Revealing cellulose pyrolysis chemistry for biofuels production

• Main Author: Mettler, Matthew S
• Format: Dissertation
• Language: English
• Subjects: chemistry, biochemistry; chemistry, organic; engineering, chemical
• ISBN: 9781267662491; 1267662492

Biomass pyrolysis is a promising technology for producing renewable fuels and chemicals from lignocellulosic feedstocks. This process utilizes moderate temperatures (400-600 °C) to depolymerize biomass to a mixture of oxygenates (i.e., bio-oil) that are liquid at room temperature. The major benefit of this "next-generation" biofuels production process over traditional biological approaches is that solid biomass is converted in only a few seconds (compared to hours or days for biological conversion). Despite the potential for pyrolysis as a future biofuels production platform, there is currently a lack of understanding of the fundamental processes (i.e., chemistry, transport phenomena).
The initial chemical reactions in pyrolysis convert solid biopolymers, such as cellulose (up to 60% of biomass), to a short-lived (less than 0.1 seconds) but high temperature liquid phase, which subsequently reacts to produce volatile products. While pyrolysis has been studied for decades, existing kinetic models lump intermediates and products by phase (e.g., vapor, solid), thereby preventing molecular-level predictions. Additionally, isothermal pyrolysis experiments capable of revealing the intrinsic kinetics have so far eluded researchers since biomass must be heated from ambient to reaction temperature in only a few milliseconds. In this work, a novel thin-film pyrolysis technique is developed which overcomes these typical experimental limitations. This technique shows that pyrolysis of the cellulose monomer (glucose) generates a very different product distribution compared to the long chain polymer (cellulose). Additional thin-film experiments reveal that the end group to monomer ratio (or degree of polymerization) is a critical descriptor in cellulose pyrolysis since short chains (degree of polymerization less than 6) generate different products than long chains (degree of polymerization greater than 100). Based on this knowledge, a small-molecule surrogate of cellulose (i.e., -cyclodextrin) is discovered which enables first-principles modeling. Ab-initio molecular dynamics simulations are then used to reveal for the first time the long-debated pathways of cellulose pyrolysis and indicate homolytic cleavage of glycosidic linkages and furan formation directly from cellulose without any small-molecule (e.g., glucose) intermediates.
After identifying how major products form from cellulose, the breakdown of these primary products within the intermediate liquid phase to form secondary compounds is investigated. A novel co-pyrolysis technique is developed and identifies secondary pyrolysis pathways where levoglucosan, the most abundant product of cellulose pyrolysis (60% of total), is deoxygenated within molten biomass to form products with higher energy content. Using isotopically labeled biomass, we co-pyrolyze levoglucosan and fructose to show that pyrans and light oxygenates are produced directly from levoglucosan. The yield of these products can be increased by a factor of six under certain reaction conditions such as high condensed-phase residence times. Finally, co-pyrolysis experiments with deuterated starting materials reveal that hydrogen exchange is a critical component of levoglucosan deoxygenation.
The final part of the thesis examines methods for influencing condensed-phase pyrolysis chemistry to improve bio-oil yield and quality (i.e., energy content, stability). The effect of temperature, sample dimensions, and heterogeneous catalysts on product yields is investigated. We find that while temperature and sample dimensions play a role in condensed-phase pyrolysis, their effect is minor compared to condensed-phase catalysts. Supported catalysts, such as Pd/Al2O3 and Pt/Al2O3, can drastically enhance decarbonylation of reactive intermediates and products (e.g., 5-hydroxymethylfurfural) which ultimately produces an 800% increase in CO yield and 100% increase in more energy dense furans (e.g., methyl furan). The increase in these low molecular weight products is concomitant with a reduction in the yield of char and char precursors (e.g., 5-hydroxymethylfurfural). In summary, this work shows that condensed-phase catalysts hold promise for reducing or eliminating char while simultaneously improving bio-oil yield and quality (in terms of energy content and stability).