Palladium-catalyzed enantioselective allylic substitutions on bifunctional substrates

Abstract

Enantioselective palladium catalyzed allylic substitutions were developed on substrates which proceed via symmetrical n3-allyl complexes. These substrates were designed with two identical functional groups on each end of the allyl system which allowed the derivatization of the resulting chiral products into attractive compounds. Furthermore, the advantage of a symmetrical substrate-palladium complex enabled an enantioselective attack of the nucleophile, resulting in a true enantioselective conversion. Di-benzylether substrates were synthesized and enabled the study of these substrates in allylic substitutions. Different parameters can influence the reaction such as the leaving group, ligand, solvent, base, catalyst and nucleophile. Therefore several screenings were performed to optimize the reaction conditions. For the leaving group screening, ethylcarbonate-, methylcarbonate, benzylcarbamate and chloracetates were chosen as leaving groups. The (E)-chloracetate substrate was found to give the best results and was chosen as a standard. It was observed that E/Z isomerization occurred, producing the same E/Z ratio, either starting from (Z) or (E)-ethylcarbonate substrates and (Z) or (E)-methylcarbonate substrates. An explanation for these results was given by the pi-sigma-pi isomerization. For the ligand screening, two classes of ligands were selected and compared on our substrate: PHOX and Trost ligands. From the results obtained, (R,R)-Trost ANDEN showed minimized E/Z isomerization as well as a good yield and ee. The choice of a good solvent is important and the screening showed three reliable solvents for our substrate: CH2Cl2, toluene, CF3Ph. The addition of a base improved the ee and the yield. When different bases were screened, the best results were obtained from the addition of 1eq. EDIPA. With (E)-chloracetate, [Pd(allyl)Cl]2 was the best suited catalyst. The catalyst loading was lowered down to 1 mol% but unfortunately lowering the catalyst loading also decreased the yield. Therefore 8 mol% of catalyst was chosen as a standard. With these optimized reaction conditions, a broad range of nucleophiles was tested on our substrate. Primary aryl amine nucleophiles were found to be excellent nucleophiles in our system with ee’s up to 97% and yields up to 93%. Good results were also obtained from not commonly used phenol nucleophiles with ee’s up to 92% and yields up to 83%. The absolute configuration was determined by derivatization of the substitution product (E)-N-benzyl-1,5-bis(benzyloxy)pent-3-en-2-amine into N-(1,5-dihydroxypenta-2-yl)benzamide. Optical activity as well as the assignment of the HPLC peaks on chiral stationary phase of the enantiomers allowed determination of the absolute configuration to be (R). The next step was focused on the derivatization of the chiral products to give potentially useful compounds. Oxazolidinone products were synthesized in a 1:1 diastereoisomeric ratio separable by column chromatography. From these products, the synthesis of various compounds could be imagined such as 2-aminopentoses. Polyfunctionalized acetamide was synthesized from (E)-N-benzyl-1,5-bis(benzyloxy)pent-3-en-2-amine. This compound offers many possibilities for further transformations such as selective epoxidation or dihydroxylation. At last, epoxidation of chiral phenol ether succeeded and further derivatization to afford uncommon phenol sugars can be envisaged. The second part of the thesis was focused on the study of dimeric substrates possessing a cyclic carbonate as leaving group. The following substrates were synthesized: meso diethylester, meso dibenzylether, racemo dimethylester and racemo dibenzylether. Allylic substitutions on the meso-diethylester substrate were performed using different reaction conditions. The desired substitution could not be obtained and instead an elimination product was isolated which appeared to be unstable over time. This product showed that beta-hydride elimination took place during the reaction. From racemo-dimethylester substrate, no product could be isolated and decomposition was observed. The more stable dibenzylether substrates provided better results. No reaction but also no beta-hydride elimination was observed from meso-dibenzylether substrate. On the other hand, product could be isolated from the racemo-dibenzylether substrate. Even though the yield was low (38%), the reaction was proven to be feasible and further improvements may be possible.