Title:	Mechanistic Study of Heterogenously Catalyzed Lignin Depolymerization
DNr:	SNIC 2016/1-315
Project Type:	SNIC Medium Compute
Principal Investigator:	Joseph Samec <joseph.samec@su.se>
Affiliation:	Uppsala universitet
Duration:	2016-06-29 – 2017-07-01
Classification:	10404 10405 10403
Keywords:

Abstract

Lignin is a low valued and a waste stream from pulping (paper making) that currently is burnt up to generate process heat. By catalytic upgrading, this waste stream has the potential to become an important renewable carbon feed-stock for chemical synthesis as well as a biofuel. Currently, the catalytic upgrading of lignin is performed with an excess of hydrogen through a hydrogenolysis reaction generating an alkanyl ether as intermediate in an initial step. This intermediate requires high activation energy for the C-O bond cleavage that is responsible for the depolymerization. Because of the harsh reaction conditions, these methodologies have not been commercialized. New experimental studies of ours have identified a unique reaction mechanism proceeding through a low energy barrier with a transfer dehydrogenation as an initial step that generates a ketone intermediate (DOI:10.1002/cssc.201500117). The current reaction can be performed at 80 degrees Celsius without addition of hydrogen gas, as compared to the previous procedures that require above 200 degrees Celsius and high hydrogen pressure. The ketone intermediate has a calculated bond dissociation energy (BDE) that is 10 kcal/mol lower in energy than the corresponding alkenyl ether for the subsequent C-O bond cleavage (depolymerization). The difference in the calculated BDE:s for the two different pathways can only explain the different requirements in reaction conditions (80 vs 200 degrees Celsius) to a degree. This was also confirmed by measuring activation energies. We propose that the ketone intermediate, that is a key species in the low energy pathway, tautomerize to its enol form and this enol adsorbs with its C=C bond with a higher affinity to the soft palladium surface. We have observed the ketone intermediate in solution by NMR spectroscopy and this supports the proposed reaction mechanism. However, the ketone intermediate could also be generated as a side reaction in another reaction mechanism. The tautomer has not been observed and this is not expected. Thereby, experiments are not sufficient to determine the reaction mechanism in these complex transformations and need to be supplemented. During the last year, we have had a medium-scale project at the NSC Triolith computer facility for this project. In the beginning of the project, we did not fully utilize all of our quota because the initial PhD training and setup of the calculations took a bit longer than expected. However, by the end of the project, when all obstacles had been overcome, we more than filled our assigned quota. The initial study has in total been very successful, and resulted in one submitted publication and been presented at two conferences with full acknowledgement to NSC and SNIC. Our initial work on the lignin/Pd system gave us insight in to important intermediates in the lignin depolymerization, and further given us ideas on how to improve the catalytic activity. However, to verify these ideas additional calculations are needed. According to our initial proposal, the enol intermediate showed a higher affinity to the palladium surface and thereby supports our experimental data. Additionally, we found that the effect was higher for low-coordinated Pd sites (in our previous study modeled using only one layer of Pd). In parallel, we have experimentally determined the palladium nanoparticles to be in the order of 2-3 nm. This means that the studied lignin derivative is approximately of the same size (20 Å) as the Pd catalyst particles. With these findings in mind, we propose that it is necessary to have a better, and more realistic, description of the Pd catalyst in order to give a good description of the experimental situation. In the continuation, we intend to study the effect of reduced Pd coordination by introducing more realistic models (such as stepped surfaces and small nanoparticles). Moreover, we also intend to study the effect the carbon support has on the catalytic activity. In these studies, we will continue to use density functional theory calculations as the main computational tool, but also work in close contact to the experimentalists working on the very same problems. More specifically, we will continue to study how the key species found in our initial study interacts with the more realistic models discussed above. When we have a satisfying model system, which is in better accord to the experimental situation. We will furthermore perform extensive analysis by for example calculating vibrational spectra for the adsorbed intermediates. These results will provide useful insight to the experimentalists on how to best prepare their catalyst for maximal efficiency. The proposed research is an interdisciplinary research project between Joseph Samec’s group (experimental part) and Peter Broqvist’s group (computational and inorganic chemistry).