Introduction

This book has been produced as a deliverable under the UK's Valid Analytical Measurement (VAM) Programme. VAM is a programme of work sponsored by the Department of Trade and Industry as part of the UK's National Measurement System. The book provides guidance and information on the application of certified reference materials (CRMs) in the context of how they can best be used to achieve valid analytical measurements and thereby improve quality in the analytical laboratory.

The main applications of CRMs in analytical chemistry, such as instrument calibration, method validation, checking laboratory performance, internal quality control and uncertainty estimation are described. Worked examples are used to illustrate key issues and to lead the analyst through the most common calculations and statistical tests employed. The first two chapters provide general information on CRMs and how they are produced, whilst in Chapter 3 the statistics relating to the use of CRMs are explained in an easy to understand manner. Chapters 4–6 describe the main applications of CRMs, and are written so that they may be read independently of each other.

1.1 Certified Reference Materials and the VAM Principles

All chemical measurements, whether qualitative or quantitative, depend upon and are ultimately traceable to, a CRM or a standard material of some sort. Qualitative measurements of identity based, for example, on gas chromatographic retention times or spectroscopic properties, require a reliable, authentic, reference material to calibrate the particular instrument or test used. Quantitative determinations have the additional requirement that the instrument is calibrated with an accurately known amount of the reference material concerned.

The key role of CRMs in analytical chemistry is recognised in the VAM Principles, a set of six statements, shown in the box, that set out the essential practical requirements for achieving valid analytical measurements.

The testing of methods and equipment referred to in VAM Principle 2 is most effectively accomplished by the use of appropriate CRMs. For example, the accuracy of the wavelength scale of UV-visible HPLC detectors can be verified by the use of a CRM comprising a solution that has absorption peaks with well characterised reference values for the wavelengths of maximum absorbance. Likewise, the accuracy of an entire analytical method, such as the determination of the proximate constituents in foods, could be checked by the use of a CRM comprising a food-type matrix with well characterised reference values for constituents such as protein (nitrogen), moisture, fat, fibre and ash.

VAM Principle 5 emphasises the importance of comparability between analytical data produced in different laboratories and locations. One effective means of achieving this is to ensure that all analytical data are traceable to reliable CRMs. A laboratory can check the repeatability of its data by setting up internal quality control (IQC) procedures to provide evidence of day-to-day consistency in its results, but a laboratory relying exclusively on IQC procedures could conceivably be producing consistent, but biased results. The use of CRMs as measurement benchmarks can provide the essential reference or anchor points against which analysts can achieve comparability of their measurements. When several laboratories can achieve the same analytical result (within the uncertainty specified) for a given CRM, they will have demonstrated the comparability of their measurements.

1.2 Definitions and Hierarchy of Reference Materials

Although this book mainly covers the uses of CRMs, they are essentially a subset of materials that are generically known as reference materials. Within the scope of the generic term there are several classifications, and the following international definitions for some of these have been published by the International Organisation for Standardisation (ISO):

• Primary Standard: Standard that is designated or widely acknowledged as having the highest metrological qualities and whose value is accepted without reference to other standards of the same quantity.

• Secondary Standard: Standard whose value is assigned by comparison with a primary standard of the same quantity.

• Certified Reference Material (CRM): Reference material, accompanied by a certificate, one or more of whose property values are certified by a procedure which establishes its traceability to an accurate realisation of the unit in which the property values are expressed, and for which each certified value is accompanied by an uncertainty at a stated level of confidence.

• Reference Material (RM): Material or substance one or more of whose property values are sufficiently homogeneous and well established to be used for the calibration of an apparatus, the assessment of a measurement method, or for assigning values to materials.

These various types of reference materials can be conveniently arranged, as shown in Figure 1.1, according to their relative metrological positions (essentially, the uncertainty associated with their measurement value) between the SI base units and routine test sample measurements.

It should be noted that the position of a particular reference material in the hierarchy is not necessarily an indication of its suitability for a particular purpose. For example, in the determination of trace copper in a soil, a matrix CRM comprising a soil with certified levels of copper would be of greater value than a primary standard consisting of ultra-pure copper.

Primary standards represent the top-tier of chemical standards and, in principle, provide a means of establishing the traceability of analytical data to the SI measurement units, such as the kilogram, the metre and the mole. The concept of primary standards in analytical chemistry has been in existence for many years, and a detailed system for establishing the traceability of a range of pure chemicals and their solutions to a primary standard of ultra-pure silver was published in 1950. Subsequently, the characteristics of primary standards were defined in more detail, and now a number of pure chemicals are recognised as primary standards.

The essential characteristics of a primary standard are that it should be:

• readily available commercially

• of high chemical purity 100% [+ or -] 0.02%

• of high stability

• homogeneous

• non-hygroscopic and non-efflorescent

• readily soluble

• of high equivalent weight (to minimise weighing errors)

• able to undergo accurate stoichiometric reaction in titration

In addition, variations in isotopic abundance should not materially affect the molecular weight. A number of pure chemicals are recognised as primary standards including pure silver, zinc, bismuth, sodium carbonate, potassium dichromate, potassium iodate, iodine, sodium chloride and sulfamic acid. The system of primary standards and the traceability that ensues from it is essentially based on classical volumetric and gravimetric analysis. However, the system does extend to certain types of trace analysis, such as trace metal analysis. It should be noted that most CRMs are not primary standards although, where practical, the characterisation of the property values of candidate CRMs should be carried out using procedures that are traceable to primary standards.

1.3 Types of Reference Material

In terms of chemical composition, there are two main types of reference material:

• single substance reference materials (including primary standards)

• matrix reference materials Single substance reference materials are pure chemicals (elements or compounds), or solutions of pure chemicals that have well-characterised reference values for such properties as chemical purity, concentration, melting point, enthalpy of fusion, viscosity, UV-visible absorbance, flash point, etc. An important use of reference materials of this type is in the calibration of analytical instrumentation and, as such, they feature in the great majority of analytical determinations. The choice of which reference material to use will depend on a variety of factors including availability, cost, suitability and the measurement uncertainty required for the measurements that are ultimately based on the use of that material.

Matrix reference materials are usually real-world materials containing the analytes of interest in their natural form and in their natural environment. Matrix reference materials should be chosen which have a matrix that closely resembles the matrix of the samples to be tested. In addition, they should ideally contain analytes with well-characterised reference values that are similar to the samples to be tested. The most important use of matrix reference materials is in the testing and validation of analytical methods. In contrast to single substance reference materials, which are primarily used in the measurement steps of an analytical process (i.e. for instrument calibration), matrix reference materials are introduced at the beginning of the analytical process. They are therefore used to assess the quality of the entire analytical process including sample extraction, clean up and concentration, as well as the final measurement step.

1.4 Uses of Reference Materials in Analytical Chemistry

Reference materials play an important role in analytical chemistry, their uses including:

• checking instrument performance (wavelength, absorbance, melting point, etc.)

• calibrating instruments

• validating methods and estimating the uncertainty of analytical measurements

• checking laboratory and analyst performance

• internal quality control

In general, CRMs should be used within the framework of a comprehensive quality assurance system. It is not acceptable to use a CRM once only and then to assume that accurate results will be produced in the future. There are, however, cases for the legitimate use of CRMs on a non-routine basis. For example, when a new method has been developed and information is needed on whether reliable and accurate results are being produced. Equally, it may be appropriate to use suitable RMs for everyday applications, once an analytical system has been properly validated and calibrated by the use of a CRM and a correlation that is fit for purpose has been derived between the results obtained with the RM and the CRM.

1.5 Interpretation of Results Obtained with CRMs

Many analysts are unsure as to how best to interpret the results obtained from the use of a CRM. This includes how to compare the actual test results obtained when using the CRM with its certified reference values and their associated uncertainties.

The most common problems voiced by analysts include:

• How many replicate measurements are required for a proper comparison of the certified reference values and the test results obtained when analysing the CRM?

• Is it necessary for the mean of the test results found for a CRM to lie within the uncertainty range of the certified value, and are any differences found actually significant?

• If the mean of the test results found for a CRM lies outside the uncertainty range, what valid conclusions can be drawn and what action should be taken (if any)?

• Is a result outside the uncertainty range acceptable and if so what are the limits that do have to be respected? What are the conclusions and necessary actions in case of non-compliance with these limits?

The answers to these questions are covered in Chapters 4–6.

1.6 Availability of Reference Materials

The huge range of different sample matrices, requiring many different analytical determinations to be carried out, results in hundreds of thousands of measurements being made every year in a wide range of laboratories. It therefore follows that the number and range of reference materials that are required is equally large and diverse.

It is the analyst's responsibility to choose the best available material for the particular requirement. Information on the availability of reference materials and CRMs can be obtained from a number of sources including:

• The COMAR certified reference material database

• Reference material producer catalogues

• Internet web sites of major reference material producers (e.g. LCC, NIST, BCR, etc.)

A list of the main CRM producers can be found in Annex A.

1.7 Quality of Certified Reference Materials

The absence of any internationally recognised classification or approval system for CRMs can make it difficult for the user to choose a reference material whose quality is appropriate for its intended use.

CRMs are supplied with certificates documenting their certified values and their associated uncertainties. The procedures by which the values were obtained and other information concerning homogeneity, stability and the correct use of the material may also be included. Some CRMs are also provided with detailed scientific reports that provide further detail. Most RMs and CRMs have well characterised property values, but some do not. It is important that a CRM has been produced and characterised in a technically valid manner. Users of CRMs need to be aware that not all materials have been validated to the same standard. Details of homogeneity trials, stability trials, method(s) used for characterisation and the uncertainties associated with the certified values are usually available from the producer and can, in part, be used to judge the pedigree and quality of the reference material. Also, guidance on CRM certificates, as well as guidance on production procedures' is available from ISO, which may provide assistance in dealing with any enquiries to producers or suppliers.

Unfortunately, at present there is no independent means by which the users of CRMs can be assured of the adequacy of the processes used to certify the materials, or of the adequacy and integrity of the information contained in CRM certificates. There is great variability in certificates issued around the world and few means exist by which users can judge for themselves those materials which are adequate and fit for purpose. This is clearly an unsatisfactory situation and one that might seriously disadvantage those producers whose materials have been rigorously characterised, Some form of independent quality assessment of reference material producers, possibly including accreditation or registration, would both help users judge the quality of the reference materials offered by different producers and allow reference material producers to demonstrate the quality of their products. The recently published IS0 Guide 34, which provides guidance on the general quality requirements for the production of reference materials, is a step in this direction.

CRM Production

The process of preparing and characterising a CRM typically consists of a number of different steps including:

• assessment of priority needs and confirmation of demand

• literature search and project plan

• raw material selection and processing

• homogeneity testing

• stability testing

• characterisation

• preparation of the certification report and certificate

The sequence of main steps in CRM production is illustrated in Figure 2.1.

Each step in the production process will be discussed briefly in the following sections in order that the user can evaluate information on the suitability and quality of a given reference material more readily.

2.1 Assessment of Priority Needs and Confirmation of Demand

This is an essential first step in the production of a new CRM. Clearly, it is important to produce CRMs that meet the priority needs of the analytical community. Assessing this need is usually carried out by the use of surveys, questionnaires, customer liaison, and discussion at technical committees and conferences. For commercial production, the demand for a new CRM needs to be sufficient to ensure that the full production and marketing costs are recovered within the life span of the material. When government funding is available, such as that under the UK's Valid Analytical Measurement (VAM) Programme, the emphasis is usually placed on producing materials which are priority needs, but for which it would not be financially viable to produce commercially.

2.2 Literature Search and Project Plan

It is important to confirm that it is technically feasible to produce the required CRM and that any potential problems are identified, especially those requiring an element of research and development prior to starting production. Producing a new CRM can take from one to five years and proper planning is essential for success. A detailed project plan for production is important in setting out exactly what needs to be done, what resources are required and when.

2.3 Raw Material Selection and Processing

Identifying a source for a particular material with the desired property values is not always straightforward and may involve the producer in much initial research and consultation with outside bodies. Choosing the potential source of a raw material is often dependent on the quantity of final material required. Commonly in the order of 100–200 kg of raw material is processed down to produce 1000–5000 units of finished material (unit sizes are typically 25–100 g for matrix materials). Sourcing and processing such large quantities of material requires specialised equipment, knowledge and expertise and is not within the normal scope of a typical chemical testing laboratory.