In Fig. 5 we reproduce one of the release profiles for MET in a NRL
matrix from our previous work [22], which points to saturation at
approximately 110 h. Upon analyzing the profiles, we concluded that
the first, fast step of burst release corresponded to the MET near or on
the surface of the NRL membrane. Therefore, the slower release
process would be associated with MET diffusing slowly through the
matrix. This conclusion was based on scanning electron microscopy
images [22], but the EDX spectra in Fig. 4 now provide direct evidence
that the drug was not present in the inner portion of the polymeric
matrix. Diffusion does not play a major role, and the release of MET
depends mainly on the amount of the encapsulated material (as a
reservoir).