Ligand optimization for the selective tetramerization of ethylene using Cr-complexes

Linear alpha olefins (LAOs) possessing a reactive C-C-double bond at the α-position are used as starting material for the production of a great variety of organic intermediates and speciali- ties. Nowadays, the use as co-monomer in polymer synthesis (C4–C8) covers more than 50 % of the global LAO demand by end use. Because of the rare number of suppliers of 1-octene, this product is most economically favourable. SASOL Ltd. commerciallized an effective process for the fabrication of 1-octene based on the chromium catalyzed selective tetramerization of ethylene. To improve the homogeneous catalyst consisting of chromium(III)-acetylacetonate, Cr(acac)3, a bis(diarylphosphino)amine ligand (PNP-ligand) and an aluminium alkyle based co-activator, the following investigations were carried out. The theory of a bimetallic mechanism for the ethylene tetramerization was probed by comparing experiments using a bifunctionalized PNP-ligand with a ferrocene-backbone and its monofunctionalized homologue. In result, both ferrocenyl-PNP-ligands give a nearly identical prod- uct distribution and no proof for the theory of a bimetallic mechanism was found. Contrarily, the results strongly support the idea of a monometallic mechanism as accepted by the majority of researchers’ community. Ligands for the ethylene tetramerization at enhanced temperatures were screened. The ligand N-bis-(2-fluorophenyl)-phosphino-N-diphenylphosphinobenzyla- mine, (2F)PN(Bn)P, was the best candidate. At ϑR = 80 °C, pC2 = 45 bar, tR,tot = 30 min and cCr = 4 µmol/l the ligand gave 60.5 % 1-octene by showing an activity of 8.72 1E+6 gproduct/gchromium/h. Kinetic studies using (2F)PN(Bn)P at ϑR = 80 °C gave the reaction orders with respect to the ethylene concentration: n1-C6 = 0.68, n1-C8 = 1.73 and noverall = 1.34. In order to determine the activation energy, the system which showed great instability at ϑR = 80 °C was stabilized by the addition of the modifier hexafluorobenzene (HFB). This caused also a drastic increase of the measured activities. The Arrhenius-plot indicated that, the reaction was already mass transfer limited above ϑR = 50 °C. That is why the value for EA = ~ 59 kJ/mol could only be estimated for the range of ϑR = 40 – 50 °C. The promoting effect of HFB on activity was probably due to a partial fluor- ination of the aluminium co-activator. It was shown that HFB could be replaced by the less volatile modifier perfluorobiphenyl, PFB. It is assumed that the formed fluoro-aluminium compounds are also prone to react with the metal surface of the reactor wall causing a modification effect by surface fluorination. By the application of optimized parameters (ϑR = 80 °C, pC2 = 70 bar, cCr = 2 µmol/l , PFB : Al = 0.5 : 1, baffles), a maximal activity of 60 E+6 gproduct/gchromium/h was reached.