ASTM D7260 Standard Practice for Optimization, Calibration, and Validation of Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES) for Elemental Analysis of Petroleum Products and Lubricants
8. Calibration
8.1 After a warm-up time of at least 30 min, operate the instrument according to the manufacturer's instructions. Some manufacturers recommend even longer warm-up periods to minimize changes in the slopes of the calibration curves.

8.2 Wavelength Profiling - Perform any wavelength profiling that is specified in the normal operation of the instrument, if so equipped. Correct profiling is important to reveal spectral interferences from high concentration of additive elements on the spectral lines used for determining low levels of other elements present.

8.3 Calibrate the instrument by aspirating the blank and standards. Allow at least 30 s aspiration beyond the point that the solution reaches the nebulizer to allow the instrument to equilibrate prior to signal integration. A total flush-out time of approximately 1.5 min to 2 min should be allowed between standards or following samples of comparably high or higher concentrations, during which diluent solvent is being aspirated. The computer establishes the slope, intercept, and correlation statistics for each element.

8.4 Often a two-point calibration is used. However, multiple calibration standards may be used for obtaining improved element to signal correlation or validate linearity of the concentration range intended for use for a given analyte.

8.5 To ensure the validity of calibration, check standards with known elemental concentration should be used after calibration but before actually analyzing the unknown samples. It is recommended that the calibration blank and standards be matrix matched if possible with the same solvent, base oil, and so forth.

8.6 To minimize physical interferences caused by changes in sample transport process (due to variations in sample viscosity and concentration), it may be necessary to use a peristaltic pump in conjunction with certain nebulizers. Internal standard(s) could also be used instead of or as well as peristaltic pump.

8.7 Background Correction:
8.7.1 Ion-electron recombination and stray radiation give rise to background continuum. Introduction of an aerosol into the plasma may affect the intensity of this background continuum. An additional recombination may also give rise to background continuum from dissolved major ions in the sample.

8.7.2 Background correction generally involves measuring the background emission at a wavelength near the analytical wavelength (free of line emission from concomitant elements), and subtracting that intensity from the total intensity measured at the analytical wavelength. Selection of these correction points should also be carefully considered so as not to fall on the concomitant emission lines of elements potentially present in the solutions being analyzed. Nor should any structured spectral peaks be used (as opposed to flat baseline) for this correction that might result from seemingly stable/consistent emission lines that may arise from matrix elements or argon.

8.7.3 In some cases, if the background continuum intensity varies with wavelength, it may be necessary to measure the background on both sides of the analytical wavelength and interpolate to estimate the continuum intensity to be subtracted.

8.8 Internal Standardization:
8.8.1 Several ICP-AES test methods (see 2.2) require the mandatory use of internal standard in analysis. This procedure requires that every test solution (sample and standard) have the same concentration (or a known concentration) of an internal standard element that is not present in the original sample. The internal standard is usually combined with the dilution solvent because this technique is common and efficient when preparing many samples. However, this assumes that the dilution solvent will be diluted with sample at a consistent ratio. For example, a tenfold dilution with an internal standard spiked solvent will render a differing final internal standard concentration solution than a one hundred-fold dilution of this sample if the same dilution solvent is used. Though mathematical corrections may be applied for the resulting disparate internal standard concentration in these solutions, analyte/internal standard free matrix additives are frequently used to prepare such greater (one hundred-fold) dilutions and thus address this concern. This practice also aids in maintaining a consistent solution matrix. Alternatively, the internal standard can be added separately from the dilution solvent as long as the internal standard concentration is constant or accurately known.

8.8.2 Internal standard compensation is typically handled in one of the two different ways. Note that blank (zero analyte concentration) calibration standard - internal standard concentrations must be addressed accordingly. This may be an issue on analysis accuracy if the internal standard contains traces of analyte contaminants (or spectral interference from the internal standard) that would be inherently subtracted with an alternate calculation (scaled) - given all standards, blanks, and samples have the same contaminant level.
8.8.2.1 Calibration curves are based on the measured intensity of each analyte divided (that is, scaled) by the measured intensity of the internal standard per unit internal standard element concentration. Concentration of each analyte in the test specimen solution is read directly from these calibration curves.

8.8.2.2 For each analyte and the internal standard element, calibration curves are based on measured (unsealed) intensities. Uncorrected concentration of each analyte in the test specimen solution is read from the calibration curves. These are multiplied by a factor equal to the actual internal standard concentration divided by the uncorrected internal standard concentration determined by analysis, to obtain corrected analyte concentrations.

8.8.2.3 Although normally the calibration curves would be linear, sometimes inclusion of a second order term can give a better fit, although in such cases it would require at least five standards for calibration.

8.8.3 The organometallic compound representing the internal standard should be dissolved in the dilution solvent. Solution stability must be monitored and prepared fresh (typically weekly) when the concentration of the internal standard element changes significantly. The concentration of the internal standard element shall be at least 100 times the detection limit. A concentration of 10 mg/kg to 20 mg/kg is typical. The internal standard must also remain stable and soluble in the sample solution, given the potential for precipitation due to the presence of unexpected concomitant species in the sample solution.

8.8.4 Ideally, the internal standard element should not be present in the sample being analyzed; its chosen spectral line should have the same excitation characteristics as the analyte line; the chosen internal standard line should occur in a similar region of the spectrum as the analyte line so that the spectrometer characteristics affect both similarly; the internal standard should not itself be interfered with by other elements; its intensity should be measured simultaneously with that of the analyte line; and its intensity should be similar to that of the analyte line so that the detector responses are similar. It should be noted that some times the internal standard may not track the lines of all species of interest present in the sample solution.
8.8.4.1 In practice, all of these criteria may not be met, and it is common to for one chosen internal standard line to serve as a reference for various analyte lines. The following internal standards have been successfully used in the laboratories: Ag, Be, Cd, Co (most common), La, Mn, Pb, Sc, and Y.

9. Calibration Standards
9.1 Multi-element Calibration Standards - Multi-element standards containing known concentrations (approximately 0.1 % by mass) of each element can be prepared from single element stock solutions at appropriate concentration levels for each element. Prior to preparing the mixed standards, each stock solution should be analyzed separately to determine possible spectral interference or the presence of impurities. Care should be taken when preparing multi-element calibration standard solutions that the elements as well as the standard solution matrix be compatible and stable, especially when diluted. Many manufacturers recommend including additives when making diluted multi-component solutions to enhance solubility/stability such as one manufacturer advising to add their stabilizer for sulfonates to be maintained at 0.6 %. The actual number of calibration standards required will be a function of both chemical compatibility and the restrictions on the computer system used to control the spectrometer. Additional calibration standards may be needed if a second, less sensitive emission line is used to extend the linear range ofone or more elements.
9.1.1 When preparing multi-element standards care should be taken to ascertain that proper mixing is achieved. Ultrasonic homogenizers and vortex mixers are recommended. Refer to Practice D4307 for a procedure for preparation of multielement component liquid blends. Commercially available multi-element blends (with known concentrations of each element at approximately 0.1 mass %) are also satisfactory. In some cases 0.01 m % for lower concentrations and 0.5 m % for higher concentrations may be appropriate.

9.1.2 Concentrations of elements in above calibration standards should be such that emission intensities measured with the working standard can be precisely measured (that is, the emission intensities are significantly greater than the background) and these standards represent the linear region of the calibration curve.

9.1.3 Some commercially available organometallic standards are prepared from metal sulfonates and therefore contain large amounts of sulfur that can be calculated to render a calibration for sulfur. However, if low concentration of sulfur determinations are desired, a separate sulfur standard should be prepared or alternate type organometallic standards be used.

9.1.4 Petroleum additives can also be used as organometallic standards if the concentration of metals in them is accurately known, and if their use does not adversely affect the precision or introduce significant bias.

9.1.5 Before use check the accuracy of elemental concentrations of commercially obtained standards either by comparing with alternative sources or analyzing by independent primary methods.

9.1.6 Concentrations of standards can change with age. Hence, fresh calibration standards should be prepared as needed when the concentration becomes suspect.

9.2 Interference Check Samples - These are prepared from single element stock standard solutions to contain elements and concentrations appropriate to the sample type.

9.3 Check Standards - These are prepared in the same manner as the working standards such that the concentrations of elements in check standards are similar to the concentration of elements in the test specimen solutions. It is advisable to prepare check standards from alternative sources of certified organometallic standards.

9.4 Reagent Blank - The reagent blank must contain all of the reagents and solvents (and internal standards if used) in the same percentages as used in the preparation of the samples. The reagent blank must be carried through the complete procedure and contain the same solvent concentration in the final solution as the sample solution used for final measurement.