Reactors, hemicellulose hydrolysis

In the present study, two mass transfer models were adapted from other applications, and preliminary comparisons were made to conventional reaction-only models to assess their abilities to describe hemicellulose hydrolysis inbatch and flowthrough reactors. Particular attention was paid to including production and diffusion of oligomers in these models with the intent of exploring whether this approach holds promise for explaining the performance of batch and flowthrough systems in a more consistent manner. 

Traditional hemicellulose hydrolysis kinetic models cannot account for a change in hemicellulose sugar yields with solids concentration and suffer from inconsistencies that bring into question their mechanistic accuracy. Thus, although current models can be usefiil for a given flow regime, their ability to describe different systems such as flowthrough reactors on a consistent basis is unproven (29). 

The C-5 sugar alcohols produced from the hydrolysis of hemicellulose are both xylitol and arabitol [6], Equivalence testing was performed with Ni/Re catalyst in the batch reactor to verily similar performance between xylitol and arabitol feedstocks. The operating conditions were 200°C and 8300kPa H2 using the procedure outlined in section Catalyst Screening section. 

Autohydrolysis enables the selective hydrolysis of hemicelluloses to a mixture mainly consisting of oligosaccharides and monosaccharides. The monosaccharide content can be increased under harsher reactor conditions, but then monosaccharides can undergo decomposition reactions, thereby increasing the content of potential fermentation inhibitors in hydrolysates. 

A brief overview of the reaction kinetics concerning the acid hydrolysis of cellulose and hemicellulose is presented here for subsequent discussion on reactor design and operation. 

Although percolation reactors have been in use extensively over several decades, it was not until 1983 that the first theoretical model of this type of reactor was introduced [5]. The model was developed for sequential first-order reactions in order to assess the performance in hydrolysis of hemicellulose. As an unsteady reactor, the model involves a partial differential equation with the following parameters kinetic parameter a = k2/kj operational parameter (3 = kiL/u, T = ut/L, where L is the bed length and u is the liquid flow velocity. 

As described in the previous section. Brown et al. obtained an aromatic yield of 1% in the catalytic fast pyrolysis of milled wood lignin with ZSM-5 at 600°C, which was considerable lower than the yields obtained from cellulose and hemicellulose [286]. Also, the coke yield from lignin (59%) was much higher. Huber et al. used the lignin residue obtained after H SO hydrolysis of maple wood as a substrate for pyrolysis in a fluidized-bed reactor with spray-dried ZSM-5 at 600°C [298]. Next to lignin, this substrate also contains humic residue, a compound also composed of phenolic entities. An aromatic and olefin yield of only 2% and 1% was obtained, compared to respective yields of 15% and 7% from pure maple wood. The coke yield from the solid residue (69%) was also much higher than the coke yield from maple wood (32%). 

---