Discrimination of pure and adulterated Heterotrigona Itama honey with high fructose corn syrup using selected vibrational spectroscopy and chemometric techniques

• Main Author: Azhar, ‘Aina Qistina Zaharatul
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: QD Chemistry
• Online Access: http://eprints.utm.my/id/eprint/102173/1/AinaQistinaZaharatulAzharMFS2020.pdf.pdf; http://eprints.utm.my/id/eprint/102173/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:146361

Honey is one of the natural food products available since in the olden days and are always in demand due to its numerous nutritional value and therapeutic properties. However, increasing in environmental pollution, spreading of diseases and world climate changes have led to a dwindling in global honeybee populations. The difficulty of this issue is becoming more complex due to increased application of sugar-based adulterants which shows similar physical appearance and carbohydrate composition to pure honey. This study presents the data from Attenuated Total Reflectance- Fourier Transform Infrared (ATR-FTIR) and Raman spectroscopy coupled with Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) to differentiate pure and adulterated stingless bee honey, Heterotrigona itama (H. itama). The pure H. itama honey was obtained in Skudai, Johor and adulterated honey samples were attained with High Fructose Corn Syrup (HFCS) in five different percentages (w/w) namely 5%, 10%, 20%, 40% and 50%. All samples were firstly subjected to physical properties (density and refractive index) analysis. Welch-ANOVA results concluded that there were significant differences (p <0.05) among groups for both density and refractive index analyses. Next, characterization of the pure and adulterated H. itama were done in full region for ATR-FTIR (4000-600 cm-1) and Raman (1430-200 cm-1) spectroscopy. For chemometric analysis, fingerprint region (1700-400 cm-1) for both spectroscopic spectra were analysed. PCA was done to understand the organization of data, while LDA was done to predict the major factors for grouping and prediction. PCA scores for first three components for both ATR-FTIR and Raman were 79.1% and 72.8% respectively. Employment of LDA had improved the groupings of the honey samples when compared with PCA, which resulted in LDA cross-validation for correct classifications rate at 97.6% for ATR-FTIR while for Raman was 100%. While PCA and LDA models appear satisfactory, utilization of ATR-FTIR can be the more preferable means of analysis for discriminating the pure and adulterated honey samples, considering its simplicity and effectiveness.