Traceability chain

Very specific forms of low pH localized corrosion may develop in large high heat-flux WT boilers. Where subsequent analysis work is undertaken, the results usually provide the basis for a traceable chain of cause and effect. An example of such a chain of cause and effect is detailed in an interesting corrosion case history provided by Maringer, Saavedra, Selby, and Haberman, and published by Tall Oaks Publishing, Inc. A synopsis is provided below (with kind permission of Tall Oaks Publishing). 

If we use a standard solution traceable to a stated reference, we ensure the traceability of our measurement by extending the traceability chain (for more details on traceability see chapter 10). 

Traceability is the relationship of a result of a measurement to a value of a standard through an unbroken chain of comparisons (traceability chain). In the case of the length measurements this chain can be realized in the way presented in the slide. When we want to measure the length of a shark, the result of oirr measurement is dependent on the tape we are using (comparison). 

Now let s see how a traceability chain can be realized for a chemical measurement. Generalizing and in an analogous way to length measurements it can be like the chain in the slide. The amount content of a compound X in a solution is compared with the amount content in a working standard. This in turn was compared with the amount content in a reference standard and after some further comparisons, in an ideal situation we end up with the SI unit, the mole. 

Let s see what is the influence for the two cases where the certified reference materials are useful in the laboratory (i.e. calibration and vahdation). In the case of calibration, the properly value of the certified reference material is used to calibrate the analytical instrument used for the measurement, thus it is used in order to obtain the measurement result. In this way the properly value of the certified reference material is part of the traceability chain, as shown in the shde, and is directly involved in the establishment of the measurement traceabihty. 

FIGURE 4 Traceability chain and relationship between traceability and uncertainty of measurements. The three possibilities for establishing traceability referred to in Figure 5 are indicated in bold [25, 28]. 

As both measurements are made in Australia and the weighing instruments (bathroom and doctor s scales) are calibrated to the mass of the Australian standard kilogram, the traceability chain could have stopped there, along with any comparability. Although the definitions do not mention the SI, and of course for some measurements there are no SI units anyway, the more universal the stated reference is, the more measurements will come... 

At the top of the traceability chain is the stated metrological reference, which for our purposes is the definition of a unit. It might not be obvious how a piece of paper from Paris giving the interesting, but rather esoteric, definitions of the SI base units can be used as a reference. The metrological reference allows the creation of a primary calibrator that embodies the value of the unit (or some multiple or submultiple of it). The international prototype... 

The top of the metrological traceability chain is given in figure 7.6. The procedure and system referred to in the figure may be either for the production of the calibrator, such as the international prototype of the kilogram, or for a primary measurement procedure governing a measuring system, as in the use of a coulometric titration. 

The top of the metrological traceability chain resides with artifacts or procedures that realize a unit. These are described in more detail below. 

Figure 7.7. Multistranded traceability chains, showing the need to establish traceability to a number of references in a chemical measurement (here titration).

Figure 7.8. A possible metrological traceability chain for the result of a breathalyzer measurement of a motorist s breath alcohol.

Figure 7.9. A possible metrological traceability chain for the result of a measurement of protein in a sample of grain. aTris = 2-amino-2-hydroxymethyl-1,3-propanediol b Dumas apparatus is calibrated using tris CRM and grain samples are certified in an interlaboratory study cthe master instruments measure grain samples to act as the grower s calibrator for field measurements.

Identify requirements for metrological traceability. In a very simple system, a single metrological traceability chain may lead to a single... 

Document metrological traceability. This requires identification of all CRMs used as calibrators, calibration certificates for equipment, and a statement of the measurement uncertainty of the measurement result. The metrological traceability chain is thus established. 

Report metrological traceability. Measurement reports may require details of the metrological traceability chain or at least a statement of the metrological reference. 

Figure 7.10. A metrological traceability chain from the point of view of an end user.

Results can only be accurate and comparable worldwide if they are traceable. In the vast majority of chemical analyses, the traceability chain is broken because in the treatment the sample is physically destroyed by dissolutions, calcinations, etc. With respect to speciation analysis, the chain is even more complex since it involves successive analytical steps (Quevauviller et al, 1996a). To approach full traceability it is necessary to demonstrate that no loss or contamination has occurred in the course of the sample treatment. The only possibility for any... 

Tentative attempts have been made to categorize chemical RMs in terms of their material composition (Table 2) and to classify them by the length and strength of their traceability chain (Table 3). This classification might in future be generalized and expanded in terms of ranges of relative uncertainties. 

Wherever possible, a traceability chain of measured values terminates in an SI unit. When the base unit for mass is appropriate, this relationship is readily achieved through a mass standard, calibrated in terms of the kilogram prototype. The concept of traceability to SI has to be more carefully considered when conformity to SI depends on the SI concepts in the definition of the SI unit itself. 

The SI traceability statement for a chemical composition of a material cannot be completed by the traceability to the mole of one entity. The statement must include reference to another quantity, which could be a mass, a length, some other quantity, or even an amount of substance of another entity. Examples of such traceability statements for chemical composition could refer to a mole and the kilogram for the concentration, say, of a known element in an ore. The source of the element in that ore is then described in terms of the ratio of SI units mol/kg. Similarly by the SI units of mole and meter one could designate the solution of a defined organic compound, that is in mol/L. For some important chemical measurements we need to find traceability to SI for the mole of one entity as well as the mole of another entity. These moles are not identical and need separate traceability chains (see next paragraph). Measurements by mol/mol ratios are appropriate, for instance, for a trace impurity of known composition in a pure compound, or for an amount of iso-tope-to-element substance ratio (abundance). 

To decide on the reference end of the examinational traceability chain, it is necessary primarily to define that which is to be examined. According to the IUPAC/IFCC Recommendation 1966 [5], the designation of a property must comprise identity of system and any pertinent components), together bearing the property, and the kind-of-property involved. Standardized templates for such designations are currently found in an IFCC/IUPAC database comprising over 10000 entries [6]. In some cases, the examination procedure becomes an integral part of the definition of the examinand. 

Depending on specified medical needs, the allowed combined examination uncertainty, also called an analytical performance goal, should be decided upon a priori because its magnitude influences the choice of metrological traceability chain. Guidance on how to choose combined uncertainty is given in the upcoming ISO/DTR 15196 [7],... 

Metrological traceability, according to its definition, is a property of a measurement result or quantity value, and the trace is towards a stated reference through a metrological traceability chain. Its links and relations between them have to be established a priori in the opposite direction from the chosen reference towards the measurement result. This structure is termed a calibration hierarchy. 

The correct sequential transfer of values down the hierarchy requires measurement specificity and selectivity of each measurement procedure and commutability of each calibrator, i.e. it has the same behaviour towards the preceding and following measuring systems, working according to their respective measurement procedures, as have the routine samples. Lack of fulfilment of these requirements breaks the traceability chain. Unfortunately, the respective causes are often subsumed under the concept of "matrix effect". 

The sequence of five types of calibration hierarchy outlined above obviously constitutes decreasingly transparent and generalized traceability chains, and international collaboration is necessary to improve comparability. 

Key words Traceability Interlaboratory comparisons Reproducibility Traceability chain Metrology in chemistry... 

This is where traceability comes in. Traceability of a measurement result means that the value of that result (each result of each laboratory) is demonstrated to be linked to an independent common stated reference [1], through an uninterrupted chain of comparisons (see Fig. 1 for the most popular traceability chain). That requires a priori knowledge of the measurement process and of the stated reference , because the... 

Fig. 1. The best known traceability chain, that of weights...

IF THINKING ABOUT TRACEABILITY, CHEMISTS THINK IN TERMS OF A TRACEABILITY CHAIN SUCH AS IN MASS WHERE A CHAIN OF REAL OBJECTS ("ARTEFACTS ) LEADS TO A PRIMARY STANDARD OF MEASUREMENT - THE kfj - WHICH ITSELF IS - STILL - A REAL OBJECT ("ARTEFACT")... 

Fig. 3. Core part of a traceability chain of the measurement of an isotope amount ratio in an isotope dilution measurement procedure...

There are two major reasons why a traceability chain may be broken and trueness lost due to the introduction of bias insufficient commutability of a calibration material and non-specificity of a measurement procedure. The effect of these separate properties are often indiscriminately lumped together as matrix effect . Commutability refers to the ability of a material, here a calibrator, to show the same relationships between results from a set of procedures as given by routine samples [16, 17]. Analytical specificity refers to the ability of a measurement procedure to measure solely that quantity which it purports to examine [16, 18]. Discrepancies between results of a reference procedure and a routine procedure applied to routine samples are often caused by non-specificity of the routine procedure. The use of a set of human samples as a manufacturer s calibrator to eliminate so-called matrix effects should only be accepted if the relationship between the results from reference and routine procedures is sufficiently constant to allow explicit correction with consequent increased uncertainty of assigned values. 

---