Fig. 10

Surface accessibility versus catalyst radius. Shown above is a semilog plot of accessibility versus probe radius for each of four datasets. Using surface accessibility maps, the area of biomass surface accessible to probes of at most a given radius are computed. In order to fairly compare datasets, accessible surface area is divided by the total volume of biomass in the segmented tomogram. Lower bound curves are obtained by seeding from the cell lumen, while upper bounds are obtained by seeding from the axial tomogram boundary. The resulting plots show that as probe radius increases, less biomass surface can be contacted. Clearly, the AFEX dataset shows the most accessibility to catalyst radii below 5 nm. However, for catalyst radii in the range of 8–20 nm, AFEX and DA/SE exhibit similar surface accessibility. The native and DA/ZC datasets have much less biomass surface overall as shown in Table 1, which is expected as AFEX and DA/SE appear to have more thoroughly deconstructed the cell wall. For catalysts above 20 nm radius it appears the accessible surface to volume ratio is not particularly increased by even the AFEX pretreatment. In fact, due to increased luminal surface irregularity in the pretreated samples (see Fig. 2, bottom row), the surface area to volume ratio at such high catalyst radii is decreased relative to native