The statistical principles given in this standard are intended to assist in the understanding and development of effective methods for assigning values ​​to the properties of reference materials, including methods for assessing the associated uncertainties and establishing metrological traceability, and to help fully exploit the potential of certified reference materials (CRMs) to ensure that measurement results are comparable, accurate and consistent to national or international standards. GB/T 15000.3-2008 Guidelines for working with reference materials (3) General principles and statistical methods for assigning values ​​to reference materials GB/T15000.3-2008 Standard download decompression password: www.bzxz.net
The statistical principles given in this standard are intended to assist in the understanding and development of effective methods for assigning values ​​to the properties of reference materials, including methods for assessing the associated uncertainties and establishing metrological traceability, and to help fully exploit the potential of certified reference materials (CRMs) to ensure that measurement results are comparable, accurate and consistent to national or international standards.

ICS71.040.30
National Standard of the People's Republic of China
GB/T15000.3—2008/IS0 Guide35:2006 replaces GB/T15000.3-1994, GB/T15000.5—1994 Directives for the work of reference materials (3)
Standard samples
General principles and statistical methods for certification
Directives for the work of reference materials(3)-Reference materials--General and statistical principles for certification(ISO Guide 35:2006,Reference materials-General and statisticalprinciples for Certification, IDT)2008-03-07 Issued
General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China Standardization Administration of China
2008-08-01 Implementation
People's Republic of China
National Standard
Guidelines for Standard Samples (3]wwW.bzxz.Net
Standard Samples
General Principles and Statistical Methods for Determination
GB/T 15000. 3—2008/IS0 Cuide 35:20c6+
Published by China Standards Publishing House
No. 16, Sanlihebei Street, Fuxingmenwai, Beijing No.
Postal code: 100045
Website: spc,net.cn
Tel: 6852394668517548
Printed by Zahuangdao Printing Factory of China Standard Press Distributed by Xinhua Bookstores in various places
Format: 880×12301/16 Printing sheet: 3.5 Words: 102,000 words First edition in June 2008
First printing in June 2008
Book number: 155066·1-31429 Price: 36.00 yuan If there is any printing error, the publishing center of our company will replace the version Proprietary rights infringement must be investigated
Report phone: (010)68533533
Normative reference documents
Terms and definitions
Project design
Project definition
Raw material collection
Feasibility study
Required service life and shelf life
Sample preparation
Homogeneity study
Stability study
Selection of measurement method·
Project design summary
Measurement uncertainty assessment
CRM/RM Evaluation of uncertainty in characteristic valuesBasic model for batch determination
Sources of uncertainty
Distribution function questions
Application of ratios
Choice of coverage factors
Re-rating
Uniformity studies·
Materials·
Concept of uniformity
Practical application·
Statistically valid sampling plans and trend analysis·Evaluation of uniformity studies.
Uniformity studies between bottles
Poor repeatability of the test method
Uniformity within bottles
GB/T 15000.3—2008/I50Guide35:2006N
GB/T 15000. 3--2008/ISO Guide 35:20068 Stability study
8.1 Types of stability...
8.2 Experimental design..
8.3 Evaluation of results
8. 4 Stability monitoring
8.5 Determination of shelf life related to long-term stability..9 Determination of characteristic values
9.1 General measurement...
Establishing and demonstrating traceability
9.3 Practical approach
Measurement design
Some issues related to characteristics-
Data and uncertainty assessment
Data format
Data review
Data assessment
Uncertainty assessment
Evaluation based on uncertainty
Some specific issues
11 Certificate
Appendix A (informative)
Appendix B (informative) Material Appendix)
References
Statistical methods
GB/T15000 "Guidelines for Standard Samples" is divided into 8 parts: Part 1: General provisions for stating standard samples in technical specifications; Part 2: Common terms and definitions for standard samples; Part 3: General principles and statistical methods for the determination of the value of standard samples; Part 4: Contents of standard sample certificates and labels; Part 6: General requirements for the packaging of standard samples; Part 7: General requirements for the competence of standard sample producers; Part 8: Use of certified standard samples; Part 9: Calibration and use of certified standard samples in analytical chemistry, GB/T 15000.3—2008/ISO Guide 35:2006 This part is Part 3 of GB/T15000, corresponding to ISO Guide 35:2006 "General principles and statistical methods for the determination of the value of standard samples". This part is consistent with ISO Guide 35:2006 to the same extent. For ease of use, the following editorial changes have been made: for the international standards in the normative references of ISO Guide 35:2006, this part uses equivalent national standards instead; - "this part\" replaces "this guide"; the informational overview elements such as the foreword in ISO Guide 35:2006 have been deleted. This part replaces GB/T 15000.3-1994 "Guidelines for Standard Samples (3) General principles and statistical methods for determining the value of standard samples" and GB/T 15000.5-1994 "Guidelines for Standard Samples (5) Technical general rules for chemical analysis standard samples". Compared with GB/T 15000.3-1994 and GB/T 15000.5-1994, the main changes are as follows: the definitions of RM and CRM have been modified;
→ 13 related to RM have been added This terminology gives new requirements for the determination of values, uniformity studies and stability studies of difficult-to-standard samples. Appendices A and B of this part are both informative appendices. This part replaces GB/T15000.3-1994 and GB/T15000.5-1994 from the date of implementation. This part is proposed and coordinated by the National Technical Committee for Standard Samples (SAC/TC118). Drafting unit of this part: Secretariat of the National Technical Committee for Standard Samples. Main drafters of this part: Hu Xiaoyan, Wu Zhongxiang, Wang Xianghong, Xu Dajun, Tang Benling, Peng Zui, Chen Hongyuan, Yang Chunmei, He Ping. This part was first issued in 1994, and this is the th revision. E
CB/T 15000.3—2008/IS0 Guide 35:2006 Introduction
The production, determination and valuation of reference materials (RMs) are key activities to improve and maintain a worldwide system of measurement consistency. As indicated in GB/T15000.9--2004 and CB/T15000.8--2003, certified reference materials (CRMs) are primarily used for calibration, quality control and method validation, and are also used to assign values ​​to other materials, which in turn can become CRMs. Furthermore, CRMs are also used to maintain or establish the traceability of agreed scales such as octane number, hardness and pH. Last but not least, Importantly, some selected pure substances can also be used to maintain international temperature scales.
There are three national standards (all equivalent to ISO guidelines) to help CRM manufacturers establish a system for producing and measuring RMs and ensuring that the quality of the produced CRMs meets the requirements of end users. GB/T15000.7-2001 outlines the requirements that CRM manufacturers should meet to demonstrate their capabilities, and this part also provides methods for how to meet these requirements. This part provides a basic model for homogeneity testing, stability testing and measurement of candidate CRMs. GB/T15000.4--2003 describes the format and content of the certificate for a CRM. This part of ISO 14001 can be seen in some respects as an application of the Guide for the Expression of Uncertainty in Measurement (GUM) to the professional field of CRM production. Whenever possible, this part of ISO 14001 refers to the GUM, as the latter describes in detail how to evaluate the measurement uncertainty of a measured value. This part of ISO 14001 supplements the GUM in the sense that it provides additional guidance on the inclusion of uncertainties due to (residual) batch heterogeneity and instability in the uncertainty of the characteristic value of a CRM and on the determination of the contribution of these uncertainties. Although this part of ISO 14001 is intended to support the production and measurement of RMs, if it is used without careful consideration of whether special circumstances apply to a particular CRM, then its characteristic values ​​(and uncertainties) may be based on an erroneous or questionable basis. Users are cautioned that it is not a substitute for "rigorous thinking, intellectual honesty and professional skill" (GUM: 1993, 3.4.8). The quality of "CRM products" depends as much on these aspects as on the use of appropriate procedures and methods. In a typical valuation project, in order to correctly perform valuations and interpret experimental data, it is necessary to have a comprehensive knowledge of the materials and their characteristics, as well as the measurement methods used in homogeneity tests, stability tests and material determinations, and a comprehensive knowledge of statistical methods. This combination of skills makes RM production and valuation very complex, and the biggest challenge is to combine these skills to ensure smooth implementation of the project plan.
Most of this part can be used for RM production. Requirements such as traceability of characteristic values ​​and the need for a comprehensive evaluation of measurement uncertainty are particularly applicable to most RMs that serve as calibrators, or tools for verifying method performance, or assign values ​​to another material. Pharmacopoeia standards and drugs are established and published by pharmacopoeial authorities in accordance with the general principles of this part, and special guidelines for the production of such RMs are available. However, it should be noted that pharmacopoeial authorities use different ways to provide information to users through analytical certificates and validity dates, and do not state the uncertainty of the assigned value because it is not allowed to be specified in the summary of the use of these RMs. Scope
GB/T 15000. 3—2008/IS0 Guide 35 :2006 Guidelines for working with reference materials (3)
General principles and statistical methods for assigning values ​​to reference materials The statistical principles given in this part of ISO 14001 are intended to assist in the understanding and development of valid methods for assigning values ​​to the properties of reference materials, including methods for assessing associated uncertainties and establishing metrological traceability. Reference materials (RMs) prepared by following all the steps described in this part of ISO 14001 are normally accompanied by a certificate and are referred to as certified reference materials (CRMs). This part of ISO 14001 will help to fully exploit the potential of CRMs to ensure comparability, accuracy and consistency of measurement results on a national or international basis. In order to be comparable in time and space, measurements need to be traceable to appropriate and specified measurement standards. CRMs play an important role in establishing traceability of measurement results in chemistry, biology and physics, particularly in disciplines involving materials and/or samples. Laboratories should use CRMs as readily available measurement standards to establish the consistency of their measurement results with international standards. The traceability of the CRM is to the SI unit or other internationally agreed units. This part describes how to develop methods that can well determine the property values ​​so that they can be traced back to appropriate and specified measurement standards: This part is applicable to a wide range of materials (matrices) from mixed gases to biological samples, and also to various properties from chemical composition to physical and immunoassay properties. The methods described in this part are not necessarily applicable to all aspects of RM production and property value determination (including related uncertainties). It can be considered that the methods given in this part are the main methods for most RM production and value assignment, and appropriate modifications may be required for some special cases. The statistical methods described in this part illustrate the outline of the methods by examples and assume that the data are normally distributed. When the data are not normally distributed, it is best to use other statistical methods to obtain valid property values ​​and corresponding uncertainties. This part also outlines the design of the production CRM plan.
2 Normative references
The clauses in the following documents become the clauses of this part through reference in this part. For any dated referenced document, all subsequent amendments (excluding errata) or revisions are not applicable to this part, however, parties reaching an agreement based on this part are encouraged to study whether the latest versions of these documents can be used. For any undated referenced document, the latest version applies to this part. GB/T3358.1…1993 Statistical terminology Part 1 - General statistical terminology GB/T6379.1-2004 Accuracy (trueness and precision) of measurement methods and results Part 1: General principles and definitions CB/T6379.2-2004 Accuracy (trueness and precision) of measurement methods and results Part 2: Basic methods for determining the repeatability and reproducibility of standard measurement methods
GB/T6379.4:2006 Validity (trueness and precision) of measurement methods and results Part 4: Basic methods for determining the trueness of standard measurement methods
GB/T15000.4-2003 Guidelines for standard sample work (4) Contents of standard sample certificates and labels GB/T15000.7-2001 Guidelines for standard sample work Working Guide (7) General requirements for the competence of standard sample producers GB/T15000.8-2003 Working Guide for Standard Samples (8) Use of certified standard samples GB/T15000.9-2003 Working Guide for Standard Samples (9) Calibration and use of certified standard samples in analytical chemistry ISO Guide 30 Terms and definitions related to standard samples ISO5725-3, 2003 Accuracy (trueness and precision) of measurement methods and measurement results Part 3: Mean average accuracy of standard test methods ISO5725-5:2002 Accuracy (trueness and precision) of measurement methods and results Part 5: Alternative methods for determining the precision of standard measurement methods CB/T15000.3—2008/IS0 Guide 35:20061S05725-62002 Accuracy (trueness and precision) of measurement methods and results Part 6: Practical application of accuracy values ​​Guide for the expression of uncertainty in measurement, BIPM, EC, IFCC, ISO, IUPAC, IUPAP, OIML, 19931 International vocabulary of fundamental and general terms in metrology, BIPMEC, IFCC, ISOIUPAC, IUPAP, OIML, 1993 Note: Hereinafter, "Guide for the expression of uncertainty in measurement" will be represented by "GUM" and "International vocabulary of fundamental and general terms in metrology" will be represented by "VIM". 3 Terms and definitions
GB/T 3358.1—1993, the terms and definitions defined in ISO Guide 30 and VIM and the terms and definitions given below apply to this part of GB/T 15000. The symbols used will be given in Chapter 4. 3.1
Reference standard sample (RM) material A material that is sufficiently homogeneous and stable for one or more specified properties and has been determined to be suitable for its intended use in a measurement process. NOTE 1: RM is a general term. NOTE 2: Properties may be either quantitative or qualitative (e.g., characteristic properties of a substance or species). NOTE 3: Use may include calibration of a measurement system, evaluation of a measurement procedure, assignment of values ​​to other materials, and quality control. NOTE 4: An RM may have only a single use in a given measurement base. 3.2 Certified reference material (CRM) A reference material for which one or more specified properties have been determined using a metrologically valid procedure and is accompanied by a certificate providing the value of the specified property and its uncertainty and a statement of metrological traceability, as in 1; the concept of value includes qualitative properties, such as characteristic attributes or sequences, the uncertainty of which may be expressed in terms of probability. NOTE 2: Metrologically valid procedures for the production and assignment of values ​​to reference materials are given in GB/T 15000.7-2001 and this part of ISO 14000. NOTE 3: Guidance on the content of certificates is given in GB/T 15000.4-2003. 3.3
property value (of a reference naterial) is the value of a physical, chemical or biological property of a reference naterial. 3.4
characterization (of a reference naterial) is the process of determining the property value of a reference naterial, which is part of the process of determining the property value. Note 1: The determination process provides the property value to be determined. Note 2: When batch determination is performed, the property value of the batch is determined. 3.5
between-bottlehomogeneity refers to the differences in sample characteristics between bottles. Note: The term "bottle homogeneity" applies to other types of packaging (e.g. vials) and other physical shapes and test pieces. 3.6
within-bottle homogeneity refers to the difference in the characteristics of a sample within a bottle. 3.7
blending
mixing of two or more matrix materials to obtain a material with specific properties. 3.8
matrix material
material sampled from nature, industrial production or other places. For example: soil, drinking water, air.
1) This version was corrected and reprinted in 1995. 2
spiking
adding a known amount of a compound or element to the matrix material. 3.10
short-term stability
short-term stability
GB/T 15000,3-2008/IS0 Gaide 3.11
long-term stability
Long-term stability
Stability of the characteristics of the reference material under the storage conditions specified by the CRM manufacturer. 3.12
Lifetime (of reference material)Time interval during which the reference material can be used. 3.13
1Shelf life (of an RM/CRM)
Time interval during which the manufacturer guarantees the stability of the CRM. Note: The time interval during which the (RM/CRM) is valid is the same as that in GB/T 15000,4-2C03 The validity period of the certificate is the same. 4 Special Number
A Offset term of ANOVA
a Number of groups of ANOVA
B Shift term of ANOVA
b Number of subgroups of ANOVA
E Error term of ANOVA2)
Inclusion factor
MS Mean square of ANOVA
Number of observations
Number of (valid) (sub)group units||tt| |Number of laboratories in collaborative research
Standard deviation of (in)homogeneity between bottles
Standard deviation due to lack of repeatability
Standard deviation of long-term (in)stability
Standard deviation of repeatability
Standard deviation due to (in)stabilityStandard deviation of short-term (in)stability
Standard deviation within bottle
Sum of squares of analysis of variance (ANOVA)
Standard uncertainty due to (in)homogeneity between bottlesStandard uncertainty due to measurement
ta gtg
Standard uncertainty of characteristic value
Standard uncertainty due to long-term (in)stabilityStandard uncertainty due to short-term (in)stabilityThis standard uses the term error in a strictly statistical sense, that is, the difference between an observed value and its mathematical expectation. 2
GB/T 15000.3—2008/IS0 Guide 35:200EUcRM
Expanded uncertainty of characteristic value
Measured characteristic value
Characteristic value of CRM
Error term for (in)homogeneity between bottles
lts: Error term for long-term (in)stability sts
Error term for short-term (in)stability
Overall mean (expected) of the results of a single measurement in an analysis of variance (ANOVA) test
Note 1: In some clauses, symbols are used to illustrate typical methods used to solve statistical problems in the specified value items. These symbols are explained in the text. The symbols MS and SS are taken from the literature and do not conform to the rules for the use of symbols in 150, but because their concepts are clear, it is considered that the symbols commonly used in scientific literature are accepted in this part.
5 Project Design
5.1 General
The actual activities of CRM production require a lot of project planning, which mainly involves the quantity of materials required and the design of homogeneity, stability and assay studies, including the selection of appropriate measurement methods for these studies. The number of samples produced is a very important variable in the planning process. The number of samples and the amount of raw materials required depend on all of the above factors. In the chapters on homogeneity studies (see Chapter 7), stability studies (see Chapter 8) and assays (see Chapters 9 and 10), guidance is provided on how to plan and implement project planning. Feasibility studies are also part of project planning. 5.2 Project Definition
Project planning begins with what CRM is to be produced. A typical example is as follows: "Prepare a soil CRM containing a range of trace elements at levels suitable for environmental analytical chemistry with an uncertainty of less than or equal to % of the certified value\
This definition is a good summary of the project. Projects related to environmental chemistry may vary, but the definition fully limits the range of materials, and "soil" also limits the choice of matrix. In general, it is very important to specify what is to be produced. The definition can be further elaborated during the project design phase. Finally, the target uncertainty is set to ensure that the material will be suitable for its intended use. For example, the uncertainty of the calibration standard should be much smaller than the uncertainty of the confirmed trace element. The uncertainty of the value of the material used in the environmental analysis method. The rational choice of a "specified measurement standard" to which the traceability of the characteristic value has been established is a major design issue, which depends primarily on what measurement standards are available, what are the specific CRMs necessary for routine laboratory measurements, and what is technically feasible. Since the CRM is primarily intended to make subsequent measurements most traceable, the choice of an appropriate measurement standard is critical both scientifically and economically for the CRM characteristic values ​​produced. The scope of use of the CRM should also be stated. In most cases, the scope of use is implicit in the project definition, but sometimes further details are required. Such a scope of use does not necessarily exclude other uses, but it should be remembered that such use does not necessarily include in the certificate or documentation provided. The scope of use of the RM may be based on regulations and/or international agreements. 5.3 Transport
Before starting the actual work, it is important to consider whether the transportation of the CRM, once it has been developed, meets the requirements of existing regulations. Many CRMs present a health and safety risk when people come into direct contact with them. Suitable packaging and correct labelling are basic requirements for meeting regulations for the transport of (potentially) dangerous goods. Sometimes, laws and regulations prohibit the transport of materials with certain properties (such as viruses, diseases), which means that such CRMs cannot be sold at all. Therefore, before starting the actual valuation project, it is recommended to conduct a comprehensive review of transportation and packaging. 5.4 Raw material collection
The first task of the project is to obtain sufficient The initial material should have the required properties. For the base material, it should be noted that there may be material property limitations. Some material/property combinations are rare, or the combination with other properties is rare, and often a compromise needs to be found. Sometimes mixing and/or dosing techniques can solve this problem. 4
The amount of material required depends on the following factors:-Number of CRM/RM samples required:
…feasibility study needs;
-Number of samples required for homogeneity studies;
-Number of samples required for stability studies;
Number of samples required for candidate CRM determination;
Amount of material required for one measurement.
GB/T 15000.3—2008/IS0 Gaide 35:2006The number of samples required for candidate CRMs is a commercial issue and should be carefully planned in advance. The important variable is the maximum number of samples to be issued during the useful life of the CRM. Since service life is a function of intrinsic stability, this variable also affects the amount of raw material required. For example, many microbial materials are expected to have a shorter service life than trace element standards in sediment due to limited intrinsic stability. If the same number of samples are sent each year, less sample will be required for microbial materials than for dry sediment. However, for microbial CRMs, more samples may be required for stability studies in the first year or throughout the service life than for dry sediment. 5.5 Feasibility Study
When there is doubt about the feasibility of producing and testing a CRM that is sufficiently homogeneous and stable, a feasibility study may be considered (see Ref. [11]). Issues such as the best method for preparing the sample, the stability of the material, or the suitability for the target can be addressed in the project feasibility study (see Refs. [11], [12]). Sometimes a feasibility study is organized to allow the participating laboratories to fine-tune the equipment and procedures. For feasibility studies with testing purposes, it is recommended to use a batch of material that is slightly different from the candidate CRM. 5.6 Required Service Life and Shelf Life
The expected service life of the reference sample is an important variable in the planning of the valuation project. Another variable related to stability is the expected service life of the reference sample. The parameter is the shelf life of the CRM. Depending on the nature of the mechanism affecting the stability of the material, various measures can be taken to extend the shelf life and/or service life. Controlling the effect of moisture is one of the first methods to consider, because excessive drying or high moisture content can make the material unstable. In many cases, moisture plays a very important role in the mechanism leading to matrix and/or parameter instability. In addition, sterilization or disinfection of the material can be considered to inhibit bacterial activity, but these measures may also have a negative effect on stability. Information on stability and storage conditions can be found in the literature or obtained from users of similar materials (e.g., factories). When preparing solutions, additives can increase the service life and/or shelf life of the sample. The shelf life of the standard sample is both the storage The bacterial count is a function of the conditions and the quality of the stability study work, which determines the extent to which the shelf life can be extrapolated (see 8.5). 5.7 Sample preparation
5.7.1 Overview
It is very difficult to give general guidelines for the preparation of standard samples. This clause attempts to provide guidance on certain specific issues and mainly summarizes some issues that are often highly relevant to the success of the valuation project and require careful consideration. 5.7.2 Synthetic materials
Synthetic standard samples, such as pure substances, solutions and mixed gases, are prepared in a completely different way from most matrix standards. The preparation of pure substances must use purification techniques to reduce the total amount of impurities. The choice of purification technology depends on the target main component and can include distillation and/or recrystallization techniques. When a batch of CRMs is prepared, after aliquoting, it should be handled as described in 5.7 to 5.9. Many solutions and gas mixtures are prepared by gravimetric methods, so uncertainty estimates are usually well established. As well as uncertainty, the purity (composition) of the raw materials also enters into the model for calculating the composition of the candidate CRM. Volumetric techniques are also widely used for batch material preparation. Volumetric methods are generally slightly easier to perform, but their uncertainty is greater than that of gravimetric preparation. 5.7.3 Mixing
If a property value is considered to be too high or too low, consider mixing two or more matrices. This process is best carried out with matrices of similar types, although "like types" may still be very different. In order to achieve good mixing, the particles in the material should be controlled for agglomeration. The moisture content of the material is often the determining factor. If the material is "air-dried", agglomeration will generally (but not always) not occur during a good mixing process. This is the case with materials such as mud. If agglomerates do not disappear when combined, there may be a problem. Agglomeration of particles is unavoidable: for example, soy flour with a moisture content of less than 2% is still sticky. 5
GB/T 15000.3—2008/IS0 Guide 35:2006 Good integration of different matrices also requires that the density and particle size distribution of the materials being mixed are sufficiently similar and that the distribution is sufficiently narrow, which will actually reduce the risk of segregation. With appropriate processing and the correct use of particle crushing and mixing techniques, materials with good homogeneity and stability can generally be obtained.
If in doubt, a quick homogeneity test can be carried out on the mixed material, that is, several portions are taken from the mixed material to study the homogeneity of the property to be determined. This test can be carried out on only a few portions, but if some information on homogeneity is to be obtained, more portions should be made. It is generally considered that 10 portions can provide meaningful results to determine issues such as whether the mixed material is suitable for further processing. 5.7.4 Additions
In some cases, additions can be used as a suitable method for the production of standard samples. For example, extracts prepared from solid materials; another example is a series of 3 PCBs CRMs in pig fat, which is liquid at high temperature; good examples of the addition method to prepare CRMs with predetermined properties include liquids, metals and alloys, oils and workplace air. The main problem with addition is to make the candidate standard sample have sufficient homogeneity and stability. The use of appropriate addition methods can make materials, even solid materials, meet the requirements of homogeneity and stability. For example, a suitable method for solid addition is the "just wetting" technology; that is, the addition components are dissolved in an appropriate amount of solvent so that they are just enough to completely wet the solid surface. When choosing a solvent, a solvent with a controllable evaporation rate should be selected. If the evaporation rate is too fast, the addition material may come out of the micropores and microbeams again, and cannot be well combined with the surface, affecting the stability of the material. If the evaporation rate is too slow, other components originally in the matrix will be dissolved or even lost. However, for some matrix CRMs, addition is definitely not a suitable method to obtain materials with expected values ​​of the properties to be determined, because it may make the CRM completely different from the properties of everyday samples. It can usually be expected that natural and added The main difference between the analytes in the spiked material is in the binding, which can lead to differences in extraction behavior, for example. Therefore, the equivalence between the spiked material and the natural (contaminated) material should be checked to ensure that the material is representative of the actual sample.
5.7.5 Homogenization and packaging
The collected material usually needs to go through several preparation steps before it can become a standard sample. Necessary steps include drying, crushing, screening, stabilization and packaging/bottling. When designing the project, it should be determined to what extent the sample preparation will be achieved. For example, the collected material can be prepared into an extract that can be measured directly. In most cases, however, it is best to keep the collected material in its original state when preparing the sample, although the sample preparation process will usually reduce heterogeneity and increase stability. The required inhomogeneity of the RM characteristic value Certainty and longevity determine the requirements for the choice of sample preparation technique. It should be borne in mind that the way in which candidate standard samples are prepared will affect their possible applications, for example, distributing extracts will make it impossible to check the accuracy of the extraction procedure in the customer laboratory. Therefore, the goals of the preparation of the CRM should be kept in mind when deciding how to prepare the raw materials into a CRM that meets the application range and is suitable for the CRM to be rated. 5.8 Homogeneity studies
Homogeneity studies must be performed in batch rating projects to demonstrate that a batch of bottles (units) is sufficiently homogeneous. Quality assurance is as important as determining the bottle-to-bottle variability of the batch, which is an uncertainty component that should be included in the estimate of the uncertainty of the CRM characteristic value. Even when the material is expected to be homogeneous as a solution sample, an assessment of bottle-to-bottle heterogeneity is required. When dealing with solid standard samples , including muds and sludges, the minimum number of samples to be taken should be determined by an in-bottle homogeneity study. In principle, this homogeneity study should not increase the uncertainty of the property value concerned. The number of additional samples required depends mainly on the in-bottle homogeneity study. The minimum number of bottles to be randomly selected is between 10 and 30 and should not normally be less than 10.
The most appropriate number of samples for a homogeneity study can be determined by statistical design techniques. The design usually takes into account, for example, the fact that no inhomogeneities can be detected due to measurement uncertainty. Furthermore, the number of bottles depends on the batch size. Therefore, the number of samples taken from a batch can be considered representative of the entire batch. This requirement should be matched to the measurement uncertainty, which is a function of the repeatability standard deviation of the measurement (under repeatability conditions) and the number of replicates. The above statistical techniques can help to balance the number of bottles and the number of replicates in order to select the best method. 5.9 Stability studies
The purpose of stability testing is to determine the instability of the candidate RM after preparation or the stability of the material. Even a "stable" material may show instability in one or more property values. Stability can be divided into: 1. Stability under specified storage conditions (long-term stability); 2. Stability under specified transportation conditions.6 Required Service Life and Shelf Life
The expected service life of the CRM is an important variable in the planning of the calibration project. Another parameter related to stability is the shelf life of the CRM. Depending on the nature of the mechanism affecting the stability of the material, various measures can be taken to extend the shelf life and/or service life. Controlling the effect of aging is one of the first methods to consider, because excessive drying or excessive moisture content can make the material unstable. In many cases, moisture plays a very important role in the mechanism leading to matrix and/or parameter instability. In addition, sterilization or disinfection of the material can be considered to inhibit bacterial activity, but these measures may also have a negative effect on stability. Information on stability and storage conditions can be found in the literature or obtained from users of similar materials (e.g., factories). When preparing solutions, additives can increase the service life and/or shelf life of the samples. The shelf life of the standard sample is a function of both the storage conditions and the bacterial count of the stability study work, which determines the extent to which the shelf life can be extrapolated (see 8.5). 5.7 Sample Preparation
5.7.1 General
It is very difficult to give general guidelines for the preparation of reference standards. This clause seeks to provide guidance on specific issues and summarizes those issues that are often highly relevant to the success of a certification project and require careful consideration. 5.7.2 Synthetic Materials
Synthetic reference standards, such as pure substances, solutions, and gas mixtures, are prepared in a completely different manner from most matrix reference standards. The preparation of pure substances requires the use of purification techniques to reduce the amount of impurities. The choice of purification technique depends on the target major component and may include distillation and/or recrystallization techniques. When a batch of CRMs is prepared, after aliquoting, it should be handled as specified in 5.7 to 5.9. Many solutions and gas mixtures are prepared by gravimetric methods and therefore uncertainty estimates are usually well established. As with uncertainty, the purity (composition) of the raw materials also enters into the model for calculating the composition of the candidate CRM. Volumetric techniques are also widely used for batch material preparation. Volumetric methods are generally slightly easier to perform, but they have greater uncertainties than gravimetric preparation. 5.7.3 Mixing
If a property value is considered to be too high or too low, then mixing of two or more matrices may be considered. This process is preferably carried out with matrices of similar types, although "similar types" may still be quite different. In order to achieve good mixing, the agglomeration of particles in the material should be controlled. The moisture content of the material is often the determining factor. If the material is "air-dried", agglomeration will generally (but not always) not occur during a good mixing process. This is the case with materials such as mud. If agglomerations do not disappear when combined, there may be a problem. Agglomeration of particles is inevitable: for example, soy flour with a moisture content of less than 2% is still sticky. 5
GB/T 15000.3—2008/IS0 Guide 35:2006 Making different matrices blend well also requires that the density and particle size distribution of the materials being mixed are sufficiently similar, and the distribution should be sufficiently narrow, which will actually reduce the risk of segregation. With the right process and the correct use of particle crushing and mixing technology, materials with good homogeneity and stability can generally be obtained.
If in doubt, a quick homogeneity test can be performed on the mixed material, that is, several aliquots are taken from the mixed material to study the homogeneity of the property to be determined. This test can be done with only a few aliquots, but more aliquots are needed to obtain some information on the homogeneity. It is generally believed that 10 aliquots can provide meaningful results to determine issues such as whether the mixed material is suitable for further processing. 5.7.4 Spiking
In some cases, spiking can be a suitable method for the production of standard samples. For example, extracts prepared from solid materials; another example is a series of CRMs of 3 PCBs in pig fat, which is liquid at high temperatures; good examples of CRMs of predetermined properties prepared by spiking include liquids, metals and alloys, oils and workplace air. The main problem with spiking is to make the candidate standard samples have sufficient homogeneity and stability. The use of appropriate dosing methods can make materials, even solid materials, meet the requirements of homogeneity and stability. For example, a suitable method for solid dosing is the "just wetting" technique; that is, the dosing components are dissolved in an appropriate amount of solvent so that they are just enough to completely wet the solid surface. When selecting a solvent, a solvent with a controllable evaporation rate should be selected. If the evaporation rate is too fast, the dosing material may come out of the micropores and microbeams again and cannot be well combined with the surface, affecting the stability of the material. If the evaporation rate is too slow, other components originally in the matrix will be dissolved or even lost. However, for some matrix CRMs, dosing is definitely not a suitable method to obtain materials with expected values ​​of undetermined properties, because it may make the properties of the CRM completely different from those of everyday samples. It can usually be expected that natural and dosing The main difference between the analytes in the spiked material is in the binding, which can lead to differences in extraction behavior, for example. Therefore, the equivalence between the spiked material and the natural (contaminated) material should be checked to ensure that the material is representative of the actual sample.
5.7.5 Homogenization and packaging
The collected material usually needs to go through several preparation steps before it can become a standard sample. Necessary steps include drying, crushing, screening, stabilization and packaging/bottling. When designing the project, it should be determined to what extent the sample preparation will be achieved. For example, the collected material can be prepared into an extract that can be measured directly. In most cases, however, it is best to keep the collected material in its original state when preparing the sample, although the sample preparation process will usually reduce heterogeneity and increase stability. The required inhomogeneity of the RM characteristic value Certainty and longevity determine the requirements for the choice of sample preparation technique. It should be borne in mind that the way in which candidate standard samples are prepared will affect their possible applications, for example, distributing extracts will make it impossible to check the accuracy of the extraction procedure in the customer laboratory. Therefore, the goals of the preparation of the CRM should be kept in mind when deciding how to prepare the raw materials into a CRM that meets the application range and is suitable for the CRM to be rated. 5.8 Homogeneity studies
Homogeneity studies must be performed in batch rating projects to demonstrate that a batch of bottles (units) is sufficiently homogeneous. Quality assurance is as important as determining the bottle-to-bottle variability of the batch, which is an uncertainty component that should be included in the estimate of the uncertainty of the CRM characteristic value. Even when the material is expected to be homogeneous as a solution sample, an assessment of bottle-to-bottle heterogeneity is required. When dealing with solid standard samples , including muds and sludges, the minimum number of samples to be taken should be determined by an in-bottle homogeneity study. In principle, this homogeneity study should not increase the uncertainty of the property value concerned. The number of additional samples required depends mainly on the in-bottle homogeneity study. The minimum number of bottles to be randomly selected is between 10 and 30 and should not normally be less than 10.
The most appropriate number of samples for a homogeneity study can be determined by statistical design techniques. The design usually takes into account, for example, the fact that no inhomogeneities can be detected due to measurement uncertainty. Furthermore, the number of bottles depends on the batch size. Therefore, the number of samples taken from a batch can be considered representative of the entire batch. This requirement should be matched to the measurement uncertainty, which is a function of the repeatability standard deviation of the measurement (under repeatability conditions) and the number of replicates. The above statistical techniques can help to balance the number of bottles and the number of replicates in order to select the best method. 5.9 Stability studies
The purpose of stability testing is to determine the instability of the candidate RM after preparation or the stability of the material. Even a "stable" material may show instability in one or more property values. Stability can be divided into: 1. Stability under specified storage conditions (long-term stability); 2. Stability under specified transportation conditions.6 Required Service Life and Shelf Life
The expected service life of the CRM is an important variable in the planning of the calibration project. Another parameter related to stability is the shelf life of the CRM. Depending on the nature of the mechanism affecting the stability of the material, various measures can be taken to extend the shelf life and/or service life. Controlling the effect of aging is one of the first methods to consider, because excessive drying or excessive moisture content can make the material unstable. In many cases, moisture plays a very important role in the mechanism leading to matrix and/or parameter instability. In addition, sterilization or disinfection of the material can be considered to inhibit bacterial activity, but these measures may also have a negative effect on stability. Information on stability and storage conditions can be found in the literature or obtained from users of similar materials (e.g., factories). When preparing solutions, additives can increase the service life and/or shelf life of the samples. The shelf life of the standard sample is a function of both the storage conditions and the bacterial count of the stability study work, which determines the extent to which the shelf life can be extrapolated (see 8.5). 5.7 Sample Preparation
5.7.1 General
It is very difficult to give general guidelines for the preparation of reference standards. This clause seeks to provide guidance on specific issues and summarizes those issues that are often highly relevant to the success of a certification project and require careful consideration. 5.7.2 Synthetic Materials
Synthetic reference standards, such as pure substances, solutions, and gas mixtures, are prepared in a completely different manner from most matrix reference standards. The preparation of pure substances requires the use of purification techniques to reduce the amount of impurities. The choice of purification technique depends on the target major component and may include distillation and/or recrystallization techniques. When a batch of CRMs is prepared, after aliquoting, it should be handled as specified in 5.7 to 5.9. Many solutions and gas mixtures are prepared by gravimetric methods and therefore uncertainty estimates are usually well established. As with uncertainty, the purity (composition) of the raw materials also enters into the model for calculating the composition of the candidate CRM. Volumetric techniques are also widely used for batch material preparation. Volumetric methods are generally slightly easier to perform, but they have greater uncertainties than gravimetric preparation. 5.7.3 Mixing
If a property value is considered to be too high or too low, then mixing of two or more matrices may be considered. This process is preferably carried out with matrices of similar types, although "similar types" may still be quite different. In order to achieve good mixing, the agglomeration of particles in the material should be controlled. The moisture content of the material is often the determining factor. If the material is "air-dried", agglomeration will generally (but not always) not occur during a good mixing process. This is the case with materials such as mud. If agglomerations do not disappear when combined, there may be a problem. Agglomeration of particles is inevitable: for example, soy flour with a moisture content of less than 2% is still sticky. 5
GB/T 15000.3—2008/IS0 Guide 35:2006 Making different matrices blend well also requires that the density and particle size distribution of the materials being mixed are sufficiently similar, and the distribution should be sufficiently narrow, which will actually reduce the risk of segregation. With the right process and the correct use of particle crushing and mixing technology, materials with good homogeneity and stability can generally be obtained.
If in doubt, a quick homogeneity test can be performed on the mixed material, that is, several aliquots are taken from the mixed material to study the homogeneity of the property to be determined. This test can be done with only a few aliquots, but more aliquots are needed to obtain some information on the homogeneity. It is generally believed that 10 aliquots can provide meaningful results to determine issues such as whether the mixed material is suitable for further processing. 5.7.4 Spiking
In some cases, spiking can be a suitable method for the production of standard samples. For example, extracts prepared from solid materials; another example is a series of CRMs of 3 PCBs in pig fat, which is liquid at high temperatures; good examples of CRMs of predetermined properties prepared by spiking include liquids, metals and alloys, oils and workplace air. The main problem with spiking is to make the candidate standard samples have sufficient homogeneity and stability. The use of appropriate dosing methods can make materials, even solid materials, meet the requirements of homogeneity and stability. For example, a suitable method for solid dosing is the "just wetting" technique; that is, the dosing components are dissolved in an appropriate amount of solvent so that they are just enough to completely wet the solid surface. When selecting a solvent, a solvent with a controllable evaporation rate should be selected. If the evaporation rate is too fast, the dosing material may come out of the micropores and microbeams again and cannot be well combined with the surface, affecting the stability of the material. If the evaporation rate is too slow, other components originally in the matrix will be dissolved or even lost. However, for some matrix CRMs, dosing is definitely not a suitable method to obtain materials with expected values ​​of undetermined properties, because it may make the properties of the CRM completely different from those of everyday samples. It can usually be expected that natural and dosing The main difference between the analytes in the spiked material is in the binding, which can lead to differences in extraction behavior, for example. Therefore, the equivalence between the spiked material and the natural (contaminated) material should be checked to ensure that the material is representative of the actual sample.
5.7.5 Homogenization and packaging
The collected material usually needs to go through several preparation steps before it can become a standard sample. Necessary steps include drying, crushing, screening, stabilization and packaging/bottling. When designing the project, it should be determined to what extent the sample preparation will be achieved. For example, the collected material can be prepared into an extract that can be measured directly. In most cases, however, it is best to keep the collected material in its original state when preparing the sample, although the sample preparation process will usually reduce heterogeneity and increase stability. The required inhomogeneity of the RM characteristic value Certainty and longevity determine the requirements for the choice of sample preparation technique. It should be borne in mind that the way in which candidate standard samples are prepared will affect their possible applications, for example, distributing extracts will make it impossible to check the accuracy of the extraction procedure in the customer laboratory. Therefore, the goals of the preparation of the CRM should be kept in mind when deciding how to prepare the raw materials into a CRM that meets the application range and is suitable for the CRM to be rated. 5.8 Homogeneity studies
Homogeneity studies must be performed in batch rating projects to demonstrate that a batch of bottles (units) is sufficiently homogeneous. Quality assurance is as important as determining the bottle-to-bottle variability of the batch, which is an uncertainty component that should be included in the estimate of the uncertainty of the CRM characteristic value. Even when the material is expected to be homogeneous as a solution sample, an assessment of bottle-to-bottle heterogeneity is required. When dealing with solid standard samples , including muds and sludges, the minimum number of samples to be taken should be determined by an in-bottle homogeneity study. In principle, this homogeneity study should not increase the uncertainty of the property value concerned. The number of additional samples required depends mainly on the in-bottle homogeneity study. The minimum number of bottles to be randomly selected is between 10 and 30 and should not normally be less than 10.
The most appropriate number of samples for a homogeneity study can be determined by statistical design techniques. The design usually takes into account, for example, the fact that no inhomogeneities can be detected due to measurement uncertainty. Furthermore, the number of bottles depends on the batch size. Therefore, the number of samples taken from a batch can be considered representative of the entire batch. This requirement should be matched to the measurement uncertainty, which is a function of the repeatability standard deviation of the measurement (under repeatability conditions) and the number of replicates. The above statistical techniques can help to balance the number of bottles and the number of replicates in order to select the best method. 5.9 Stability studies
The purpose of stability testing is to determine the instability of the candidate RM after preparation or the stability of the material. Even a "stable" material may show instability in one or more property values. Stability can be divided into: 1. Stability under specified storage conditions (long-term stability); 2. Stability under specified transportation conditions.9. Many solutions and gas mixtures are prepared by gravimetric methods and therefore uncertainty estimates are usually well established. As well as uncertainty, the purity (composition) of the raw materials also enters into the model for calculating the composition of the candidate CRM. Volumetric techniques are also widely used for batch material preparation. Volumetric methods are generally slightly easier to perform, but the uncertainty is larger than that of gravimetric preparation. 5.7.3 Mixing
If a property value is considered too high or too low, consider mixing two or more matrices. This process is best carried out with matrices of similar types, although "similar types" may still be very different. In order to achieve good mixing, the particles in the material should be controlled for agglomeration. The moisture content of the material is often the determining factor. If the material is "air-dried", agglomeration will generally (but not always) not occur during a good mixing process. This is the case with materials such as mud. If agglomerates do not disappear when combined, there may be a problem. Agglomeration of particles is inevitable: for example, soy flour with a moisture content of less than 2% is still sticky. 5
GB/T 15000.3—2008/IS0 Guide 35:2006 A good combination of different matrices also requires that the density and particle size distribution of the mixed materials are sufficiently similar and that the distribution is sufficiently narrow, which will actually reduce the risk of segregation. With the use of suitable processes and the correct application of particle crushing and mixing techniques, materials with good homogeneity and stability can generally be obtained.
If in doubt, a quick homogeneity test can be carried out on the mixed material, that is, several portions are taken from the mixed material to study the homogeneity of the properties to be determined. This test can be carried out on only a few portions, but if some information on homogeneity is to be obtained, more portions should be made. It is generally believed that 10 portions can provide meaningful results to determine issues such as whether the mixed material is suitable for further processing. 5.7.4 Addition
In some cases, addition can be used as a suitable method for the production of standard samples. For example, extracts prepared from solid materials; another example is a series of 3 PCBs CRMs in pig fat, which is liquid at high temperature; good examples of the addition method to prepare CRMs with predetermined properties include liquids, metals and alloys, oils and workplace air. The main problem with addition is to make the candidate standard sample have sufficient homogeneity and stability. The use of appropriate addition methods can make materials, even solid materials, meet the requirements of homogeneity and stability. For example, a suitable method for solid addition is the "just wetting" technology; that is, the addition components are dissolved in an appropriate amount of solvent so that they are just enough to completely wet the solid surface. When choosing a solvent, a solvent with a controllable evaporation rate should be selected. If the evaporation rate is too fast, the addition material may come out of the micropores and microbeams again, and cannot be well combined with the surface, affecting the stability of the material. If the evaporation rate is too slow, other components originally in the matrix will be dissolved or even lost. However, for some matrix CRMs, addition is definitely not a suitable method to obtain materials with expected values ​​of the properties to be determined, because it may make the CRM completely different from the properties of everyday samples. It can usually be expected that natural and added The main difference between the analytes in the spiked material is in the binding, which can lead to differences in extraction behavior, for example. Therefore, the equivalence between the spiked material and the natural (contaminated) material should be checked to ensure that the material is representative of the actual sample.
5.7.5 Homogenization and packaging
The collected material usually needs to go through several preparation steps before it can become a standard sample. Necessary steps include drying, crushing, screening, stabilization and packaging/bottling. When designing the project, it should be determined to what extent the sample preparation will be achieved. For example, the collected material can be prepared into an extract that can be measured directly. In most cases, however, it is best to keep the collected material in its original state when preparing the sample, although the sample preparation process will usually reduce heterogeneity and increase stability. The required inhomogeneity of the RM characteristic value Certainty and longevity determine the requirements for the choice of sample preparation technique. It should be borne in mind that the way in which candidate standard samples are prepared will affect their possible applications, for example, distributing extracts will make it impossible to check the accuracy of the extraction procedure in the customer laboratory. Therefore, the goals of the preparation of the CRM should be kept in mind when deciding how to prepare the raw materials into a CRM that meets the application range and is suitable for the CRM to be rated. 5.8 Homogeneity studies
Homogeneity studies must be performed in batch rating projects to demonstrate that a batch of bottles (units) is sufficiently homogeneous. Quality assurance is as important as determining the bottle-to-bottle variability of the batch, which is an uncertainty component that should be included in the estimate of the uncertainty of the CRM characteristic value. Even when the material is expected to be homogeneous as a solution sample, an assessment of bottle-to-bottle heterogeneity is required. When dealing with solid standard samples , including muds and sludges, the minimum number of samples to be taken should be determined by an in-bottle homogeneity study. In principle, this homogeneity study should not increase the uncertainty of the property value concerned. The number of additional samples required depends mainly on the in-bottle homogeneity study. The minimum number of bottles to be randomly selected is between 10 and 30 and should not normally be less than 10.
The most appropriate number of samples for a homogeneity study can be determined by statistical design techniques. The design usually takes into account, for example, the fact that no inhomogeneities can be detected due to measurement uncertainty. Furthermore, the number of bottles depends on the batch size. Therefore, the number of samples taken from a batch can be considered representative of the entire batch. This requirement should be matched to the measurement uncertainty, which is a function of the repeatability standard deviation of the measurement (under repeatability conditions) and the number of replicates. The above statistical techniques can help to balance the number of bottles and the number of replicates in order to select the best method. 5.9 Stability studies
The purpose of stability testing is to determine the instability of the candidate RM after preparation or the stability of the material. Even a "stable" material may show instability in one or more property values. Stability can be divided into: 1. Stability under specified storage conditions (long-term stability); 2. Stability under specified transportation conditions.9. Many solutions and gas mixtures are prepared by gravimetric methods and therefore uncertainty estimates are usually well established. As well as uncertainty, the purity (composition) of the raw materials also enters into the model for calculating the composition of the candidate CRM. Volumetric techniques are also widely used for batch material preparation. Volumetric methods are generally slightly easier to perform, but the uncertainty is larger than that of gravimetric preparation. 5.7.3 Mixing
If a property value is considered too high or too low, consider mixing two or more matrices. This process is best carried out with matrices of similar types, although "similar types" may still be very different. In order to achieve good mixing, the particles in the material should be controlled for agglomeration. The moisture content of the material is often the determining factor. If the material is "air-dried", agglomeration will generally (but not always) not occur during a good mixing process. This is the case with materials such as mud. If agglomerates do not disappear when combined, there may be a problem. Agglomeration of particles is inevitable: for example, soy flour with a moisture content of less than 2% is still sticky. 5
GB/T 15000.3—2008/IS0 Guide 35:2006 A good combination of different matrices also requires that the density and particle size distribution of the mixed materials are sufficiently similar and that the distribution is sufficiently narrow, which will actually reduce the risk of segregation. With the use of suitable processes and the correct application of particle crushing and mixing techniques, materials with good homogeneity and stability can generally be obtained.
If in doubt, a quick homogeneity test can be carried out on the mixed material, that is, several portions are taken from the mixed material to study the homogeneity of the properties to be determined. This test can be carried out on only a few portions, but if some information on homogeneity is to be obtained, more portions should be made. It is generally believed that 10 portions can provide meaningful results to determine issues such as whether the mixed material is suitable for further processing. 5.7.4 Addition
In some cases, addition can be used as a suitable method for the production of standard samples. For example, extracts
Tip: This standard content only shows part of the intercepted content of the complete standard. If you need the complete standard, please go to the top to download the complete standard document for free.