A statistic comparison of multi-element analysis of low atmospheric fine particles (PM2.5) using different spectroscopy techniques

Abstract

Over the past few decades, the metal elements (MEs) in atmospheric particles have aroused great attention. Some well-established techniques have been used to measure particle-bound MEs. However, each method has its own advantages and disadvantages in terms of complexity, accuracy, and specific elements of interest. In this study, the performances of inductively coupled plasma–optical emission spectrometry (ICP-OES) and total reflection X-ray fluorescence spectroscopy (TXRF) were evaluated for quality control to analyze data accuracy and precision. The statistic methods (Deming regression and significance testing) were applied for intercomparison between ICP-OES and TXRF measurements for same low-loading PM2.5 samples in Weizhou Island. The results from the replicate analysis of standard filters (SRM 2783) and field filters samples indicated that 10 MEs (K, Ca, V, Cr, Mn, Fe, Ni, Cu, Zn, and Pb) showed good accuracies and precision for both techniques. The higher accuracy tended to the higher precision in the MEs analysis process. In addition, the interlab comparisons illustrated that V and Mn all had good agreements between ICP-OES and TXRF. The measurements of K, Cu and Zn were more reliable by TXRF analysis for low-loading PM2.5. ICP-OES was more accurate for the determinations for Ca, Cr, Ni and Pb, owing to the overlapping spectral lines and low sensitivity during TXRF analysis. The measurements of Fe, influenced by low-loading PM2.5, were not able to determine which instrument could obtain more reliable results. These conclusions could provide reference information to choose suitable instrument for the determination of MEs in low-loading PM2.5 samples.