Selective dehydration of polyols over solid acids and metal-acid bifunctional catalysts: Towards a catalytic toolbox for rational catalyst design
Interest in biorenewable alternatives to existing petrochemicals has lead to increased interest in catalytic means of upgrading sugar-derived molecules into useful commodities. While some biorenewable chemicals aim to break into the existing market as a new product (polylactic acid, or PLA, is one example), there is also interest in developing catalytic routes to existing commodities. In order to do this, a "catalytic toolbox" will need to be developed, which can enable a chemical engineer to rationally design a catalytic upgrading pathway from a given starting molecule to a desired end product. While the petrochemical industry dealt primarily in carbon and hydrogen, biorenewables deals in carbon, hydrogen, and oxygen, and requires a new suite of catalysts to handle selective oxygen removal.
Polyols are a particular class of compounds which consist of a carbon backbone and multiple hydroxyl groups. Most polyols are derived from biorenewable sources, ranging from glycerol, which comes from triglycerides, up to 6-carbon sugar alcoohols. Selective removal of these hydroxyl groups is desired, either in the form of eliminating undesired hydroxyls in order to end at a linear molecule (such as 1,6-hexanediol), or in the form of selectively forming a ring structure by converting to hydroxyl groups on a polyol to a cyclic ether (such as the conversion of sorbitol to isosorbide). Both types of reactions are called dehydration reactions, as they lead to the removal of water. Since most polyols can undergo either reaction, and since either reaction may be desired for a given polyol in order to reach different products from the same reagent, an understanding of selective dehydration in either direction is necessary.
The current work focuses on discovering the catalytic properties needed to selectively dehydrate triols (polyols with three hydroxyl groups) into linear or ring dioxygenates. The major variables investigated in the current work include the dehydration reaction conditions, catalyst acid stength, catalyst deactivation, and the modification of acid catalysts with metals. From these tests, a tradeoff between ring/linear selectivity and selectivity to dioxygenates and mono-oxygenates was found, as pyran-selective systems tended to remain as dioxygenates, while linear-selective systems were more likely to continue dehydrating to mono-oxygenates. However, there were cases found in which 1,2,6-hexanetriol was up to 50% selective toward linear products instead of a more typical 25% selectivity, with little mono-oxygenate production. In these cases, the key was to modify a given catalyst's acid site distribution, either by deactivating the catalyst or by modifying the catalyst with added metals.