Organosolv switchgrass lignin is depolymerized and hydrodeoxygenated with a formic acid hydrogen source, 20 wt % Pt/C catalyst, and ethanol solvent. The combination of formic acid and Pt/C is found to promote production of higher fractions of lower molecular weight compounds in the liquid products. After 4 h of reaction, all of the switchgrass lignin is solubilized and 21 wt % of the biomass is shown to be converted into seven prominent molecular species that are identified and quantified. Reaction time is shown to be an important variable in affecting changes in product distributions and bulk liquid product properties. At 20 h of reaction, the lignin is significantly depolymerized to form liquid products with a 76 % reduction in the weighted average molecular weight. Elemental analysis also shows that the resultant liquid products have a 50 % reduction in O/C and 10 % increase in H/C molar ratios compared to the switchgrass lignin after 20 h.

No abstract

No abstract

The deployment of in situ analytics for monitoring chemical reactions in process chemistry development and scale-up is facilitated by advanced instrumentation such as Raman spectrometry. Furthermore, greater process understanding can be engendered by coupling in situ Raman data with multivariate chemometrics analyses and kinetics modeling. Such information is important for devising science-based process control strategies along the concept of quality by design (QbD) initiated through the U.S. FDA process analytical technology (PAT) framework. A series of experiments using varied glass reactors, stirring speeds, and isothermal reaction temperatures were designed with acetic anhydride hydrolysis as the model reaction to successfully demonstrate the efficacy of combining in situ Raman spectroscopy, multivariate analyses, and kinetics modeling. Two different Raman measurement methods, using immersion and noncontact probe optics, were tested through a process Raman spectrometer with multiplexing capability. Information-theoretic multivariate chemometrics were applied to elicit pure component spectra and transient concentrations of chemical species, and two differential-algebraic equations modeling approaches were adopted for elucidating chemical and dissolution kinetics information. The variations in reactor vessel type and sizes, stirring speeds, Raman measurements, and kinetics models were compared in this study.