This behavior was also observed by Forssell et al. (1997), who suggested that a phase separation
has occurred between a starch rich phase and glycerol rich phase. This phenomenon may explain the high
standard deviation on data of elongation at break of films with 30 and 45% of glycerol. In fact, these films
presented less workability than those with 0 or 15% of glycerol due to a “sticky” character.