Anal. Chem. 1989, 6 1 , 701-708
701
Comparison of Pneumatic Nebulization and Hydride Generation Inductively Coupled Plasma Mass Spectrometry for Isotopic Analysis of Selenium Morteza Janghorbani* and Bill T. G. Ting Department of Medicine, The University of Chicago, Chicago, Illinois 60637
A comparative investigation between pneumatic nebulization and continuous hydride generation as sample introduction methods for inductively coupled plasma ma88 spectrometry was carried out for isotopic analysis of selenium in blologkal samples of interest to human metabolic studies. Experimental parameters known to affect the analytkai performance of the system were evaluated: instrument operating parameters, analyte soiution/NaBH, flow rate, and NaBH, concentration. Sgnal-to-background ratb was examined for the three stable Isotopes ‘,Se, “Se, and %e. WMe background count rates for the hydride system were 3-5 times larger than those for the nebulization method, the signal-to-background ratios, normalized for Se concentration, were 30-50 times greater for the hydride system. Absolute detectkn limits (38) for the two systems were 20-60 (nebullzatkn) and 0.6-1.8 (hydride) ng of Se. Overall memory of the hydride system was evaiuated. Measurable effects were observed wlthin 400 s from swltchlng to analyte solution wlh differing isotopic composition, only H the sequence of analysis was from high to low ratio ( 1 4 % bias). However, If the sequence was from low to high ratio, precise and linear calibration plots could be obtalned over the isotope ratk range of an order of magnnude or hlgher. Whle further improvements might lead to potential enhancement of sensitivity and precision of as much as an order of magnlude, the present performance of the hydride system was satisfactory in relation to the requirements of isotopic analysis for metaboiic investigations employ~ng“56 as the in vivo stable isotope tracer.
INTRODUCTION Recent development of inductively coupled plasma mass spectrometry (ICP-MS) for application to stable isotope tracer studies has shown an exciting potential for a broad isotopic capability (1-6). This new method of isotopic analysis provides the experimental base for a large number of mineral/ trace element investigations not previously possible or improves significantly on the previously available methodology (7). An important trace element of concern to human health and disease is selenium (8),for which a concerted effort has recently been directed toward development and application of the stable isotope tracer approach (9). The existing stable isotope measurement methods for this trace element have been based on neutron activation analysis (10)or gas chromatography/mass spectrometry (11). Accurate measurement of stable isotopes of selenium with ICP-MS in matrices derived from metabolic studies constitutes a challenging methodological development. This is due to two major characteristics of the ICP-MS process: relatively low inherent ionization efficiency of Se compared to such other elements as Fe, Cu, or Zn (12), and the abundance of argonrelated ions (e.g. “Ar2+) in the spectral region of interest. 0003-2700/89/0361-0701$01.50/0
These factors, coupled with the relatively low concentrations of Se present in certain matrices of interest (e.g. blood plasma), have precluded rapid development of ICP-MS for this purpose. Powell et al. (13)have recently shown that, in comparison with pneumatic nebulization, use of a hydride generator provides a substantial enhancement in signal intensity for most hydride-forming elements in an ICP-MS system. In the present study, the analytical performance of a pneumatic nebulizer and a continuous hydride generator were compared with an otherwise identical instrument. These studies were directed toward precise and accurate measurement of stable isotopes of selenium for human metabolic investigations. An important element of these applications is the requirement for quantitative analysis of at least two stable isotopes of selenium including the least-abundant isotope, ?‘Se. When coupled with the method of in vitro stable isotope dilution, necessary for accurate quantitative isotopic analysis, a third stable isotope is also needed. EXPERIMENTAL SECTION Instrumentation. The ICP-MS instrument employed in these studies was an Elan Model 250 system (SCIEX, Thornhill, Ontario, Canada). A Meinhard concentric glass type nebulizer, TR-30C(Meinhard Associates,CA), was used. The distance from the load coil to the sampler was fixed at 27 mm. The length of the outer coolant tube of the torch (distance between the end of the auxiliary gas tube to the end of the outer coolant tube) was 36 mm. The torch was placed in fixed position as close to the sampler as possible. Argon gas was from liquified argon (industrialhigh purity, A&R Welding Supply, h i p , IL). It supplied all the argon requirements of the instrument via stainless steel and plastic tubing. Two modes of sample introduction were employed: (a) a pneumatic nebulizer (PN) and (b) a hydride generation (HG) system operating in the continuous mode (P.S. Analytical Automatic Hydride Generator, P.S. Analytical Ltd., Princeton, NJ). For the PN system, sample solution was introduced into the argon plasma via a peristaltic pump (Rabbit, Rainin Instrument Co., Inc., Woburn, MA) using approximately 150 cm of tubing (poly(vinylchloride),0.76 mm i.d., Rainin Instrument Co., Inc., Woburn, MA). The total calculatedvolume of sample introduction tubing was 0.68 mL. Solution flow rate, controlled by the peristaltic pump settings, was set at 0.96 mL/min. This system has been described in detail previously (14). For the HG system, the peristaltic pump described above was used to introduce the reagent (Na13H4)and sample solutions directly into the mixing chamber of the hydride generator, as shown in Figure 1. The hydride generator pumping and control system were not used. The output of the gas-liquid separator of the HG system was connected to the ICP spray chamber inlet (nebulizer removed), using approximately a 120 cm length of Teflon tubing (2.5mm i.d.). All tubing used between the reagent-sample containers and the gas-liquid separator consisted of flexible Teflon of small inside diameter ( alwnlnum oxlde > NaCl = carbon. I n case of aluminum oxide and Aerosil200 a strong chemisorption of the PAHs Is supposed to be the reason for the differences in the observed desorption temperatures.
INTRODUCTION Polycyclic aromatic hydrocarbons (PAHs) are one of the most important and ubiquitous environmental pollutants. They are formed by the pyrolysis of carbonaceous materials at high temperatures or during oxidant-deficient combustion processes ( I , 2). Many of them are identified as carcinogens or mutagens (3-5). Within the last years the particulate emissions from diesel engines are recognized as one source of particle-bound PAHs, which contribute to the atmospheric
pollution in specific regions (6-9). During cooling down of the gas/aerosol mixture streaming from the center of combustion to the outlet of the exhaust pipe, the gaseous PAHs are enriched by adsorption and/or condensation on the surface of particles, especially on carbon-containing particles (10). Only PAHs with a low vapor pressure, e.g. PAHs with four or more condensed rings are enriched on particle surfaces at ambient temperatures (11,12). On the other hand the four-, five-, and six-ring PAHs are expected to be the most carcinogenic or mutagenic substances (13,14). They are therefore relevant for human health, because during breathing, the PAH-contaminated particles are transported into the human lung and are deposited there. Insoluble particles like aluminosilicates which are accumulated in the bronchial or alveolar regions possess a long residence time within the lung ranging from several days up to years, as shown by Bailey and Hodgson (15). A t present it is not clear whether such a long clearance time will influence the carcinogenic activity of possibly adsorbed PAHs. Sun et al. (16) compared the deposition, retention, and biological fate of inhaled benzo[a]pyrene (BaP) as a pure aerosol and BaP adsorbed onto ultrafhe particles. Particle adsorption significantly increased the retention of the BaP in the respiratory tract. Experiments with pure BaP aerosol and BaP adsorbed onto diesel particles showed an approximately 200-fold long-term lung retention compared to pure BaP (17). The primary aerosols generated were carbon, sodium chloride, Aerosil200 (spherical SiOz particles), and aluminum oxide. The most important material for these studies was the carbon aerosol because carbon particles are always formed by combustion processes which produce PAHs. Especially the diesel exhaust contains carbon particles (18). Sodium chloride is teated as a model aerosol for nonspherical and ionic particles. Aerosil200 and aluminum oxide aerosols are tentatively used as model aerosols for inert fly ash particles. The work presented here contains the results of PAH-adsorption and -desorption studies of well-defined PAH-coated
0003-2700/89/0361-0708$01.50/00 1989 American Chemical Society
