120
3 Traceability: Reference Materials
The most thorough approach possible involves realizing traceability via equipment and
method calibration, using primary standards allegedly linked to 51 units. However, what is
usually available is one or more substances of purity or composition certified exclusively by
the supplier that must be linked to CRMs or 51 units via transfer standards in much the same
way as in physical metrology.
The use of primary methods (gravimetry, titrimetry, coulometry, isotope dilution mass
spectrometry). which are at the top of the metrological ranking, to check the results of routine
CMPs is one alternative to realizing traceability.
The results obtained by subjecting certified reference materials (CRMs) to CMPs provide
a means for realizing global traceability in the latter.
Sample (matrix) standards can be prepared by individual institutions (e.g.
a scientific society, a laboratory group) or by international bodies. Those
standards that meet the following stringent requirements are endorsed as CRMs:
(a) The matrix should be as similar as possible to that of the real sample. Natural
samples are better than partially natural ones (e.g. samples spiked with the
target analytes), which in turn are preferable to artificial samples.
(b) The material should be homogeneous and stable.
(c) The reference quantities should all have been determined with a high accuracy and a low uncertainty by different laboratories using various CMPs;
also, they should be traceable to base standards.
(d) A complete "history" of the procedures used to prepare the standard and
measure its certified uncertainties should be included in the accompanying
documentation.
Obtaining a standard stock (e.g. 1000-3000 samples containing 100 mL or 100 g
of a CRM) usually involves the following operations:
(1) Selecting the natural starting material.
(2) Preparing the selected material by freeze-drying, grinding, sieving, blending, etc., depending, among others, on the state of aggregation of the raw
material; thus, some types of liquid samples (e.g. milk) are freeze-dried
while others (e.g. sea water) are not.
(3) Preparing an artificial material. Artificial samples are simply spiked with the
analyte as uniformly as possible - both on the surface and inside the sample
if it is a solid (e. g. soil soaked in a pesticide solution in a volatile solvent that
is allowed to freely evaporate with time). Liquid artificial samples are easier
to prepare.
(4) Checking for sample stability in various containers under different conditions (e.g. a variable temperature, presence and absence of light) over a
period of a least one year, using the same CMP throughout.
(5) Distributing the material among small containers of the selected type in
order to check for homogeneity (both in the bulk material prior to splitting
and in its portions) by using statistical criteria to process the data provided
by CMPs applied to different aliquots.
(6) Conducting a preliminary campaign to check the suitability of the laboratories that are to take part in the certification of the reference parameters.
3 Traceability: Reference Materials
The most thorough approach possible involves realizing traceability via equipment and
method calibration, using primary standards allegedly linked to 51 units. However, what is
usually available is one or more substances of purity or composition certified exclusively by
the supplier that must be linked to CRMs or 51 units via transfer standards in much the same
way as in physical metrology.
The use of primary methods (gravimetry, titrimetry, coulometry, isotope dilution mass
spectrometry). which are at the top of the metrological ranking, to check the results of routine
CMPs is one alternative to realizing traceability.
The results obtained by subjecting certified reference materials (CRMs) to CMPs provide
a means for realizing global traceability in the latter.
Sample (matrix) standards can be prepared by individual institutions (e.g.
a scientific society, a laboratory group) or by international bodies. Those
standards that meet the following stringent requirements are endorsed as CRMs:
(a) The matrix should be as similar as possible to that of the real sample. Natural
samples are better than partially natural ones (e.g. samples spiked with the
target analytes), which in turn are preferable to artificial samples.
(b) The material should be homogeneous and stable.
(c) The reference quantities should all have been determined with a high accuracy and a low uncertainty by different laboratories using various CMPs;
also, they should be traceable to base standards.
(d) A complete "history" of the procedures used to prepare the standard and
measure its certified uncertainties should be included in the accompanying
documentation.
Obtaining a standard stock (e.g. 1000-3000 samples containing 100 mL or 100 g
of a CRM) usually involves the following operations:
(1) Selecting the natural starting material.
(2) Preparing the selected material by freeze-drying, grinding, sieving, blending, etc., depending, among others, on the state of aggregation of the raw
material; thus, some types of liquid samples (e.g. milk) are freeze-dried
while others (e.g. sea water) are not.
(3) Preparing an artificial material. Artificial samples are simply spiked with the
analyte as uniformly as possible - both on the surface and inside the sample
if it is a solid (e. g. soil soaked in a pesticide solution in a volatile solvent that
is allowed to freely evaporate with time). Liquid artificial samples are easier
to prepare.
(4) Checking for sample stability in various containers under different conditions (e.g. a variable temperature, presence and absence of light) over a
period of a least one year, using the same CMP throughout.
(5) Distributing the material among small containers of the selected type in
order to check for homogeneity (both in the bulk material prior to splitting
and in its portions) by using statistical criteria to process the data provided
by CMPs applied to different aliquots.
(6) Conducting a preliminary campaign to check the suitability of the laboratories that are to take part in the certification of the reference parameters.