Abstract
Ethylene glycol (EG) is a bulk chemical that can be produced from sugar via tungstate catalysed hydrogenolysis with high atom efficiency (100%), as an alternative to the conventional fossil-based production process. In this route, sugar is cleaved to glycolaldehyde by a tungstate species, after which the resulting glycolaldehyde is hydrogenated to EG over a heterogeneous catalyst. This thesis has investigated the conversion of non-edible lignocellulosic biomass to ethylene glycol (EG), with a particular emphasis on the identification and quantification of catalyst deactivation phenomena. In addition, pretreatment strategies that prevent catalyst deactivation have been developed.
We have studied the poisoning effect of lignin (Chapter 3), inorganics (Chapter 4) and sulphur/nitrogen (Chapter 5) on the catalysts. It turned out that inorganics, in particular divalent cations, precipitate the homogenous tungstate catalyst. These poisons can be leached from the biomass by an acetic acid (10 wt.%) water mixture. Sulphur, a known metal catalyst poison, indeed deactivates the hydrogenation catalyst. However, we found that the sulphur content of biomass greatly exceeds the desired target specification. Lignin appeared to retard the hydrogenation activity, but not critically in the batch tests performed, as the hydrogenation of glycolaldehyde to EG was not affected. Moreover, we have studied the impact of biomass structure, i.e. particle size (Chapter 6), on the hydrogenolysis outcome and found that there is no effect on the glycol yields under the conditions studied.
In parallel, we studied an integrated process concept of biomass pretreatment and subsequent hydrogenolysis that allowed significant removal of lignin (>80%) while preserving the cellulose fraction (~100% retention) in the solid residue (Chapter 7). In this process woody biomass is fractionated in a solvent composed of water and at least 50 wt.% of a mixture of light alcohol and organic acid (e.g., ethanol/acetic acid). We chose organic molecules that are by-products from the hydrogenolysis and pretreatment steps and can thus be recycled over the process to minimize the need for solvent make-up.
The removal of the contaminants, i.e. lignin and ashes, can be run more efficiently by operating the washing section counter currently (Chapter 8). The spent solvent (after washing) should then be used in the reaction step to minimize solvent use.
We have studied the poisoning effect of lignin (Chapter 3), inorganics (Chapter 4) and sulphur/nitrogen (Chapter 5) on the catalysts. It turned out that inorganics, in particular divalent cations, precipitate the homogenous tungstate catalyst. These poisons can be leached from the biomass by an acetic acid (10 wt.%) water mixture. Sulphur, a known metal catalyst poison, indeed deactivates the hydrogenation catalyst. However, we found that the sulphur content of biomass greatly exceeds the desired target specification. Lignin appeared to retard the hydrogenation activity, but not critically in the batch tests performed, as the hydrogenation of glycolaldehyde to EG was not affected. Moreover, we have studied the impact of biomass structure, i.e. particle size (Chapter 6), on the hydrogenolysis outcome and found that there is no effect on the glycol yields under the conditions studied.
In parallel, we studied an integrated process concept of biomass pretreatment and subsequent hydrogenolysis that allowed significant removal of lignin (>80%) while preserving the cellulose fraction (~100% retention) in the solid residue (Chapter 7). In this process woody biomass is fractionated in a solvent composed of water and at least 50 wt.% of a mixture of light alcohol and organic acid (e.g., ethanol/acetic acid). We chose organic molecules that are by-products from the hydrogenolysis and pretreatment steps and can thus be recycled over the process to minimize the need for solvent make-up.
The removal of the contaminants, i.e. lignin and ashes, can be run more efficiently by operating the washing section counter currently (Chapter 8). The spent solvent (after washing) should then be used in the reaction step to minimize solvent use.
| Original language | English |
|---|---|
| Qualification | Doctor of Philosophy |
| Awarding Institution |
|
| Supervisors/Advisors |
|
| Award date | 3 Jun 2022 |
| Place of Publication | Enschede |
| Publisher | |
| Print ISBNs | 978-90-365-5364-3 |
| DOIs | |
| Publication status | Published - 3 Jun 2022 |