Search

A novel near-infrared spectroscopy (NIRS) method has been researched and developed for the simultaneous analyses of the chemical components and associated properties of mint ( Mentha haplocalyx Briq.) tea samples. The common analytes were: total polysaccharide content, total flavonoid content, total phenolic content, and total antioxidant activity. To resolve the NIRS data matrix for such analyses, least squares support vector machines was found to be the best chemometrics method for prediction, although it was closely followed by the radial basis function/partial least squares model. Interestingly, the commonly used partial least squares was unsatisfactory in this case. Additionally, principal component analysis and hierarchical cluster analysis were able to distinguish the mint samples according to their four geographical provinces of origin, and this was further facilitated with the use of the chemometrics classification methods—K-nearest neighbors, linear discriminant analysis, and partial least squares discriminant analysis. In general, given the potential savings with sampling and analysis time as well as with the costs of special analytical reagents required for the standard individual methods, NIRS offered a very attractive alternative for the simultaneous analysis of mint samples.

An analytical method for the classification of complex real-world samples was researched and developed with the use of excitation-emission fluorescence matrix (EEFM) spectroscopy, using the medicinal herbs, Rhizoma corydalis decumbentis (RCD) and Rhizoma corydalis (RC) as example samples. The data set was obtained from various authentic RCD-A and RC-A, adulterated AD, and commercial RCD-C and RC-C samples. The spectra (range: λex=215∼395 nm and λem=290∼560 nm), arranged in two- and three-way data matrix formats, were processed using principal component analysis (PCA) and parallel factor analysis (PARAFAC) to produce two-dimensional component-by-component plots for qualitative data classification. The RCD-A and RC-A object groups were clearly discriminated, but the AD and the RCD-C as well as RC-C samples were less well separated. PARAFAC analysis produced somewhat better discrimination, and loadings plots revealed the presence of the marker compound Protopine—a strongly fluorescing substance—as well as at least two other unidentified fluorescent components. Classification performance of the common K-nearest neighbors (KNN) and linear discrimination analysis (LDA) methods was relatively poor when compared with that of the back propagation- and radial basis function-artificial neural networks (BP-ANN and RBF-ANN) models on the basis of two- and three-way formatted data. The best results were obtained with the three-way fingerprints and the RBF-ANN model. Subsequently, the quality of the commercial samples (RCD-C and RC-C) was classified on the best optimized RBF-ANN model. Thus, EEFM spectroscopy, which provides three-way measured data, is potentially a powerful analytical technique for the analysis of complex real-world substances provided the classification is performed by the RBF-ANN or similar ANN methods.

Raw and roasted Semen Cassiae seeds, a complex traditional Chinese medicine (TCM), are used as examples to research and develop a method of classification analysis based on measurements of Fourier transform infrared (FT-IR) spectral fingerprints. Eighty samples of the TCM were measured in the mid-infrared range, 400–2000 cm−1(KBr pellets), and the complex overlapping spectra were submitted for interpretation to a principal component analysis least squares support vector machine (PC-LS-SVM), kernel principal component analysis least squares support vector machine (KPC-LS-SVM), and radial basis function artificial neural networks (RBF-ANN). The LS-SVM models were developed with an RBF kernel function and a grid search technique. Training models were constructed with the use of raw and first-derivative spectra and these were then verified by another data set containing both raw and roasted spectral objects. It was demonstrated that the first-derivative data set produced the best separation of the spectral objects. In general, satisfactory analytical performance was obtained with the PC-LS-SVM, KPC-LS-SVM, and RBF-ANN training models and with the classification of the verification spectral objects. With regard to chemometrics modeling, the performance of KPC-LS-SVM was somewhat more economical than that of the PC-LS-SVM model. It would appear that the latter relatively simple model would be sufficient for application to most small to medium sized FT-IR fingerprint data sets, but with larger matrices the more complex models, such as the RBF-ANN and KPC-LS-SVM, may be more advantageous on a computational basis.

This paper demonstrates how inferential measurements or indirect methods using near-infrared (NIR) methodology and chemometrics can be used to predict sugarcane clonal performance. Fiji leaf gall resistance is used in this study as an example. Fiji leaf gall is one of Australia's most serious sugarcane diseases, representing a significant problem in almost half of the total area under production. Traditional rating of sugarcane clones for resistance/susceptibility is difficult and expensive because of the nature of field-based methods and variable infection levels of the trials. Thus, the aim of this work was to investigate the potential of NIR spectroscopy as an alternative means to rate clones from direct measurement of sugarcane leaf spectra and to examine its ability to successfully predict traditional resistance ratings using a calibration model based on a chemometrics method such as partial least squares (PLS). A scanning electron microscopy (SEM) study of the leaf substrate was undertaken to elucidate the nature of the NIR sample site. In addition, an NIR study of freeze-dried sugarcane leaf samples resolved the heavily overlapping O–H bands present in the NIR spectrum due to water/cellulose interaction. A significant decrease in the spectral intensity between 5205 and 5393 cm−1 was observed and a similar decrease was noted in the OH stretching overtone (7114 cm−1) with an accompanying shift to lower wavenumbers. PLS modeling based on traditional ratings as the dependent variable and the corresponding NIR spectra showed satisfactory results with standard error of validation (SEV) and standard error of prediction (SEP) values being 0.98 ( R2 = 0.97) and 1.20 ( R2 = 0.88), respectively. This methodology has now been recommended for more extensive field trials.

The Chinese yam ( Dioscorea opposita) is a basic food in Asia and especially China. Consequently, an uncomplicated, reliable method should be available for the analysis of the quality and origin of the yams. Thus, near-infrared (NIR) and mid-infrared (mid-IR) spectroscopic methods were developed to discriminate among Chinese yam samples collected from four geographical regions. The yam samples were analyzed also for total sugar, polysaccharides, and flavonoids. These three analytes were used to compare the performance of the analytical methods. Overlapping spectra were resolved using chemometrics methods. Such spectra were compared qualitatively using principal component analysis (PCA) and quantitatively using partial least squares (PLS) and least squares-support vector machine (LS-SVM) models. We discriminated among the four sets of yam data using PCA, and the NIR data performed somewhat better than the mid-IR data. We constructed the PLS and LS-SVM calibration models for the prediction of the three key variables, and the LS-SVM model produced better results. Also, the NIR prediction model produced better outcomes than the mid-IR prediction model. Thus, both infrared (IR) techniques performed well for the analysis of the three key analytes, and the samples were qualitatively discriminated according to their provinces of origin. Both techniques may be recommended for the analysis of Chinese yams, although the NIR technique would be preferred.

Studies of kaolinite surfaces are of industrial importance. One useful method for studying the changes in kaolinite surface properties is to apply chemometric analyses to the kaolinite surface infrared spectra. A comparison is made between the mechanochemical activation of Kiralyhegy kaolinites with significant amounts of natural quartz and the mechanochemical activation of Zettlitz kaolinite with added quartz. Diffuse reflectance infrared Fourier transform (DRIFT) spectra were analyzed using principal component analysis (PCA) and multi-criteria decision making (MCDM) methods, the preference ranking organization method for enrichment evaluations (PROMETHEE) and geometrical analysis for interactive assistance (GAIA). The clear discrimination of the Kiralyhegy spectral objects on the two PC scores plots (400–800 and 800–2030 cm−1) indicated the dominance of quartz. Importantly, no ordering of any spectral objects appeared to be related to grinding time in the PC plots of these spectral regions. Thus, neither the kaolinite nor the quartz are systematically responsive to grinding time according to the spectral criteria investigated. The third spectral region (2600–3800 cm−1, OH vibrations), showed apparent systematic ordering of the Kiralyhegy and, to a lesser extent, Zettlitz spectral objects with grinding time. This was attributed to the effect of the natural quartz on the delamination of kaolinite and the accompanying phenomena (i.e., formation of kaolinite spheres and water). The mechanochemical activation of kaolinite and quartz, through dry grinding, results in changes to the surface structure. Different grinding times were adopted to study the rate of destruction of the kaolinite and quartz structures. This relationship (i.e., grinding time) was classified using PROMETHEE and GAIA methodology.

Two sets of cotton fabrics—one consisting of different types of fabric and the other of a fabric sampled sequentially during its normal processing stages—were investigated by DRIFT spectroscopy. The band assignments were consistent with previous work but, in general, the spectra could not be differentiated on visual examination. However, they were successfully qualitatively and quantitatively distinguished with the use of chemometrics; a chemical basis for fabric clustering was also indicated. This approach is a significant improvement on previously reported work with KBr disk IR sampling.

DRIFT spectra were collected from dye mixtures extracted from small yarn samples taken from several areas of a worn test garment. Raw and pretreated data matrices were interpreted with the use of chemometrics (PCA, SIMCA, fuzzy clustering). This approach showed that spectra cluster and match according to their sampling area on the test textile. The DRIFT results provide detailed information that is not available from the application of the usual techniques (e.g., TLC) with such samples and therefore should be of use in forensics or similar work where comparisons of unknown, small dye mixtures are required.

A rapid NIR method for the determination of octane number of reformate samples of Australian refined motor gasoline is described. Results of predictions using calibration sets based on ASTM full blends and actual reformate samples are presented. Interpretation of spectral data using principal component analysis facilitates the selection of a suitable analytical spectral range and demonstrates that the reformate samples may be distinguished on the basis of the geographical origins of the parent crude oils. The described method forms a basis for practical application and is a possible alternative to the standard CFR engine test method.