News from IUPAC
Report on International Bureau of Weights and Measures (BIPM)– Consultative Committee for Amount of Substance (CCQM) Working Group Meeting and Workshop on Measurement Uncertainty, 29 November–3 December 1999, Paris, France
Dr. Ales Fajgelj [Quality Assurance Supervisor, International Atomic Energy Agency (IAEA) Laboratories, A-2444 Seibersdorf, Austria; E-mail: A.Fajgelj@iaea.org], Chairman of the IUPAC Interdivisional Working Party on Harmonization of Quality Assurance Schemes for Analytical Laboratories, has submitted the following report:
As a successor to Prof. Folke Ingman in the position of IUPAC representative to the Consultative Committee for Amount of Substance (CCQM), International Bureau of Weights and Measures (BIPM), Paris, France, I attended a working group meeting and a workshop on measurement uncertainty held 29 November- 3 December 1999 at BIPM.
General Information and Observations
CCQM is a technical committee that operates as a part of BIPM, a central international metrological organization. CCQM was established in 1993 to support the BIPM mandate in
• establishing fundamental standards and scales for measuring principal physical quantities and maintaining international prototypes,
• carrying out comparisons of national and international standards,
• ensuring coordination of corresponding measurement techniques, and
• carrying out and coordinating measurements of fundamental physical constants relevant to these activities in the field of chemical measurements.
Working Groups (WGs) carry out the technical work of CCQM. Present WGs are grouped in two fields, as follows: i) primary methods, such as isotope dilution mass spectrometry, coulometry, static and dynamic analysis of gas mixtures, titrimetry, determination of freezing point depression, and NMR spectroscopy as a primary method; and ii) international comparisons, including key comparisons, organic analysis, inorganic analysis, gas analysis, and pH. In one of the most important events related to international metrological harmonization in the last decade, 49 countries signed the Mutual Recognition Arrangement (MRA) in October 1999. The MRA provides a formal basis for mutual acceptance of national measurements standards and of calibration certificates issued by national metrology institutes.
To assure and demonstrate the comparability (reproducibility of measurement results) between the measurements carried out by respective national metrological institutions, many international comparisons are being organized. Each international comparison–key comparison–is first organized as a pilot study. Several metrological institutions take part in the characterization of a selected material and establish the best estimate of a "true value" for analytes of interest and a target value for the associated measurement uncertainty. Although metrological comparisons organized by CCQM are not intended for production of reference materials, the technical principles are exactly the same.
For this reason, such work carried out by CCQM is of great interest to IUPAC in general and especially to IUPAC’s Analytical Chemistry Division and its Interdivisional Working Party on Harmonization of Quality Assurance Schemes for Analytical Laboratories. In most cases, those institutions participating in pilot studies are also the main reference materials-producing organizations, e.g., National Institute of Standards and Technology (NIST), Institute for Reference Materials and Measurements (IRMM), LGC, etc. The results of pilot studies are discussed within WGs, and all the technical reasons for eventual discrepancies are investigated before any outliers are identified and before any decision on acceptance of results is taken. I participated in the work of the WGs on Organic Analysis and Inorganic Analysis and at the workshop on measurement uncertainty organized in the framework of this CCQM meeting. Some of the following points raised during a discussion in the WGs or at the measurement uncertainty workshop might be of interest to IUPAC members; for example:
At the WGs discussion of international laboratory comparisons:
• For complicated analyses or when the complete material has to be used, more than one sample bottle (vial) is provided to the participants. One bottle is provided for training purposes.
• There is no general guidance on how participants’ data should be statistically treated. Arithmetic mean, weighted mean, median, and total median are used to express the best estimate of a "true value" on the basis of the organizers’ decision in each comparison separately.
• Criteria for data acceptance in international key comparisons are set up during pilot studies. They are established by a small number of laboratories (with demonstrated quality). The measurement capability of these laboratories has to be demonstrated at regular intervals (at least once per year). Pilot studies should define "what is reasonably achievable".
• Laboratories participating in pilot studies can withdraw their data at any time. Laboratories participating in key comparisons cannot withdraw data after submission.
• Large differences in perception between the theoretical and practical approaches related to the traceability of analytical results exist even between metrological institutions.
• Problems with inhomogeneity of intercomparison samples are often observed.
• Problems with shipment of some types of materials (liquid samples on trans-Atlantic flights, customs regulations, etc.) are often reported.
• The extent of instructions given to the participants was discussed. Regardless of the instructions given, there are always some laboratories that are not following the instructions.
At the measurement uncertainty workshop:
• What are the consequences of the given mean value and associated uncertainty? In many cases, uncertainties of results reported from different laboratories do not coincide with those given as target values. Which data are still acceptable?
• How should we combine results and associated uncertainties from a single laboratory, as well as sets of data from different laboratories?
Most of the discussion points are still open questions. No general answers were provided, and there is still a lot to do in the harmonization of these open questions. At this CCQM meeting, an important change in the perception and classification of primary methods of analysis was observed. The potential of all techniques to demonstrate traceability of measurement results obtained to the International System of Units (SI) should be reinvestigated. It was pointed out that analytical techniques could not be declared as primary per se. Their potential should be demonstrated for each sample/ measurand/analyte/technique combination. For this reason, as a first step in this direction, an international symposium was planned for April 2000 at BIPM. In connection with the "single-laboratory method validation principle" proposed and discussed at the AOAC/FAO/ IAEA/IUPAC Workshop in Budapest in November 1999, this new perception of primary methods of analysis is very promising for different analytical techniques. Being declared as primary and accepted as fully validated, analytical techniques might be applied to a wide range of analysis required for legislative and international trading purposes.
As agreed before the meeting, I had the opportunity to meet individually with Dr. Terry Quinn, Director of BIPM; Dr. R. Kaarls, CCQM President; and Dr. R. Davis, CCQM Executive Secretary.
In all cases, the importance of international cooperation in the field of chemical metrology was pointed out. The relationships between BIPM, CCQM, and IUPAC were found to be successful. National and international needs relating to metrology are well elaborated in the report prepared by BIPM for governments of Member States of the Convention of the Metre in 1998. (Copies are available free of charge from BIPM.) It was pointed out that IUPAC input into metrology in chemistry might even be larger.
The next meeting of CCQM took place 4-7 April 2000 at BIPM in Sèvres, France. It was combined with the working group meetings and with the international symposium on primary methods. For more information about BIPM and CCQM, visit the BIPM web site at http://www.bipm.fr/.
About the article
Published Online: 2009-09-01
Published in Print: 2000-05-01