3.5 Analytical Chemical Standards
121
Tests can be based on standard solutions of the pure analyte or on matrix
solutions (extracts). While not indispensable, this step provides highly
valuable preliminary information about the identity of the participating
laboratories and the different methodologies (CMPs) they use, and helps
establish mutual relations that will no doubt favour collaborative work in
subsequent steps of the process. If the results of this campaign differ
markedly among laboratories, the exercise is repeated until closer agreement is reached.
(7) Conducing the certification campaign proper. Sample standards stored in
appropriate containers are sent to the participating laboratories for subjection to their preferred CMPs. The more different the CMPs used and the
more similar the interlaboratory results are, the higher will be the quality of
the ensuing CRM. The results of each laboratory are submitted to the
supervising body for statistical processing. Subsequently, officials from the
participating laboratories meet to reach a consensus on the certified
quantities and their uncertainties, which occasionally entails discarding
disparate results from one or more of them. Finally, a certified value to be
held as the true value, and its uncertainty, X' ± U x' , are issued.
The need for a competent, independent body to scrutinize, supervise and
sanction all these operations is self-obvious - otherwise, the CRM in question
will be subject to generic uncertainty. Consequently, not all CRMs are equally
reliable.
__ Box3. 12
Preparation of two simulated natural water sample standards (CRMs) with
certified nitrate contents (CRM 479 and CRM 480 from BCR)
Obj ective: To obtain CRMs for the determination of nitrate at concentrations below and
slightly above the maximum tolerated level in the European Union directives for underground and drinking water.
Preliminary studies: Three solutions with an NaNO J concentration of 1, 10 and 50 ~lg /L in
ultrapure water were supplied with variable amounts of Na1CO J and HCI to obtain a fina l pH
of 6.8, CaClJ and MgS04 to simulate the hardness of natural water, solid lauryl sulphate to
simulate the presence of humic acids and phenylmercury acetate as sterilizer. The stability of
the three solutions was checked by storing them at - 20, +20 and + 40 °C in conta iners of
different materials (glass, plastic). The stability of the three solutions over a period of
3 months was also examined. A preliminary interlaboratory study showed that the solutions
were suitable for certifi cation; two of them were selected to prepare as many CRMs.
Pre paration of the CRMs: Two solutions of 150 L each containing different nitrate concentrations and the same ingredients used in the preliminary experiments at thei r stated levels
were prepared in PVC containers. Under continuous stirring, each solution was used to fi ll
1200 glass vials, wh ich were heat -sea led after loading. All these operations were carried out
under strictly clean conditions. By using a highly reproducible FIA method, the samples were
shown to be homogeneous and to remain stable for at least one year.
121
Tests can be based on standard solutions of the pure analyte or on matrix
solutions (extracts). While not indispensable, this step provides highly
valuable preliminary information about the identity of the participating
laboratories and the different methodologies (CMPs) they use, and helps
establish mutual relations that will no doubt favour collaborative work in
subsequent steps of the process. If the results of this campaign differ
markedly among laboratories, the exercise is repeated until closer agreement is reached.
(7) Conducing the certification campaign proper. Sample standards stored in
appropriate containers are sent to the participating laboratories for subjection to their preferred CMPs. The more different the CMPs used and the
more similar the interlaboratory results are, the higher will be the quality of
the ensuing CRM. The results of each laboratory are submitted to the
supervising body for statistical processing. Subsequently, officials from the
participating laboratories meet to reach a consensus on the certified
quantities and their uncertainties, which occasionally entails discarding
disparate results from one or more of them. Finally, a certified value to be
held as the true value, and its uncertainty, X' ± U x' , are issued.
The need for a competent, independent body to scrutinize, supervise and
sanction all these operations is self-obvious - otherwise, the CRM in question
will be subject to generic uncertainty. Consequently, not all CRMs are equally
reliable.
__ Box3. 12
Preparation of two simulated natural water sample standards (CRMs) with
certified nitrate contents (CRM 479 and CRM 480 from BCR)
Obj ective: To obtain CRMs for the determination of nitrate at concentrations below and
slightly above the maximum tolerated level in the European Union directives for underground and drinking water.
Preliminary studies: Three solutions with an NaNO J concentration of 1, 10 and 50 ~lg /L in
ultrapure water were supplied with variable amounts of Na1CO J and HCI to obtain a fina l pH
of 6.8, CaClJ and MgS04 to simulate the hardness of natural water, solid lauryl sulphate to
simulate the presence of humic acids and phenylmercury acetate as sterilizer. The stability of
the three solutions was checked by storing them at - 20, +20 and + 40 °C in conta iners of
different materials (glass, plastic). The stability of the three solutions over a period of
3 months was also examined. A preliminary interlaboratory study showed that the solutions
were suitable for certifi cation; two of them were selected to prepare as many CRMs.
Pre paration of the CRMs: Two solutions of 150 L each containing different nitrate concentrations and the same ingredients used in the preliminary experiments at thei r stated levels
were prepared in PVC containers. Under continuous stirring, each solution was used to fi ll
1200 glass vials, wh ich were heat -sea led after loading. All these operations were carried out
under strictly clean conditions. By using a highly reproducible FIA method, the samples were
shown to be homogeneous and to remain stable for at least one year.