OXSAS helps improve data quality for QC labs through calibration setup, drift correction, type standards, and measurement of uncertainty. It allows calibration of elements, drift correction using stable samples, and type standardization for similar alloy analysis. Measurement uncertainty can be determined using certified reference materials, calibration certificates, and a top-down method accounting for reference value differences and analysis errors.
OXSAS helps improve data quality for QC labs through calibration setup, drift correction, type standards, and measurement of uncertainty. It allows calibration of elements, drift correction using stable samples, and type standardization for similar alloy analysis. Measurement uncertainty can be determined using certified reference materials, calibration certificates, and a top-down method accounting for reference value differences and analysis errors.
OXSAS helps improve data quality for QC labs through calibration setup, drift correction, type standards, and measurement of uncertainty. It allows calibration of elements, drift correction using stable samples, and type standardization for similar alloy analysis. Measurement uncertainty can be determined using certified reference materials, calibration certificates, and a top-down method accounting for reference value differences and analysis errors.
OXSAS helps improve data quality for QC labs through calibration setup, drift correction, type standards, and measurement of uncertainty. It allows calibration of elements, drift correction using stable samples, and type standardization for similar alloy analysis. Measurement uncertainty can be determined using certified reference materials, calibration certificates, and a top-down method accounting for reference value differences and analysis errors.
QC Lab (from the calibration setup, drift correction, type standard and Measurement of Uncertainty) Ling Yok Ung Application Specialist Southeast Asia & Taiwan Bulk Elemental Analyzer Business
The world leader in serving science
1 Content • Calibration Setup • Drift Correction • Type Standard • Measurement Uncertainty
2 Step for Calibration
• 1st – Select Elements for Analysis
• Not only elements of interest but also any interfering ones present
• 2nd – Select Measurement Parameters and Conditions
• Crystal, Collimator, Detector, kV, mA, etc
• 3rd – Run Scans and Energy Profiles
• Overlap, Backgrounds, Constraints, etc
• 4th – Define and Measure Drift Corrections
• Setting Up Samples for instrument drift
• 5th - Measure Standards and Create Calibrations
• Overlap and inter-elemental correction 3 Analysis Considerations
Common peak overlaps
Element Overlaps V Ka Ti Kb Cr Ka V Kb Mn Ka Cr Kb Fe Ka Mn Kb Co Ka Fe Kb Pb Ma S Ka, Mo La Pb La As Ka Si Ka W, Ta Ma Ti Ka Ba La Al Ka Br La
4 Energy Profiles • Escape Peaks – Ti, V, Cr, Mn, Fe, Co
5 Energy Profile- Second and Third Order • Second order overlay from Sb on S
6 Instrument Drift • What is drift; Loss of intensity over time received by the instrument detector
• X-ray source loss due to Rh filament tube decay
• Think if a light bulb filament
• Powder sample dusting of breaking
• Can block or reduce intensity of X-ray tube
• Crystal decay and decomposition
• Chemical attacks
7 What should be used as drift standards? • Stable Solid Sample • Glass • Polish metal
• Drift samples do not have to be same
as standards • Only concerned about elemental intensities
8 Type Standardization
• After setting up a method and computed the corrections, calibration curves etc., it is frequently the case that analysis is required of alloys of materials, which possess slightly different matrix. • Corrections can be derived and inserted into the method data (probably set to be dependent on the range of certain elements) which will enable these different alloys to be handled automatically. • However, determination of corrections can involve a major evaluation of a large number of analysis results and require the use of a large number of special standards. • An alternative approach is available within the system using type standards. • A type standard is a standard of known composition, which is very close to the unknown sample to be analyzed. The procedure used is then to store the known (nominal) concentrations of the elements together with the concentrations calculated by the associated method. The differences obtained can be used for correction in subsequent analysis of samples of this "type".
9 Measurement uncertainty in XRF & Spark-OES analysis Norms and standards • DIN EN ISO/IEC 17025: General requirements for the competence of testing and calibration laboratories • GUM (Guide to the Expression of Uncertainty in Measurement) • Eurachem/CITAC Guide: Determining Measurement Uncertainty in Analytical Measurements • BAM Guidelines for determining measurement uncertainties in quantitative test results • ISO Guide 30-35: Reference samples: production, certification, storage • DIN 32632: Guidelines for the determination of measurement uncertainties in quantitative test results • ... and a variety of other publications
10 Measurement uncertainty in XRF & Spark-OES analysis Definition
• According to GUM (Guide to the Expression of
Uncertainty of Measurement) and VIM (International Vocabulary of Metrology) the Uncertainty of Measurement is a parameter, associated with the result of a measurement, that characterizes the dispersion of the values that could reasonably be attributed to the measurand.
• Another definition of uncertainty could be
Measurement uncertainty is a range of values, usually centered on the measurement value, which contains the true value with a stated probability.
Definition and Evaluation of the Uncertainty of Measurement - Calibrate
11 Sources of uncertainty External influencers Error sources in analysis
• Homogeneity • Dimensions • Measurement strategy • Resources (adaptors) • Measuring method Sources of uncertainty Sources of uncertainties from the instrument • Standardization (drift correction) • Measurement reference materials for calibration (repeatability) • Certificate of reference materials (uncertainty) • Calibration model and corrections • Sampling of unknown sample • Variance of the of unknown sample‘s measurement (repeatability)
• Possible to directly evaluate and combine the calibration uncertainty uc and the measurement repeatability Sd • Other error sources may be estimated by the user (e.g., sampling, sample preparation) Measurement uncertainty in XRF & Spark-OES analysis
1. Reference materials • SUS (Setting Up Sample): Samples for control cards, recalibration or homogeneity tests • RM (Reference material): Samples with information about concentration and uncertainty • CRM (Certified reference material): Samples produced according to relevant standards • Traceable CRM: Reference materials from national or accredited institutes as BAM, NIST, PTB, BAS… • Producer accredited according DIN EN ISO 17034 or similar • Production and certification of the sample within the scope of accreditation
2. Calibration certificate with DAkkS symbol by TAZ Servicetechnik GmbH & CoKG • Their offering is to validate the calibration with a set of CRMS and certifying each element with comparing the certificate value with the measured
3. Possibilities for the determination of the measurement uncertainty with an
Excel file: Top-down method
15 Measurement uncertainty in Spark-OES analysis
1. Reference materials – CRM reference material
16 Measurement uncertainty in Spark-OES analysis
1. Reference materials – CRM reference material
17 Measurement uncertainty in Spark-OES analysis
2. Calibration certificate
18 Measurement uncertainty in Spark-OES analysis
2. Calibration certificate
19 Measurement uncertainty in Spark-OES analysis
3. Measurement uncertainty according to
TOP DOWN
20 Measurement uncertainty in Spark-OES analysis
3. Measuring uncertainty according to Top Down
• Advantage: • Relatively easy to calculate • Global recording of all influencing factors by measuring one or two reference sample • Recording also „hidden“ systematic influences
• Disadvantages: • Reference and test object are not always comparable • Unsuitable for very high quality requirements (e.g. CRM production) • Consideration of the difference target – actual is often unclear
21 Measurement uncertainty in Spark-OES analysis
3. Measuring uncertainty according to Top Down
1. Analysis of the unknown sample
2. Selection of one or two reference samples with similar composition and same matrix Requirements for the samples: • Certified (traceable) reference sample • „similar“ analyte content of control sample and unknown sample
3. Analysis of the control sample(s) under identical conditions
4. Calculation of the 4 uncertainty contributions 1. Error from analysis of the unknown sample 2. Error form the certificate of the reference sample 3. Error from analysis of the reference sample 4. Difference of measured and certified value of the reference sample
22 Measurement uncertainty in Spark-OES analysis Top-Down method
23 Measurement uncertainty in Spark-OES analysis Top-Down method
24 Measurement uncertainty in Spark-OES analysis
3. Summary and conclusion
- 4 different type of reference materials are available (setting up
samples, RM, CRM and traceable reference materials)
- Traceable calibration certificates are necessary for determining
measurement uncertainties and a great help in audits
- Measurement uncertainties can be determined by Bottom Up and
Top Down method
- If you are interested in the Excel file just let us know!
25 Botton up method with OXSAS
26 Potential Sources of uncertainty in analytical Measurements focus on the instrument Measurement uncertainty in spark OES analysis Integration in OXSAS analytical software example 1: Grade 304L Measurement uncertainty in spark OES analysis Integration in OXSAS analytical software Example 2: Medium alloyed steel
Advantages • User doesn’t need statistical skills • Is calculated for the average concentration • Sample related information included, mainly homogeneity • Can be activated or not • Can be further expanded by adding other sources of uncertainty Questions 30
Técnicas Estadísticas para la Ciencia de Datos a través de R. Aprendizaje Supervisado: Análisis Discriminante, Árboles de Decisión, Redes Neuronales y Modelos Lineales Generalizados
Difference Between Aac and CLC - Foam Generator Suppliers in Pune, CLC Block Making Plant in Mumbai, Pune, Gujarat, Nashik, Akola, Nellore - Foam Generator Manfacturer India
Técnicas Estadísticas para la Ciencia de Datos a través de R. Aprendizaje Supervisado: Análisis Discriminante, Árboles de Decisión, Redes Neuronales y Modelos Lineales Generalizados
Difference Between Aac and CLC - Foam Generator Suppliers in Pune, CLC Block Making Plant in Mumbai, Pune, Gujarat, Nashik, Akola, Nellore - Foam Generator Manfacturer India
