Preparation of Extremely Active Ethylene Tetramerization Catalyst [iPrN(PAr 2 ) 2 − CrCl 2 ] + [B(C 6 F 5 ) 4 ] – (Ar = − C 6 H 4 - p -SiR 3 )

: Numerous efforts have been made to achieve “on-purpose” 1-octene production since Sasol discovered a Cr-based selective ethylene tetramerization catalyst in the early 2000s. By preparing a series of bis(phosphine) ligands iPrN(PAr 2 ) 2 where the Ar contains a bulky –SiR 3 substituent (Ar = − C 6 H 4 - p -Si(nBu) 3 ( 1 ), − C 6 H 4 - p -Si(1-hexyl) 3 ( 2 ), − C 6 H 4 - p -Si(1-octyl) 3 ( 3 ), − C 6 H 4 - p -Si(2-ethylhexyl) 3 ( 4 ), − C 6 H 4 - p -Si(3,7-dimethyloctyl) 3 ( 5 )), we obtained an extremely active catalyst that meets the criteria for commercial utilization. The Cr complexes [iPrN(PAr 2 ) 2 − CrCl 2 ] + [B(C 6 F 5 ) 4 ] – , obtained by reacting ligands 1 – 5 with [(CH 3 CN) 4 CrCl 2 ] + [B(C 6 F 5 ) 4 ] – , showed high activity exceeding 6000 kg/g-Cr/h, when combined with the inexpensive iBu 3 Al, thus avoiding the use of expensive modiﬁed methylaluminoxane (MMAO). The bulky –SiR 3 substituents played a key role in the suc-cess of catalysis by blocking the formation of inactive species (Cr complexes coordinated by two iPrN(PAr 2 ) 2 ligands, that is, [(iPrN(PAr 2 ) 2 ) 2 − CrCl 2 ] + [B(C 6 F 5 ) 4 ] – ). Among the complexes prepared, [ 3 − CrCl 2 ] + [B(C 6 F 5 ) 4 ] – exhibited the highest activity (11,100 kg/g-Cr/h, 100 kg/g-catalyst) with high 1-octene selectivity (75 wt%) and, moreover, mitigated the generation of undesired > C10 fractions (10.7 wt%). A 10-g-scale synthesis of 3 was developed, as well as a facile and low-cost synthetic method for [(CH 3 CN) 4 CrCl 2 ] + [B(C 6 F 5 ) 4 ] – .


Introduction
Currently, 1-Octene is produced on a large scale (~1 million ton/y), consumed mainly as a comonomer in the ethylene polymerization processes, and its demand is increasing with the increase in the production of polyolefins with high 1-octene content, e.g., polyolefin elastomer (POE) in which the 1-octene content is substantial (~40 wt%). There are several manufacturing routes for 1-octene. It can be obtained in the Fischer-Tropsch process through either direct fractionation or chemical conversion of the heptene fraction [1]. In 2007, the Dow Chemical Company commercialized a route from butadiene, consisting of Pd-catalyzed telomerization of butadiene with methanol to yield 1-methoxy-2,7-octadiene, which is subsequently transformed to 1-octene [2,3]. The main route to 1-octene is the oligomerization of ethylene, using a nickel-or zirconium-based catalyst, in which a wide range of linear α-olefins are generated, with the 1-octene fraction being less than 10% [4][5][6].
Catalysts that can selectively convert ethylene to 1-octene were discovered by Sasol in the early 2000s [7][8][9], while catalyst system that can selectively generate 1-hexene had been discovered~10 years earlier, and were commercialized in the early 2000s [10][11][12]. made to develop so-called "on-purpose 1-octene production technology" companies that use 1-octene in large quantity (e.g., POE producers) [13-1 Sasol catalyst system is composed of Cr(acac)3, a bis(phosphine) PNP-ty iPrN(PPh2)2) and the substance known as modified methylaluminoxane there have been several issues raised in the commercialization process. A the use of the expensive MMAO in large excess (Al/Cr, ~500), which is a b of catalyst cost. Many attempts have been made to replace MMAO with amounts of [Ph3C] + [B(C6F5)4] -or [PhN(H)Me2] + [B(C6F5)4] -, but most of these unsuccessful [16][17][18][19][20][21][22][23], although the replacement of MMAO with non-coor based salts has been successful in olefin polymerization catalysis [24]. closed, in a patent, a catalyst system that worked fairly efficiently with MMAO [25]. Another critical issue is the generation of insoluble polyeth side product. Even though the amount is typically small (<1%), the generat the operation of a large scale continuous commercial process, causing a so problem'. There is also a selectivity issue. Some 1-hexene is concomita which is, however, a useful coproduct. Some useless compounds, such methylene-cyclopentane (cy-C6), and cotrimers containing two molecules one of 1-hexene or 1-octene (>C10), are also generated. The amount of the is small (2-3%), but the amount of the cotrimers is substantial (10-15%).
The second step, the conversion of 1-bromo-4-(R3Si)-benzene to Et2N-PAr2 (Ar = −C6H4-p-SiR3), was rather tricky. In the conversion of p-(R3Si)-C6H4-Br to p-(R3Si)-C6H4-Li with nBuLi, nBu-Br is inevitably formed as a byproduct, and this could react with the product Et2N-PAr2, since nBu-Br is an electrophile and Et2N-PAr2 is a good nucleophile. In order to prevent such side reactions, the reaction temperature should be maintained at −78 °C , and the work-up procedure should be carefully controlled. However, it was found possible to avoid this awkward situation by performing the reaction of p-(R3Si)-C6H4-Li with Et2NPCl2 after removal of byproduct nBu-Br. The boiling point of nBu-Br is 102 °C , and it could be removed by evacuation after the generation of p-(R3Si)-C6H4-Li in toluene/ether (w/w 3:1) at −30 °C . Some excess nBuLi (1.3 eq per p-(R3Si)-C6H4-Br) was added  The second step, the conversion of 1-bromo-4-(R 3 Si)-benzene to Et 2 N-PAr 2 (Ar = −C 6 H 4 -p-SiR 3 ), was rather tricky. In the conversion of p-(R 3 Si)-C 6 H 4 -Br to p-(R 3 Si)-C 6 H 4 -Li with nBuLi, nBu-Br is inevitably formed as a byproduct, and this could react with the product Et 2 N-PAr 2 , since nBu-Br is an electrophile and Et 2 N-PAr 2 is a good nucleophile. In order to prevent such side reactions, the reaction temperature should be maintained at −78 • C, and the work-up procedure should be carefully controlled. However, it was found possible to avoid this awkward situation by performing the reaction of p-(R 3 Si)-C 6 H 4 -Li with Et 2 NPCl 2 after removal of byproduct nBu-Br. The boiling point of nBu-Br is 102 • C, and it could be removed by evacuation after the generation of p-(R 3 Si)-C 6 H 4 -Li in toluene/ether (w/w 3:1) at −30 • C. Some excess nBuLi (1.3 eq per p-(R 3 Si)-C 6 H 4 -Br) was added for complete conversion of p-(R 3 Si)-C 6 H 4 -Br to p-(R 3 Si)-C 6 H 4 -Li because when the stoichiometric amount of nBuLi was used, some portion of reactant p-(R 3 Si)-C 6 H 4 -Br (~5%) remained unconverted. The nBuLi that remained unreacted owing the excess addition was destroyed during the course of the evacuation process; it reacted with nBu-Br at room temperature and was converted to inert nBu-H and LiBr. After dissolving the p-(R 3 Si)-C 6 H 4 -Li in cold THF (−30 • C), Et 2 NPCl 2 was added to obtain the desired Et 2 N-PAr 2 (Ar = −C 6 H 4 -p-SiR 3 ). A slightly excess of Et 2 NPCl 2 (0.52 eq per p-(R 3 Si)-C 6 H 4 -Br) was added for complete consumption of p-(R 3 Si)-C 6 H 4 -Li, and the side product Et 2 N-P(Cl)Ar (Ar = −C 6 H 4 -p-SiR 3 ) formed in the presence of the slight excess of Et 2 NPCl 2 was removed using a small amount of silica gel. While the product Et 2 N-PAr 2 (Ar = −C 6 H 4 -p-SiR 3 ) was intact on the silica surface, Et 2 N-P(Cl)Ar was chemically attached to the silica surface by forming a ≡SiO-P(Ar)(NEt 2 ) bond. Et 2 N-PAr 2 (Ar = −C 6 H 4 -p-SiR 3 ) was cleanly converted to Cl-PAr 2 when it was dissolved in neat PCl 3 (5.5 eq) and then heated at 70 • C for 2 h [44]. Unreacted PCl 3 and byproduct Et 2 N-PCl 2 were removed by distillation under reduced pressure; the boiling points of PCl 3 and Et 2 N-PCl 2 are 76 • C and 179 • C, respectively, and they were easily separated for reuse. The target ligand iPrN(PAr 2 ) 2 (Ar = −C 6 H 4 -p-SiR 3 ) was obtained by the routine method of reacting Cl-PAr 2 with iPrNH 2 in CH 2 Cl 2 in the presence of Et 3 N (10 eq per iPrNH 2 ). A slight excess of Cl-PAr 2 (2.2 eq per iPrNH 2 ) was used to ensure complete conversion of iPrNH 2 to iPrN(PAr 2 ) 2 , without leaving intermediate iPrN(H)(PAr 2 ). The unreacted Cl-PAr 2 remaining due to the excess addition could be removed by addition of a small amount of silica gel, to which Cl-PAr 2 was attached by forming a ≡ SiO−PAr 2 bond.
In the 1 H NMR spectrum of iPrN(PAr 2 ) 2 (Ar = −C 6 H 4 -p-Si(1-octyl) 3 ) (Figure 1a), a signal corresponded to N-CHMe 2 was distinctly observed at 3.90 ppm as a multiplet coupled with methyl protons and the two phosphorous atoms. Impurity signals were observed at 2.51 ppm as a triplet (J = 7.8 Hz) and at 7.23 and 7.41 ppm as doublets overlapped with the product signals, which were assigned to CH 3 CH 2 CH 2 CH 2 -C 6 H 4 -p-SiR 3 generated at the stage of conversion of Br-C 6 H 4 -p-SiR 3 to Li-C 6 H 4 -p-SiR 3 . There was no way to remove the impurities; the product was highly soluble in most organic solvents, thus not permitting its crystallization, and its molecular weight was too high to perform vacuum distillation (1894 Da). Even though the signal intensity at 2.51 ppm assigned to the impurity was substantial (8.8%) relative to the product signal at 3.90 ppm, the amount of impurity as a weight percentage was negligible (1.2 wt%). The signal corresponding to the ortho-protons on −C 6 H 4 -p-SiR 3 moieties was typically very broad at 7.0-8.0 ppm while that corresponding to the meta-protons was sharp at 7.56 ppm as a doublet coupled with the ortho-protons (J = 6.6 Hz). Two signals were distinctly observed at 55.5 ppm and 42.1 ppm in the 31 P NMR spectrum (Figure 1c), which we attribute to restricted rotation around the P-N bonds; the two phosphorus atoms are inequivalent in the most stable resting state [27].

General Remarks
All manipulations were performed in an inert atmosphere using a standard glove box and Schlenk techniques. CH 2 Cl 2 and acetonitrile were stirred over CaH 2 and transferred to the reservoir under vacuum. Toluene, hexane, diethyl ether and THF were distilled from benzophenone ketyl. Methylcyclohexane (anhydrous grade) used for the oligomerization reactions was purchased from Aldrich and purified over a Na/K alloy. Ethylene was purified by contact with molecular sieves and copper for more than 12 h under a pressure of 48 bar. The 1 H NMR (600 MHz), 13 C NMR (150 MHz) and 31 P NMR (243 MHz) spectra were recorded using a JEOL ECZ 600 spectrometer. Gas-chromatography with flame-ionization detection (GC-FID) analysis was performed using a YL6500 GC system equipped with an HP-PONA (50 m × 0.200 mm × 0.50 µm) column. [CrCl 2 (NCCH 3 ) 4 ] + [B(C 6 F 5 ) 4 ]and [CrCl 2 (µ-Cl)(thf) 2 ] 2 were prepared according to previously reported methods [48].

(1-Octyl) 3 SiCl
(1-Octyl) 3 SiH (18.3 g, 49.6 mmol) and FeCl 3 (0.0805 g, 0.496 mmol, 1mol%) were dissolved CH 2 Cl 2 (70 mL) and acetyl chloride (4.28 g, 54.6 mmol) dissolved in CH 2 Cl 2 (9 mL) was added dropwise. After stirring for 1 d, all volatiles were removed using a vacuum line. The residue was dissolved in hexane (90 mL), and the insoluble fractions were removed by Celite-aided filtration. Removal of the solvent afforded a yellow oil (19.9 g, 99%), which was used for the next step without further purification. 1 3 1,4-Dibromobenzene (6.63 g, 28.1 mmol) was dissolved in THF (70 mL). After cooling to −78 • C, nBuLi (10.2 mL, 2.50 M in hexane, 25.5 mmol) was added dropwise. The resulting solution was stirred at −78 • C for 2 h, and then (1-octyl) 3 SiCl (10.0 g, 24.8 mmol) dissolved in THF (14 mL) was added dropwise. The solution was allowed to warm to room temperature and then stirred at room temperature for 2 h. After all volatiles were removed using a vacuum line, the residue was treated with hexane (30 mL). The insoluble fractions were removed by Celite-aided filtration, and the solvent was removed using a rotary evaporator. The residue was dissolved in hexane (60 mL), and the resulting solution was passed through a short pad of silica gel (13 g). Removal of the solvent afforded a colorless oil, which was evacuated at 70 • C to remove unreacted 1,4-dibromobenzene (yield 11.7 g, 90%). 1

Et 2 NP[C 6 H 4 -p-Si(1-octyl) 3 ] 2
BrC 6 H 4 -p-Si(1-octyl) 3 (13.1 g, 25.1 mmol) was dissolved in a mixed solvent of diethyl ether (41 mL) and toluene (98 mL). After cooling to −30 • C, nBuLi (13.0 mL, 2.50 M in hexane, 32.6 mmol) was added dropwise. The solution was allowed to warm to room temperature and then stirred for 2 h. All volatiles were completely removed using a vacuum line. THF (135 mL) was added and the resulting solution was cooled to −30 • C. Et 2 NPCl 2 (2.29 g, 13.2 mmol) dissolved in THF (13 mL) was added dropwise. The solution was allowed to warm to room temperature and stirred for 2 h. After the solvent was removed using a vacuum line, the residue was treated with hexane (190 mL). The insoluble fractions were removed by Celite-aided filtration, and the resulting solution was passed through a short pad of silica gel (12.4 g). Removal of the solvent afforded a light yellow oil (9.14 g, 74%), which was used for the next step without further purification. 1
3.9. [CrCl 2 (NCCH 3 ) 4 ] + [B(C 6 F 5 ) 4 ] -2,2-Dimethoxypropane (45.7 g, 452 mmol) was added to a flask containing CrCl 3 ·6H 2 O (9.30 g, 34.9 mmol). Initially, CrCl 3 ·6H 2 O was insoluble in 2,2-dimethoxypropane, but a purple solution was obtained after stirring for 4 h. The volatiles were removed using a vacuum line to obtain a brown, gummy residue. 2,2-Dimethoxypropane (45.7 g, 452 mmol) was added again and the mixture stirred at 70 • C for 1.5 h to precipitate green solids, which were collected by filtration and washed with 2,2-dimethoxypropane (17 mL). The isolated solids were dried under vacuum to obtain [(MeOH) 4 CrCl 2 ] + Cl -(9.4 g, 94%) [49]. The prepared [(MeOH) 4 CrCl 2 ] + Cl -(2.39 g, 8.35 mmol) was dissolved in EtOH (38 mL), and K + [B(C 6 F 5 ) 4 ] -(6.00 g, 8.35 mmol) dissolved in EtOH (38 mL) was added to the resulting solution. A white solid (KCl) precipitated immediately and was removed by Celite-aided filtration. Volatiles in the filtrate were removed using a vacuum line. The residue was dissolved in acetonitrile (12 mL), and the resulting solution was heated to 60 • C with heat gun. The volatiles were removed using a vacuum line while heated with heat gun. This procedure of dissolution in acetonitrile and solvent removal was repeated five times more time. Finally, the residue was dissolved in acetonitrile (12 mL) and the resulting solution was stored at −30 • C, resulting in the precipitation of a green solid (6.46 g). Solvent was removed from the mother liquor, and the residue was dissolved in acetonitrile (2.2 mL). Storing the solution at −30 • C precipitated the second crops (0.670 g, total yield 88%). The isolated solid (14.0 mg) and 9-methylanthracene (10.0 mg) as an external standard were dissolved in THF-d 8 , and the 1 H NMR spectrum was recorded. Integration values at 1.94 ppm (CH 3 CN signal) and 3.08 ppm (methyl signal in the external standard) agreed with the formula [CrCl 2 (NCCH 3 ) 4 . 02 ] + [B(C 6 F 5 ) 4 ] -.

Ethylene Tetramerization
Methylcyclohexane (200 mL) was placed in a bomb reactor, and after the temperature was raised to 40 • C, iBu 3 Al (0.149 g, 750 µmol) and [3−CrCl 2 ] + [B(C 6 F 5 ) 4 ] -(7.55 mg, 1.0 wt% solution in methylcyclohexane, 2.8 µmol) were successively added. The reactor was immediately charged with ethylene gas to a pressure of 35 bar. The temperature was controlled at 40 • C, and ethylene gas was continuously fed to maintain the pressure at 35 bar. The oligomerization was allowed to proceed for 30 min, and the ethylene gas was then vented off. The weight percentages of the oligomers generated (1-octene (1-C8), 1hexene (1-C6), methylcyclopentane + methylenecyclopentane (cy-C6) and higher oligomers above C10 (>C10)) were calculated from the analysis of GC data, using nonane as an external standard. The solid PEs generated were isolated by filtration at room temperature.