• 2
• 2-oxindole N-allylate
• 2-oxindoles
• 2-trichloroethyl) azodicarboxylate
• AG-041R
• asymmetric α-amination
• bis(2
• BTCEAD
• Cinchona
• enantiomeric excess
• enantiomers
• HPLC-DAD
• HPLC-FSC
• LeBlanc-Fitzsimmons conditions
• optimisation
• p-methoxybenzyl azodicarboxylate
• quinidine
• quinine
• stereochemistry
• α-amino carbonyl compounds

In recent years, the synthesis of α-amino carbonyl compounds has attracted increasing interest due to their role in pharmaceutical chemistry. These molecules can be used as non-natural amino acids that modify the biological properties of the drugs in which they are incorporated. In particular, 2-oxindoles functionalised at the C3 position with an amine group form the core of numerous bioactive molecules, including enzyme inhibitors and pharmaceutical candidates for the treatment of diseases such as anxiety, depression and malaria. The biological activity of these chiral derivatives is strictly dependent on their absolute configuration, thus necessitating the development of synthetic methods that exert good stereochemical control.
A particularly effective approach to obtain these structures is asymmetric α-amination (αAA), a reaction that introduces an amine group at the α-position with respect to a carbonyl using electrophilic nitrogenous reagents such as azodicarboxylic esters. The reaction can be conducted via metal or organic catalysis, but the literature mainly reports the use of chiral organocatalysts with a low catalytic load (5-20 mol%), among which the alkaloid derivatives of Chincona, such as quinine and quinidine, stand out. These catalysts allow the two opposite enantiomers of the same product to be obtained, effectively regulating the stereochemistry of the reaction.
This thesis work focused on optimising the αAA of 2-oxindoles, with the aim of synthesising AG-041R, a gastrin/CCK-B receptor antagonist with potential therapeutic applications in cartilage repair and the treatment of gastric ulcers. This compound, initially developed by the Chugai pharmaceutical group, has stimulated numerous synthetic studies due to its interesting biological properties. However, its enantioselective synthesis presents several difficulties, including moderate yields and problems in functionalising the endocyclic nitrogen atom with the 2,2-diethoxyethyl fragment.
In our laboratory, the synthesis of AG-041R was previously studied in the theses of Spadoni, Bondanza and Bonciolini, which provided promising results but with some limitations. Spadoni used p-methoxybenzyl azodicarboxylate (Moz) as the aminating agent, while Bonciolini explored the use of bis(2,2,2-trichloroethyl) azodicarboxylate (BTCEAD). Both strategies achieved good enantioselective control. Bonciolini's approach, however, showed low yields in the diethoxyethyl fragment introduction step via N-alkylation of the 2-oxindole substrate with free NH, whereas Spadoni's methodology required a long sequence of reactions and the use of osmium salts, in contrast to the synthetic choice of using organocatalysis.
To overcome these problems, this work developed an alternative synthetic strategy, exploiting the N-allylated substrate and incorporating some of the procedures described by Bondanza. The α-amine compound was subjected to deprotection and reductive fission of the N-N bond based on LeBlanc-Fitzsimmons conditions (Zn/AcOH in the presence of acetone). For the processing of the allylic fragment, an ozonolysis reaction, followed by acetalisation, was carried out to resolve the difficulties encountered in previous studies.
The synthesis of AG-041R was carried out from both the aminated substrate with BTCEAD racemate and the enantioenriched substrate. The racemic intermediates were used to develop chiral-phase enantiomeric separation methods (HPLC-FSC), which were necessary to determine the enantiomeric purity of the corresponding enantioenriched intermediates.
The first part of the thesis focused on the optimisation study of the αAA of N-allylated 2-oxindole with BTCEAD, in which several parameters were varied, including solvent, temperature, catalyst and mode of addition of the amine reagent. The optimum conditions were found to be toluene at 70°C, using (DHQN)₂PHAL as the catalyst and adopting a slow addition of the azodicarboxylate, which made it possible to reduce the decomposition of the reagent and improve the reaction efficiency. This strategy resulted in an α-amine product with high conversion and an enantiomeric excess of 58%. The possibility of further enriching the product, which behaves as a true racemate, by recrystallisation in MeOH was also investigated, achieving an ee of up to 94%.
The next step involved the reduction of α-amine according to LeBlanc-Fitzsimmons, as already tested in the work of Bondanza and Bonciolini. The progress of the reaction was monitored using an HPLC method with a diode array detector (DAD), showing the formation of the amine within a few hours. The amine intermediate was then converted to urea, whose enantiomeric excess was not altered by the reductive fission step. In addition, recrystallisation in EtOH/H2O showed true racemate behaviour of the urea, allowing the compound to be enriched from 56% up to 77% ee.
For the introduction of the diethoxyethyl fragment, an ozonolysis reaction followed by acetalisation was performed, avoiding the problems associated with N-alkylation reactions. The use of a commercial ozonator and an organic dye (Sudan II) capable of accurately signalling the end of the reaction improved the yield of (64%) compared to that reported by Bondanza (34%). The obtained acetal was finally processed to obtain AG-041R with the same enantiomeric purity as the starting urea (77%).
In parallel, αAA was tested with Azodicarboxylate Moz on the N-allylated substrate as performed by Spadoni. The product obtained was deprotected with trifluoroacetic acid (TFA) and anisole, reduced by means of LeBlanc-Fitzsimmons conditions and converted to urea, showing that this strategy is also promising for the α-amino product Moz.
In conclusion, the present study has led to a significant improvement in the synthesis of AG-041R, overcoming the critical issues raised in previous work. The optimisation of the asymmetric α-amination reaction with BTCEAD on the N-allylated substrate resulted in a product with high selectivity. This product was efficiently processed in urea, and the allyl fragment present in the starting molecule allowed the diethoxyethyl group to be introduced via ozonolysis and acetalisation, making the synthesis of AG-041R easier and more efficient. The results obtained open the way for further developments, such as the synthesis of AG-041R from α-amino Moz using this synthetic strategy.