Analytical precision concentration

Precision When the analyte s concentration is well above the detection limit, the relative standard deviation for fluorescence is usually 0.5-2%. The limiting instrumental factor affecting precision is the stability of the excitation source. The precision for phosphorescence is often limited by reproducibility in preparing samples for analysis, with relative standard deviations of 5-10% being common. 

The inverse calibration regresses the analytical values (concentrations), x, on the measured values, y. Although with it a prerequisite of the GAussian least squares minimization is violated because the y-values are not error-free, it has been proved that predictions with inverse calibration are more precise than those with the classical calibration (Centner et al. [1998]). This holds true particularly for multivariate inverse calibration. 

Recall from our previous chapter [1] how Horwitz throws down the gauntlet to analytical scientists stating that a general equation can be formulated for the representation of analytical precision based on analyte concentration (reference [2]). He states this as equation 72-1 ... 

Boyle and Edmond [679] determined copper, nickel, and cadmium in 100 ml of seawater by coprecipitation with cobalt pyrrolidine dithiocarba-mate and graphite atomiser atomic absorption spectrometry. Concentration ranges likely to be encountered and estimated analytical precisions (lcr) are l-6nmol/kg ( 0.1) for copper, 3-12nmol/kg ( 0.3) for nickel, and 0.0-1.1 nmol/kg ( 0.1) for cadmium. 

Analytical Precision. Additionally, the analytical method was tested to assure that its precision was acceptable. This test was performed at three different levels as before, except that the spiking was done directly into known amounts of the Freon 113/perchloroethylene mixture. Six samples at each of the three concentrations were used to form the basic statistical set of data. The samples were compared to previously prepared standards. 

Initially, the analytical method was tested to assure that it was acceptable for analyte recovery as well as for precision. The sampling medium was spiked with known amounts of the test chemical at three levels corresponding to one-half, one, and two times the occupational PEL for a given air volume. Six spiked samples for each level were analyzed. The success of this portion of the validation assured that the analytical precision was acceptable for the desired concentration range. 

Table 25.1 Analyte Precision and Recovery at Different Concentrations...

Included is a graph (Figure 5) from an article by William Horwitz which relates analytical precision to concentration. It shows that the analytical variability increases as the concentration decreases. The Horwitz data were generated fran collaborative studies where methodology was exactly defined. The data should be... 

Figure 5 Graph relating analytical precision to concentration. (Reproduced from Ref. 3. Copyright 1981 American Chemical Society.)...

Laboratories establish analytical precision for each method using a laboratory control sample (LCS) and laboratory control sample duplicate (LCSD). These samples are made at the laboratory with interference-free matrices fortified (spiked) with known amounts of target analytes. Interference-free matrices are analyte-free reagent water or laboratory-grade (Ottawa) sand. Precision is then calculated as the RPD between the results of the LCS and LCSD. Analytical precision depends on analytical method and procedure the nature of the analyte and its concentration in the LCS and LCSD and the skill of the chemist performing analysis. The RPD for interference-free laboratory QC samples is typically below 20 percent for soil and water matrices. 

Insufficient analytical accuracy may be a cause of poor field duplicate precision. For example, low analytical accuracy that is typical for samples with low contaminant concentrations reduces analytical precision. 

We have described elsewhere the nature of the MI FTIR spectra of PAHs and their derivatives (1, 12, 13, 1J>, 18, 21-24) the following points are especially significant. First, FTIR spectra devoid of rotational structure and having individual bandwidths on the order of 2-7 cm are obtained both for PAHs and for polar derivatives thereof (such as nitrogen heterocycles) These spectra are sufficiently characteristic to enable identification of individual isomers to be made in mixtures [e.g., the six methylchrysenes (12) and the various mono- and dimethyl naphthalenes and biphenyls (24)]. Second, detection limits for individual PAHs can be as low as 50 ng, if special "micro samp ling 1 deposition apparatus is used (23) Third, Beer s law plots typically are linear over 1.5-2 decades in PAH concentration by the complementary use of two different deposition cells, linearity over 3 decades in PAH concentration for Beer s law plots can be approached (23). Finally, for both MI FTIR and MI fluorescence spectrometry, analytical precision of ca. 3-7 % relative standard deviation can be achieved. 

Because geochemically different clay sources may have been used by potters to produce ceramics for both domestic and trade purposes, neutron activation analysis (NAA) has been used as an independent means of ceramic characterization. Because of the relatively good analytical precision possible with NAA, statistical patterning of NAA data for major, minor, and trace element concentrations may be used as a powerful provenancing tool. 

To measure E over a 10 C interval to =h0.5 per cent will generally require a temperature error of less than 0.03 0 and a measurement of k and k to within zbO.3 per cent. This latter in turn will, as we have seen, require analytical precision of dbO.l per cent over an extended range of concentration changes. 

When the sample contains interferents, their influence on the analyte in the sample is progressively diluted and usually changes. As a consequence, the apparent concentrations are also changed and approaches nonlinearly at a certain concentration value, which is assumed as the final analytical result (Fig. 3.8b). This value is calculated by extrapolation of the nonlinear function fitted to the experimental points. Such a method is not favorable for good analytical precision and accuracy, which is the most serious drawback of the method. On the other hand, if the interference effect is diminished in the course of sample dilution (which is often the case), the analytical result has a chance to be free of this effect and to be accurate without any other additional efforts. 

Estimation of Uncertainty in The determination of the total ion molal concentration of calcium from salinity measurement is relatively precise with a probable error of less than 0.3% under open ocean conditions. Dickson and Riley (37) have recently discussed the effect of analytical errors on the evaluation of the components of the aquatic carbon-dioxide system for seawater at 25°C and 1 atmosphere total pressure. Their conclusions Indicate that if alkalinity and total carbon dioxide are the measured parameters a probable combined uncertainty in the total carbonate ion molal concentration from 3 to 6 percent results, depending on Fco2 If pH and alkalinity are the measured parameters the uncertainty is approximately 4 percent. In addition to the probable error introduced by analytical precision, the absolute accuracy of the measurements introduces an error which is difficult to evaluate. The results of the GEOSECS intercalibration study (38) were indicative of this problem. A conservative guess is that accuracy introduces at least a one percent further uncertainty. It is also difficult to determine exactly what error is introduced through temperature and pressure corrections to situ conditions. For the deep sea this may introduce a further uncertainty of at least... 

Sample Treatments. Blood (1 ml) and fecal samples (1 g dry-matter) were ashed on hot plates by sequential treatment with concentrated nitric acid and 30% hydrogen peroxide. The white residue of each sample was dissolved in 3-5 ml of 6 N HCl, and the final volume was brought up to 25 ml with 6 N HCl. Several 0.1 ml aliquots were transferred to test tubes, and iron concentrations were determined by a colorimetric method using the Batho-reagent (17) which contains hydroxylamine hydrochloride (10%), sodium acetate (1.5 M), and bathophenan-throline disulfonate (0.5 mM). The analytical precision of iron quantification was evaluated by measuring the iron concentrations of 13 replicates of one imenriched fecal sample. The mean of these measurements was 365.7 ug per gram of dry feces, with a relative standard deviation of 2. 8%. 

Final points to consider include the degree of replication and treatment of the initial, or time zero, concentrations and enrichment of both the product and source pools. The ability to replicate analyses has been greatly enhanced by the advent of automated CF-IRMS so that there is no longer a reason not to measure duplicate samples as a minimum. However, given the high analytical precision of modem mass spectrometers, the replication should be from the field by incubating replicate bottles... 

---