Date of Award


Document Type

Open Access Dissertation


Chemical Engineering

First Advisor

John R. Regalbuto


With worldwide petroleum resources dwindling and greenhouse gas emissions rising, it is urgent to find renewable replacements for petroleum-derived products. A biomass-derived chemical with high potential as a platform intermediate, γ-valerolactone (GVL) can be readily synthesized by hydrogenation of levulinic acid (LA), itself a common biomass intermediate, using supported Ru catalysts. Overall, vapor phase hydrogenation is more energy sensitive as the higher boiling point of LA (~245°C) and requires a high energy input, comparable to liquid phase hydrogenation, which is more economical. To date the literature on many novel biomass conversion processes such as the hydrogenation of LA to GVL has focused more on the process and less on the catalyst. Therefore, many efforts have been given to develop the novel catalysts, to improve the activity, selectivity and stability for the hydrogenation of LA to GVL.

The purpose of this dissertation is to systematically study the effects of nanoparticle size, support, dopants including alkali metal, alkaline earth metal and Rhenium on ruthenium activity on the levulinic acid (LA) hydrogenation to γ- valeroactone (GVL), which could provides insight into optimizing commercial process for hydrogenation of LA to GVL by studying the structure-function relationships. The reactions were evaluated in a stainless steel EZE-Seal batch reactor with 100ml capacity. The first part of this work is to derive fundamental synthesis-structure-function relationships of Ru catalysts for LA hydrogenation using carbon and alumina supported Ru nanoparticles which have been synthesized in a rational, repeatable, scalable way. We have demonstrated that the method of strong electrostatic adsorption (SEA) yields well dispersed, homogeneously distributed Ru particles with tight particle size distributions over both types of supports. SEA synthesis of well dispersed nanoparticles results in higher activity than commercial Ru catalysts with higher Ru loadings. The dramatic, beneficial effect of potassium doping is reported for the first time. The carbon support yields higher inherent activity than alumina. Activity as a function of particle size appears to go through a maximum at about 1.5 nm for both supports and suggests hydrogenation of LA is structure sensitive on Ru particle size. Catalyst deactivation after 24 h occurs to significant extents (8 – 58%) mainly by nanoparticle sintering, but also by minor amounts of K loss in K-doped samples.

Secondly, a systematic study with alkali and alkaline earth elements was carried out to verify the origin of the promotion effect and the optimal ratio of dopant to Ru. Series of catalysts were prepared by impregnating various amounts of alkali onto 2% Ru/alumina (itself synthesized with high Ru dispersion by strong electrostatic adsorption) and were evaluated for LA hydrogenation. With the ratio of dopant to Ru constant, strong promotional effects of alkali and alkaline earth metals on the activity in terms of turnover frequency (TOF) were observed in the following order Na+

Furthermore, we report the rational synthesis of bimetallic RuRe catalysts supported on VXC-72R carbon and γ-Al2O3 comparing co-strong electrostatic adsorption (co-SEA) to a more traditional method, co-dry impregnation (co-DI). We have found that the bimetallic catalyst exhibits an optimum activity with a Ru:Re atomic ratio of 2:1. In order to establish a correlation between catalytic properties and structure, selected catalysts were characterized by TPR-H2, XRD, STEM/EDXS elemental mapping, XPS and CO-FTIR. The formation of a RuRe alloy along with segregated ReOx particles suggests that the geometric effect on activity promotion is more significant. XRD data of post reaction bimetallic catalysts indicate particle sintering and separated two metal particles for catalysts prepared by co-DI, but not for catalysts prepared by co-SEA.