Two water-extraction schemes (Method A and Method B) were investigated, as
described in the experimental section. Method A was different from Method B in that it
had one additional water-extraction step to ensure that all remaining unwanted polarcompounds were separated from the pyrolytic lignin. Phase separation took place, as it
usually does when the amount of water added directly to pyrolysis oil exceeds its
maximum water content, which is typically about 30–45 wt.% [11]. The aqueous phase
on top (the water-soluble fraction) was rich in polar carbohydrate-derived compounds,
while the viscous phase at the bottom (pyrolytic lignin) was dominated by less polar
lignin-derived chemicals [3]. Table 1 shows that both washing methods improved the
properties of the pyrolytic lignin by removing the light compounds, leading to lower
acidity, lower oxygen content, higher carbon content, and higher heating value than
those in the starting bio-oil. Method B produced a higher yield of pyrolytic lignin (49.0
wt%) than Method A. This indicated that the additional water extraction in Method A
effectively dissolved more light, polar compounds in the bio-oil into the aqueous, watersoluble
fraction. This result was confirmed by the chromatograms of starting bio-oil and pyrolytic lignin extracted from both methods, shown in Figure 2 and Table 2. Most
light-compound peaks, especially peaks 1-10, disappeared or weakened. The reduction
of peak intensity was more significant in Method A. However, heavy compounds still
remained in the pyrolytic lignin (WIF) from both methods. The bigger loss of lightphase
compounds in Method A caused a higher increase in viscosity of the WIF.
However, the density, oxygen content, and heating value of the extracted pyrolytic
lignin from both methods were not much different.