Principal component analysis and cluster analysis for the characterisation of marbles by capillary electrophoresis

Discrimination of Natural Mature Acacia Honey Based on Multi-Physicochemical Parameters Combined with Chemometric Analysis

Honey maturity is an important factor in evaluating the quality of honey. We established a method for the identification of natural mature acacia honey with eighteen physicochemical parameters combined with chemometric analysis. The analysis of variance showed significant differences between mature and immature acacia honey in physicochemical parameters. The principal component analysis explained 82.64% of the variance among samples, and indicated that total phenolic content, total protein content, and total sugar (glucose, fructose, sucrose) were the major variables. The cluster analysis and orthogonal partial least squares-discriminant analysis demonstrated that samples were grouped in relation to the maturity coinciding with the results of the principal component analysis. Meanwhile, the 35 test samples were classified with 100% accuracy with the method of multi-physicochemical parameters combined with chemometric analysis. All the results presented above proved the possibility of identifying mature acacia honey and immature acacia honey according to the chemometric analysis based on the multi-physicochemical parameters.

Characterization of Hatay honeys according to their multi-element analysis using ICP-OES combined with chemometrics

Chemical characterisation has been carried out on 45 honey samples collected from Hatay region of Turkey. The concentrations of 17 elements were determined by inductively coupled plasma optical emission spectrometry (ICP-OES). Ca, K, Mg and Na were the most abundant elements, with mean contents of 219.38, 446.93, 49.06 and 95.91 mg kg(-1) respectively. The trace element mean contents ranged between 0.03 and 15.07 mg kg(-1). Chemometric methods such as principal component analysis (PCA) and cluster analysis (CA) techniques were applied to classify honey according to mineral content. The first most important principal component (PC) was strongly associated with the value of Al, B, Cd and Co. CA showed eight clusters corresponding to the eight botanical origins of honey. PCA explained 75.69% of the variance with the first six PC variables. Chemometric analysis of the analytical data allowed the accurate classification of the honey samples according to origin.

A review of atmospheric fine particulate matter and its associated trace metal pollutants in Asian countries during the period 1995-2005

Many studies have monitored atmospheric particulates and gaseous phases of PM(2.5) in Asia over the past 10 years. This work also compared and discussed different sample collection, pretreatment and analytical methods in Asia countries in past decade. The results indicated that the main PM(2.5) sources are traffic exhausts. PM(2.5) concentrations are also ranked highest in the areas of traffic, followed by the urban sites, and lowest in rural sites in Asian countries. This work elucidates the sources, analytical tools, and the average concentrations for PM(2.5) and related metallic elements during 1995-2005. The results indicated that the average highest concentrations order of metallic elements for PM(2.5) were Fe>Mg>Zn, and the average concentrations of lowest metallic elements was Pb>Cu>Mn>Cr>Cd. The results also indicated that the concentration of metallic element Cu increased as the averaged concentrations of metallic element Zn and Mn increased during the past 10 years in Asian countries.

Experimental design for a basic mixture on a fluorinated packing

An optimization methodology is introduced for investigating the separation and the retention behavior of analytes on a new fluorinated reversed-phase packing. Ten basic compounds were selected as test probes to study the predictive models developed by using SPSS and MATLAB software. A two-level orthogonal array design (OAD) was used to extract significant parameters. The significant factors were optimised using a central composite design to obtain the quadratic relationship between the dependent and the independent variables. Using this strategy, response surfaces were derived as the 3D and contour plots, and mathematical models were defined for the separation. The models had a satisfactory coefficient (R(2) > 0.97, n = 16). For the test compounds, the best separation condition was: MeCN/30 mM phosphate buffer pH 7.1(55.5:44.5, v/v) and 10 basic solutes were resolved in 22 min. The significant influence of the concentration of buffer shows that different mechanisms of separation for basic compounds on the fluorinated packing exist compared with a common ODS stationary phase.