Lignocellulosic biomass will be a preferred feedstock for the production of fuels and chemicals in second-generation biorefineries. One of the first steps in its chemical processing is hydrolysis, a process that can be conducted hydrothermally (without catalyst), using chemical catalysts (acids) or using enzymes. Each approach has unique advantages and disadvantages; the undesired formation of humins is a major disadvantage of the two non-enzymatic routes, because at present humins must be considered to be a low-value waste product. Surprisingly little is known about how these materials form and grow during acid-catalyzed hydrolysis or about their resulting chemical structure. This situation motivates an on-going research program within the Chemical and Biological Engineering (CBE) Department to develop catalytic hydrolysis processes where (a) the amount of humins that form is minimized and (b) those humins that do form are modified in ways that increase their value.

Cellulose is a polymer that consists of long chains of linked glucose molecules. When cellulose is exposed to a mineral acid, such as H₂SO₄, the chains are cleaved producing glucose along with a number of other minor products. The acid further catalyzes the conversion of the glucose, leading to the generation of 6-hydroxymethylfurfural (HMF). It is possible that fructose is produced as an intermediate in the conversion of glucose to HMF. The acid also catalyzes the conversion of HMF into levulinic acid. The levulinic acid is stable at reaction conditions, and so can be recovered as a product of the hydrolysis process. Levulinic acid is considered to be a platform chemical because once recovered, it can be used to produce a variety of other chemicals, fuel additives or fuels. At the same time this series of reactions is taking place to form levulinic acid, the undesirable humins are being formed. It is not known conclusively whether humins form from all of the intermediates in the process, or from just some of them. It is known for certain, though, that humins do form from HMF.

Dr. Lund's research group is working to understand the mechanism by which humins form during acid-catalyzed conversion of carbohydrates. The group is using reactor studies, computational chemistry and an array of materials characterization tools to identify the chemical reactions by which humins form and grow. Their strategy is to start at the tail-end the study of the process (acid-catalyzed conversion of HMF) and work their way backwards through the process (conversion of fructose, then conversion of glucose and so on) building mechanistic knowledge of how humins form at each stage. At the same time they are studying the humin growth mechanism, they are simultaneously seeking to discover ways to control and modify it.

In the case of HMF conversion, the group has shown that the reaction of a previously proposed intermediate, 2,5-dioxo-6-hydroxy-hexanal (DHH), with additional HMF molecules is responsible for the growth of humins. They argued that if this reaction is important for humin growth, then as a consequence one would expect to find that the furan ring and the hydroxymethyl group of HMF are also present in humins, but the carbonyl group of HMF would not be incorporated. Their infrared spectra of humins formed from HMF were consistent with their argument. On that basis, they further argued that if other aldehydes or ketones were added to the system during the acid-catalyzed conversion of HMF, then those added reagents would also be incorporated into the humins, but with the loss of their carbonyl groups. Infrared spectra again were consistent with their prediction. For example, when benzaldehyde was added during the acid-catalyzed conversion of HMF, the resulting humins contained benzene rings from the benzaldehyde, but not its carbonyl group.

Computational chemistry has proven to be a valuable complement to the group's experimental studies. Pure samples of many of the intermediates in the mechanism, such as DHH, are not readily available. As a consequence their infrared spectra are not available. The group has been able to use computational chemistry to assist in assigning infrared spectral features species such as DHH. In addition, the computational studies allow the group to assess the free energy changes associated with different reaction steps and pathways. This helps to identify the steps that are most likely to be important during humin growth.

Presently the research group is investigating the acid-catalyzed conversion of glucose and fructose. HMF is produced during both of these processes, so one focus is to determine whether there are differences in the resulting humins and to explain their origins. In separate work, they are examining the use of added aldehydes as a means of imparting specific functionality to humins and thereby increasing their value. One particular focus is the determination of whether the functionalization using aldehydes must take place simultaneously during the carbohydrate conversion or whether the humins can be recovered and functionalized in a separate process after the carbohydrate conversion is complete. The group also is pursuing a number of other exciting leads for meeting the objectives of minimizing humin formation and increasing the value of those humins that do form. Being able to do so will improve the eventual commercial viability of acid-catalyzed conversion of biomass to fuels and chemicals.