URGENT (Instrumental analysis subject)

The selenium contents of steel can cover a wide range of concentrations (0.0001-0.3%). It is generally only present as an impurity, and although it has a detrimental effect on the mechanical properties of steel, the machinability of corrosion resistant steels may be improved with its intentional addition. Atomic absorption spectrometry (AAS) has been used for the determination of selenium in steel and nickel alloys using flame,l electrothermal atomisation2-5 and hydride generation techniques.68 The accuracy of these methods, except for the last one, may be affected by the spectral interference of iron at the 196.0-nm line, which cannot be eliminated by the use of deuterium background correction.4 Adequate correction is only possible using the Zeeman effect. Interference from iron can be eliminated by measuring at a wavelength of 204.0 nm, but the sensitivity at this line is about five times less than at 196.0 nm. This paper describes the direct determination of selenium by flame AAS. The optimum operating conditions for the determination are verified, and the method of steel sample decomposition is discussed.

Apparatus:

A Perkin-Elmer Model 4000 atomic absorption spectrometer with a 10-cm single slot air - acetylene burner head and a nebuliser with flow spoiler was used. The rate of aspiration was 7.3 ml min-1. A selenium electrodeless discharge lamp, from the same manufacturer, was also used

Reagents and Samples:

All acids used were of analytical-reagent grade. A standard selenium solution, 1 mg ml-1, was prepared by dissolving 1.0000 g of selenium in 10 ml of concentrated nitric acid and diluting to 1000 ml with water. To investigate the sample decomposition procedure, two samples of carbon steel containing the following were prepared: 0.24% of C, 0.58% of Mn, 0.18% of Si, 0.02% of P and S, 0.17% of Cr, 0.11% of Ni and 0.15% of Cu with selenium contents of 0.014 and 0.096%, respectively.

Procedure:

Weigh 1.000 g of steel containing up to 0.2% of selenium and dissolve it in 15 ml of HC1- HN03 (3 + 1). To accelerate the dissolution of high-silicon steel and cast iron, add 1 ml of hydrofluoric acid. After the sample dissolution, evaporate the solution to about half of its original volume and dilute to 50 ml in a calibrated flask. Measure the selenium absorbance at 196.0 nm, with a slit width of 2.0 nm, using deuterium background correction and an air - acetylene flame with the gases adjusted to give the maximum reading. Determine the selenium content of the sample from a calibration graph constructed by means of standard additions of selenium to a selenium-free steel; the graph does not pass through the origin (see Results and Discussion).

Results and Discussion:

Using the conditions described here, the absorbance of selenium remains practically constant under the acid concentrations chosen: hydrochloric acid up to 2 M, nitric and hydrofluoric acids up to 1 M. There is only a small decrease (1.3% relative) in the selenium absorbance due to the iron that is present (20 mg ml-1). The sensitivity is influenced very little by the composition of the flame, by the height of observation or by the rate of aspiration. The 196.026-nm selenium line is overlapped by the 196.013- nm iron line so that if measurements are taken without background correction there is a positive error, and on using background correction there is a negative error. The quantitative data are summarised in Table 1. It is evident from Table 1 that the sensitivity is ca. 10% lower when using a wider slit width but the standard deviations of the slopes and the intercepts remain identical. A small part of the error is due to the interference from iron and occurs when the measurements are carried out with background correction and with a slit width of 2.0 nm. The remainder of the negative error can be corrected for by measuring the blank values, i.e., a steel solution that is free from selenium. During the investigation on sample decomposition, it was found that some recommended procedures are inefficient. For example, the selenium recoveries from the samples (see Reagents and Samples) were 64, 85 and 73% for sample decomposition with 20% V/V HN03,3 HC1- HF - H2025 and 15% V/V H2S04,8 respectively. There is a loss of selenium due to volatilisation of H2Se if the sample is dissolved with non-oxidising acids or due to incomplete oxidation of selenium in steel samples. For example, the selenium losses from steel containing 0.096% of selenium were 62 and 25% if the steel was dissolved in HC1 (1 + 1) and H2S04 (1 + S), respectively. However, no loss was observed by evaporating the sample solution to dryness in the HCl - HN03 medium. It was observed that HN03 alone is not sufficient to decompose samples with selenium contents of 0.014 and 0.096%, the selenium recoveries were 45,67 and 89% for 1 + 3,1+ 2 and 1 + 1 HN03, respectively. Accurate results are only obtained if sample dissolution is carried out using HCI - HN03 mixtures or by the oxidation of the sample solution in HN03 with a strong oxidising agent, i. e., permanganate, chlorate or by evaporation with perchloric acid to fumes. The interferences in the proposed method were tested by standard additions of selenium to a variety of selenium-free steels. None of the elements occurring in these materials interfered in the determination of selenium. The relative standard deviations (n = 8) were found to be 5.12 and 1.03% for 0.014 and 0.094% selenium, respectively. The quantitative detection limit (criterion, 10 sblank) is 0.007% selenium.

Conclusions:

The determination of selenium in steel by the proposed flame AAS method is reliable when using the optimised sample decomposition procedure and correcting for the spectral interference of iron , which is eliminated by calibration. The method is convenient for the determination of selenium in the range 0.01-0.2%.

According to the article, What is meant by the rate of aspiration? The height of observation? The composition of the flame?