Optimization and comparison of chemical and electrochemical hydride generation for optical emission spectrometric determination of arsenic and antimony using a novel miniaturized microwave induced argon plasma exiting the microstrip wafer

Continuous flow (CF) chemical hydride generation (CHG) and electrochemical hydride generation (ECHG) directly coupled to a novel 40 W, atmospheric pressure, 2.45 GHz microwave microstrip Ar plasma exiting a microstrip, wafer has been developed for the emission spectrometric determination of As and Sb using a miniaturized optical fiber spectrometer and a CCD-array detector. The experimental conditions for both procedures were optimized with respect to the relative net intensities of the As 1228.8 nm and Sb I 252.8 nm lines and their signal-to-background intensity ratios. Additionally, the susceptibility to interferences from Cd, Co, Cr, Cu, Fe, Ni, Pb and Zn and other hydride-forming elements in the determination of As and Sb using the CHG and ECHG techniques was investigated in detail. Under the optimized conditions, it was found that ECHG is more prone to interferences compared to CHG. The detection limits (3 sigma) of As (6 ng mL(-1)) and Sb (7 ng mL(-1)) obtained for the ECHG-MSP-OES method are about three times lower than in the case of the CHG-MSP-OES method due to a two-fold lower amount of H-2 introduced into the MSP in case of the ECHG, resulting in a better plasma stability and reduced background level. The linearity ranges for both calibration curves to a concentration of up to 5 mu g mL(-1) and a precision between 2% and 7% (2 mu g mL(-1) and 0.050 mu gmL(-1) of As and Sb, respectively) were found for both methods. The developed ECHG-MSP-OES method was validated for As through the analysis of a certified coal fly ash standard reference material (NIST SRM 1633a) after sample dissolution. The derived concentration (140 +/- 8 mu g g(-1)) was found to agree well with the certified data (145 +/- 15 mu g g(-1)). The method was also successfully applied to the analysis of both a galvanic bath sample, which contained Sb and was spiked with As, and a tap water sample spiked with both analytes. Recovery rates of 99-101% and a Sb concentration of 6.6 mu g mL(-1) in the galvanic bath sample were revealed. The latter value showed a good agreement with the data obtained from ICP-OES analysis, which was also used for validation purpose. (c) 2007 Elsevier B.V All rights reserved.