The experimental design is shown in Table 1. Experiments were performed according to a 23 factorial design: two different inlet air-temperatures (Ti ¼ 150 and 170 C), two inlet flows (Ff ¼ 1.5 and 1.7 L/h) and two atomization speeds (As ¼ 23,000 and 27,500 rpm). Outlet (To) and wall temperatures (Tw) were measured but not
controlled in the SD process. A freeze-dried sample (T9) and fresh mucilage (T10) are used for comparison (see Table 1). Three replications were performed for each test for a total of 24 SD runs. Data were analyzed by ANOVA with statistical significance of P < 0.05.