Dialkyl carbonates (DAC) are an important class of compounds for a variety of industries, with demand currently outstripping production for their use as solvents, electrolytes, and fuel additives. Current industrial processes rely on atom inefficient methods of production, such as transesterification or utilise hazardous starting materials such as carbon monoxide or phosgene that come with a large energy and carbon footprint.
The direct synthesis of carbonates from CO2 and alcohols promises to be a more atom economic route, though the unfavourable thermodynamics of these reactions make the development of efficient processes challenging.
In this work, a number of high surface area pure cerium oxide materials have been produced using polymer templates achieving surface areas exceeding 400 m2.g-1, compared to 50 m2.g-1 for commercial samples. Mixed metal oxides, containing cerium, zirconium, and aluminium, have also been produced by precipitation. These have been evaluated for their activity and stability in the direct synthesis of dimethyl carbonate (DMC) with chemical dehydrating agents and were shown to be stable across five batchwise reuse cycles.
The direct synthesis of DMC using different dehydrating agents has been investigated. The use of diisopropyl carbodiimide (DIC) allows for 20–40% conversion to be achieved in 2–4 h. The use of 2-cyanopyridine allows for similar conversions over 6 h, while also exceeding 80% methanol conversions in 20 h. We observe that the reaction with DIC has an onset temperature of 100 °C, below which the reaction progresses much more slowly. We also observe that the reaction progresses more rapidly at pressures below the critical point of the methanol/CO2 mixture. Above the mixture critical point, reaction rates are decreased by up to 90% in batch, due to liquid phase expansion by CO2.
The thermodynamics have been calculated for both the formation of DMC and diethyl carbonate (DEC) from published enthalpies of formation and standard entropies. These show the reaction to be endergonic at 298 K by over 25 kJ.mol-1, and the negative entropy of these reactions also shows that the reactions become more endergonic as temperature increases. An iterative multivariate model has been written to allow for the calculation of equilibrium concentrations of each reaction component. This has been used to determine the best-case equilibrium conversions, as well as in scenarios where wet starting material is utilised. Our model has also been applied in several scenarios where products are continually removed from the reactor, in which we demonstrate that a strategy where both the DAC and the water by-product is removed from the reactor, in a non-reactive manner, allows for greater cumulative conversion than water sequestration alone.
The thermodynamic model has been validated by performing reactions without dehydrating agents with a <5% discrepancy between the experimental and calculated values. A productive flow system for synthesis of DMC and DEC has been demonstrated, with our optimised system showing equal productivity for both DMC and DEC. Evaluation of catalyst stability has also been performed in flow, where a less active ceria-zirconia mixed metal oxide catalyst produced in this work showed superior stability and productivity compared to the commercial cerium oxide catalyst which showed higher activity in batch mode. Further development of the system could allow for continuous removal of products, and recycling of reactants, allowing for a more sustainable production of DACs.
|Date of Award||26 May 2021|
|Supervisor||Ulrich Hintermair (Supervisor), Meenakshisundaram Sankar (Supervisor) & Antoine Buchard (Supervisor)|
- carbonate synthesis
- CO2 utilisation
- continuous-flow processing
- supercritical fluids