(741b) Reaction Condition Optimization of Sorbitol Simultaneous Hydrodeoxygenation over a ReOx-Pd/CeO2 Catalyst Via Design of Experiments | AIChE

(741b) Reaction Condition Optimization of Sorbitol Simultaneous Hydrodeoxygenation over a ReOx-Pd/CeO2 Catalyst Via Design of Experiments

Authors 

MacQueen, B. - Presenter, University of South Carolina
Lauterbach, J., University of South Carolina
Lignocellulosic biomass is rich in sugars that contain hydroxyl groups. These sugars can be upgraded to shorter chain sugar alcohols, which can undergo a simultaneous hydrodeoxygenation (S-HDO) to remove vicinal hydroxyl groups. S-HDO is a two-step process in which a deoxydehydration occurs to form a double bond immediately followed by a hydrogenation step1. This S-HDO removes two hydroxyl groups and leaves a single bond while having the capacity to remove multiple vicinal hydroxyl groups over time under reaction conditions2. ReOx-Pd/CeO2 has been shown to be a superior catalyst for S-HDO in literature1. Our previous work studied the pressure, temperature, and catalyst loading effects on the S-HDO of 1,4-anhydroerythritol and xylitol utilizing an L9 Taguchi design. Through the Taguchi design optimal conditions for these S-HDO reactions were found along with factor relations to yield.

This methodology was extended to the S-HDO of D-sorbitol. D-sorbitol is produced through a hydrogenation process of cellulose3. D-sorbitol can be chemically upgraded to various products including 1,2-, 2,3-, and 3,4-dideoxyhexitol, 1,2-, 1,4- and 1,6-hexanediol, and hexane using a S-HDO. Sorbitol S-HDO was previously reported in literature at 160°C and 80bar H2 with a 72 hour reaction time while utilizing a 2wt% ReOx-Pd/CeO2 catalyst1. However, the temperature, pressure, and catalyst loading effects on conversion and selectivity of sorbitol S-HDO are unknown. To investigate these factor effects, a Taguchi design and a response surface design of experiments were implemented to investigate the sorbitol S-HDO. The Taguchi design offers a significant reduction in the necessary experimental runs but only gives linear correlations. The response surface gives higher order terms to model the design space and gives insight into cofactor interactions but requires more experimental runs. The two designs were compared to see if the general conclusions were the same and to elucidate optimal and milder reaction conditions.

References

(1) Ota, N.; Tamura, M.; Nakagawa, Y.; Okumura, K.; Tomishige, K. Hydrodeoxygenation of Vicinal OH Groups over Heterogeneous Rhenium Catalyst Promoted by Palladium and Ceria Support. Angew. Chemie - Int. Ed. 2015, 54 (6), 1897–1900.

(2) Ota, N.; Tamura, M.; Nakagawa, Y.; Okumura, K.; Tomishige, K. Performance, Structure, and Mechanism of ReO X –Pd/CeO 2 Catalyst for Simultaneous Removal of Vicinal OH Groups with H 2. ACS Catal. 2016, 6 (5), 3213–3226.

(3) Ding, L.; Wang, A.; Zheng, M.; Zhang, T. Selective Transformation of Cellulose into Sorbitol by Using a Bifunctional Nickel Phosphide Catalyst. 2010, No. entry 3, 818–821.