Polyhedral oligomeric silsesquioxane-conjugated bis(diphenylphosphino)amine ligand for chromium(III) catalyzed ethylene trimerization and tetramerization

Applied Catalysis A, General 560 (2018) 21–27

Contents lists available at ScienceDirect

Applied Catalysis A, General journal homepage: www.elsevier.com/locate/apcata

Polyhedral oligomeric silsesquioxane-conjugated bis(diphenylphosphino) amine ligand for chromium(III) catalyzed ethylene trimerization and tetramerization Hoseong Leea,b, Soon Hyeok Honga, a b

T

⁎

Department of Chemistry, College of Natural Sciences, Seoul National University, Gwanak-gu, Seoul, 08826, Republic of Korea Department of Chemical R&D Center, SK Innovation, Yuseong-gu, Daejeon, 34124, Republic of Korea

A R T I C LE I N FO

A B S T R A C T

Keywords: Ethylene trimerization Ethylene tetramerization Chromium Ethylene oligomerization Polyhedral oligomeric silsesquioxane Homogeneous catalysis

Polyhedral oligomeric silsesquioxanes (POSSs) were attached to conventional bis(diphenylphosphino)amine (PNP) ligand as solubility-enhancing materials for catalytic ethylene trimerization and tetramerization. Differently functionalized arylphosphine ligands of the type (Ph)2PN(POSS)P(Ph)(ArR) (R = functional groups) were systematically developed, and their corresponding chromium(III) complexes were formed. The developed precatalysts exhibited excellent tolerance in solvents, including even low-carbon-number hydrocarbons such as n-pentane, n-hexane, or cyclohexane. In particular, the ortho-fluorophenyl-substituted complex showed higher stability even at higher temperatures above 120 °C. The ortho-OCF3–phenyl-substituted complex showed outstanding catalytic activity, which reached 2287 kg/g Cr/h at 30 bar.

1. Introduction Selective oligomerization of ethylene for linear α-olefins (LAOs), such as 1-hexene and 1-octene, is a highly challenging subject in catalysis and has been researched extensively [1]. LAOs, which are valuable co-monomers in the production of linear low-density polyethylene (LLDPE), are produced industrially via a generally less-selective oligomerization of ethylene. The conventional oligomerization process produces not only 1-hexene and 1-octene, but also higher-molecularweight oligomers according to the Schulz-Flory or Poisson distribution [2–4]. To address the selectivity issue, various catalytic systems that utilize chromium complexes have been developed [5–12]. Generally, an active chromium catalyst can be generated in situ using a ligand and Cr (acetylacetonate)3 or Cr(ethylhexanoate)3 or CrCl3(THF)3 [13–18]. The generated precatalyst, however, has poor solubility in many non-polar solvents, which often causes lower catalytic activities, so that several attempts have been reported to overcome the low solubility. For instance, K. Blann et al. demonstrated significant improvement on catalytic activity by increasing carbon number and solubility with results of methyl-(26 kg/g Cr/h), n-pentyl-(43 kg/g Cr/h), and n-decyl-(50 kg/g Cr/h)PNP precatalysts [19]. Meanwhile, hydrocarbon solvents with fewer carbon atoms have not been evaluated adequately, and most selective oligomerization reactions have conventionally been carried out in toluene or methylcyclohexane (MCH). The use of light

⁎

Corresponding author. E-mail address: [email protected] (S.H. Hong).

https://doi.org/10.1016/j.apcata.2018.04.030 Received 9 December 2017; Received in revised form 22 April 2018; Accepted 23 April 2018 Available online 24 April 2018 0926-860X/ © 2018 Elsevier B.V. All rights reserved.

hydrocarbon solvents, however, has a potential advantage since they dissolve ethylene well [20–22]. The solubility of ethylene in some hydrocarbon solvents has been calculated by Aspen Plus (Table S1). It is higher in light hydrocarbon solvents at lower temperatures, and also in a linear aliphatic solvent than in a cyclic one with the same number of carbon atoms. Furthermore, the processing temperature for the oligomerization is still restrictive, although numerous modifications and developments have been made to improve the catalytic systems. Because of the intrinsic activity of conventional catalytic systems, operation at 45–60 °C is encouraged; otherwise, at elevated temperatures, the performance drops significantly, resulting in a drastic decrease in the productivity [23]. The advantage of operating at higher temperatures is that viscous byproduct polymers can be dissolved, preventing post-process plugging. In order to enable operation at high temperatures, more innovative catalytic systems are required. Polyhedral oligomeric silsesquioxanes (POSSs) have been highlighted as advanced materials with excellent solubility and thermal stability [24–29]. Guenthner et al. showed that POSSs have excellent solubility in non-polar solvents such as n-hexane by measuring the Hansen solubility parameters [30]. Fina et al. reported that POSSs have high thermal stability; for instance, isobutyl-substituted POSS has a pyrolysis temperature above 260 °C [31]. Owing to such advantageous properties, POSSs have been widely used as scaffolds for organic/inorganic hybrid catalysts [32], microfluidic formats [33], monolithic

Applied Catalysis A, General 560 (2018) 21–27

H. Lee, S.H. Hong

to yield 4.32 g (70.2%) of the colorless solid.; m/z (APPI) [M + H]+ calcd for C31H72NO12Si8+: 874.3203; found: 874.3178. v (CHCl3)/ cm−1: 2953s, 2926 w, 2902 w, 2868 w, 1463s, 1398 w, 1388 w, 1366 w, 1332 w, 1229s, 1096s, 907 s, 852 s, 732 s. δH(CDCl3): 2.65 (t, 2 H, -CH2N), 1.84 (m, 7 H, CH), 1.51 (m, 2 H, -CH2-), 1.14 (b, 2 H, NH2), 0.94 (m, 42 H, CH3), 0.58 (m, 16 H, Si-CH2). δC(CDCl3): 44.7, 27.1, 25.7, 23.9, 22.5, 9.2.

structures [34–36], tailored internal chemistries of porous materials [37], conjugates to nanoparticles [38–40] as well as stiffeners for transparent electronic devices with inherent thermal stability [41]. Duchateau et al. reported the use of titanium- and zirconium-tethered POSSs as catalysts for ethylene polymerization [42]. However, the use of chromium-tethered POSSs for ethylene oligomerization has not been investigated thus far. In this study, versatile chromium complexes based on POSS-conjugated ligands were developed as catalysts with excellent solubility and thermal stability for selective ethylene oligomerization.

2.3.2. Preparation of (Ph)2PN(POSS)P(Ph)2 (L1) POSS-NH2 (0.6 g, 0.68 mmol) and triethylamine (1.0 mL, 7.20 mmol) were dissolved in dichloromethane (10 mL), and then chlorodiphenylphosphine (0.317 g, 1.44 mmol) was added. The solution was stirred at ambient temperature for 1 h. The volatile solvent was evaporated and the product was washed with methanol (2 × 3 mL). Further purification was performed by recrystallization in n-hexane/ methanol, recovered, and dried to yield 0.71 g (83%) of (Ph)2PN(POSS) P(Ph)2 as a white solid.; m/z (APPI) [M + H]+ calcd for C55H90NO12P2Si8+: 1242.4087; found: 1242.4033. v (CHCl3)/cm−1: 2952s, 2915 w, 2868 w, 2848 w, 1464s, 1434s, 1401 w, 1382 w, 1365 w, 1331 w, 1229s, 1096s, 866 s, 837 s, 741 s, 725 s. δH(CDCl3): 7.35 (b, 8 H, aromatics), 7.26 (b, 12 H, aromatics), 3.17 (m, 2 H, -CH2N), 1.80 (m, 7 H, -CH-), 1.24 (b, 2 H, -CH2-), 0.92 (m, 42 H, CH3), 0.56 (m, 14 H, Si-CH2), 0.15 (t, 6 H, Si-CH2). δC(CDCl3): 139.8, 132.7, 128.6, 127.9, 55.5, 45.7, 25.6, 23.8, 22.5, 9.2. δP(C6D6): 61.8. δSi(C6D6): -67.9.

2. Experimental 2.1. General conditions All reactions were performed under an inert atmosphere using standard Schlenk techniques. All solvents and gases were dried and degassed using standard procedures. Chemicals were purchased from Sigma Aldrich or Strem and used without further purification unless otherwise stated. mMAO-3A was obtained from Akzo Nobel Corporation as a 7% w/w solution in heptane. The IR spectra were recorded with a Nicolet 6700 FT-IR Spectrometer from Thermo Scientific. 1H, 19F, 29Si and 31P nuclear magnetic resonance (NMR) spectra were recorded on a Bruker AVANCE III HD 500 MHz spectrometer in CDCl3 or C6D6. Chemical shifts are reported in ppm with reference to tetramethylsilane. 13C NMR spectra were recorded on a Bruker AVANCE III HD 600 MHz spectrometer in CDCl3. Chemical shifts are reported in ppm with reference to internal chloroform. Bruker Daltonics (Billerica, MA, USA) APPI 7T FT-ICR MS was used for (+) mode atmospheric pressure photoionization analysis. Quantitative chromatographic analysis of the oligomerization products was performed using an Agilent 7890 A GC-FID with an HP-PONA column (50 m × 0.20 mm). The reaction solvent was used as an internal standard.

2.3.3. Preparation of (Ph)2PN(n-Bu)P(Ph)2 (L2) (Ph)2PN(n-Bu)P(Ph)2 was prepared according to a modified literature method [12]. N-Butylamine (0.1 g, 1.37 mmol) and triethylamine (1.72 mL, 8.4 mmol) were dissolved in dichloromethane (5 mL), and then chlorodiphenylphosphine (0.618 g, 2.8 mmol) was added. The solution was stirred at ambient temperature for 1 h. The volatile solvent was evaporated and the product was washed with methanol (2 × 3 mL), recovered, and dried to yield 0.5 g (83%) of (Ph)2PN(n-Bu) P(Ph)2 as a white solid.; m/z (APPI) [M + H]+ calcd for C28H30NP2+: 442.1848; found: 442.1843. δH(CDCl3): 7.39 (b, 8 H, aromatics), 7.29 (b, 12 H, aromatics), 3.23 (t, 2 H, -CH2N), 1.07 (b, 2 H, -CH2-), 0.92 (m, 2 H, -CH2-), 0.60 (t, 3 H, CH3). δC(CDCl3): 139.7, 132.7, 128.6, 128.0, 52.8, 33.4, 19.9, 13.6. δP(C6D6): 62.2.

2.2. Ethylene oligomerization All runs were carried out in a 50-mL stainless steel Parr autoclave with a magnetic stirrer. In a glovebox, a glass vial was charged with the ligand (0.5 μmol) and CrCl3(THF)3 (0.5 μmol) followed by 1 mL of dichloromethane, and then the solution was stirred for 10 min. After this, the solvent was removed under reduced pressure and the resultant solid was suspended in 20 mL of the reaction solvent. The solution was placed in the autoclave and mMAO (0.14 mL, 0.25 mmol, 500 equivalents) was added; then, the solution was pressurized with ethylene and stirred at 600 rpm. Ethylene was fed on demand to keep the reactor pressure constant, and the uptake was monitored using a flow meter. After 15 min, the autoclave was cooled to 0 °C and depressurized slowly to atmospheric pressure. The product was quenched by adding 2ethylhexanol (1 mL). The crude products were filtered and analyzed using GC-FID. The polymeric products were recovered by filtration and dried overnight in an oven at 100 °C.

2.3.4. Preparation of N,N-diethylaminochlorophenylphosphine N,N-Diethylaminochlorophenylphosphine was prepared according to a literature method [47]. Pyridine (13.05 g, 165 mmol) was added dropwise to a solution of dichlorophenylphosphine (14.77 g, 82.5 mmol) in n-hexane (80 mL) at −78 °C, and this was followed by the dropwise addition of diethylamine (12.07 g, 165 mmol). The reaction mixture was warmed to room temperature, stirred for 3 h, and then filtered to remove the precipitated diethylammonium chloride salt. Removal of the solvent under reduced pressure gave a pale yellow oil (17.75 g, 99.75%), which was distilled to yield 13.53 g (76.05%) of pure N,N-diethylaminochlorophenylphosphine as colorless oil. b.p. 110 °C/150 mbar.

2.3. Ligand preparation 2.3.5. Typical procedure for ClPPh(ArR) for L3–L8, L10, L11, L13, and L14 Magnesium turnings (30 mmol) were activated in anhydrous THF (40 mL), and then functionalized aryl bromide (20 mmol) was added dropwise. The reaction mixture was stirred overnight at 45 °C. After it had cooled to room temperature, the reaction mixture was separated from excess magnesium via decantation to obtain a Grignard reagent (0.5 M). A portion of the reagent (8 mL, 4 mmol) was added to a dilute solution of N,N-diethylaminochlorophenylphosphine (3.2 mmol), and the reaction mixture was refluxed for 3 h. After it had cooled to room temperature, the volatile solvent was removed in vacuo, and then the crude mixture was slurried in n-hexane (20 mL). The slurry was filtered using an activated alumina pad, and the filter cake was washed with n-

2.3.1. Preparation of POSS-NH2 3-Aminopropyl-substituted heptaisobutyl-POSS (POSS-NH2) was synthesized as described in the literature [43–46]. IsobutyltrisilanolPOSS (5.57 g, 7.0 mmol) was dissolved in THF (47 mL) and then (3aminopropyl)trimethoxysilane (1.64 g, 9.1 mmol) was added. The reaction mixture was vigorously stirred for 24 h at 25 °C and then the solvent was removed in vacuo. The crude product which was further dissolved with 30 mL of n-hexane was filtered out to remove some of insoluble residue. The obtained clear solution was added to the same amount of acetonitrile to precipitate the desired product, filtered, washed with acetonitrile (2 × 20 mL), and dried over in vacuo. The product was further purified by recrystallization in n-hexane/acetonitrile 22

Applied Catalysis A, General 560 (2018) 21–27

H. Lee, S.H. Hong

Scheme 1. Synthesis of a POSS-conjugated PNP ligand.

2.3.8. Typical procedure for (Ph)2PN(POSS)P(Ph)(ArR) for L3–L15 Prepared POSS-PNH (0.2 g, 0.19 mmol) and triethylamine (0.05 mL, 0.38 mmol) were dissolved in dichloromethane (2 mL), and then prepared ClPPh(ArR) (0.23 mmol) was added. The solution was stirred at ambient temperature for 1 h. The volatile solvent was evaporated and the product was washed with methanol (2 × 2 mL), recovered, and dried to afford (Ph)2PN(POSS)P(Ph)(ArR) in 72%–95% yield as a white solid. Yields and characterization data are summarized in Table S4.

hexane (2 × 4 mL). Afterwards, the volatile solvent was removed in vacuo to afford pure (Et)2NPPh(ArR) in 71.8%–99.3% yield as colorless oil or white solid. Yields and characterization data are summarized in Table S2. The obtained (Et)2NPPh(ArR) was dissolved in Et2O (10 mL), and then a 2 M HCl/Et2O solution (2.1 equiv.) was added; then, the mixture was stirred for 1 h at room temperature. White ammonium salt precipitated and was removed using an activated alumina pad to afford pure ClPPh(ArR) in 99.9% yield as colorless oil or a white solid. Yields and characterization data are summarized in Table S3.

3. Results and discussion R

2.3.6. Typical procedure for ClPPh(Ar ) for L9, L12, and L15 Functionalized aryl bromide (2 mmol) was dissolved in Et2O (8 mL), and then a 1.6 M n-BuLi solution (1.25 mL, 2 mmol) was added dropwise at −78 °C. After 1 h of stirring, N,N-diethylaminochlorophenylphosphine (1.8 mmol) in Et2O (2 mL) was added dropwise. The mixture was allowed to warm to ambient temperature and stirred for another 3 h. The volatile solvent was removed in vacuo and the crude mixture was slurried in n-hexane (10 mL). The slurry was filtered using an activated alumina pad, and the filter cake was washed with nhexane (2 × 2 mL). The volatile solvent was removed in vacuo to afford pure (Et)2NPPh(ArR) in 75.8%–85.3% yield as colorless oil. Yields and characterization data are summarized in Table S2. The ClPPh(ArR) products were obtained in the same manner described above in 99.9% yield. Yields and characterization data are summarized in Table S3.

3.1. Synthesis of POSS-conjugated catalysts and solvent screening An amino-POSS scaffold (POSS-NH2), which can be synthesized by tethering an aminosilane to an incompletely condensed POSS triol, was chosen for conjugation to bis(diphenylphosphino)amine (PNP)-ligandbased chromium complexes. A POSS-PNP ligand (L1) was obtained by adding chlorodiphenylphosphine to POSS-NH2 (Scheme 1). Absence of –OH moiety and formation of P-N-P bond were confirmed by FT-IR analysis (Figure S1). A conventional ligand system, Ph2PN(n-Bu)PPh2 (L2), was also synthesized in a similar way for control experiments. The corresponding Cr complexes were prepared via the reaction between the ligands and CrCl3(THF)3 in-situ. Prior to ligand screening, a pretest was carried out to investigate an appropriate ratio of ligand/Cr for L1. As expected, when the same equivalent of ligand/Cr, the best activity and α-olefins selectivity were observed. In case of a higher equivalence of chromium, binuclear chromium species with bridging diphosphine ligands would be formed which would lower the activity [48]. With a higher equivalence of ligand, on the other hand, a stable four ligand coordinated chromium complex could be formed, which would retard the activation of the catalyst. The yield of each product was determined using gas chromatography with a flame ionization detector (GC-FID) (Figure S2 for calibration curves). The results are summarized in Table 1 (Figure S3). Further screening was examined to check the activity and the solubility. L1 and L2 complexes were dissolved in various reaction solvents. The solubility of the L1 complex in various hydrocarbon solvents was excellent. Ligand L2, however, did not dissolve in most solvents. Subsequently, ethylene oligomerization was carried out and the results are summarized in Table 2 (Figure S4). L1 showed excellent activities in most solvents tested, and the best activity was observed in n-pentane. No clear difference was observed between C6 and C7 solvents, and cyclic hydrocarbon solvents such as cyclohexane and MCH led to slightly better activities compared to

2.3.7. Preparation of POSS-PNH Synthesized POSS-NH2 (1.98 g, 2.26 mmol) and 1 equiv. of chlorodiphenylphosphine (0.50 g, 2.26 mmol) were dissolved in dichloromethane (30 mL). The solution was cooled to 0 °C, and then triethylamine (1.43 mL, 10.3 mmol) was added dropwise. The mixture was warmed to room temperature and stirred for 1 h. The volatile solvent was evaporated and the solid product was washed with acetonitrile several times (5 × 3 mL). Further purification was performed by recrystallization in n-hexane/acetonitrile, recovered, and dried to yield 2.22 g (92%) of POSS-PNH as a white solid.; m/z (APPI) [M + H]+ calcd for C43H81NO12PSi8+: 1058.3645; found: 1058.3612. v (CHCl3)/ cm−1: 2952s, 2923 w, 2887 w, 2867 w, 1464s, 1439 w, 1401 w, 1382 w, 1366 w, 1332 w, 1229s, 1096s, 836 s, 739 s, 700 s. δH(C6D6): 7.55 (m, 4 H, aromatics), 7.23 (m, 4 H, aromatics), 7.18 (m, 2 H, aromatics), 2.98 (m, 2 H, -CH2N), 2.16 (m, 7 H, -CH-), 1.75 (b, 2 H, -CH2-), 1.16 (m, 42 H, CH3), 0.92 (t, 14 H, Si-CH2), 0.79 (t, 2 H, Si-CH2). δC(CDCl3): 141.8, 131.3, 128.9, 128.1, 25.7, 23.8, 22.5, 9.3. δP(C6D6): 41.8. δSi(C6D6): -67.3. Table 1 Ethylene oligomerization results according to ligand/Cr ratio.a Ligand/Cr molar ratio

Productivity (kg/g Cr/h)

C6 (%)

1-C6/C6 (%)

C8 (%)

1-C8/C8 (%)

C10-14 (%)

Polymer (%)

Selectivityb

0.5 1.0 1.5

434 1049 995

18.6 23.3 21.3

30.2 42.2 33.9

55.6 63.0 62.0

93.2 95.5 94.7

12.0 10.3 11.7

13.8 3.4 5.0

3.0 2.7 2.9

a b

Standard conditions: 50 mL autoclave, 0.5 μmol of CrCl3(THF)3, 500 equiv. of mMAO-3A, 15 min, 20 mL of MCH, 30 bar of ethylene, 45 °C. C8/C6 selectivity. 23

Applied Catalysis A, General 560 (2018) 21–27

H. Lee, S.H. Hong

Table 2 Ethylene oligomerization results of solvent screening.a Ligand

Solventb

Productivity (kg/g Cr/h)

C6 (%)

1-C6/C6 (%)

C8 (%)

1-C8/C8 (%)

C10-14 (%)

Polymer (%)

Selectivityc

L1

Pen nHx cHx Hep MCH Pen nHx cHx Hep MCH

1250 889 1021 869 1049 125 161 214 228 446

21.2 20.6 21.7 20.8 23.3 20.2 19.2 11.5 11.5 13.1

39.5 36.9 37.2 37.3 42.2 45.9 51.3 30.5 33.5 34.2

64.6 64.5 64.7 63.9 63.0 69.8 69.3 71.5 73.6 70.8

96.9 96.0 95.4 95.6 95.5 97.2 96.1 95.8 96.1 96.5

11.1 10.5 9.9 11.2 10.3 8.8 9.9 15.8 14.5 15.9

3.2 4.4 3.7 4.1 3.4 1.2 1.6 1.2 0.4 0.2

3.1 3.1 3.0 3.1 2.7 3.5 3.6 6.2 6.4 5.4

L2

a b c

Standard conditions: 50 mL autoclave, 0.5 μmol of ligand, 0.5 μmol of CrCl3(THF)3, 500 equiv. of mMAO-3A, 15 min, 20 mL of solvent, 30 bar of ethylene, 45 °C. Abbreviations for solvents: Pen; n-pentane, nHx; n-hexane, cHx; cyclohexane, Hep; n-heptane, MCH; methylcyclohexane. C8/C6 selectivity.

Unlike the experiments in Table 2, these reactions were performed at 60 °C to enable comparison with previously reported results for related PNP systems. The activity was slightly lower at 60 °C than at 45 °C for L1, which seems to be because of reduced ethylene solubility (1049 kg/g Cr/h at 45 °C, 997 kg/g Cr/h at 60 °C). The productivity of L2 was only approximately 60% of that of L1, despite an increase of temperature to 60 °C (Figure S5). Various substituents (Me, OMe, F, CF3, OCF3) were introduced at the ortho-, meta-, and para-positions. ortho-Substituted L3, L6, and L12 were noticeably superior to the others in terms of C6 fraction selectivity. This result is in agreement with previously reported experimental results and can be attributed to steric effects [52,53]. A sterically bulky substituent disturbs the formation of a metallacycle, and consequently, a metallacycloheptane intermediate is in a more advantageous geometry than a metallacyclononane intermediate. The inductive effect was observed as well for the methyl-substituted phosphoamino ligands. The use of L5, which has an electron donating methyl group at the paraposition, resulted in higher oligomers productivity and less polymer productivity compared to the use of L4. In the case of the methoxyl-substituted ligands (L6–L8), the L6based catalytic system remarkably produced C6 and C10–14 products with high activity and a high selectivity of 99.9% for 1-hexene. The formation of C10–14 products with a trace amount of 1-decene is attributed to co-trimerization of ethylene with 1-hexene. Therefore, most of the products are branched olefins of C10–14 compounds rather than αolefins [55]. L6 led to the production of only a trace amount of C8 products without any polymer generation. This can be explained by the pendant effect of the hemilabile ether moiety, which may coordinate to the metal center to form a typical trimerization catalyst (Fig. 1a) [54]. It is noteworthy that the use of the L6-based catalytic system resulted in higher productivity than that reported in the previous literature under the same pressure (1977 kg/g Cr/h at 30 bar) [53] with excellent selectivity for C6 products and with no polymerization. L8, which has a methoxyl substituent at the para-position, showed slightly higher activity compared to L7. Fluoro-substituted ligands L9–L11 exhibited unique characteristics. The activity with the ortho-F ligand was twice as high as those with the other fluoro-substituted ligands, and the C8/C6 selectivity was maintained at approximately 2.3. As previously reported, steric hindrance limits metallacycle formation, and thus C6 production dominates over C8 production when there is an ortho substituent; however, the selectivity with L9 did not follow this trend. This result can be explained as follows. First, the relatively low steric hindrance from a fluoro substituent may not interfere with metallacycle formation. In addition, a bulky POSS structure can limit the movement and rotation of the aryl group, and a repulsive force from the lipophobicity of the fluorine atom may affect the ligand orientation during the formation of the lipid metallacycle [56]. In conclusion, lowering the flexibility and crowding

linear hydrocarbon solvents such as n-hexane and n-heptane. In the case of L2, no increased activity was observed in low-carbon-number solvents, seemingly owing to its low solubility. After the reaction in npentane an insoluble greenish material which was likely derived from residual chromium(III) complex was observed inside the reactor. A high 1-C8 fraction of the total C8 products (> 95%) and a moderate 1-C6 fraction of the total C6 products (< 52%) were obtained with both precatalysts. It is known that alpha branching on the N atom is important in determining the proportion of the 1-C6 fraction of C6 products; with tert-, sec-, and prim-alkyl alpha branching, the proportions of 1-C6 in the C6 products were approximately 97%, 85%, and 45%, respectively [19]. Since a methylene group is at the alpha position to the N atom in the case of L1, it is consistent with the above tendency. In other words, improving the selectivity by adjusting the alpha branching group is possible. It is well reported that the major C6 byproducts are methylcyclopentane and methylene cyclopentane [49]. It is also known that the bulkiness of substituents on the N atom increases the C6 fraction and decreases the C8 fraction [19,50,51]. Similarly, the reason that the C8/C6 selectivity of L1 (2.7–3.1) is somewhat lower than that of L2 (3.5 to 6.4) may be the bulkiness of the POSS structure, even though it is far away from the PNP moiety. To summarize the first screening, POSS-containing ligand L1 showed consistent activities in whole solvents, whereas ordinary ligand L2 was behind. It is noteworthy that a fine powdery polymer was produced rather than a viscous polymer when a solvent with six or fewer carbons was used. The powdery polymer can be separated easily using a filter, thereby making it probably suitable for a continuous process. 3.2. Synthesis of functionalized POSS-conjugated catalysts and their catalytic activities Control of activity and selectivity by adjusting the functional groups on the aryl substituent in the PNP ligand system has already been reported by Sasol Technology [52,53] and British Petroleum [54]. However, many of the studies only focused on alkyl and methoxyl functional groups. It has also been reported that the catalytic activity and selectivity depend slightly on the number of substituents on the four aryl groups in a diphosphinoamine ligand. The C6 fraction, for instance, can be increased gradually by reducing the number of substituents. To examine electronic and steric effects simultaneously, we conducted further studies with POSS-conjugated ligands with various functional groups on the aryl substituents of the diphosphinoamine, as shown in Scheme 2. The POSS-PNH precursor was reacted with ClPPh (ArR) containing various functional groups. The ClPPh(ArR) compounds were synthesized using Grignard reagents or via lithiation of the relevant aryl compounds with N,N-diethylaminochlorophenylphosphine. Using the synthesized ligands, ethylene oligomerization was carried out (Table 3). 24

Applied Catalysis A, General 560 (2018) 21–27

H. Lee, S.H. Hong

Scheme 2. Synthesis of POSS-conjugated PNP ligands. [a] ClPPh2/triethylamine (TEA), dichloromethane (DCM), 0 °C → rt, 1 h. [b] ClPPh(ArR)/TEA, DCM, rt, 1 h. [c] Diethylamine/pyridine, n-hexane, −78 °C to rt, 3 h. [d] 1) Grignard reagent/tetrahydrofuran (THF), reflux, 3 h, or lithiation/Et2O, −78 °C, 3 h, 2) HCl/Et2O, 1 h.

produced mainly C6 and C10–14 products. The higher activity and selectivity with L15 can be attributed to the interference of the coordination of the pendant ether to the metal center by the strongly electron-withdrawing CF3 group (Fig. 1b). As a result, the co-trimerization to produce > C10–14 did not occur. In terms of energy efficiency, the methoxyl group in anisole tends to be coplanar with the phenyl group, whereas the OeCF3 bond in trifluoromethoxybenzene prefers to be perpendicular to the phenyl group [57]. This tendency can be explained through the stereoelectronics of the OCF3 group. First, there is a steric effect caused by the longer CeF bond length (1.41 Å) compared to that of a CeH bond (1.09 Å) [57]. In addition, the CeF bond is advantageous since its energy is minimized when oriented antiperiplanar

at the active catalytic site aids in the formation of an intact metallacycle. Notably, the 1-C6/C6 selectivity (70.4%) with L9 increased despite the lack of bulky N-alkyl groups, which was also probably because of the prevention of the obstruction of metallacycle formation. Neither the meta- (L10) nor the para-substituted ligand (L11) was beneficial to productivity. L12–L14, which have CF3 substituents, showed lower activities than the others. C6 production with ortho-CF3-substituted L12 was higher than C8 production. L15 showed much higher activity than the other ligands. It is also noteworthy that the properties of L15 were quite different from those of L6 despite structural similarities. For instance, L15 led to a common distribution of C6, C8, and C10–14, whereas L6 25

Applied Catalysis A, General 560 (2018) 21–27

H. Lee, S.H. Hong

Table 3 Ethylene oligomerization results with functionalized POSS-PNP.a Ligand

Productivity (kg/g Cr/h)

C6 (%)

1-C6/C6 (%)

C8 (%)

1-C8/C8 (%)

C10-14 (%)

Polymer (%)

Selectivityb

L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 L12 L13 L14 L15

997 575 1254 606 892 1977 579 717 1745 799 809 730 437 562 2275

24.8 23.0 41.6 23.9 29.6 60.0 29.1 32.2 27.7 25.7 25.6 54.4 27.2 27.8 37.2

47.4 51.7 65.5 46.9 59.3 99.9 57.2 65.1 70.4 44.9 44.6 65.9 51.8 54.2 67.8

64.4 68.6 50.5 65.7 60.1 1.5 60.4 57.4 63.6 63.3 63.9 38.8 60.7 62.5 56.8

95.3 97.0 97.9 96.3 96.5 93.2 96.3 96.6 98.1 95.6 95.8 97.1 95.7 96.3 96.7

8.8 7.6 7.6 9.1 9.7 38.5 9.9 9.7 8.3 10.2 9.7 6.5 11.0 8.8 5.6

2.0 0.8 0.3 1.3 0.6 0.0 0.6 0.7 0.4 0.8 0.8 0.3 1.1 0.9 0.4

2.6 3.0 1.2 2.7 2.0 0.0 2.1 1.8 2.3 2.5 2.5 0.7 2.2 2.2 1.5

a b

Standard conditions: 50 mL autoclave, 0.5 μmol of ligand, 0.5 μmol of CrCl3(THF)3, 500 equiv. of mMAO-3A, 15 min, 20 mL of MCH, 30 bar of ethylene, 60 °C. C8/C6 selectivity.

with the nonbonding orbital of oxygen. In other words, the nonbonding orbital of oxygen and σ* orbital of the CeF bond are hyperconjugated, lowering the electron density of oxygen and the coordination bond strength. These hypotheses could be clarified through single crystal xray diffraction and structure analysis, but it was unsuccessful to form a single crystal due to the good solubility of the complexes. 3.3. Thermal stability Considering the thermal stability of the catalysts, the oligomerization reactions with the POSS-containing catalytic systems were conducted at higher temperatures (Table 4). Like with L2, the activity of L1 decreased significantly at above 80 °C (Figure S6). It was observed that

Fig. 1. Proposed possible interactions in active catalysts: [a] coordination of pendant ether to metal center, [b] interference of coordination of pendant ether by electron withdrawing character of trifluoromethoxyl group. Table 4 Ethylene oligomerization results with different reaction temperatures.a Ligand

Temp. (°C)

Productivity (kg/g Cr/h)

C6 (%)

1-C6/C6 (%)

C8 (%)

1-C8/C8 (%)

C10-14 (%)

Polymer (%)

Selectivityb

L1

45 60 80 100 120 140 45 60 80 100 120 140 45c 45 60 80 100 120 140 45c 45 60 80 100 120 140

1049 997 865 632 513 316 446 575 396 239 218 128 1841 1793 1745 1638 1619 1617 1464 2275 2287 2275 2030 1595 1371 636

23.3 24.8 26.1 30.7 34.2 40.7 13.1 23.0 24.5 29.0 35.5 36.8 22.7 25.1 27.7 34.4 42.7 46.7 47.3 32.6 35.5 37.2 46.9 51.2 56.2 62.3

42.2 47.4 54.6 63.9 71.0 80.0 34.2 51.7 57.1 67.6 77.5 82.0 58.8 62.8 70.4 80.3 87.5 89.1 89.3 54.8 58.5 67.8 81.7 84.8 88.9 92.4

63.0 64.4 63.9 60.0 57.6 48.6 70.8 68.6 66.8 61.8 56.3 48.3 65.9 64.2 63.6 54.9 43.8 42.4 44.9 59.2 56.8 56.8 46.1 42.5 37.0 32.3

95.5 95.3 96.5 96.9 96.4 96.2 96.5 97.0 97.4 97.2 96.5 96.0 98.2 98.0 98.1 98.3 98.5 98.8 98.7 97.9 97.9 96.7 97.9 97.8 97.9 97.6

10.3 8.8 8.2 7.5 6.5 9.0 15.9 7.6 7.7 7.7 6.0 8.2 11.1 10.3 8.3 10.3 13.4 10.8 7.7 8.0 7.5 5.6 6.5 6.2 6.6 5.0

3.4 2.0 1.8 1.8 1.7 1.7 0.2 0.8 1.0 1.5 2.2 6.7 0.3 0.4 0.4 0.4 0.1 0.1 0.1 0.2 0.2 0.4 0.5 0.1 0.2 0.4

2.7 2.6 2.4 2.0 1.7 1.2 5.4 3.0 2.7 2.1 1.6 1.3 2.9 2.6 2.3 1.6 1.0 0.9 0.9 1.8 1.6 1.5 1.0 0.8 0.7 0.5

L2

L9

L15

a b c

Standard conditions: 50 mL autoclave, 0.5 μmol of ligand, 0.5 μmol of CrCl3(THF)3, 500 equiv. of mMAO-3A, 15 min, 20 mL of MCH, 30 bar of ethylene. C8/C6 selectivity. n-Pentane as solvent.

26

Applied Catalysis A, General 560 (2018) 21–27

H. Lee, S.H. Hong

the amount of 1-C6 in the C6 products increased as the temperature increased, and the C8/C6 selectivity decreased gradually (L1: 2.7 at 45 °C, 1.2 at 140 °C). These results were similar to those of L2. To our surprise, L9 exhibited excellent activity even at a higher temperature (1464 kg/g Cr/h at 140 °C). The ortho-F substituent may stabilize the active catalytic intermediate. It is industrially important for the catalytic activity to remain high at high temperatures, since this allows the dissolution of byproduct polymers during the process. As previously mentioned, fouling can be prevented and continuous operation can be enabled. Although the activity with L15 was twice that of L1 at every temperature, it decreased sharply at high temperatures like with the other catalysts. However, it still reached up to 2287 kg/g Cr/h at 45 °C. It is surprising to see activity comparable to a previously reported value under the same pressure (30 bar) [58]. Both of ligands, L9 and L15, also showed good activities in n-pentane at 45℃.

Limited, S. Afr.). US, (2015), p. 0080629. [12] M.S. Mokhadinyana, M.C. Maumela, M.M. Mogorosi, M.J. Overett, J.-A. Van Den Berg, W. Janse Van Rensburg, K. Blann, (Sasol Technology (Proprietary) Limited, S. Afr.). US, (2016), p. 0075616. [13] E.H. Kim, H.M. Lee, M.S. Jeong, J.Y. Ryu, J. Lee, B.Y. Lee, ACS Omega 2 (2017) 765–773. [14] J. Wang, N. Zhang, W. Shi, Y. Shang, C. Li, J. Coord. Chem. 70 (2017) 2708–2721. [15] A.H. Ulbrich, J.L. Milani, P.R. Livotto, J.-F. Carpentier, O.L. Casagrande Jr, J. Braz. Chem. Soc. 25 (2014) 2466–2474. [16] L.H. Do, J.A. Labinger, J.E. Bercaw, ACS Catal. 3 (2013) 2582–2585. [17] T. Jiang, Y. Tao, X. Gao, G. Mao, H. Chen, C. Cao, Y. Ning, Chin. Sci. Bull. 57 (2012) 1510–1515. [18] S. Licciulli, I. Thapa, K. Albahily, I. Korobkov, S. Gambarotta, R. Duchateau, R. Chevalier, K. Schuhen, Angew. Chem. Int. Ed. 49 (2010) 9225–9228. [19] K. Blann, A. Bollmann, H. de Bod, J.T. Dixon, E. Killian, P. Nongodlwana, M.C. Maumela, H. Maumela, A.E. McConnell, D.H. Morgan, J. Catal. 249 (2007) 244–249. [20] J.A. Waters, G.A. Mortimer, J. Chem. Eng. Data 17 (1972) 156–157. [21] A. Sahgal, H. La, W. Hayduk, Can. J. Chem. Eng. 56 (1978) 354–357. [22] T. Zhuze, A. Zhurba, Russ. Chem. Bull. 9 (1960) 335–337. [23] R. Walsh, D.H. Morgan, A. Bollmann, J.T. Dixon, Appl. Catal. A Gen. 306 (2006) 184–191. [24] G. Li, L. Wang, H. Ni, C.U. Pittman, J. Inorg, Organomet. Polym. 11 (2001) 123–154. [25] D.B. Cordes, P.D. Lickiss, F. Rataboul, Chem. Rev. 110 (2010) 2081–2173. [26] T. Luanphaisarnnont, S. Hanprasit, V. Somjit, V. Ervithayasuporn, Catal. Lett. 148 (2018) 779–786. [27] L.A. Bivona, F. Giacalone, E. Carbonell, M. Gruttadauria, C. Aprile, ChemCatChem 8 (2016) 1685–1691. [28] V. Ervithayasuporn, K. Kwanplod, J. Boonmak, S. Youngme, P. Sangtrirutnugul, J. Catal. 332 (2015) 62–69. [29] W. Zheng, C. Lu, G. Yang, Z. Chen, J. Nie, Catal. Commun. 62 (2015) 34–38. [30] A.J. Guenthner, K.R. Lamison, L.M. Lubin, T.S. Haddad, J.M. Mabry, Ind. Eng. Chem. Res. 51 (2012) 12282–12293. [31] A. Fina, D. Tabuani, F. Carniato, A. Frache, E. Boccaleri, G. Camino, Thermochim. Acta 440 (2006) 36–42. [32] F. Carniato, C. Bisio, G. Gatti, M. Guidotti, L. Sordelli, L. Marchese, Chem. Asian J. 6 (2011) 914–921. [33] P. Scholder, M. Hafner, A.W. Hassel, I. Nischang, Eur. J. Inorg. Chem. 2016 (2016) 951–955. [34] J. Su, L. Yang, Q. Wang, J. Chromatogr. A 1533 (2018) 136–142. [35] Z. Liu, J. Ou, Z. Liu, J. Liu, H. Lin, F. Wang, H. Zou, J. Chromatogr. A 1317 (2013) 138–147. [36] M. Wu, Ra. Wu, R. Li, H. Qin, J. Dong, Z. Zhang, H. Zou, Anal. Chem. 82 (2010) 5447–5454. [37] F. Alves, P. Scholder, I. Nischang, ACS Appl. Mater. Interfaces 5 (2013) 2517–2526. [38] C.-e. Zhao, J. Chen, Y. Ding, V.B. Wang, B. Bao, S. Kjelleberg, B. Cao, S.C.J. Loo, L. Wang, W. Huang, ACS Appl. Mater. Interfaces 7 (2015) 14501–14505. [39] Y. Mi, K. Li, Y. Liu, K.-Y. Pu, B. Liu, S.-S. Feng, Biomaterials 32 (2011) 8226–8233. [40] Y. Xiao, C. He, X. Lu, X. Zhang, Int. J. Nanosci. 8 (2009) 185–190. [41] D. Chattopadhyay, D.C. Webster, Prog. Polym. Sci. 34 (2009) 1068–1133. [42] R. Duchateau, Chem. Rev. 102 (2002) 3525–3542. [43] C. Di Iulio, M.D. Jones, M.F. Mahon, D.C. Apperley, Inorg. Chem. 49 (2010) 10232–10234. [44] J. Zhou, Y. Zhao, K. Yu, X. Zhou, X. Xie, N. J. Chem. 35 (2011) 2781–2792. [45] S. Spoljaric, A. Genovese, R.A. Shanks, J. Appl. Polym. Sci. 123 (2012) 585–600. [46] V. Ervithayasuporn, X. Wang, Y. Kawakami, Chem. Commun. (2009) 5130–5132. [47] M. Oliana, F. King, P.N. Horton, M.B. Hursthouse, K.K. Hii, J. Org. Chem. 71 (2006) 2472–2479. [48] T.W. Hey, D.F. Wass, Organometallics 29 (2010) 3676–3678. [49] M.J. Overett, K. Blann, A. Bollmann, J.T. Dixon, D. Haasbroek, E. Killian, H. Maumela, D.S. McGuinness, D.H. Morgan, J. Am. Chem. Soc. 127 (2005) 10723–10730. [50] E. Killian, K. Blann, A. Bollmann, J.T. Dixon, S. Kuhlmann, M.C. Maumela, H. Maumela, D.H. Morgan, P. Nongodlwana, M.J. Overett, J. Mol. Catal. A Chem. 270 (2007) 214–218. [51] T. Jiang, S. Zhang, X. Jiang, C. Yang, B. Niu, Y. Ning, J. Mol. Catal. A Chem. 279 (2008) 90–93. [52] K. Blann, A. Bollmann, J.T. Dixon, F.M. Hess, E. Killian, H. Maumela, D.H. Morgan, A. Neveling, S. Otto, M.J. Overett, Chem. Commun. (2005) 620–621. [53] M.J. Overett, K. Blann, A. Bollmann, J.T. Dixon, F. Hess, E. Killian, H. Maumela, D.H. Morgan, A. Neveling, S. Otto, Chem. Commun. (2005) 622–624. [54] A. Carter, S.A. Cohen, N.A. Cooley, A. Murphy, J. Scutt, D.F. Wass, Chem. Commun. (2002) 858–859. [55] L.H. Do, J.A. Labinger, J.E. Bercaw, Organometallics 31 (2012) 5143. [56] B. van der Westhuizen, P.J. Swarts, L.M. van Jaarsveld, D.C. Liles, U. Siegert, J.C. Swarts, I. Fernández, D.I. Bezuidenhout, Inorg. Chem. 52 (2013) 6674–6684. [57] K. Müller, C. Faeh, F. Diederich, Science 317 (2007) 1881. [58] S.-K. Kim, T.-J. Kim, J.-H. Chung, T.-K. Hahn, S.-S. Chae, H.-S. Lee, M. Cheong, S.O. Kang, Organometallics 29 (2010) 5805–5811.

4. Conclusions POSS-conjugated PNP ligands were introduced as new ligands for chromium-based catalysts for ethylene trimerization and tetramerization. These ligand-based catalytic systems showed high solvent tolerance and thermal stability. They exhibited good activities in lowcarbon-number solvents with good solubility. The activity was higher when an aryl group in the ligand contained a para substituent than when it contained a meta substituent. Ligands with ortho substituents led to higher selectivity, and steric and inductive effects influence the catalytic performance. The most surprising results were that the orthoF-substituted aryl phosphine ligand exhibited excellent thermal tolerance and that the ortho-OCF3-substituted ligand showed high activity even at low temperatures owing to the enhanced solubility from POSS. Acknowledgements This research work was supported by SK Innovation Co., Ltd. for the doctoral training program and the National Research Foundation of Korea (NRF-2014R1A5A1011165, Center for New Directions in Organic Synthesis; NRF-2015M3D3A1A01065480). The author would like to thank Dr. Yongnam Cho for NMR analysis and Chansaem Park for Aspen plus calculation. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.apcata.2018.04.030. References [1] T. Agapie, Coord. Chem. Rev. 255 (2011) 861–880. [2] J.T. Dixon, M.J. Green, F.M. Hess, D.H. Morgan, J. Organomet. Chem. 689 (2004) 3641–3668. [3] G.R. Lappin, L.H. Nemec, J.D. Sauer, J.D. Wagner, Kirk-Othmer Encyclopedia of Chemical Technology, (2010), pp. 1–20. [4] P.W.N.M. van Leeuwen, N.D. Clément, M.J.L. Tschan, Coord. Chem. Rev. 255 (2011) 1499–1517. [5] A. Bollmann, K. Blann, J.T. Dixon, F.M. Hess, E. Killian, H. Maumela, D.S. McGuinness, D.H. Morgan, A. Neveling, S. Otto, J. Am. Chem. Soc. 126 (2004) 14712–14713. [6] O.L. Sydora, T.C. Jones, B.L. Small, A.J. Nett, A.A. Fischer, M.J. Carney, ACS Catal. 2 (2012) 2452–2455. [7] Y. Shaikh, K. Albahily, M. Sutcliffe, V. Fomitcheva, S. Gambarotta, I. Korobkov, R. Duchateau, Angew. Chem. Int. Ed. 51 (2012) 1366–1369. [8] Y. Yang, Z. Liu, B. Liu, R. Duchateau, ACS Catal. 3 (2013) 2353–2361. [9] J.Y. Jeon, D.S. Park, D.H. Lee, S.C. Eo, S.Y. Park, M.S. Jeong, Y.Y. Kang, J. Lee, B.Y. Lee, Dalton Trans. 44 (2015) 11004–11012. [10] M.J. Overett, M.C. Maumela, M.M. Mogorosi, H. Maumela, M.S. Mokhadinyana, (Sasol Technology (Proprietary) Limited, S. Afr.). US, (2015), p. 0087873. [11] M.J. Overett, E. Grobler, S.J. Evans, K. Blann, (Sasol Technology (Proprietary)

27