Infoscience

Thesis

Structured Non-Noble Metal Catalysts for the Liquid-Phase Reduction of Nitroaromatics

The industrial production of aromatic amines, essential intermediates for the synthesis of several drugs, dyes, pigments and agrochemicals, requires the use of heterogeneous catalysis. These molecules were formerly produced by the Béchamp reduction, a non-catalytic batch process requiring costly downstream separation steps. Catalytic approaches afford the implementation of continuous processes producing less waste. While it is generally not the primary goal when improving processes to make them more environmentally friendly, it is often one of the consequences, thereby improving the image of the chemical industry. The development of new catalysts is one of the major axes for optimizing industrial production of chemicals. Often multidisciplinary, it involves a better understanding of existing systems, advances in characterization and modelling techniques for the discovery of new catalytic materials. This thesis aims at designing new catalysts based on non-noble metals (e.g. Fe, Co, Ni) for the chemoselective hydrogenation of nitroaromatics in the liquid phase. The use of activated carbon fibres (ACFs) as a structured support has been at the centre of the strategy applied during this work in order to develop new catalysts capable of surpassing the best systems based on Fe, Co or Ni. The high porosity of ACFs, composed of very thin pores (  2 nm) and a high specific surface area, has been exploited to generate and stabilize highly dispersed crystallites of active phase. The formation of  2 nm metal oxides (MOx) nanoparticles (NPs) resulting from the impregnation/pyrolysis of Fe, Co or Ni precursors within the ACFs was performed. The reduction of nitroaromatics with different functional groups was carried out by catalytic hydrogenation with molecular hydrogen and by catalytic transfer hydrogenation (CTH) with hydrazine as reducing agent. The MOx/ACF catalysts were shown to be active and selective under mild conditions in the CTH of para-chloronitrobenzene (p-CNB) to para-chloroaniline (p-CAN). The FeOx/ACF was tolerating many substituents of nitroarenes during the -NO2 hydrogenation. The CoOx/ACFHNO3 catalyst was less active compared to the Fe-based catalyst but afforded a meta-vinylaniline (m-VA) yield above 99% during the reduction of the meta-nitrostyrene (m-NS), exceeding, to the best of our knowledge, the best performance reported to date. The very low activity of the NiOx/ACF catalyst in CTH has prompted us to test Ni-based catalyst in the hydrogenation of nitroarenes with gaseous hydrogen. Raney Nickel is a well-known catalyst used in industry for this reaction. First, the existence of particles size effects was evaluated using unsupported Ni PVP-stabilized NPs (2¿14 nm). An antipathetic structure sensitivity of the meta-dinitrobenzene (m-DNB) hydrogenation to meta-nitroaniline (m-NAN) was demonstrated, showing a 4.6-fold increase in the turnover frequency over the 14 nm NPs. Once supported on ACFs and cleaned from PVP by a UV-ozone treatment, the 2 nm Ni NPs afforded an increase of selectivity towards m-NAN from 34 to 96% at close to full conversion (Xm-DNB = 99%). The selective transformation favouring m-NAN has been attributed to reactions taking place at the edges and vertices of the NPs. The Langmuir-Hinshelwood two-site kinetic model was consistent with the experimental data. The incorporation of gold to form Ni-Au (1:1) bimetallic NPs led to a significant increase in the selectivity (99%), possibly by mimicking the electronic

Related material