(475e) Tuning Pathways for the Diversification of Biomass-Derived Coumalic Acid- Insights from First-Principles

Functionalized aromatics used in the production of value-added chemicals and materials can be readily synthesized from biomass-derived coumalic acid and its ester derivatives via a cascade of reactions involving Diels-Alder addition of a dienophilie to form a cycloadduct, followed by decarboxylation and dehydrogenation.^1-4 Some of these compounds have the potential to directly replace petroleum-based aromatics, while others have novel functionalities, traditionally inaccessible through petroleum sources.⁵ In-situ NMR experiments, together with density functional theory (DFT) calculations are used herein to show that for a series of electron-donating dienophiles, decarboxylation of the cycloadduct product is rate-limiting during non-catalytic synthesis of aromatics in solution phase.^6,7,8 We find that γ-Al₂O₃,aLewis acid catalyst considerably accelerates decarboxylation, while Brønsted acid catalysts result in ring opening of the cycloadduct to form isophthalic intermediates. Over model γ-Al₂O₃ surfaces, theoretical results suggest that five-coordinate Al sites are the active sites for decarboxylation. While lower coordinate Al sites are present, they are blocked by surface hydroxyl (-OH) groups. These hydroxyl surface intermediates, however, act to assist in the adsorption of the cycloadduct. The theoretical results are fully consistent with results from solid-state NMR experiments. Rigorous microkinetic modeling of DFT-derived free energy diagrams and NMR kinetic data reveal that the Lewis acid catalyzed route is ~50 kJ/mol lower than the non-catalytic solution phase reaction, thus, explaining the significantly improved rates of decarboxylation. The lower barriers from the Lewis-acid catalyzed route enable low temperature one-pot synthesis of aromatics with high selectivity (~ 99.7 %), directly from coumalic acid using a bifunctional Pd/γ-Al₂O_3, where decarboxylation is catalyzed over γ-Al₂O₃ and dehydrogenation takes place on Pd.

Understanding the factors that control the regio-chemistry of Diels-Alder addition is very crucial so that theoretical yields for different regio-isomers can be predicted prior to synthesizing these compounds. Theoretical results find that electron-donating dienophiles highly favor para addition over meta. While electron-withdrawing dienophiles also favor para addition, the ratio of para to meta is much closer to unity. This is explained using frontier orbitals interactions as well as steric hindrance during addition. We further explore the influence of different solvents on the formation and reactivity of the cycloadduct. We find that the Diels-Alder addition reaction as well as decarboxylation is accelerated in polar protic solvents such as water and methanol via stabilization of the charged transition state. However, protic solvents are found to result in a slight degradation of coumalic acid. In addition, weakly acidic protic solvents such as methanol result in ring opening similar to the Brønsted acid catalyzed ring opening reaction instead of decarboxylation. Water degrades coumalic acid, but, does not induce ring opening by itself. Dry polar aprotic solvents such as dioxane were found to result in decarboxylation alone. Thus, the choice of solvent is found to critically alter product selectivity and is very likely the result of rapid proton shuttling offered by protic solvents. Therefore, through the choice of solvent and catalyst (Lewis acid or Brønsted acid) we can functionalize coumalic acid to yield different aromatics, dihydro-aromatics, and isophthalic intermediates. Our results have direct implications towards high-throughput and optimized manufacture of conventional compounds as well as a plethora of unique compounds inaccessible via petroleum.

References

1. Kraus et. al, Green Chem. 2011, 13, 2734−2736.
2. Kraus et. al, RSC Adv. 2013, 3, 12721−12725.
3. Lee et. al, RSC Adv. 2014, 4, 45657−45664.
4. Lee et. al, Green Chem. 2014, 16, 2111−2116.
5. Shanks and Keeling, Green Chem. 2017, 19, 3177−3185.
6. Pfennig et. al, Green Chem. 2017, 19, 4879−4888.
7. Pfennig et. al, ACS Catal. 2018, 8, 2450−2463.
8. Pfennig et. al, Submitted manuscript. Green Chem.