M.Sc. / The development of synthetic lubricants was the logical result of high demands being made by modern technology, the exhaustion of natural mineral oil sources and the search for environmentally friendly substances. Alpha olefins are products of the Sasol Fischer-Tropsch process and are currently part of the wider petrol pool. Substantial value can be added to the alpha olefins by the production of high value polyalphaolefins (PADs). This project was aimed at finding explanations for previous unsuccessful attempts to produce acceptable PADs from Sasol feedstreams. High priority was given to the modification of experimental procedures to produce a synthetic oil that conforms to specific standards. Oxygenates, aromatics and "other than alpha" olefins (branched, internal and cyclic olefins) were identified as problem components in the Sasol feedstream. These compounds led to early termination of oligomerisation and in the presence of BF3 , unwanted compounds were even incorporated into the PAD products. The feedstream was thus purified by distillation and a MeDH/H2D extraction procedure. As a result, better reaction control (with regard to oligomer distribution) was gained and the product quality improved. Reaction conditions were optimised and basic reaction models (to predict oligomer distributions) were developed. It became clear however, that product quality did not only depend on oligomer distribution, but also on the specific structures present in the various oligomers. GC analyses showed that each oligomer consists of a great number of isomers - even more than can be explained in terms of the classic cationic mechanism for oligomerisation. GC-MS analyses could not successfully distinguish between different branched isomers, because of the complexity of the mixture. Summary The question was asked whether the complex mixture has already been formed during the primary oligomerisation process, or at a later stage by skeletal rearrangements. Research work indicated that one can distinguish between a primary (normal oligomerisation) process, and a secondary process (skeletal rearrangement). It became clear that the secondary process is favoured by specific reaction conditions (e.g. long reaction times and high co-catalyst concentrations). The formation of isomers increased under these conditions and product quality was adversely affected. The possibility to produce PAD products from Sasol feedstreams identical to commercial products still exists, and necessitate further research work.

No description available.

No description available.

No description available.

This work involves a three part study concerning the synthesis and reactivity of quinodimethanes, the determination of reduction potentials of electrophilic olefins, and the crystal structure determination of a new charge transfer complex. Treatment of p-xylylene dicyanide (10) with NaH and dimethyl carbonate in glyme yielded a,a'-bis(methoxycarbonyl)-p-xylylene dicyanide (11a), 82% yield. Oxidation of 11a with N-chlorosuccinimide and triethyl amine in acetonitrile at 0° gave 7,8-di(methoxycarbonyl)-7,8-dicyanoquinodimethane (12a), 64% yield, d. 268°. This quinodimethane homopolymerizes spontaneously and forms a Wurster complex with tetramethyl-p-phenylenediamine and a blue-black CT-complex with tetrathiafulvalene. Cyclic voltammetry reduction potentials (E^(Red)(,p)) UV absorptions and ¹H-NMR chemical shift data of 22 olefins compounds were measured. The sequence for the ability to stabilize the ethylenic radical anion was found to be (diagram omitted) > -CN >-CO₂CH₃ > Cl > Br > H as expected. An inverse relationship between chemical reactivity and reduction potential was found for tetramethyl ethylenetetracarboxylate (25) and the triester derivative (33). Although 25 has a lower E^(Red)(,p) (-1.30v) than 33, it is less reactive toward polymerization. A linear correlation of E^(Red)(,p) and Hammett substituent constants was seen for the substituted maleic anhydrides and cyanofumarates. No correlations were seen for sterically crowded derivatives of 33. The crystal structure of tetrathiafulvalinium dimethyl dicyanofumarate was determined from a single crystal X-ray study. The crystal belongs to the monoclinic space group P2₁/C with two complexes per unit cell with cell constants a = 11.075(2) Å, b = 11.615(3) Å, c = 6.623(4) Å, α = 95.7°(16), V = 847.6(9) ų. The structural parameters have been refined to convergence, R = 0.0492 and R(,w) = 0.0614.

The spontaneous polymerization behavior of an acceptor substituted quinodimethane was investigated. High molecular weight homopolymers of 7,8,-di(ethoxycarbonyl)-7,8-dicyanoquinodimethane (DEDCQ) or copolymers of DEDCQ-styrene (St) and DEDCQ-p-methoxystyrene (p-MeOSt) were formed spontaneously in chloroform containing acetic acid. All polymerizations occurred by a radical mechanism. High molecular weight polymers formed at low conversion. Radical homopolymerization of DEDCQ was promoted by ambient light or heat. DEDCQ-p-MeOSt copolymerizations yielded alternating copolymers, which formed rapidly relative to the DEDCQ homopolymers and copolymers (rich in styrene). All of these polymerizations are thought to begin with a bond forming mechanism and propagate by polyaddition. A new acceptor group substituted quinodimethane was synthesized, characterized and copolymerized with electron donating monomers. Oxidation of the dianion of α, α'-dicyano-α, α'-diphenyl-p-xylene yielded 45% 7,8-dicyano-7,8-diphenylquinodimethane. In solution this compound exists in both the trans and cis forms but forms a sharply melting crystalline solid. It was resistant to homopolymerization but did form low molecular weight copolymers with electron rich aromatic monomers.

Donor-Acceptor tetramethylenes have been studied by polymerizations. 1,4-Zwitterionic intermediates are indicated when reactive tetramethylenes initiate homopolymerization. Alternately, 1,4-diradical intermediates initiate copolymerization. This basis for studying intermediates has led to an empirical table for predicting the zwitterionic and diradical nature of addition and polymerization reactions of tetramethylenes. Here we attempted to extend this work to trimethalylenes by studying the thermal ring opening of ethyl chrysanthemate, ethyl 1-cyano-2-(4-methoxyphenyl)-cyclopropane-corboxylate, ethyl 1-cyano-2-(2-methoxyphenyl)-cyclopro-panecroboxylate, and diethyl 1,3-dicyano-w,r-di(2-methoxyphenyl)-cyclobutanedicarboxylate. These compounds were found to be thermally stable to 150°C and did not initiate polymerization in styrene, methyl methacrylate, a series of high boiling acrylates, and dimethyl fumarate. Free radicals were trapped in dimethyl fumarate to give oligomers at temperatures above 110°C. Even though the compounds studied did not initiate polymerization at decomposition temperatures of 175°-200°C, dimethyl fumarate may prove useful in these studies in the future.

published_or_final_version / Chemistry / Doctoral / Doctor of Philosophy

published_or_final_version / Chemistry / Doctoral / Doctor of Philosophy

No description available.