Page 94 - rise 2017

Table I
Configurations of samples.
Sample Configuration
1 Chicken breast
2 Chicken breast and plastic
3 Chicken breast and straw
4 Onion
5 Onion and plastic
6 Onion and straw
7 Banana
8 Banana and plastic
9 Banana and straw
10 Plastic
11 Straw

Differential scanning calorimetry
Thermal analysis method was done according to Yanty, Marikkar, Man, and Long [16]. DSC was
carried out on a Mettler Toledo differential scanning calorimeter (DSC 823 Model) equipped with a
thermal analysis data station (STARe software, Version 9.0x, Schwerzenbach, Switzerland). The
purge gas used was Nitrogen (99.999% purity) at a rate of ~20mL/min. Approximately 4-8 mg of
sample was placed in a standard DSC aluminum pan and then hermetically sealed. The reference used
was an empty, hermetically-sealed DSC aluminum pan. The oil samples were subjected to the
following temperature program: 70°C isotherm for 1 min, cooled at 5°C/min to -70°C. The samples
were held at -70°C isotherm for 1 min, and heated at 5°C/min to reach 70°C.

Statistical analysis
Principal component analysis (PCA) was carried out using Unscrambler 9.7 (Camo, USA).
Exothermic and endothermic regions of DSC were used as variables. PCA is a technique which
reduces the original data to acquire a new smaller set of data called principal components (PC) [18].
There are usually two outcomes of PCA: (i) the loading plot and (ii) the score plot. The loading plot
infers to the relationships between the variables, while the score plot indicates the sample patterns,
grouping differences and similarities [17].

Results and discussion
Principal component analysis (PCA)
PCA is a mathematically defined as an orthogonal linear transformation that transforms the data to a
new coordinate system such that the greatest variance by any projection of the data comes to lie on the
first coordinate called the first principal component (PC1), the second greatest variance on the second
coordinate called the second principal component (PC2), and so on [15]. In this study, PC1 and PC2
were calculated from the twelve variables of the DSC. The explained variances of PC1 and PC2 of
chicken, onion, banana, plastic, straw and control are 86% and 14%, respectively. Both PC1 and PC2
account for 100% of variance, which explains the total holder of variance in the data matrix of the
samples. The score of PC1 (horizontal) versus the score of PC2 (vertical) was obtained in Figure 1 to
acquire the distribution chart of samples. It is observable that plastic and straw samples’ data points
appear near the center line of PC2, while control, chicken, onion and banana samples’ data points
appear far from the center line of PC2. Clearly, plastic and straw samples can be separated from other
three edible samples by PCA based on differential scanning calorimeter, due to the different material
contents of the samples.