The combined effects of enzyme concentration (X1), hydrolysis temperature (X2) and hydrolysis time (X3) on total sugar content (Y1), score of color (Y2) and score of appearance (Y3) are presented in Table 4.3. Table 4.4 shows the coefficients of variables in the models calculated using the least square technique and their statistical significances were judged using an ANOVA test at a significant level of 0.05. For any of the terms in the model, a large regression coefficient and a small p-value would indicate a more significant effect on the respective response variables (Khuri and Cornell, 1987). ANOVA showed that the resultant second-order polynomial model adequately represented the experimental data with the coefficient of multiple determination (R2) of 0.9501;
0.9334 and 0.9401 for the total sugar content, score of color and score of appearance, respectively.
Table 4.3. Responses of the dependent variables to the hydrolysis parameters.
Run
Technical parameters
Total sugar content (mg/ml)
Score of color
Score of appearance Enzyme
concentration (%)
Temperature (0C)
Time (min)
1 0.05 80 80 88.37 5.09 5.31
2 0.05 100 80 28.50 4.56 4.46
3 0.15 80 80 91.89 3.49 3.55
4 0.15 100 80 22.44 9.01 8.82
5 0.10 80 60 77.15 4.61 4.50
6 0.10 100 60 47.77 7.29 7.47
7 0.10 80 100 99.50 2.56 2.52
8 0.10 100 100 34.37 1.90 1.63
9 0.05 90 60 68.10 6.99 7.18
10 0.15 90 60 76.66 7.56 7.61
11 0.05 90 100 57.93 2.65 2.64
12 0.15 90 100 83.49 1.96 1.81
13 0.10 90 80 101.10 7.14 7.08
14 0.10 90 80 84.59 8.79 8.81
15 0.10 90 80 91.95 6.58 6.62
Table 4.4. Regression coefficients of the predicted second-order polynomial models for the total sugar content, score of color and score of appearance.
Total sugar content Score of color Score of appearance Regression
coefficients
F-value Regression coefficients
F-value Regression coefficients
F-value
a0 92.55 7.50 7.50
Liner term
a1 3.95 1.31 0.34 0.83 0.27 0.57
a2 -27.98 65.90*** 0.88 5.46 0.81 5.01
a3 0.70 0.041 -2.17 33.55** -2.27 39.11**
Quadratic term
a11 -13.95 7.56* -0.63 1.31 -0.59 1.24
a22 -20.80 16.80** -1.33 5.83 -1.37 6.61*
a33 -7.05 1.93 -2.08 14.20* -2.10 15.44*
Interaction term
a12 -2.39 0.24 1.51 8.13* 1.53 8.88*
a13 4.25 0.76 -0.31 0.35 -0.32 0.38
a23 -8.94 3.36 -0.83 2.48 -0.97 3.53
R2 0.9501 0.9334 0.9401
Note: *p<0.05; **p<0.01; ***p<0.001
4.2.2. Response surface analysis of total sugar content
Response surface analysis (RSA) of the data in table 4.3 demonstrated that the relationship the total sugar content and enzymatic hydrolysis parameters is quadratic with good regression coefficient (R2 = 0.9501). Eq. (2) shows the relationship between total sugar content and enzymatic hydrolysis parameters
Y1 = 92.55 + 3.95×X1 – 27.98×X2 + 0.70×X3 – 2.40×X1×X2 + 4.25×X1×X3
– 8.94×X2×X3 – 13.95×X12 - 20.08×X22 – 7.05×X32 (2)
Fig. 4.1 (A) is a response surface plot showing the effect of enzyme concentration and temperature on the total sugar content at the fixed time of 80 min. The enzyme concentration was shown as a negative quadratic effect on the total sugar content (p<0.05). The total sugar content first increased and then lightly decreased when enzyme concentration increased.
(A) (B)
Fig. 4.1. Surface plot of the total sugar content (Y1) as a function of concentration of enzyme and temperature at time of 80 min and a function
of temperature and time at concentration of 0,15% enzyme.
Hydrolysis temperature had more significant effect on the total sugar content which showed negative linear (p<0.001) and quadratic effects (p<0.01).
The total sugar content first increased and then decreased when temperature increased (Fig. 4.1(B)) which revealed that medium temperature of enzyme was favorable for hydrolyzing quinoa starch. In contrast, the interaction between variables had not any effect on the total sugar content (p>0.05). The maximal total sugar content predicted by RSA was 70.54 mg/ml under the following hydrolysis condition: the enzyme concentration of 0.15%, hydrolysis temperature of 90°C and time of 60 min.
4.2.3. Response surface analysis of score of color
The RSA in Table 4.2 also demonstrated a high regression value (R2 = 0.9334) and Eq. (3) showed the relationship between the score of color and hydrolysis parameters of enzyme concentration, temperature and time
Y2 = 7.50 + 0.34×X1 + 0.88×X2 – 2.17×X3 + 1.51×X1× X2 – 0.31×X1×X3 – 0.83×X2×X3 – 0.63×X12 – 1.33×X22 – 2.08×X32 (3)
The time showed a negative liner effect ( p < 0.01) in Fig. 4.2(B), while the interaction between enzyme concentration and hydrolysis temperature had a positive effect (p < 0.05, Table 4.4) on the score of color (Fig. 4.2(A)).
(A) (B)
Fig. 4.2. Surface plot of the score of color (Y2) as a function of concentration of enzyme and temperature at time of 80 min and as a function of
temperature and time at concentration of 0.15% enzyme.
The software suggested that the optimal hydrolysis parameters were predicted to be the enzyme concentration of 0.15%, hydrolysis temperature of 100°C and time of 64 min. The maximal score of color was 9.59.
4.2.4. Response surface analysis of score of appearance
The RSA (Table 4.4) also demonstrated a high regression value (R2 = 0.9401) for the score of appearance and Eq. (4) shows the relationship between the score of appearance and enzymatic hydrolysis parameters of enzyme concentration, hydrolysis temperature and time
Y3 = 7.50 + 0.27×X1 + 0.81×X2 – 2.27×X3 + 1.53×X1×X2 – 0.32×X1×X3 – 0.97×X2×X3 – 0.59×X12 – 1.37×X22 – 2.10×X32 (4)
The temperature and hydrolysis time both had negative quadratic effects on the score of appearance (Fig. 4.3(A)). In addition, enzyme concentration and temperature had positive linear effects on score of appearance (Fig. 4.4(B), p values are listed in Table 4.4). In term of interaction variables, the only interaction between enzyme concentration and temperature had significant effect on score of appearance (p<0.05, table 4.4). The highest score of appearance (9.65) determined by RSA was obtained at a critical value of 0.15%
(enzyme’concentration), 100°C (hydrolysis temperature) and 63.1 min (time hydrolysis).
(A) (B)
Fig. 4.3. Surface plot of the score of appearance (Y3) as a function of time and temperature at concentration of 0.15% enzyme and as a function of
temperature and enzyme concentration at time of 63.1 min.
The optimum hydrolysis conditions were achieved using the desirability function and determined to obtain highest score in color (8.47) and appearance (8.5) and the optimum sugar content (83.95 mg/ml). The optimal parameters of enzyme were 0.12 % of enzyme dose, 91.90C of temperature and 69.6 minutes of time. (Fig. 4.4)
Fig. 4.4. Contour plot showing optimal values of responses
4.2.5. Verification of predicted enzymatic hydrolysis parameters
To confirm the validity the model developed in this work, an experiment under the optimal conditions for enzymatic hydrolysis was conducted. The experimental value (Table 4.5) was 84.93 mg/ml, 8.35 and 8.29 for total sugar content, score of color and score of appearance, respectively which were approaching to values predicted by the regression models.
Table 4.5. Experimental data of verification of predicted technical parameters of enzyme
Predicted value Experimental value Y1 (total sugar content)(mg/ml) 83.99 84.93±0.43
Y2 (score of color) 8.47 8.35±0.14
Y3 (score of appearance) 8.49 8.29±0.21