Microwave-assisted sample preparation of Hungarian raw propolis in quartz vessels and element analysis by ICP-OES and ICP-MS for geographical identification.

One of the aims of this study was to improve the sample throughput of a microwave-assisted closed vessel digestion system by using small quartz vials in polytetrafluoroethylene (PTFE) vessels for the sample preparation of raw propolis samples in small amounts. The digested samples were measured by inductively coupled plasma optical emission spectrometry (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS) analyzing 36 elements. Limit of detection was low enough to measure all the elements, with the exception of La, in all raw propolis samples. There were no cross-contamination between the quartz vials, therefore independent samples can be prepared in the same PTFE vessel. Accuracy of the method was checked by spike recoveries and by analyzing BCR 189 wholemeal flour and two plant samples from a collaborative study. The means of RSDs were 5.3%-14.4% in the case of measured elements. The sample throughput was increased by three times using quartz vials in PFTE vessels besides matching with the requirements of green chemistry. Another goal was the characterization of the element content and thereby geographical identification of Hungarian raw propolis. In total, 252 samples were analyzed and their statistical characteristics were described. We cannot find globally such results of propolis element content, which is representing one country and with such a number of elements and samples. All the elements have positive skew and positive kurtosis. Concentration range is above two orders of magnitude in the case of Ba, Zn, V, Cr, Ni, Cd and Eu elements. The decimal logarithm of element concentrations was used for geographical identification of raw propolis samples originating from seven regions of Hungary by linear discriminant analysis (LDA). Grouping of the samples of the Northern Great Plain was the most effective with 96.3% and 77.8% based on the original method and the cross validation, respectively. The same indicators for all the groups are 76.6% and 61.5%.