First Interlaboratory Study

The stability of solutions 1 and 2 was verified on the content of 10 bottles stored at 4 °C in the dark over a period of 8 weeks at the National Food Agency. Analyses were performed by packed column gas chromatography followed by electron capture detection. Determinations were performed in triplicate on each of three bottles. 

The first intercomparison of the solutions did not reveal any major discrepancies in the results and hence in the final methods of determination used. The 

Range ratio of higher value/lower value. 

---

In most cases, the pretreatment techniques used were based on complexation, GC separation and emplyed various different detection techniques (e.g. MIP-AES, AAS, ICP-MS). No specific comments were made during the meeting to discuss the first interlaboratory study, except that the participants considered it impossible to correct the results for impurities in the calibrant matrix. 

A systematic difference was observed by one laboratory between two different sets of calibrants (calibrant solution made with a newer calibrant from the same producer). This highlighted the need to thoroughly verify the calibrant, i.e. not to rely on calibrants from one producer of which the quality could vary from one set to another. Most of the laboratories actually used their own calibrants, which were not verified for purity and stoichiometry. Only one laboratory used the calibrant previously verified and distributed in the first interlaboratory study [8]. It was stressed that calibration was an important issue and that more efforts should be put on the verification of calibrants in future exercises. It was agreed that the coordinator of the project would purchase calibrant from a chemical company and establish its purity sets of verified primary calibrants would then be made available to participants in a further exercise to characterise their own calibrants. 

Table 8.18 summarises the results obtained in the first interlaboratory study. From this table, it is clear that improvements were strongly required in the trace element determinations in estuarine water, particularly for Cd, Co, Mn, Pb and Zn. A comparison with the CVs obtained in the certification of trace elements in seawater [24] also highlighted the enhanced difficulties in the analysis of estuarine water these CVs ranged from 9.5% (Cu) to 21.4 % (Pb) for concentrations three times lower than that found in the estuarine water samples. 

SUMMARY OF THE RESULTS OF THE FIRST INTERLABORATORY STUDY. THE RESULTS ARE GIVEN IN nmol kg ... 

A comparison with the results obtained in the first interlaboratory study (Table 8.18) showed a clear improvement in the quality of the determination of most of the elements, particularly cadmium (considering the much lower concentrations determined in the second exercise), copper, lead and zinc. For iron and nickel, the CVs between laboratories were approximately the same. 

In the first interlaboratory study, the examination of the raw data (14 sets of results of which 12 involved CVAAS, one RNAA and one MIP-AES) revealed a high spread of results due to two outliers. The mean obtained was 12.6 pg L of Hg with a coefficient of variation (CV) between laboratories of 33%. The two high results were attributed to a laboratory contamination. The accepted values showed a picture which was found more acceptable, i.e. the mean obtained was 10.8 pg L with a CV between laboratories of 6.6"/n [8]. At this stage, the agreement between the laboratories was found to be satisfactory however, the Hg content in this (spiked) sample was considered much too high for being representative of natural samples which justified the organization of a second interlaboratory exercise for which results are described below. 

The conclusions of the overall study were that a CT concentration of 2000 mg L or more is suitable to stabilize selenite but samples have to be opened only at the time of analysis to ensure a complete stability. This recommendation was clearly stressed to participants in the first interlaboratory study (see below). 

The choice of the extraction procedures is described in detail elsewhere [18]. On the basis of the results of the first interlaboratory study [19], the choice of EDTA and DTPA for certification was discussed. Whereas EDTA was widely accepted, the choice of DTPA was more criticised because of its operational difficulties the wide use of the latter would, however, justify the certification of DTPA extractable trace element contents, providing that its limitations in comparison with EDTA were clearly identified. 

The project to improve the quality control of lead speciation analysis was started in 1990 by a feasibility study on the stability of alkyllead species in solution [118], and was concluded in 1991. The first interlaboratory study was conducted in 1992 [119] and was followed by a second exercise carried out in 1993 [120]. The certification campaign of trimethyllead in artificial rainwater and urban dust was conducted in 1995-96. 

Table 11.8 Coefficients of variation of the mean of laboratory means of EDTA and acetic acid extractable contents as obtained in the first interlaboratory study...

Only in a few cases was there a distinction made between both concepts by their use in intra laboratory studies (robustness) and interlaboratory studies (ruggedness) see USP 23-NF18 [1995] USP 24-NF19 [2000] Rodriguez et al. [1998] Zeaiter et al. [2004]. Wahlich and Carr [1990] seem to be the first to use robustness and ruggedness in a hierarchical sense but in a reverse meaning as given above. 

To our knowledge, in the first approach described, Youden and Steiner introduced the term ruggedness test for a setup in which by means of an experimental design the influences of minor but deliberate and controlled changes in the method parameters or factors are evaluated in order to detect non-robust factors, i.e., with a large influence on the response. Controlling the latter factors avoids problems in the following interlaboratory study. 

Robustness-based approach Based on robustness tests as intralaboratory simulations of interlaboratory studies Simple, time-efficient approach Some sources of uncertainly may be overlooked method must first show to be robust Hund et al. [39]... 

Validation of a new analytical method is typically done at two levels. The first is the level of prevalidation, aiming at fixing the scope of the validation. The second level is an extensive, full validation performed through a collaborative trial or interlaboratory study. The objective of full validation, involving a minimum number of laboratories, is to demonstrate that the method performs as was stated after the prevalidation. 

There are two aspects to a report of an interlaboratory study. First the nature, organization, and treatment of results of the trial must be specified,... 

The first two approaches using definitive or primary or reference methods within one single laboratory require that in this laboratory everything is done to eliminate sources of systematic errors. Experience has demonstrated that it is very difficult to achieve 100% certainty and that within the laboratory a systematic bias does not remain. An additional confirmation through an — even limited — interlaboratory study is therefore advisable. Such an approach is used by NIST the single laboratory certification complies with the demand of US law that results and certificates must be NIST traceable . 

As interlaboratory studies imply that data are generated in various locations, it is necessary to bring the data producers together so that they can examine and judge the value of the measurements. Before any further treatment, all data used to certify the material must be validated by the certification group. Therefore, they must first be collected by the organiser, and presented in a way in which they can be easily compared. Statistical representations may be used to highlight trends or systematic errors within or between sets of data. 

One of the most critical points in organometallic chemistry analysis is the availability of calibrants of suitable purity and verified stoichiometry. This aspect was recognised at an early stage of the project and the purity of alkyllead compounds used in the feasibility study was carefully verified [10]. Additional experiments were performed on calibrants in the frame of the first interlaboratory exercise as described below. 

For the first intercomparison exercise, participants were asked to dilute the solution 1000 fold, i.e. to determine levels of TriML of ca. 40 pg L. Some laboratories also analysed the solutions after a 10,000-fold dilution. The participating laboratories in the second interlaboratory study received two sets of solutions containing ca. 50 and 5 pg L of TriML respectively. They were requested to perform five replicate analyses of, respectively ... 

Two interlaboratory studies were organised prior to the certification campaign. The first one dealt with the analysis of artificial seawater and the second exercise concerned the analyses of natural and spiked seawater. The results obtained for Cd, Cu, Pb and Zn are compared in Table 8.23. The CV between all the laboratories appeared to be quite high in the second round-robin exercise. However, the participation in such intercomparison combined with critical discussions of methods and results was found to be a most useful tool in obtaining a high level of accuracy (which is reflected in the... 

Previous exercises have shown the difficulties of determining Hg in seawater. As an example, coefficients of variation of 26.7 and 11.6% between eleven laboratories (CV of the mean of laboratory means) were found for Hg levels of respectively 6.0 and 24.2 ng L [31]. Therefore, it was chosen to consider seawater samples containing much higher Hg concentrations (coastal seawater samples spiked with mercury) in a first method performance study and to use natural coastal seawater, along with a spiked sample, in the second interlaboratory study. 

The stability was tested at -18°C, +20°C and +37°C by analysing the material at the beginning of the storage period and at the end of the certification campaign (36 months after the first analysis). The same analytical method as the one used for the homogeneity study was applied. No instabililty could be demonstrated for any of the CB determined. It was observed that the repeatability of the methods was significantly better after three years, which was the result of the participation in interlaboratory studies [21,23]. 

---