Duku (
Food processing such as thermal and nonthermal processes could affect changes in structure and composition of the food (
The use of infrared has been carried out in previous research for drying the skin of fresh duku. The design of this research was a multivariate process involving many factors that affected the efficiency and the effectiveness of drying process on duku fruit. Previously, classical method was used to determine the optimum condition by using one parameter and it was timeconsuming, laborious and the result mostly could not be guaranteed (
The use of the RSM method has been widely used in food processing in optimizing various processes such as extraction, drying, blanching, enzymatic hydrolysis and clarification, production of microbial metabolites. RSM also has been used in the formulation of food products such as the optimization of food drying process including spray drying of guava powder (
The hypothesis of this research was that the response surface methodology (RSM) could optimize the drying process of the skin of the whole of duku fruit.
The material used in this research was the local exotic fruit
The fruit was exposed using infrared radiation which has two emitters (245 mm x 60 mm size; 1000w for each). The emitter was adjusted for 6 cm and 10 cm (X_{1}) and the temperature of infrared radiation was turned into 200 °C, 300 °C and 400 °C (X_{2}), while the temperature of emitter was arranged to 50, 60, 70 and 80 seconds according to the exposure time (X_{3}). After exposing time, the fruits were stored in controlled temperature (15 °C ±2 °C) and then the chemical and physical effects of infrared were analysed continuously (every two days). The chemical and physical properties have been discussed in previous research (
A Box and Behnken design (BBD) on Design Expert Program (DEP)TM Version 11 (StatEase, Inc., Minneapolis, US) with three variables was used in this research to obtain the optimum IR heating to maximize the process. The independent variables in this design were the Infrared emitter distance (X_{1}, cm), the temperature of infrared emitter (X_{2}, °C), exposure time (X_{3}, s), storage time (X_{4}, days), while the response variables were weight loss (Y_{1}, g), fruit firmness (Y_{2}, N), total soluble solid (Y_{3}, %), titratable acidity (Y_{4}, °brix), and browning index (Y_{5}). The preliminary single factor test was used to determine the range for each variable. To analyse the response pattern and to establish the models for this research, seventeen experiments was conducted randomly. A secondorder polynominal and regression coefficients were calculated for the experimental data. The quadratic equation to predict the optimum point of this research was explained as follows (Eq.
where Y is the dependent variable; a_{o}, a_{i}, a_{ii} and a_{ij} are coefficients estimated by the model, X_{i}, X_{j} are levels of the independent variables. They represent the linear, quadratic and cross product effects of the X_{1}, X_{2}, X_{3}, and X_{4} factors on the response, respectively.
The point of optimization in this research was employed in order to optimize the independent variables to maximize the result of the response variables. The goals of the response variables were chosen from the sensory test of consumers (Table
The goals of the response variable of duku’s quality.
Name  Goals  Lower Limit  Upper Limit 



A:Temperature  in the range of  200  400 
B:Exposure Time  in the range of  50  80 
C:Storage Time  in the range of  1  25 
Weight loss  minimize  6.97  19.41 
Texture  maximize  19.26  56.25 
Titratable Soluble Solid  maximize  8.9  20.2 
Total Acidity  minimize  0.35  0.67 
Index Browning  minimize  0.06  0.26 
The statistical analysis of the optimation were analysed statistically by analysis of variance (ANOVA) for each response and significance was judge by Pvalue on model which calculated from the data by (DEP)^{™} software.
RSM is a collection of statistical and mathematical methods used to improve, develop and optimize a process with multivariable data simultaneously. The result of ANOVA is shown in Table
Result of the model design on the response variables Weight loss (Y_{1}), Fruit firmness (Y_{2}), Titratable acidity (Y_{3}), Total soluble solid (Y_{4}), and Browning index (Y_{5}).
Response Variables  



Sources  Y_{1}Weight loss (gram)  Y_{2}Fruit firmness (N)  Y_{3}Titratable acidity (TA) (%)  Y_{4}Total soluble solid (TSS) (brix)  Y_{5}Browning index (BI)  


6cm  10cm  6cm  10cm  6cm  10cm  6cm  10cm  6cm  10cm  


Model ( 
0.0002 
<0.0001 
<0.0001 
0.15  0.10  0.005 
0.0002 
0.0006 
0.0098 
0.14 
0.99  0.98  0.82  0.54  0.73  0.91  0.97  0.96  0.90  0.75  
0.99  0.97  0.78  0.27  0.38  0.80  0.93  0.90  0.76  0.43  
5.41  12.25  13.99  28.81  11.69  13.48  4.2  11.44  16.86  97.37  
8.74  103.38  751.88  2515.2  0.45  2.84  56.22  262.80  0.08  10.32  
66.35  23.91  15.53  5.67  5.35  11.24  18.48  12.45  9.35  6.35 
Note: (*) Significant effect (
A small P value shows that the data is significant (
High stability and insignificant variability of the model were implied by low value of
The determination of the optimum value of response variables on the drying process by exposing to IRE is performed by plotting the response surface of variables against the independent variables. The purposes of the optimization process in this research are selected based on the consumers and seller point of view. For the consumers, the most important sequence in choosing fruit is a visual appearance (good texture), a minimum browning colour, a relatively high sweetness (TSS) and a relatively low sour taste (TA). For the seller, in order to get high profit, the fruit with a relatively low in weight loss is more desirable.
The desirability value for the optimization process were in the range of 0.80 at an IRE distance of 6 cm and 0.92 at an IRE distance of 10 cm (Table
Optimization result of independent variables and response variables on duku with design expert program.
IRE distance (cm)  Independent variables  Response variables 
Desirability  



Temperature of IRE ( C)  Exposure time of IRE (s)  ML  FF  TSS  TA  BI  


6  400  80  2.2  40.9  17.4  0.3  0.09  0.80 
10  300  80  1.3  31.8  16.8  0.6  0.30  0.92 
Note:^{1}ML = Weight Loss (%); FF= fruit firmnes (N); TSS = Total Soluble Solid, (^{o}Brix); TA = Titratable Acidity (%); BI = Browning Index (Abs).
The response surface had an optimum point within the experimental range of the independent variables. The coordinates of the three independent variables were obtained by numerical optimization analysis.
By using the secondorder polynominal shown on equation (
Where:
Y_{1} = Weight loss,
Y_{2} = Fruit firmness,
Y_{3} = Total soluble solid,
Y_{4} = Titratable acidity, and
Y_{5} = Browning index.
The results of the above equation were derived from equation (
All the equation (
Based on the equation and Figure
Response surfaces (3D) showing the desirability at an IRE distance of 6 cm. Note: The correlation between exposure time and IRE temperature (a), storage time and IRE temperature (b), and storage time and exposure time (c), respectively.
A response surface could also shown on a 2D contour plot (Figure
Response surfaces (2D) at 6 cm distance of emitter infrared for the optimization points of exposure time, and temperature on response variables.
Based on the equation and Figure
A response surface could also shown on a 2D contour plot (Figure
It could be concluded from the optimization performed that drying duku skin a whole fruit by exposing the fruit to the infrared emitter resulted in a duku fruit with a relatively good physical and chemical conditions and still be consumable. The IRE distance of 6 cm gave a desirability value of 0.80 while the IRE distance of 10 cm gave a desirability value of 0.92 however the IRE distance of 6 cm gave a better storage time. The IRE distance of 6 cm has an optimum value of weight loss 2.2%; optimum value of fruit firmness of 40.92 N; optimum value of total soluble solid of 17.48 brix; optimum value of titratable acidity of 0.33%; and optimum value of browning index of 0.9. The fitting model base on RSM resulted from this research indicated that this study could be used as the basis for alternative process in food processing of duku but still need further research to increase the shelf life and a better result in the chemical and physical characteristics of duku.
This research was supported by the Ministry of Research, Technology and Higher Education, Indonesia through the PMDSU Project BatchII.