful [45,46]. The action of MeLi on 6 did not afford the desired dilithium compound (MeC5(Li)-Si(Me)(CH2CH=CH2)-N(Li)(CH2CH=CH2)), but instead resulted Scheme 1. Construction of the ligand frameworks for the targeted complexes.  The metalation of 6 using the methods introduced by Resconi et al. (treatment of 4 equiv. of MeLi in diethyl ether then with TiCl4) was unsuccessful [45,46]. The action of MeLi on 6 did not afford the desired dilithium compound (MeC5(Li)-Si(Me)(CH2CH=CH2)-N(Li)(CH2CH=CH2)), but instead resulted The metalation of 6 using the methods introduced by Resconi et al. (treatment of 4 equiv. of MeLi in diethyl ether then with TiCl 4 ) was unsuccessful [45,46]. The action of MeLi on 6 did not afford the desired dilithium compound (MeC 5 (Li)-Si(Me)(CH 2 CH=CH 2 )-N(Li)(CH 2 CH=CH 2 )), but instead resulted in destruction of the ligand framework. However, treatment of 6 with 4 equiv. of MeMgCl for a rather long reaction time (three days) in tetrahydrofuran (THF) at room temperature cleanly generated the desired dianion (MeC 5 (MgCl)-Si(Me)(CH 2 CH=CH 2 )-N(MgCl)(CH 2 CH=CH 2 )). Addition of TiCl 4 (DME) to the resulting solution containing the dianionic species and remaining 2 equiv. of MeMgCl afforded the desired complex 13 (Chart 2) in moderate yield (63%). The 1 H-NMR spectrum showed Ti-Me and Si-Me signals at 0.59, 0.46, and 0.43 ppm as singlets, while C 5 Me 4 -signals were observed at 1.98, 1.96, 1.94, and 1.87 ppm as singlets. The two protons on -NCH 2 -are diastereotopic to each other and their signals were widely separated at 5.47 and 4.63 ppm with a large geminal coupling constant ( 2 J = 19.5 Hz). The two =CH signals were observed at 5.7 and 5.5 ppm as multiplets ( Figure 1). Complex 14 was also successfully prepared from 9 (hydrogenated compound of 6) by the same method under the same conditions (treatment with 4 equiv. of MeMgCl and subsequent addition of TiCl 4 (DME)). The 1 H-NMR spectrum of 14 ( Figure 1) shows signals of the two diastereotopic -NCH 2 -protons at 5.18 and 3.65 ppm, indicating a notable up-field shift relative to the corresponding chemical shifts of 13 (5.47 and 4.63 ppm). When 7 (RCM precursor bearing a methyl substituent at the α-position of the amino group) was treated with MeMgCl, deprotonation was sluggish. Treatment of 2 equiv. of nBuLi in diethyl ether cleanly precipitated the dilithium compound, which was subsequently treated with TiCl 4 (DME) and 2 equiv. of MeMgCl simultaneously to afford the desired titanium complex 15. Having two chiral centers, 15 was obtained as a mixture of two diastereomers and the 1 H-and 13 C-NMR signals were rather complex. Titanium complexes 16 and 17 were successfully generated from the respective RCM precursors 3 and 10. In these cases, treatment of nBuLi in diethyl ether cleanly precipitated the corresponding dilithium compounds, from which the desired complexes were generated in good yields (84% and 88%) by treatment with TiCl 4 (DME) and 2 equiv. of MeMgCl. The vinyl groups in 17 were hydrogenated using a Pd/C catalyst to obtain 18 ([Me(Et)Si(η 5 -Me 4 C 5 )(κ 1 -NPr)]TiMe 2 ). All complexes, except for 14, were oily in nature with high solubility in hexane, and resistant to recrystallization, but the crude products extracted using hexane were pure based on 1 H-and 13 C-NMR analyses. Complex 14 was crystalline, but the crystals were formed as thin needles that were unsuitable for X-ray crystallography. The two =CH signals were observed at 5.7 and 5.5 ppm as multiplets ( Figure 1). Complex 14 was also successfully prepared from 9 (hydrogenated compound of 6) by the same method under the same conditions (treatment with 4 equiv. of MeMgCl and subsequent addition of TiCl4(DME)). The 1 H-NMR spectrum of 14 ( Figure 1) shows signals of the two diastereotopic -NCH2-protons at 5.18 and 3.65 ppm, indicating a notable up-field shift relative to the corresponding chemical shifts of 13 (5.47 and 4.63 ppm). When 7 (RCM precursor bearing a methyl substituent at the α-position of the amino group) was treated with MeMgCl, deprotonation was sluggish. Treatment of 2 equiv. of nBuLi in diethyl ether cleanly precipitated the dilithium compound, which was subsequently treated with TiCl4(DME) and 2 equiv. of MeMgCl simultaneously to afford the desired titanium complex 15. Having two chiral centers, 15 was obtained as a mixture of two diastereomers and the 1 H-and 13 C-NMR signals were rather complex. Titanium complexes 16 and 17 were successfully generated from the respective RCM precursors 3 and 10. In these cases, treatment of nBuLi in diethyl ether cleanly precipitated the corresponding dilithium compounds, from which the desired complexes were generated in good yields (84% and 88%) by treatment with TiCl4(DME) and 2 equiv. of MeMgCl. The vinyl groups in 17 were hydrogenated using a Pd/C catalyst to obtain 18 ([Me(Et)Si(η 5 -Me4C5)(κ 1 -NPr)]TiMe2). All complexes, except for 14, were oily in nature with high solubility in hexane, and resistant to recrystallization, but the crude products extracted using hexane were pure based on 1 H-and 13 C-NMR analyses. Complex 14 was crystalline, but the crystals were formed as thin needles that were unsuitable for X-ray crystallography. In the case of 8 (compounds bearing two methyl substituents at the α-position), treatment of nBuLi in diethyl ether cleanly precipitated the desired dilithium compound as a white solid, but metallation was unsuccessful. In the cases of 11 and 12 (five-membered ring compounds), various metallation attempts were also unsuccessful. In another route, RCM of the titanium complex 17 bearing vinyl-Si and allyl-N groups was also unsuccessfully attempted. In contrast, RCM of the titanium complex 16 bearing allyl-Si and allyl-N groups using the Mo-catalyst successfully generated six-membered ring complex 13 (Scheme 2). The bite angle in the cases of the five membered ring species 11 and 12 may not be practicable for chelation.

Chart 2. Prepared Ti-complexes.
In the case of 8 (compounds bearing two methyl substituents at the α-position), treatment of nBuLi in diethyl ether cleanly precipitated the desired dilithium compound as a white solid, but metallation was unsuccessful. In the cases of 11 and 12 (five-membered ring compounds), various metallation attempts were also unsuccessful. In another route, RCM of the titanium complex 17 bearing vinyl-Si and allyl-N groups was also unsuccessfully attempted. In contrast, RCM of the titanium complex 16 bearing allyl-Si and allyl-N groups using the Mo-catalyst successfully generated six-membered ring complex 13 (Scheme 2). The bite angle in the cases of the five membered ring species 11 and 12 may not be practicable for chelation.

Polymerization Studies
The newly prepared complexes 13-18 along with the prototype CGC [Me2Si(η 5 -Me4C5)(N t Bu)]TiMe2 as a comparison were screened for ethylene/1-octene copolymerization in aliphatic hydrocarbon solvent (methylcyclohexane) after activation with [Ph3C][B(C6F5)4]. Triisobutylaluminum (iBu3Al) was used as a scavenger and the temperature was controlled in the range of 130-150 °C. The polymerization results are summarized in Table 1. None of the newly prepared complexes exhibited activity superior to that of CGC (entries 1-6 versus entry 7). The highest activity was observed with 15, but the activity was ~2/3 that of CGC (3.1 × 10 6 versus 5.0 × 10 6 g/mol-Ti). The activities of 13 and 14 were ~1/3 that of CGC, while the activity of 18 was ~1/2 that of CGC. Complexes 16 and 17 containing acyclic olefin units showed low activity (<1/4 that of CGC). Advantageously, the molecular weights of the products were higher than that of the polymer generated with CGC. Under the polymerization conditions employing methylcyclohexane at high temperature (120-150 °C), CGC produced relatively low-molecular-weight polymer (Mn, 17 kDa) with a rather broad molecular weight distribution (Mw/Mn 4.41), in contrast with the narrow molecular Scheme 2. Ring closure metathesis (RCM) of the titanium complexes.

Polymerization Studies
The newly prepared complexes 13-18 along with the prototype CGC [Me 2 Si(η 5 -Me 4 C 5 )(N t Bu)]TiMe 2 as a comparison were screened for ethylene/1-octene copolymerization in aliphatic hydrocarbon solvent (methylcyclohexane) after activation with [Ph 3 C][B(C 6 F 5 ) 4 ]. Triisobutylaluminum (iBu 3 Al) was used as a scavenger and the temperature was controlled in the range of 130-150 • C. The polymerization results are summarized in Table 1. None of the newly prepared complexes exhibited activity superior to that of CGC (entries 1-6 versus entry 7). The highest activity was observed with 15, but the activity was~2/3 that of CGC (3.1 × 10 6 versus 5.0 × 10 6 g/mol-Ti). The activities of 13 and 14 were~1/3 that of CGC, while the activity of 18 was~1/2 that of CGC. Complexes 16 and 17 containing acyclic olefin units showed low activity (<1/4 that of CGC). Advantageously, the molecular weights of the products were higher than that of the polymer generated with CGC. Under the polymerization conditions employing methylcyclohexane at high temperature (120-150 • C), CGC produced relatively low-molecular-weight polymer (M n , 17 kDa) with a rather broad molecular weight distribution (M w /M n 4.41), in contrast with the narrow molecular weight distributions (M w /M n 1.7-3.4) obtained when the polymerization was performed in toluene at lower temperature (70-90 • C) [25,27,47]. In this study, a bimodal molecular weight distribution was observed for CGC with the low molecular weight portion comprising a main modal and the high molecular weight portion being a shoulder ( Figure 2). All other complexes (13)(14)(15)(16)(17)(18)   It was postulated that the low activities might be attributed to facile formation of "Ti(μ-R)AlR2"type heterobimetallic species due to the highly accessible reaction site. Such heterobimetallic species have been recognized to be generated as a dormant state in the olefin polymerization process [48,49]. In order to suppress formation of such dormant species, Busico and coworkers treated MAO with 2,6-di-tert-butylphenol with the aim to trap free Me3Al contained in MAO, and Soares and coworkers used highly bulky trioctylaluminum as a scavenger [50][51][52]. In this study, we explored the use of tetrabutylaluminoxane ([iBu2Al]2O) instead of iBu3Al as a scavenger.
[iBu2Al]2O is a commercially available chemical that is easily synthesized from iBu3Al by treatment with water [53,54]. By employing [iBu2Al]2O as a scavenger instead of iBu3Al, the activity was improved in all cases (entries 1-7 versus 8-16). The improvement was most significant in the case of 14, where the activity improved 2.7-fold to reach 4.7 × 10 6 g/mol-Ti, which is comparable to that of CGC/[Ph3C][B(C6F5)4]/iBu3Al (5.0 × 10 6 g/mol-Ti). In the other cases, the activity improved ~1.5-fold. Both 14 and CGC were deactivated during the course of high temperature polymerization; the yields were marginally increased from 2.36 g (entry 9) to 2.56 g and from 3.58 g (entry 15) to 3.76 g for 14 and CGC, respectively, by lengthening the reaction time from 3 min to 6 min.
The Mn values declined somewhat with the use of [iBu2Al]2O instead of iBu3Al, but the general trend of the molecular weight versus the catalyst structure was the same (Figure 2). In the cases of 14-18, the Mn values decreased by a factor of ~0.65 upon replacement of the iBu3Al with [iBu2Al]2O, while the decrease was marginal in the case of 18, which invariably produced high molecular weight polymers (Mn, ~70 kDa). The most significant decline of the Mn value was observed in the case of CGC, resulting in the production of a low-molecular-weight sticky polymer (Mn, 9.3 kDa) with bimodal molecular weight distribution.
Contrary to the expectation, the 1-octene contents of the samples generated using 13-17 were lower (12 mol %-15 mol %) than that of CGC (21 mol %) (Entries 8-13 versus 15). In the case of 18, the It was postulated that the low activities might be attributed to facile formation of "Ti(µ-R)AlR 2 "type heterobimetallic species due to the highly accessible reaction site. Such heterobimetallic species have been recognized to be generated as a dormant state in the olefin polymerization process [48,49]. In order to suppress formation of such dormant species, Busico and coworkers treated MAO with 2,6-di-tert-butylphenol with the aim to trap free Me 3 Al contained in MAO, and Soares and coworkers used highly bulky trioctylaluminum as a scavenger [50][51][52]. In this study, we explored the use of tetrabutylaluminoxane ([iBu 2 Al] 2 O) instead of iBu 3 Al as a scavenger. [iBu 2 Al] 2 O is a commercially available chemical that is easily synthesized from iBu 3 Al by treatment with water [53,54]. By employing [iBu 2 Al] 2 O as a scavenger instead of iBu 3 Al, the activity was improved in all cases (entries 1-7 versus 8-16). The improvement was most significant in the case of 14, where the activity improved 2.7-fold to reach 4.7 × 10 6 g/mol-Ti, which is comparable to that of CGC/[Ph 3 C][B(C 6 F 5 ) 4 ]/iBu 3 Al (5.0 × 10 6 g/mol-Ti). In the other cases, the activity improved~1.5-fold. Both 14 and CGC were deactivated during the course of high temperature polymerization; the yields were marginally increased from 2.36 g (entry 9) to 2.56 g and from 3.58 g (entry 15) to 3.76 g for 14 and CGC, respectively, by lengthening the reaction time from 3 min to 6 min.
The M n values declined somewhat with the use of [iBu 2 Al] 2 O instead of iBu 3 Al, but the general trend of the molecular weight versus the catalyst structure was the same (Figure 2). In the cases of 14-18, the M n values decreased by a factor of~0.65 upon replacement of the iBu 3 Al with [iBu 2 Al] 2 O, while the decrease was marginal in the case of 18, which invariably produced high molecular weight polymers (M n ,~70 kDa). The most significant decline of the M n value was observed in the case of CGC, resulting in the production of a low-molecular-weight sticky polymer (M n , 9.3 kDa) with bimodal molecular weight distribution.
Contrary to the expectation, the 1-octene contents of the samples generated using 13-17 were lower (12 mol %-15 mol %) than that of CGC (21 mol %) (Entries 8-13 versus 15). In the case of 18, the 1-octene content was as high as that of CGC (19 mol % and 21 mol %, respectively). Strangely, whereas the melting signals observed in the differential scanning calorimeter (DSC) thermogram of the samples generated with 13-18 were extremely weak, the melting signal was observed around 100 • C for the sample generated with CGC, although the 1-octene content was even higher (Figure 3). This observation and the bimodal molecular weight distribution together indicate that the comonomer composition may not be uniform for the sample generated with the prototype CGC under the present polymerization conditions. When the polymerization temperature was lowered (80-120 • C), a unimodal distribution was observed in the GCP curve and the 1-octene content was significantly higher (36 mol %, entry 16). The melting signal was neither observed at DSC thermogram. These observations indicate that the prototype CGC generates another type of active species at high temperature, which produce high-molecular-weight polymer chains with less 1-octene content. may not be uniform for the sample generated with the prototype CGC under the present polymerization conditions. When the polymerization temperature was lowered (80-120 °C), a unimodal distribution was observed in the GCP curve and the 1-octene content was significantly higher (36 mol %, entry 16). The melting signal was neither observed at DSC thermogram. These observations indicate that the prototype CGC generates another type of active species at high temperature, which produce highmolecular-weight polymer chains with less 1-octene content.

General Remarks
All manipulations were performed under an inert atmosphere using standard glovebox and Schlenk techniques. Toluene, hexane, THF and diethyl ether were distilled from benzophenone ketyl. Methylcyclohexane was purified using a Na-K Alloy. NMR spectra were recorded on a Varian Mercury plus 400 instrument (Varian, Inc., Palo Alto, CA, USA). Mass spectra were obtained with a Micromass VG Autospec apparatus. Gel permeation chromatograms (GPC) were obtained at 140 °C in trichlorobenzene using a Waters Model 150-C+ GPC instrument (Waters, Milford, MA, USA) and the data were analyzed using a polystyrene analysis curve. 1-Methylallylamine and 1,1-dimethylallylamine were prepared by the procedure reported in the literature [55,56].

General Remarks
All manipulations were performed under an inert atmosphere using standard glovebox and Schlenk techniques. Toluene, hexane, THF and diethyl ether were distilled from benzophenone ketyl. Methylcyclohexane was purified using a Na-K Alloy. NMR spectra were recorded on a Varian Mercury plus 400 instrument (Varian, Inc., Palo Alto, CA, USA). Mass spectra were obtained with a Micromass VG Autospec apparatus. Gel permeation chromatograms (GPC) were obtained at 140 • C in trichlorobenzene using a Waters Model 150-C+ GPC instrument (Waters, Milford, MA, USA) and the data were analyzed using a polystyrene analysis curve. 1-Methylallylamine and 1,1-dimethylallylamine were prepared by the procedure reported in the literature [55,56].

Compound 3
Allyl(dichloro)methylsilane ((CH 2 =CHCH 2 )(Me)SiCl 2 ) (1.63 g, 10.5 mmol) was added dropwise to a stirred solution of tetramethylcyclopentadienyllithium (0.897 g, 7.0 mmol) in THF (40 mL) at −78 • C. The solution was stirred overnight at room temperature. All the volatiles were removed under vacuum, and hexane (30 mL) was then added in a glove box. The insoluble fraction was removed by filtration with the aid of Celite. Solvent in the filtrate was removed under vacuum and the resulting crude oil was distilled under full vacuum at 60 • C to obtain (CH 2 =CHCH 2 )(Me)Si(C 5 Me 4 )Cl (1.49 g, 88%). 1

Compound 6
Compound 3 (0.600 g, 2.29 mmol) in toluene (10 mL) was cooled to −30 • C and the Schrock catalyst (35 mg, 0.046 mmol) dissolved in toluene (1.0 mL) was then added. During the course of the reaction, ethylene gas was generated and was removed through a bubbler. After stirring for 3 h at room temperature, all volatiles were removed under vacuum. The resulting crude oil was distilled under full vacuum at 60 • C (0.507 g, 95%). 1

Compound 7
It was prepared by using the same procedure applied to preparation of 6 by using 4. The product was purified by distillation under full vacuum at 70 • C (0.900 g, 79%). The 1 H-NMR spectrum indicated a mixture of two diastereomers in ca. 1:1 ratio. 1 13

Compound 8
It was prepared by the same procedure as that used for 6 by using 5. The product was purified by distillation under full vacuum at 80 • C (0.903 g, 70%). 1

Compound 9
A bomb reactor was charged with 6 (0.500 g, 2.14 mmol) dissolved in THF (15 mL) and Pd/C (10 wt %, 0.20 g); H 2 gas was introduced to a pressure of 10 bar. After stirring for 5 h, the remaining H 2 gas was vented off. Pd/C was removed by filtration over Celite. Solvent in the filtrate was removed under vacuum and the resulting crude oil was pure based on 1 H-and 13 C-NMR analyses (0.502 g, 99%).

Compound 12
It was prepared by employing the same procedure as that used for 9 by using 11 (1.53 g, 97%). 1

Compound 13
Compound 6 (0.500 g, 2.14 mmol) dissolved in THF (40 mL) was added to a solution of MeMgCl in THF (3.0 M, 2.88 mL, 8.65 mmol). The resulting solution was stirred for three days. During the course of the reaction, methane gas evolved and was vented through a bubbler. TiCl 4 (DME) (0.600 g, 2.14 mmol) was added at -30 • C and the resulting solution was stirred at room temperature for 3 h. All volatiles were removed under vacuum, and hexane (10 mL) was added. The insoluble fraction was removed by filtration with the aid of Celite. Solvent in the filtrate was removed under vacuum. The resulting dark brown oil was redissolved in hexane (5 mL) and filtration was performed another time. A dark brown oil was obtained by removal of the solvent from the filtrate, which was pure based on 1 H-and 13 C-NMR analyses (0.437 g, 63%). 1

Compound 14
It was prepared by employing the same procedure used for 13 by using 9. A dark brown oil was obtained by removal of the solvent from the filtrate, which was pure based on 1 H-and 13 C-NMR analyses (0.403 g, 58%). 1

Compound 16
It was prepared by using the same procedure as that employed for 15 by using 3. A dark brown oil was obtained by removal of the solvent from the filtrate, which was pure based on 1 H-and 13 C-NMR analyses (1.27 g, 84%). 1

Compound 17
It was prepared by using the same procedure as employed for 15 by using 10. A dark brown oil was obtained by removal of the solvent from the filtrate, which was pure based on 1 H-and 13 C-NMR analyses (1.39 g, 88%). 1

Compound 18
It was prepared by using the same procedure as employed for 9 by using 17. A dark brown oil was obtained by removal of the solvent from the filtrate, which was pure based on 1 H-and 13 C-NMR analyses (1.26 g, 98%). 1

Ethylene/1-Octene Copolymerization
Methylcyclohexane (25 mL) containing 1-octene (3.0 g) was introduced into a bomb reactor in a glove box. The reactor was assembled, removed out from the glove box, and heated to 130 • C using an oil bath. An activated catalyst was prepared by mixing the complex (0.5 µmol), iBu 3 Al (or [iBu 2 Al] 2 O) (0.20 mmol-Al), and [Ph 3 C][B(C 6 F 5 ) 4 ] (2.0 µmol) for 2 min. After adding the activated catalyst using a syringe, the system was immediately charged with ethylene gas (30 bar). The temperature increased rapidly due to the exothermic reaction, and the temperature profile is summarized in Table 1. The polymerization was conducted for 3 min using a continuous feed of ethylene at 30 bar and the reaction was quenched by venting the ethylene gas and introducing methanol through a syringe. After adding methanol (10 mL), the sticky polymer lump that was deposited was washed with methanol (10 mL) three times. The polymer was dried in a vacuum oven at 80 • C. The 1 H-NMR spectra showed the methyl (CH 3 ) signal (0.93-1.02 ppm) that was well isolated from the methine (CH) and methylene (CH 2 ) signals (1.30-1.40 ppm), and the 1-octene content could be calculated by the signal intensities of the two regions. 1 H-NMR spectra of the polymers (5 mg) were obtained at 70 • C after dissolution in deuterated toluene. It was proved that the 1-octene contents calculated from the 1 H-NMR and