Nitrogen containing biobased chemicals produced using enzymes

Today, nitriles are produced by introducing the nitrogen functionality in hydrocarbons originated from fossil resources under high pressure and temperature. Environmental concerns associated with the use of fossil resources, as shown in chapter 1, stimulate the need to produce chemicals in a more sustainable way. Renewable resources such as amino acids from biomass, that already contain nitrogen in their molecule, are investigated as alternative starting materials to produce biobased nitriles. In this thesis the enzyme vanadium chloroperoxidase (VCPO) was used as catalyst to produce biobased nitriles from amino acids via the oxidative decarboxylation reaction. Industrially relevant nitriles such as acrylonitrile and succinonitrile, can be prepared starting from amino acids. For example, glutamic acid (Glu) – the most abundant non-essential amino acid in biomass – can be converted to acrylonitrile via the intermediate 3-cyanopropanoic acid (GluCN). The oxidative decarboxylation reaction of other biomass derived amino acids was investigated as well. Glu can be fully converted into GluCN with high selectivity using the enzyme VCPO, H2O2 and catalytic amounts of NaBr. In contrast, under the same reaction conditions the oxidative decarboxylation of aspartic acid (Asp) resulted in low conversion and selectivity towards the nitrile. In chapter 2, it was investigated how two chemically similar amino acids, Glu and Asp, react differently towards the oxidative decarboxylation. For this, the conversion of Glu and Asp was investigated as a function of bromide concentration. In presence of catalytic amount of bromide (0.1 equiv.), Glu resulted in full conversion and high selectivity. It was shown that by increasing the amount of bromide present in the reaction mixture to 2 equiv., the conversion of Asp was increased from 15% to 100% and its selectivity towards 2-cyanoacetic acid (AspCN) from 45% to 80%. It was concluded that the difference in reactivity must be due to the difference of one carbon atom in the side chain between Asp and Glu and the proximity of the side chain to the reactive alpha functionalities of the amino acids. It was hypothesised that the alpha functionalities in Asp are stabilised in intra- or intermolecular interactions with the side chain carboxyl functionality which prevents Asp to react in a similar manner as Glu. The influence of the side chain functionality and the side chain length of amino acids towards the reactivity of alpha functionalities with respect to oxidative decarboxylation was further investigated for different amino acids (chapter 3). It was shown that the conversion can be modified as a function of the concentration of NaBr for all amino acids tested. Only two amino acids, Glu and aminoadipic acid, can be fully converted into nitriles with catalytic amounts of NaBr (0.04 equiv.). For all other amino acids with aliphatic, hydroxy, carboxyl and methyl ester functionalities tested, a minimum amount of NaBr present in the solution (≥ 0.4 equiv.) is required to reach full conversion. It was concluded that the length of the side chain does not make a significant difference for the selectivity, as previously proposed. However, the position of the functionality on the side chain (β-carbon) in relation to the bromination centre could hinder the production of nitriles by oxidative decarboxylation by reducing the reaction rate of the bromination. It was shown that while functional groups like aliphatic, hydroxyl or methyl ester show no significant influence on the reactivity of amino acids, the carboxyl functionality has a positive effect during the oxidative decarboxylation reaction. An addition to the known reaction mechanism was proposed for the amino acids with carboxyl functionality at the side chain. It is proposed that the side chain carboxyl functionality is involved in a self-catalysis mechanism. The elucidation of the exact reaction mechanism could enable reactions of mixtures of amino acids at lower concentration of NaBr for the production of biobased nitriles. To further enhance the sustainability of the oxidative decarboxylation of amino acids, the in situ production of H2O2 was explored in chapter 4. The direct use of oxygen by alcohol oxidase (AOX) was investigated as alternative to the hydrogen peroxide originated from the energy-intensive anthraquinone process. The conversion of ethanol to the volatile acetaldehyde was selected for the half redox reaction of AOX due to the easiness of the downstream processing, e.g. by pervaporation of acetaldehyde. The cascade AOX-VCPO was used for in situ production of hydrogen peroxide for fast halogenation reactions and oxidation reactions via halogenation. For the first time, the oxidative decarboxylation of glutamic acid - an oxidation reaction via halogenation - was shown to be possible using the cascade AOX-VCPO. For this reaction, the two enzymes had to be separated in two reactors due to inhibition of AOX caused by HOBr – the product of VCPO. However, the fast halogenation reactions such as the bromination of monochlorodimedone, using the cascade AOX-VCPO was possible in one reactor. Oxygen availability in aqueous solutions, scaling up as well as the reaction kinetics need to be further addressed. The feasibility of the conversion of Glu into GluCN – an intermediate in the production of biobased acrylonitrile, was evaluated in chapter 5. The production of GluCN by VCPO and H2O2 (Scenario 3) was compared in a techno-economic assessment with other alternative biobased routes i.e. the production of GluCN by NaOCl (Scenario 1), by the cascade AOX-VCPO (Scenario 4) and by oxygen with a Ru catalyst (Scenario 5). It was found that by replacing NaOCl with VCPO-H2O2 the energy requirements of the process is reduced by a factor of 1.5 for the production of 1 t GluCN. This is mainly as a result of performing the reaction at 25°C, eliminating the need of cooling below room temperature (4°C) as in the case of NaOCl. The mass balance is slightly improved as selectivity close to 100% can be achieved by VCPO-H2O2 system and a significant reduction in waste was achieved. By further replacing NaOCl with oxygen in Scenario 4 and 5 the cost-benefit margin was increased significantly. Based on the cost-benefit analysis the only scenario with a positive cost-benefit margin of 194 €/t GluCN is Scenario 4, owed to the co-production of acetaldehyde which is a valuable product. The sensitivity analysis of Scenario 4 and 5 where the price of different compounds was changed, shows that the price of Glu and GluCN are the parameters that influence the economics of the process the most. At the moment, the price of the substrate, Glu, and the price of the product, GluCN – the intermediate in the production of acrylonitrile or succinonitrile – are too high to be competitive with the fossil based nitriles. As the price of Glu is already a best case scenario the use of cheaper sources of amino acids, e.g. crude mixtures of amino acids, should be tested. To produce the biobased nitriles, constraints should be applied to polluting industries to increase the price of fossil-based nitriles and as a result make the biobased nitriles more competitive. In chapter 6, the results presented in chapters 2-5 and their implications are discussed. Suggestions for future research and concluding remarks are also provided.