Time-resolved fluorescent immunochromatographic assay-based on three antibody labels for the simultaneous detection of aflatoxin B1 and zearalenone in Chinese herbal medicines

ABSTRACT

A time-resolved fluorescent immunochromatographic assay (TRFICA) was successfully developed for the sensitive, simultaneous, and quantitative detection of aflatoxin B₁ (AFB1) and zearalenone (ZEN) in Chinese herbal medicines. Eu-nanospheres (EuNPs) with unique optical properties increased the stability and sensitivity of the immunochromatographic assay. To obtain stable quantitative results, we applied a three-label system in which monoclonal antibodies for AFB1 and ZEN were conjugated to the EuNPs as detection probes on the test line (T line), and EuNP-labelled chicken IgY conjugates acted as the reference on the control line (C line). The fluorescence intensities of the T and C lines were recorded, and the T/C ratio was employed as the quantitative signal for the elimination of strip variation and matrix effects. The parameters that affected the TRFICA were optimised. Under optimal conditions, the established TRFICA gave good linear ranges from 0.60 μg/kg to 3.92 μg/kg for AFB1 and from 0.40 μg/kg to 1.28 μg/kg for ZEN. The limits of detection for AFB1 and ZEN were as low as 0.60 and 0.40 μg/kg, respectively, in Chinese herbal medicines Semen coicis, Rhizoma dioscoreae, and Platycodon grandiflorus, respectively. The average recoveries of the spiked samples were 73%–95% for AFB1 and 75.83%–90% for ZEN, both with a relative standard deviation of < 9.08%. The results of 15 actual samples detected by the developed TRFICA showed a satisfactory correlation with those of ultra-performance liquid chromatography tandem mass spectrometry. Therefore, the TRFICA is a simple, rapid, and sensitive approach to quantitatively detect mycotoxins in Chinese herbal medicines.     

Plant‐oil‐based lubricants and hydraulic fluids

It is estimated that, at present, approximately 50% of all lubricants sold worldwide end up in the environment via total loss applications, volatility, spills or accidents. More than 95% of these materials are currently mineral oil based. In view of their high ecotoxicity and low biodegradability, mineral oil‐based lubricants constitute a considerable threat to the environment. In contrast, most lubricants and hydraulic fluids based on plant oils are rapidly and completely biodegradable and are of low ecotoxicity; moreover, lubricants based on plant oils display excellent tribological properties and generally have very high viscosity indices and flashpoints. However, in order to compete with mineral‐oil‐based products, some of their inherent disadvantages must be corrected, such as their sensitivity to hydrolysis and oxidative attack, and their behaviour at low temperatures. Various methods to improve the undesirable properties of native plant oils will be discussed. In parallel, government regulations that encourage or enforce the use of bio‐based fluids, at least in ecologically sensitive areas, will help to increase their market share. Using the numerous possibilities for selective breeding and/or chemical improvement of the double bond systems of natural fatty acids by increased R&D, the major obstacles regarding the use of plant‐based raw materials for the production of lubricant base fluids can be overcome and bio‐based fluids should expect a future with increasing market shares. Copyright © 2006 Society of Chemical Industry     

Evaluation of the performance of a multi-fuel domestic boiler with respect to the existing European standard and quality labels: Part-1

Emissions from a multi-fuel domestic boiler (40 kW), fired with nine different agro-biomass pellets have been compared. The pellets include apple pomace (Malus domestica), reed canary grass (Phalaris arundinacea), pectin waste from citrus shells (Citrus reticulata), sunflower husk (Helianthus annuus), peat, two types of straw pellets and two types of wood pellets. The measurements of emissions comprised carbon monoxide (CO), nitrogen oxides (NO_x), unburned hydrocarbons (C_xH_y), sulphur oxides (SO_x) and flue dust mass concentration (by DINplus and isokinetic sampling methods). Comparison of experimental emission values with relevant quality labels (Blue Angel and Swan Mark) and standard (EN-303-5) showed that the boiler satisfied the emissions requirements of Blue Angel, Swan Mark and EN-303-5 when using wood pellets-1 (except CO emission), reed canary grass and citrus pectin waste pellets as fuel at nominal load. The wood pellets-1 yielded the highest boiler efficiency of 92.4%. Dusts emission varied as a function of fine content and elemental constituent of the pellets and was the highest with sunflower husk. CO and C_xH_y emissions were maximum with peat pellets. NO_x emissions were below the concerned permissible values with all experimental pellets. Emissions of NO_x and SO_x were found maximum with straw pellets.For agro-pellets, statistical differences in ash contents were significant. High ash contents and low ash melting temperature made straw pellets less suitable for domestic applications. Reed canary grass, citrus pectin and apple pellets were the most suitable agro-pellets for small scale boilers with reasonable less ash contents and less emissions as compared to others.     

Combination of multiple diverse classifiers using belief functions for handling data with imperfect labels

This paper addresses the supervised learning in which the class memberships of training data are subject to ambiguity. This problem is tackled in the ensemble learning and the Dempster-Shafer theory of evidence frameworks. The initial labels of the training data are ignored and by utilizing the main classes’ prototypes, each training pattern is reassigned to one class or a subset of the main classes based on the level of ambiguity concerning its class label. Multilayer perceptron neural network is employed to learn the characteristics of the data with new labels and for a given test pattern its outputs are considered as basic belief assignment. Experiments with artificial and real data demonstrate that taking into account the ambiguity in labels of the learning data can provide better classification results than single and ensemble classifiers that solve the classification problem using data with initial imperfect labels.