Trace Metals Project

T he chapters in this book have been prepared from papers presented at a symposium on the Analysis of Petroleum for Trace Metals held at the 169th National Meeting of the American Chemical Society. The symposium reported on the Trace Metals Project carried out by five petroleum companies to develop methods for 13 elements of environmental interest (arsenic and selenium, which are not metals in the usual sense, are included). For the present purpose, the format of the papers has been changed to include detailed procedural instructions. 

The objectives of the Trace Metals Project, in which the Atlantic Richfield Co., Chevron Research Co., Exxon Research and Engineering Co., Mobil Research and Development Corp., and Phillips Petroleum Co. participated, were fourfold to define the current state of the art, to develop analytical methods for application at the part-per-billion level using techniques available at most laboratories (10 ppb, i.e., 10 ng/g, was chosen as a reasonable attainable improvement of existing technology), to determine the precision of the methods developed, and to publish the procedures and results. 

Figure 1.2. Comparison of results obtained by NAA with those obtained by trace metals project procedures...

Fluorocarbon materials are available as laboratory equipment in a variety of shapes, sizes, and forms and are considered to provide an ideal containment and storage material. Teflon containers have been used satisfactorily by the Trace Metals Project for containing, shipping, and storing a variety of petroleum samples. Epoxy-lined metal cans and caps were also used satisfactorily for bulk storage and shipment. 

In the absence of a standard sample, spiked samples prepared by adding a known concentration of an organometallic standard to a petroleum matrix must be used. Because the form of most native metal species in petroleum is unknown and may be variable, the Trace Metals Project investigated the availability of organometallic materials to determine those best suited for the preparation of spiked samples and calibration standards in petroleum at the nanogram/gram level. 

The Trace Metals Project conducted a study to identify the type of container which would provide minimum losses of arsenic and mercury by precipitation, volatilization, adsorption, or diffusion. Solutions of organomercury and organoarsenic compounds added to petroleum feedstock were used. Because of the relative ease with which mercury and arsenic can be determined at sub-parts-per-million levels in a hydrocarbon matrix by instrumental neutron activation analysis (INAA), this technique was used for the analytical measurements. The solutions were stored in five different types of glass and/or plastic containers and sampled periodically over eight months (12). The results of the study are summarized in Tables 2.III and 2.IV. 

One of important aspects of the Trace Metals Projects was the cross-check program of proposed analytical methods. For example, evaluation of the wet digestion flame atomic absorption method for cadmium determination (13) involved the preparation of petroleum samples spiked with cadmium cyclohexanebutyrate. Before these samples were shipped to cooperating laboratories, a check of the cadmium content showed unexpectedly low recoveries. To confirm this apparent loss, portions of gasoline were spiked at the 30 ng/g level with cadmium cyclohexanebutyrate and cadmium sulfonate. The samples were stored in Teflon bottles and were analyzed over a month. The results are shown... 

A number of published papers discuss the effect of contamination on the analytical blank and some of the extraordinary means which are often necessary to minimize this effect. To control contamination, it is necessary to understand its sources. Basically, anything which contacts the sample between and including the actual sampling process and the completion of the analysis must be considered a potential source of contamination. These sources, along with methods used in the Trace Metals Project to minimize their effects, are summarized in Table 2.VIII. 

In the Trace Metals Project standard additions were used in all heated vaporization atomic absorption (HVAA) procedures. These procedures use sample aliquots of 10 /J or less consequently, a microstandard addition technique was developed. In this technique successive microliter aliquots of a /xg/ml standard are added to the sample solutions which have also been prepared on weight/volume basis. Since the volumes added or removed from the solution are negligible, the entire analysis is done on a single solution as detailed in the individual procedures. 

The development of extremely sensitive instrumental detection techniques has further exposed deficiencies in sample preparation schemes presently available to the petroleum analyst. It is hoped that recognition of this situation will provide some impetus for extensive research efforts on sample preparation. During the Trace Metals Project various sample preparation schemes were investigated, some only briefly because serious deficiencies became immediately evident. The criteria set by the Project for sample preparation techniques were that for any given element the total concentration would be measured—i.e., any destruction of the petroleum matrix had to take place without loss of any form of the constituent to be measured. 

Heated vaporization atomic absorption (HVAA) has been described extensively. HVAA differs from conventional atomic absorption in that an electrically heated device replaces the flame. The characteristics of HVAA are microliter sample consumption, sensitivity down to picogram quantities, and applicability to a wide variety of solutions. These characteristics have led to its widespread use. In petroleum analyses, this technique has been used to determine relatively high levels (ppm) of lead in gasoline, metals in used oils, and nickel and vanadium in crude oils (18). The Trace Metals Project has extended application of this technique to the determination of Be, Cd, Co, Cr, Mn, Mo, Sb, and V at the 10-ng/g level. 

Use of INAA in Trace Metals Project Validation Studies. As, Co,... 

The results of cross-checks comparing the INAA and the chemical methods developed by the Trace Metals Project are discussed in the chapters dealing with the specific elements. The chemical analysis data in most part represent the average of triplicate results from two or more laboratories, and the neutron activation results are the average of two or more irradiations and varying counting conditions. Overall results are summarized in Figure 1.2. 

The method developed by the Trace Metals Project involves a decomposition step in which the matrix is destroyed, followed by treatment to ensure that antimony is present as Sb(III) before measurement by atomic absorption. 

Of the readily available measurement techniques, atomic absorption in combination with arsine generation has gained great popularity and widespread use in recent years. This popularity is primarily the result of the ease of applicability and high sensitivity of the technique. However, prior to the studies carried out in the Trace Metals Project, no systematic study of its application to petroleum matrices had been made. 

During the Trace Metals Project two procedures based on measurement by HVAA were developed for determining beryllium in petroleum matrices (20). One procedure utilizes the CRA-63 carbon rod atomizer,... 

The Trace Metals Project investigated both flame and heated vaporization atomic absorption. These studies included an evaluation of decomposition techniques and ultimately a program of independent cross-checks of the methods that were devised. 

A number of methods exist for the determination of parts-per-billion (ng/g) levels of chromium in aqueous media (Table 8.1). These are repeatedly reviewed as new techniques are introduced (4,5,6). Potentially all these techniques could be applied to petroleum samples after matrix destruction, but in practice, only a few have been utilized. After wet oxidation of a large sample (> 100 g), 10 to 50 fig of chromium may be determined by a colorimetric procedure with 1,5-diphenylcarbohydrazide after iron, copper, molybdenum, and vanadium are extracted as the cup-ferrates (3). In survey analyses, Cr levels as low as 5 ng/g have been measured by optical emission spectroscopy after ashing (2,3) or directly by neutron activation with extended irradiation and counting times (1). Concentrations of chromium above 100 ng/g in used lubricating oils have been measured directly by flame atomic absorption (8) for lower concentrations, heated vaporization atomic absorption (HVAA) has been utilized (9). In the Trace Metals Project, two procedures using this latter technique were evaluated for the determination of 10 ng Cr/g in a variety of petroleum matrices. 

During the Trace Metals Project it was established that some lots of acids were relatively free of selenium and could be used without further purification. Other lots of the same acid contained unacceptable concentrations of selenium and had to be purified before use. A sub-boiling distillation procedure (9) was found suitable for purifying the acids that are required for the sample decomposition and selenium measurement. Although contamination from sulfuric acid was expected because of the frequent association of selenium with sulfur, most lots tested could be used without purification. Since others have reported severe contamination (10), it is advisable to screen sulfuric acid lots carefully before use. Selenium could not be generated when one lot of unpurified perchloric acid was used, but it could be quantitatively generated after that lot had been purified by the sub-boiling distillation technique. 

Selenium has been reported to be completely lost from organic samples when subjected to dry ashing techniques (11). As part of the Trace Metals Project, ashing studies were carried out which confirmed the loss for dry ashing of petroleum matrices. When ashing aids such as sulfur or magnesium oxide were used, some selenium was retained in the ash, but the recovery was not complete. 

During the Trace Metals Project several sample preparation techniques were evaluated for the determination of selenium in petroleum and petroleum products. Wet oxidation under reflux was the technique adopted because it was the only one that gave quantitative recovery of selenium from various petroleum-type samples. Some of the techniques that were evaluated are discussed below. 

---