School of Graduate Studies, SPS, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia.
Department of Physics, Faculty of Science, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia.
Food Res Int. 2018 Mar;105:453-460. doi: 10.1016/j.foodres.2017.11.012. Epub 2017 Dec 20.
In this study, we propose an easy approach by combining the Fourier transform infrared and attenuated total reflectance (FTIR-ATR) spectroscopy together with chemometrics analysis for rapid detection and accurate quantification of five adulterants such as fructose, glucose, sucrose, corn syrup and cane sugar in stingless bees (Heterotrigona itama) honey harvested in Malaysia. Adulterants were classified using principal component analysis and soft independent modeling class analogy, where the first derivative of the spectra in the wavenumber range of 1180-750cm was utilized. The protocol could satisfactorily discriminate the stingless bees honey samples that were adulterated with the concentrations of corn syrup above 8% (w/w) and cane sugar over 2% (w/w). Feasibility of integrating FTIR-ATR with chemometrics for precise quantification of the five adulterants was affirmed using partial least square regression (PLSR) analysis. The study found that optimal PLSR analysis achieved standard error of calibrations and standard error of predictions within an acceptable range of 0.686-1.087% and 0.581-1.489%, respectively, indicating good predictive capability. Hence, the method developed here for detecting and quantifying adulteration in H. itama honey samples is accurate and rapid, requiring only 7-8min to complete as compared to 3h for the standard method, AOAC method 998.12.
在这项研究中,我们提出了一种简单的方法,将傅里叶变换红外和衰减全反射(FTIR-ATR)光谱与化学计量学分析相结合,用于快速检测和准确量化马来西亚采集的无刺蜜蜂(Heterotrigona itama)蜂蜜中的五种掺杂物,如果糖、葡萄糖、蔗糖、玉米糖浆和甘蔗糖。使用主成分分析和软独立建模分类分析对掺杂物进行分类,其中利用了光谱在波数范围 1180-750cm 的一阶导数。该方案可以满意地区分掺有浓度高于 8%(w/w)玉米糖浆和 2%(w/w)以上甘蔗糖的无刺蜜蜂蜂蜜样品。通过偏最小二乘回归(PLSR)分析证实了 FTIR-ATR 与化学计量学相结合用于精确量化这五种掺杂物的可行性。研究发现,最佳 PLSR 分析在可接受的范围内实现了校准和预测标准误差,分别为 0.686-1.087%和 0.581-1.489%,表明具有良好的预测能力。因此,与标准方法(AOAC 方法 998.12)相比,该方法用于检测和量化 H. itama 蜂蜜样品中的掺假是准确和快速的,仅需 7-8 分钟即可完成,而标准方法需要 3 小时。