Figure 4 compares the FTIR spectra of the purified PHB granules and the original oven-dried PHB-containing cell mass. The peaks of amide I band at ~1650 cm-1 and amide II band at ~1540 cm-1 (N-H bend) are characteristic infrared radiation absorption of the proteins in cell mass [25,26]. They disappeared in the spectrum of purified PHB granules. This was confirmed with a pure PHB prepared with solvent extraction (the spectrum not shown here). It was also noticed that the native amorphous PHB granules became crystallized during the process of purification, which will be further discussed. This structural change in PHB matrix was also reflected in the absorption of infrared radiation at wave numbers of 1180, 1210, and 1280 nm-1 [27].

The purified PHB granules were subjected to repeated heating and cooling in a differential scanning calorimeter (DSC), and the results are presented in Figure 5. In the first heating (solid blue line), two melting peaks were observed, around 156 oC and 167 oC, respectively, indicating that the PHB powder had two types of crystalline structures. The melted polymer was re-crystallized again during the first cooling (dotted blue line), starting at 100 oC and ending at 80 oC. When the crystallized PHB was subjected to the second heating (solid black line), it was noticed that the relatively small melting peak at 156 oC (or crystalline structure) observed in the first heating disappeared. Only one melting peak (crystalline structure) was observed, starting at 160 oC and ending at 181 oC with a peak around 174 oC. This phenomenon indicates that the first melting peak in the first heating represents a type of crystalline structure that could not be formed at a cooling rate of 5 oC/min. The whole endothermic event in the second heating absorbed 87J/g PHB. Based on a theoretical melting enthalpy of 100% crystalline PHB (146 J/g) [28], it can be estimated that about 60% of PHB matrix was crystallized during the first cooling (Eq. 1).

Where Xc is the PHB crystallinity, AHm and AHt are the melting enthalpies of PHB powder and a theoretical PHB crystal [28].

In contrast to the thermal plastic behavior of PHB powders, the oven-dried cell mass containing 73% PHB could not be melted till carbonization. This fact reveals a complicated interaction between the biopolyester and the residual biomass. It also shows that a composite of 73% PHB and 27% cellular mass is not a thermoplastic material, but a rigid composite. The role of cellular mass in the PHB composite is not clear yet.