Palladium as chemical modifier for the stabilization of volatile nickel and vanadium compounds in crude oil using graphite furnace atomic absorption spectrometry Isabel C. F. Damin, a Maria Goreti R. Vale, a Ma´rcia M. Silva, a Bernhard Welz,* b Fa´bio G. Lepri, b Walter N. L. dos Santos c and Se´rgio L. C. Ferreira c a Instituto de Quı´mica, Universidade Federal do Rio Grande do Sul, Av. Bento Gonc ¸ alves 9500, Porto Alegre—RS, 91501-970, Brazil b Departamento de Quı´mica, Universidade Federal de Santa Catarina, Floriano ´polis—SC, 88040-900, Brazil. E-mail: welz@qmc.ufsc.br; Fax: (55) 48 331-6850 c Instituto de Quı´mica, Universidade Federal de Bahia, Salvador—BA, 40170-290, Brazil Received 8th June 2005, Accepted 13th September 2005 First published as an Advance Article on the web 11th October 2005 It has been observed in previous work using high-resolution continuum-source atomic absorption spectrometry that up to 50% of the nickel and vanadium in crude oil, most likely non-polar nickel and vanadyl porphyrin complexes, was already lost at pyrolysis temperatures 4400 1C, whereas the rest of the analyte was stable up to at least 1200 1C. In this work we were investigating the use of a chemical modifier in order to avoid these low-temperature losses and also to make possible a differential determination of the volatile analyte fractions. Oil-in-water emulsions, using Triton X-100 as surfactant, were used for easy sample handling. A mass of 20 mg of Pd, introduced into the graphite tube and thermally pretreated prior to the injection of the emulsion, was found to efficiently prevent any low-temperature losses of nickel and vanadium from crude oil samples up to pyrolysis temperatures of 1200 1C and 1450 1C, respectively. The best characteristic mass obtained was m 0 ¼ 16 pg for Ni and m 0 ¼ 33 pg for V. The limits of detection and quantification were 0.02 mgg 1 and 0.065 mgg 1 , respectively, for Ni, and 0.06 mgg 1 and 0.17 mgg 1 , respectively, for V in oil, based on an emulsion of 2 g of oil in 10 mL. The precision expressed as relative standard deviation (n ¼ 9) for a crude oil containing about 5 mgg 1 Ni and 3 mgg 1 V was 2.0% and 3.1% for Ni and V, respectively. The accuracy of the procedure was verified by analyzing the certified reference material NIST SRM 1634c, trace metals in fuel oil, and the research material RM 8505, vanadium in crude oil. Introduction Nickel and vanadium in crude oil occur as nickel and vanadyl porphyrins and non-porphyrins, 1–3 which originate largely from the formation of crude oil, the former ones from chlo- rophylls, through the substitution of Mg with the trace ele- ment. 1 Their concentration, which is typically in the mgg 1 range, provides information about the origin of the oil, and the ratio between these two elements is actually ‘bore-hole speci- fic’, and is transported proportionally through crude oil refin- ing procedures, even if in markedly reduced concentrations. Nickel and vanadium are both serious catalyst poisons, and may cause undesirable side reactions in refinery operations. Vanadium, in addition, causes corrosion problems that derive from the formation, for example, in the combustion chamber of power plants, of sodium vanadates, which have low melting points; the molten vanadates react with the metal surface of the superheaters and form the metal oxide. 4 Inhalable dusts or aerosols of nickel and its compounds, which might be gener- ated during combustion of oil, are classified as hazardous substances owing to their carcinogenic and mutagenic effects, so that the release of this element has to be controlled accord- ing to various national and international regulations. 5 All these properties make it necessary that the content of nickel and vanadium in crude oil has to be determined routinely, independent of the fact that the content of such elements might have an influence on the price of the crude oil on the market. A variety of spectrometric techniques have been used for the determination of nickel and vanadium in crude oil and petro- leum products, such as inductively coupled plasma (ICP) optical emission spectrometry, 6–9 ICP mass spectrometry, 10,11 X-ray fluorescence spectroscopy, 12,13 and even high-perfor- mance liquid chromatography with UV detection. 14 However, flame atomic absorption spectrometry (FAAS) is the technique of choice in most publications. 5,15–22 The main advantages of that technique are that firstly, a direct determination of the analyte in the hydrocarbon matrix appears to be possible after corresponding dilution with an organic solvent, and secondly, the flames used in FAAS, in contrast to an ICP, are quite tolerant to most organic solvents. Nevertheless, direct deter- mination is not free of problems since the solvent and the compounds used for calibration have a substantial influence on the sensitivity. 15,23,24 The most reliable procedure for the determination of nickel in crude oil and petroleum products is hence ashing of the sample and analysis of the residue after taking it up in hydrochloric acid, 17,21 a procedure which is quite tedious and which requires careful control to avoid analyte losses. 20 Graphite furnace (GF) AAS appears to be an alternative since the problem of differing sensitivity for various com- pounds apparently does not exist; the results from a simple dilution with xylene and ashing were found to be practically identical. 25 The formation and analysis of an oil-in-water emulsion or microemulsion instead of a dilution of the crude oil sample with an organic solvent has been proposed for ARTICLE www.rsc.org/jaas DOI: 10.1039/b508099a 1332 J. Anal. At. Spectrom., 2005, 20 , 1332–1336 This journal is & The Royal Society of Chemistry 2005