2580 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Anal. Chem. 1984, zyxwvut 56, 2580-2585 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON Determination of Rare Earth Elements by Liquid Chromatography/Inductively Coupled Plasma Atomic Emission Spectrometry zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Kazuo Yoshida and Hiroki Haraguchi* zyxwvu JIHGFE Department zyxwvutsrq DCBA of Chemistry, Faculty of Science, University of Tokyo, Bunkyo-ku, Tokyo 113, Japan Inductively coupled plasma atomic emlsslon spectrometry (ICP-AES) Interfaced wlth hlgh-performance liquld chroma- tograhy (HPLC) has been applied to the determination of rare earth elements. ICP-AES was used as an element-selective detector for HPLC. The separatlon of rare earth elements wlth HPLC helped to avold erroneous analytlcal results due to spectral Interferences. Fiiteen rare earth elements (Y and 14 ianthanldes) were determlned selectively with the HPLCIICP-AES system using a concentratlon gradient me- thod. The detectlon llmits wlth the present HPLCIICP-AES system were about 0.001-0.3 MgImL wlth a 1OO-hL sample Injectlon. The callbratlon curves obtalned by the peak helght measurements showed llnear reiatlonshlps In the concentra- tlon range below 500 pg/mL for all rare earth elements. A USGS rock standard sample, rare earth ores, and hlgh-purity ianthanlde reagents (>99.9 % ) were successfully analyzed wlthout spectral Interferences. Recently rare earth elements have received much attention in the fields of geochemistry and industry (1-3). Rapid and accurate determinations of them are increasingly required as industrial demands expand. However, the determination of rare earth elements at the trace level is still very difficult because of interelement interferences or lack of sensitivities, when instrumental analytical techniques are directly used for analysis without prior sample treatment. Serious interference problems are encountered when using direct determination instrumental techniques, especially when the samples contain large amounts of several rare earth elements at different concentration levels. Isotope dilution mass spectrometry (3,4) and neutron ac- tivation analysis (5-7) are commonly used for the determi- nation of rare earth elements. Isotope dilution mass spec- trometry provides fairly high sensitivities for rare earth ele- ments. However, the selectivity of mass spectrometry is not sufficient to analyze the samples directly. Therefore some separation is required before measurement. Such techniques for separation are sequential precipitation, solvent extraction, and chromatographic methods. The isotope dilution mass spectrometric technique has another substantial problem, that is, Pr, Tb, Ho, and T m cannot be detected because these elements do not have plural stable isotopes. As for neutron activation analysis, several rare earth ele- menta can be determined with high sensitivities. The neutron activation technique, however, suffers from the following problems: high cost and cumbersome instrumentation, slow analysis time, poor precision, and interelement interferences. In the determination of rare earth elements by neutron ac- tivation analysis, separation techniques similar to those em- ployed in isotope dilution mass spectrometry are also com- monly performed because of the interferences with other coexisting elements in analysis. Recently inductively coupled plasma atomic emission spectrometry (ICP-AES) has been extensively applied to the determination of metallic elements (8-10). As is well known, ICP-AES has advantages such as low detection limits, good precision and/or accuracy, wide dynamic ranges of calibration curves, and capability of simultaneous multielement analysis. In the determination of rare earth elements, ICP-AES provid extremely higher sensitivities (1 1-13) than atomic absorption spectrometry (14), flame emission spectrometry (8), and at- omic fluorescence spectrometry (13). In the determination with ICP-AES, however, a number of emission lines are ex- cited, compared to other chemical flames, and so spectral interferences due to rare earth elements themselves are pron to be serious problems (11, 12). Crock et al. (13) corrected the spectral interferences from other coexisting rare earth elements in addition to selecting the analytical wavelengths. They separated rare earth elements from other elements with ion exchange techniques using cation and anion exchange resins and determined them in geological materials which we contained at almost similar concentration levels. The ap- plications of ICP-AES to the analysis of such samples that contain very different concentrations of rare earth elements are remarkably difficult even when the ion exchange method is used for separation. This is because the spectral inter- ferences are too large to make exact experimental correction Therefore, development of a combined system of ICP-AES with separation techniques for rare earth elements is desired for the rapid and accurate determination. In recent years, ICP-AES interfaced with liquid chroma- togrpahy has been investigated because of the advantages of both instrumental techniques (16-24). However, the analytical applications to those samples which require relatively high salt concentration of mobile phase for column separation of chemical species are still limited. This is because the large content of salts in the mobile phase causes clogging of the ICP-AES torch, which prevents sample introduction into the plasma. In the present experiment, ICP-AES interfaced with high-performance liquid chromatography (HPLC) is applied to the determination of rare earth elements. The problem caused by direct and long-term introduction of the high salt solution for the concentration gradient separation of rare ear elements has been overcome by using the modified plasma torch and adjusting the suitable torch position. Consequently rare earth elements in a USGS rock standard sample, rare earth ores, and high-purity lanthanide reagents (>99.9%) have been determined withoutspectral interferences by the HPLC/ICP-AES system. EXPERIMENTAL SECTION Apparatus. The HPLC system consisted of two solvent de- livery pumps, an injection valve (Model 7125 from Rheodyne Co., Cotati, CA), and an associated column. The solvent flow rates from pump A and pump B were controlled with a gradient pro- grammer (Model GRE-3A from Shimadzu Co., Japan). The separation was performed by using a 4 mm id. X 250 mm long stainless steel column packed with a strong cation exchange (IEX-21OSC from Toyo Soda Co., Japan). The column tem- perature was maintained at 50 O C with a column oven (Model CTO-SA from Shimadzu Co., Japan). The sample volume injected 0003-2700/84/0356-2580$01.50/0 0 1984 American Chemical Society