The Analysis Of Chemometrics On Volatile Components Of Lavandula Angustifolia Mill.Distirbuted In Yili Xinjiang

In this thesis, the Lavandula angustifolia Mill. samples (L. angustifolia Mill.)were collected from Xinjiang Yili as the research objects; some chemometric methodswere used to analyze the measured data in order to discriminate the lavender and itsessential oil. The aim of this study was to develop our own identity capability andquality control standards, and enhance competitiveness of lavender oil.Specific experiments are as follows:1. The essential oil of Lavender was extracted by microwave extraction (MWE)and the extraction technique parameters were optimized with response surfacemethodology (RSM). Volatile components from lavender were firstly extracted bysteam distillation extraction (SDE), MWE, head space-solid phase microextraction(HS-SPME), and then analyzed by gas chromatography-mass spectrometry (GC-MS).The results of three extraction methods were compared to established the characteristicdatabase of lavender.2. The chemical constituents of L. angustifolia Mill. were extracted by SDE andanalyzed by gas chromatography-mass spectrometry (GC-MS). The relative content ofthe components was determined by peak area normalization method. By using theresult of GC-MS as the original input data, principal component analysis (PCA), partialleast squares discriminant analysis (PLS-DA) method and support vector machine(SVM) were employed to identify the different varieties of lavender essential oil and toevaluate the quality of oil. The result showed it is an effective and high reliabilitymethod for establishing a classification model for the discrimination of varieties oflavender oil by combining GC-MS with PLS-DA. These results can provide atheoretical basis for the lavender essential oil from Xinjiang area to their originidentification and oil quality standard.3. HS-SPME-GC has been applied for the identification of the characteristicvolatile profile of lavender to study its varieties from Yili Xinjiang areas. Quantitative determinations of lavender samples’ common peaks were carried out by GC with flameionization detection (GC-FID) and qualitative analysis by GC with mass spectrometry(GC-MS). PLS-DA and SVM models had excellent classification sensitivity both forthe calibration set and for the test set samples. With the selective classification systems,the information of the component can be extracted from the data matrix. Thesetechniques will give reference information for quality control and for selecting the bestcultivations for lavender.4. Lavender essential oils were analyzed by using Attenuated Total ReflectanceFourier Transform Infrared spectroscopy (ATR-FTIR). PCA and orthogonal partialleast squares (OPLS-DA) were used to discriminate lavender oil cultivars usingspectral data as the original input data, and more, OPLS-DA regression model wasvalidated under the crossing-test verification. SVM models had excellent classificationsensitivity both for the calibration set and for the test set samples. The result shows thatthe model has the very good recognition capability to the lavender essential oil varietyquality. The results indicated that a quick, intuitive and feasible model wassuccessfully proposed to discriminate lavender oils by using ATR-FTIR methodcombined with OPLS-DA techniques.