Thread Content
1. Blind sample testing and types of quality control assessments. Generally, these fall into: internal quality management； Inter-laboratory comparison ； Assessment by the certification authority ； Capability verification. The main forms include: standard sample measurement, retesting of retained samples, personnel comparison, instrument comparison, and method comparison. The main methods of quality control include: duplicate samples, spiked samples, quality control samples, quality control charts, blank experiments, verification of standard curves, calibration of instruments and equipment as part of periodic checks, recovery rates, code samples, hypothesis testing, minimum detectable limit, uncertainty, etc. Control charts can be constructed using reagent blank values, intermediate values from standard curves, or quality control samples with traceability. The preparation process of standard solutions and similar steps need to be recorded. 2 Quality control for blind sample assessment 1. Preparation before the assessment: Fully consider the five influencing factors of people, machinery, materials, methods, and environment. Relevant personnel clearly understand their respective responsibilities; operators are proficient in the methods, principles, and procedures; and report reviewers are familiar with the testing processes, key control points for assessment, and the ability to issue reports ； The instruments and their accessories, auxiliary equipment such as pre-treatment devices, and measuring tools are all within their valid calibration period and in good working condition ； Clean the measuring tools and containers for storing solutions; soak them in acid if necessary ； Reference materials, reagents, laboratory water, gas cylinders, quality control samples with known concentrations, etc., must all be within their valid period, in sufficient quantity, of appropriate purity, and in good condition; the type of sample to be analyzed and its possible concentration range ； Selection of pretreatment, detection, and quality control methods ； Control of environmental factors such as temperature, humidity, and ventilation. Consider possible types of blind samples, concentration ranges, pretreatment methods, etc., and prepare contingency measures. 2. For blind sample assessments within the laboratory, from purchase and storage to dilution and distribution, care must be taken to keep the concentration and batch numbers confidential. It should be stored in strict accordance with the requirements specified on its certificate, usually at low temperature and in a dark place. Substances such as formaldehyde, on the other hand, need to be stored at room temperature in a dark place to prevent polymerization at low temperatures. Pay attention to the physical condition and expiration date. The standard samples to be tested are generally diluted in accordance with the requirements specified in the certificate, or they can also be diluted by a multiplicative factor. Samples with poor stability should preferably be diluted according to the certificate, in order to avoid increasing the uncertainty in the dilution process. For the blind sample concentration, physical and chemical laboratories generally choose a range where the linearity is relatively stable. The parameters for measurement are generally selected from those that are tested frequently on a daily basis. For the internal assessment conducted prior to expert evaluation, parameters that are difficult to detect, difficult to characterize, or require relatively complex calculations due to their lack of stability and repeatability should be selected, in order to improve the accuracy of the results and the capabilities for data processing. Those with the necessary conditions and ample time can cover all parameters in multiple batches. Generally, the means of the paired samples are taken. If the test results fall outside or on the borderline of the acceptable range, the tested individual is required to identify the reasons and retest until the results meet the requirements. 3. For the expert evaluation in the blind sample assessment conducted by the certification authority, it is generally required that the laboratory follow the testing procedures throughout the entire process – from receiving the samples, taking them, conducting tests, issuing reports, to reviewing and approving those reports – and keep proper records of everything. The testing process should have corresponding procedural documents. Such types of assessments mainly involve parameters that are tested frequently, are unstable and difficult to measure accurately, require complex preprocessing, involve many steps, or have low threshold values; they thus place a heavy strain on the laboratory’s testing capabilities. When conducting internal assessments, the laboratory should pay attention to the difficulty level of the assessments. Upon receiving the blind sample, the laboratory should carefully verify its name, storage solvent, physicochemical state, pretreatment methods, and testing procedures. If there are any discrepancies with the requirements of the laboratory’s testing standards, it should promptly communicate with the review experts. 4. Laboratory proficiency testing – Blind sample assessment: Proficiency testing generally evaluates a laboratory’s performance using the robust statistical method specified in ISO13528, Inter-laboratory Comparative Tests, namely the classical Z-value method. General: |Z|≤1, very satisfied ； 1＜|Z|≤2, satisfactory ； 2＜|Z|≤3, suspicious, still acceptable ； |Z|＞3: Unsatisfactory, not acceptable. In the EPA water proficiency testing, to help laboratories gain a better understanding of their actual performance levels, more detailed and stringent criteria are used: |Z|≤0.15, which is considered excellent ； 0.15＜|Z|≤0.32, good ； 0.32＜|Z|≤1.65, satisfactory ； 1.65＜|Z|≤2.00, suspicious (opportunity) ； |Z|＞2, unsatisfactory (concern/unacceptable). 3 Case analysis of blind sample testing: Determination of arsenic using spectrophotometry for the assessment of arsenic in water. Since it is a metal mixed sample, dilution with dilute nitric acid is required as specified. When the reaction in the hydrogen arsenide generation device is complete, the color and absorbance values of the blank samples and quality control samples in each absorption tube are close to those of the blank, indicating that the formation of hydrogen arsenide is suppressed, likely due to interference from nitric acid. Considering metal cross-contamination, arsenic is also relatively stable in nitric acid; when the blank samples were diluted with nitric acid at mass fractions of 2‰ and 1‰ respectively, the reaction was still affected by interference. Use distilled water to directly dilute the blank sample. After the reaction in the blind sample absorption bottle, the color deepened, with a Z value of 1.32. Spectrophotometry is used to detect arsenic in cosmetics, and nitric acid also interferes with the measurement. According to national standards, nitric acid remaining after sample digestion is difficult to remove. After repeated experiments, without adding any masking agents, the solution was digested until it became clear; then 0.5 to 1 mL of sulfuric acid was added, and the temperature was increased appropriately until no brown smoke appeared. Finally, 10 to 15 mL of water was added and the mixture was heated to boiling. This improvement can effectively eliminate nitric acid interference. Arsenic is determined by atomic fluorescence method, with no interference from nitric acid. Moreover, spectrophotometry has low experimental costs and its instruments are easy to maintain. For the determination of low-concentration lead, direct sampling is employed using the lead flame atomic absorption spectrophotometry method, enabling the measurement of lead concentrations in the 10-6 range. The best method for measuring low concentrations of lead is a graphite furnace ； In the extraction-enrichment flame photometry method, the standard curve preparation and the analysis of samples involve numerous extraction steps, resulting in relatively large errors at low concentrations. To ensure the accuracy of the test results, instrument comparison was conducted using graphite furnace atomic absorption spectrophotometry and flame photometry after extraction and enrichment. By following each step carefully, the results obtained using the flame method will also be within the acceptable range. For the determination of low-concentration benzene, the FID gas chromatography method is used; there is a single peak component in the air, and quantification is carried out based on the peak area, which provides relative stability. During the assessment and measurement, it was found that the concentration of the blind sample was lower than the lowest value in the standard curve series, thus falling outside the reliable linear range. When using the thermal desorption direct injection method, an enrichment trap is required; once the sample has been completely desorbed, the enriched sample is injected all at once. If no hydrazine enrichment is added between the thermal desorber and the injection port and sampling is performed directly after desorption, the benzene peak will exhibit severe tailing and a flat peak shape. The standard curve method was used directly; since the sample concentration was lower than the lowest concentration point on the standard curve, the linearity of the standard curve was poor, which inevitably led to large errors in quantitative analysis using that standard curve. In most cases, the error amounted to over 30%. The multi-point calibration curve quantitative method specified by the national standard has been changed to a single-point calibration method. Insert the peak area of the blind sample into the standard curve equation to determine the approximate concentration range. A standard sample with a concentration close to that of the blind sample is used for comparison (it is generally required that the concentration of the standard sample be within ±100% of that of the blind sample). The peak area of the blind sample is compared with that of the standard solution to quantify the concentration of the blind sample. The test results obtained using this method can be kept within the range of ±20% of the standard value (the uncertainty range for low-concentration blind samples is also relatively wide). The GCl22 chromatograph is used to analyze benzene in workshop air; benzene desorbed from carbon disulfide onto activated carbon tubes is separated using a polyethylene glycol 6000 stainless steel column, followed by detection with a FID. Carrier gas: N2, 40 mL/min; column temperature: 80 ℃; vaporization chamber temperature: 140 ℃; detection chamber temperature: 140 ℃. Multiple quality control methods are employed: significance tests are conducted on the intercept, slope, and correlation coefficient of the calibration curve ； Use the concentrations from the standard curve to determine the recovery rate ； Six replicate tests of the blind sample were conducted, along with a Q-test ； Calculate the coefficient of variation ； Report the test value after evaluating the uncertainty. 4 Conclusion When testing blind samples, certified reference materials or standard samples of the same form and known concentration are generally tested simultaneously. For parameters with complex compositions or instability, various measurement methods or quantitative approaches can be employed, and the measurement results are determined through comprehensive comparison. The known quality control samples fall within their permissible ranges; in particular, when they are close to the target values, the test results of blind samples generally also fall within their permissible ranges. In the event of abnormalities such as instability, low concentrations, or loss of control of known quality control samples, the reasons should be analyzed, and various quality control methods should be applied flexibly to improve the accuracy of the results. Blind sample testing is a process that thoroughly evaluates and assesses the organizational management, quality control, and testing capabilities of a testing laboratory; it can effectively measure the overall level of quality control and operational performance of the laboratory ； It is also possible to compare the testing capabilities with those of domestic and international external laboratories, allowing the laboratory to more clearly understand its own level and enhancing external confidence in the laboratory.