**3.2.2 Influence of extraction temperature and time**

The influence of extraction conditions on the PCO, capsaicin and capsathin are presented by the coefficients of the proposed model. As indicated by *p* value, positive linear (*p* < 0.05) effect of time is only confirmed to be significant for PCO yield, while positive linear (*p* < 0.05) effect of temperature is noticed for capsaicin content present in methanolic PCO. Furthermore, it is found that interactive influence of both variables has the prominent positive effect (*p* < 0.001) for capsaicin content. On the other hand, a negative quadratic effect (*p* < 0.01) has been verified for both variables for capsaicin.

Fig. 4 and 5 show the response surface and contour map for PCO yield and capsaicin. It was observed that the capsaicin content rises as the temperature and time increase, but prolonged phase contact at increased temperature will not be acceptable due to the negative quadratic terms at *p* < 0.01. Generally speaking, when a higher extraction temperature was applied to the process, a higher velocity and extraction efficacy were achieved. However, some degradation processes can easily occur at high temperature, resulting in lower analyte recovery.

Extraction of Oleoresin from Pungent Red Paprika Under Different Conditions 121

(a)

(b) Fig. 5. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and

time on capsaicin in methanolic PCO.

Fig. 4. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and time on PCO yield (%) in methanol.

(a)

(b) Fig. 4. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and

time on PCO yield (%) in methanol.

Fig. 5. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and time on capsaicin in methanolic PCO.

Extraction of Oleoresin from Pungent Red Paprika Under Different Conditions 123

Consequently, Fig. 6 shows that the conditions for capsanthin extraction with methanol are

**3.3 Extraction of pungent capsicum oleoresin, capsaicin and capsanthin with hexane** 

The data obtained by these models demonstrated how the independent variables in the models influenced the extraction efficiency of the analytes of interest when using *n*-hexane. Thus, the liner, quadratic and interactive coefficients of the independent variables in the models and their

Yield (%) Capsaicin (mg/100g) Capsanthin (mg/100g)

corresponding R2 when *n*-hexane was used as extraction solvent presented in Table 3.

bo (intercept) 3.922869\* - 27.952500 - 1912.489400 b1 0.040339 3.445100\* 88.014500\*\* b2 0.007445 0.300400\* 10.158300\*\*\* b12 - 0.000234 - 0.028000 - 0.712100\* b22 - 0.000013 - 0.000600\* - 0.001500 b12 0.000105 - 0.001900 - 0.159800\*\*\* *R2* 0.9482 0.7890 0.9013 adjusted R2 0.9223 0.6836 0.8519 *p* or probability 0.0000 0.0037 0.0001

Table 3. Regression coefficients, *R2*, adjusted R2 and *p* for three dependent variables for

the model. As expected, non-polar components are present in *n*-hexane extracts.

According to the *p*-value, the models appeared to be adequate for the observed data at a 99.9% confidence level for PCO yield and capsanthin when extraction process was carried out with *n*-hexane. The R2 values, as a measure of the degree of fit, for these response variables, are higher than 0.90 where PCO and capsanthin are concerned, confirming that the regression models adequately explained the extraction process with *n*-hexane. Hence, the R2 values are 0.9482 and 0.9013, respectively, for PCO yield and capsanthin. However, the R2 value of capsaicin is low (R2=0.7890) showing lack-of fit and has the less relevant dependent variable in

The effect of extraction conditions on the PCO, capsaicin and capsathin are shown by the coefficients of the proposed model and confirmed by assessing the significance of the variables. As can be seen for capsanthin, both time (*p* < 0.001) and temperature (*p* < 0.01) are significant, being affected by the positive sign, while the interaction between temperature and time is significant (*p* < 0.001) with a negative sign. However, it is evident that negative quadratic effect (*p* < 0.05) of temperature is confirmed to be significant for capsanthin indicating that extended phase contact at increased temperature will be inappropriate. Obtained results also confirmed that *n*-hexane is the appropriate choice of solvent for capsanthin extraction. Fig. 7 and 9 show the response surface and contour map for PCO yield and capsanthin. Higher temperature and

C); 2 = time (min); \*Significant at 0.05 level; \*\*Significant at 0.01 level; \*\*\*Significant at 0.001 level.

pungent capsicum oleoresin obtained by *n*-hexane.

**3.3.2 Influence of extraction temperature and time** 

a longer phase contact decrease the capsanthin content in PCO.

unsuitable.

**3.3.1 Model fitting** 

Subscripts: 1 = temperature (°

Fig. 6. 3-D mesh plot (A) and contour plot (B) of the effects of extraction temperature and time on capsanthin in methanolic PCO.

(a)

(b) Fig. 6. 3-D mesh plot (A) and contour plot (B) of the effects of extraction temperature and

time on capsanthin in methanolic PCO.

Consequently, Fig. 6 shows that the conditions for capsanthin extraction with methanol are unsuitable.

### **3.3 Extraction of pungent capsicum oleoresin, capsaicin and capsanthin with hexane 3.3.1 Model fitting**

The data obtained by these models demonstrated how the independent variables in the models influenced the extraction efficiency of the analytes of interest when using *n*-hexane. Thus, the liner, quadratic and interactive coefficients of the independent variables in the models and their corresponding R2 when *n*-hexane was used as extraction solvent presented in Table 3.


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

\*Significant at 0.05 level; \*\*Significant at 0.01 level; \*\*\*Significant at 0.001 level.

Table 3. Regression coefficients, *R2*, adjusted R2 and *p* for three dependent variables for pungent capsicum oleoresin obtained by *n*-hexane.

According to the *p*-value, the models appeared to be adequate for the observed data at a 99.9% confidence level for PCO yield and capsanthin when extraction process was carried out with *n*-hexane. The R2 values, as a measure of the degree of fit, for these response variables, are higher than 0.90 where PCO and capsanthin are concerned, confirming that the regression models adequately explained the extraction process with *n*-hexane. Hence, the R2 values are 0.9482 and 0.9013, respectively, for PCO yield and capsanthin. However, the R2 value of capsaicin is low (R2=0.7890) showing lack-of fit and has the less relevant dependent variable in the model. As expected, non-polar components are present in *n*-hexane extracts.

### **3.3.2 Influence of extraction temperature and time**

The effect of extraction conditions on the PCO, capsaicin and capsathin are shown by the coefficients of the proposed model and confirmed by assessing the significance of the variables. As can be seen for capsanthin, both time (*p* < 0.001) and temperature (*p* < 0.01) are significant, being affected by the positive sign, while the interaction between temperature and time is significant (*p* < 0.001) with a negative sign. However, it is evident that negative quadratic effect (*p* < 0.05) of temperature is confirmed to be significant for capsanthin indicating that extended phase contact at increased temperature will be inappropriate. Obtained results also confirmed that *n*-hexane is the appropriate choice of solvent for capsanthin extraction. Fig. 7 and 9 show the response surface and contour map for PCO yield and capsanthin. Higher temperature and a longer phase contact decrease the capsanthin content in PCO.

Extraction of Oleoresin from Pungent Red Paprika Under Different Conditions 125

(a)

(b) Fig. 8. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and

time on capsaicin in *n*-hexane PCO.

Fig. 7. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and time on PCO yield (%) in *n*-hexane.

(a)

(b) Fig. 7. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and

time on PCO yield (%) in *n*-hexane.

Fig. 8. 3-D mesh plot (a) and contour plot (b) of the effects of extraction temperature and time on capsaicin in *n*-hexane PCO.

Extraction of Oleoresin from Pungent Red Paprika Under Different Conditions 127

Fig. 8 clearly shows that *n*-hexane is not the best solvent of choice for extraction of capsaicin.

RSM plays a key role in an efficient identification of the optimum values of the independent variables, under which depend variable could achieve a maximum response. In line with this, the set of optimum extraction conditions were determined by superimposing the contour plots of all the responses (Montgomery, 2001). The criteria applied for the optimization included maximum PCO yield and capsaicin in ethanol and methanol as well as maximum PCO yield and capsanhin in *n*-hexane. Data obtained from the profiles for predicted values and desirability are shown in Table 4. The desirability was calculated by simultaneous optimization of multiple responses, and ranges from low (0) to high (1). The optimum combined condition for PCO yield and capsaicin in ethanol was found to be at 68C for 165 min. When methanol is used as extraction solvent, the lower temperature for protracted time contributes to maximum PCO yield and capsaicin. Therefore, the optimum combined condition in methanol is confirmed to be at 57C for 256 min. The instability of capsanthin at increased temperature is again confirmed by optimum combined condition in

> **Independent variable**  Temperature (C) Time (min) Low limit High limit Value

PCO yield (%) 68 165 11.28 21.63 19.12

(mg/100g) 68 165 118.45 290.71 269.00

(mg/100g) 35 256 195.85 303.75 293.46

PCO yield (%) 57 256 12.38 26.23 23.73

(mg/100g) 57 256 158.04 297.82 283.10

(mg/100g) 45 165 178.93 250.71 210.65

PCO yield (%) 56 256 5.14 8.41 8.00

(mg/100g) 50 165 59.27 100.14 92.84

(mg/100g) 45 256 351.32 1554.66 1054.92

Table 4. The optimum combined condition predicted values for dependent variables at

**3.4 Optimization of extraction conditions** 

*n*-hexane at 45C for 256 min.

**variable Ethanol** 

**variable Methanol** 

**variable** *n***-Hexane** 

**Dependent** 

Capsaicin

Capsanthin

**Dependent** 

Capsaicin

Capsanthin

**Dependent** 

Capsaicin

Capsanthin

optimal values of variables.

Fig. 9. 3-D mesh plot (a) and contour plot (a) of the effects of extraction temperature and time on capsanthin in *n*-hexane PCO.

Fig. 8 clearly shows that *n*-hexane is not the best solvent of choice for extraction of capsaicin.
