**3.2 Extraction of pungent capsicum oleoresin, capsaicin and capsanthin with methanol**

### **3.2.1 Model fitting**

118 Mass Transfer in Chemical Engineering Processes

(a)

(b) Fig. 3. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and

time on capsanthin in ethanolic PCO.

The liner, quadratic and interactive coefficients of the independent variables in the models and their corresponding R2 when methanol was used as extraction solvent are presented in Table 2.


Subscripts: 1 = temperature (°C); 2 = time (min);

\*Significant at 0.05 level; \*\*Significant at 0.0l level; \*\*\*Significant at 0.001 level.

Table 2. Regression coefficients, *R2*, adjusted R2 and *p* for three dependent variables for pungent capsicum oleoresin obtained by methanol.

Table 2 clearly shows that the R2 values for these response variables are higher than 0.93 for both PCO and capsaicin, indicating that the regression models adequately explain the process. Hence, the R2 values are 0.9702 and 0.9391, respectively, for methanolic PCO yield and capsaicin. The *p* values of regression models for PCO yield and capsanthin show no lack-of-fit. However, as expected, the R2 value of capsanthin is low, (R2 = 0.7228) confirming that a high proportion of variability is not explained by the model. We therefore conclude that this regression model cannot offer a satisfactory explanation of the extraction process for capsanthin.
