Renewable aromatic hydrocarbons for fuel blending from lignin by novel homogenous catalysis

Transport accounts for 25% of the energy consumption. To improve sustainability, transport increasingly runs on biofuel, and new environmentally friendly routes to sustainable biofuel are needed. For example, biofuels should be produced from resources not used for food or animal feed. This is true of wood, and Nordic forests, with a long history of sustainable management, are an attractive biomass resource for biofuel production. About one third of dry wood consists of lignin, which is the only natural feedstock for so-called aromatic hydrocarbons, a required component of combustion-engine fuels. Unfortunately, no sustainable process from lignin to aromatic hydrocarbons has been developed, despite 40 years of research. Much of this effort has involved the use of catalysts, materials that speed up chemical reactions without themselves being consumed, but so far, the catalysts have required high temperatures and have not given the desired aromatic products. Two recent developments offered a promising starting point for the project. Firstly, a process, termed the Bergen lignin-to-liquid (LtL) process, chops up lignin, a polymer, into small pieces. The resulting molecular fragments are aromatic but remaining oxygen atoms must be removed before these fragments can be used as fuel. Secondly, new and promising catalysts for selective removal of oxygen atoms left in aromatic compounds have been reported in recent years. In this project, researchers at the University of Bergen, together with partners at EPFL-Lausanne, IRCELyon, and University of Leeds, will develop the newly reported oxygen-removal catalysts further and optimize them for converting LtL oil to hydrocarbons. The optimization will proceed in cycles, each starting with computational design of catalysts, followed by synthesis and testing of the most promising candidates serving as feedback to improve the computational design of the next cycle. Finally, with the thus improved catalysts, the overall process from lignin to aromatics for liquid biofuel will be developed. The above-mentioned optimization of catalysts requires insight into the mechanism by which these catalysts remove oxygen from the aromatic compounds. This mechanistic information is mainly being obtained computationally, via molecular-level modeling, and for the most selective, state-of-the-art oxygen-removal catalysts, many possible reaction mechanisms have been followed computationally. Surprisingly, the calculations show that the structurally elaborate and difficult-to-make state-of-the art catalysts undergo severe structural changes before catalysis starts, and that the catalytically active species may, in fact, be much simpler and easier to make. An example of such a simple compound has been synthesized and shown to give selectivities in oxygen removal similar to those of the more elaborate, difficult-to-make state-of-the-art catalysts. This, and other experimental observations, suggest that the computationally predicted mechanism is correct. To exploit these mechanistic insights to make industrially compatible catalysts, simple, oxide-supported, solid-state (heterogeneous) catalysts have been made and tested. The best such catalysts obtained in the project are at least 10-fold more active than state-of-the-art catalysts for oxygen removal from lignin model compounds, work at mild reaction conditions, and match the best reported selectivities for aromatic products reported to date. The project team will continue to investigate the extent to which these breakthrough catalysts may be recovered and recycled, and to what extent they may convert lignin-derived bio-oils, which contain complex mixtures of oxygen-containing compounds, to pure aromatic compounds.

The interdisciplinary approach taken in this project, with the experimental catalyst development being guided by computationally obtained mechanistic insight and predictions, has proved to be very fruitful. An important outcome of the project is thus that the project team will continue to use this strategy in other projects. The approach is also expected to inspire other research groups to adopt similar strategies. The project has offered clear insights into the factors governing catalytic activity and selectivity for aromatic products in hydrodeoxygenation of lignin-relevant phenolic substrates. These factors are expected to be relevant not only for the catalysts studied in the project but for hydrodeoxygenation of phenolic substrates in general. The mechanistic insight generated in the project may therefore impact on the valorization of lignin far beyond the lifetime of the project. The catalysts developed in the project are also expected to have an impact. Whereas a simple-to-make molecular catalyst is likely to be further explored as a model catalyst, some of the oxide-supported catalysts are expected to be picked up by the scientific community, and perhaps also by the industry. Finally, the optimized LtL oil will be very useful when the Bergen team continues catalytic upgrading of this oil. The lessons learned during its optimization should also be useful for the wider community working on lignin-derived bio-oils.

Lignin is abundant, is not used for food or feed, and is the only natural and direct resource for aromatic hydrocarbons, a required component of motor fuels. However, sustainable routes from the polymeric, highly functionalized lignin to the depolymerized, reduced aromatic hydrocarbons have yet to be established. Any such route will have to be based on catalysis, but, in spite of immense efforts over many years, the best heterogeneous hydrodeoxygenation (HDO) catalysts still need harsh conditions and are not sufficiently selective for cleavage of the carbon-oxygen bonds. In contrast, high selectivity for these bonds has recently been reported for homogeneous iridium-based catalysts. We will design water-soluble versions of these new catalysts and develop a new, green, and economically viable route from lignin-to-liquid (LtL) fuels via a two-phase process in which the organic phase is more uniform, stable, and tunable than the standard bio-oils from pyrolysis of lignocellulosic biomass, namely the LtL oil obtained from the University of Bergen-developed solvolytic conversion. The LtL oil is dominated by depolymerized arenols and arene esters that will be hydrodeoxygenated by the recyclable and water soluble catalysts, and the resulting aromatic products will follow the organic phase and be easily separable from the catalyst-containing water phase. The water soluble catalysts will be designed using a recently developed artificial evolution method. Thus obtained computational predictions will be synthesized, characterized, and evaluated for efficiency and product composition to give an overall iterative catalyst development loop with feedback from experiments. Similarly, information from the catalytic experiments on single compounds, blends and LtL oil will help tune the oil to the needs of the catalysts in order to achieve an overall efficient and sustainable process that may help substitute fossil by renewable carbon resources for fuel production.