Detection of Nutrient Elements and Contamination by Pesticides in Spinach and Rice Samples Using Laser-Induced Breakdown Spectroscopy (LIBS)

The laser-induced breakdown spectroscopy (LIBS) technique was applied to quantify nutrients (Mg, Ca, Na, and K) in spinach and rice and to discriminate pesticide-contaminated products in a rapid manner. Standard reference materials (spinach leaves and unpolished rice flour) were used to establish a...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of agricultural and food chemistry 2012-01, Vol.60 (3), p.718-724
Hauptverfasser: Kim, Gibaek, Kwak, Jihyun, Choi, Jeunghwan, Park, Kihong
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The laser-induced breakdown spectroscopy (LIBS) technique was applied to quantify nutrients (Mg, Ca, Na, and K) in spinach and rice and to discriminate pesticide-contaminated products in a rapid manner. Standard reference materials (spinach leaves and unpolished rice flour) were used to establish a relationship between LIBS intensity and the concentration of each element (Mg, Ca, Na, and K) (i.e., calibration line). The limits of detection (LODs) for Mg, Ca, Na, and K were found to be 29.63, 102.65, 36.36, and 44.46 mg/kg in spinach and 7.54, 1.76, 4.19, and 6.70 mg/kg in unpolished rice, respectively. Concentrations of those nutrient elements present in spinach and unpolished rice from a local market were determined by using the calibration lines and compared with those measured with ICP-OES, showing good agreement. The data also suggested that the LIBS technique with the chemometric method (PLS-DA) could be a great tool to distinguish pesticide-contaminated samples from pesticide-free samples in a rapid manner even though they have similar elemental compositions. Misclassification rates were found to be 0 and 2% for clean spinach and pesticide-contaminated spinach, respectively, by applying the PLS-DA model established from the training set of data to predict the classes of test samples.
ISSN:0021-8561
1520-5118
DOI:10.1021/jf203518f