Flashcards
Sampling, calibration, quality control and validation
Flashcards do conjunto (45)
- 1Never forget need to knowPrecise analytical measurements on a non-representative sample are, at best, a waste of time.
Every process leaves a trace.
Never trust an analytical result unless there is supporting evidence to show that it is correct.
All analytical measurements are wrong – it is just a question of by how much. - 2Precise analytical measurements on a non-representative sample are, at best, a waste of timeProject design
Sample collection
Sample preparation
Instrument analysis - 3Project designEvaluate the requirements of the project e.g
What do we need to know?
Why do we need this information?
What else is in the sample?
What are the required limits of detection?
What happens to the results?
What interpretation is required – e.g. pass / fail quality control? Something more?
What actions may follow?
Review: is the approach I am proposing fit for purpose? - 4Project designWhen working with an external client it is essential to have a good dialogue with the primary client avoids misunderstanding and can reduce the amount of effort needed (hence £)
- 5Not everything comes in a little bottle…Even when it does..
Is the content of all the bottles the same…
How pure is the product, are there contaminants in the product and are they evenly distributed or present as nuggets?
Is there size segregation in the bottle which affects sampling and possibly product performance - 6Analytes and methodsAny sampling plan must define the analytes
The analytical methods to be used must be defined
-Preferably use ones that are already proven within the laboratory.
-If a new method (to the laboratory) is required then a published method should be sought and its performance evaluated. Do not assume that simply following a published method guarantees a good result – experience is invaluable. - 7Sample collectionIn many cases sample collection introduces the greatest error into the final result.
As a general rule uncertainty in analysis constitutes about ⅓rd of the total uncertainty incurred by sampling and analysis, hence ⅔rds is associated with sampling.
A representative sample must possess the essential characteristics of the bulk.
Must design a sampling strategy to ensure that a representative sample is collected.
Often the sampler is the most junior member of the team – without proper guidance and training this can compromise the result before the sample even gets to the laboratory - 8Sampling locations and incrementsNeed to define the material to be sampled in an unambiguous way – e.g. lot number etc.
Need to define the number of incremental samples to be taken from the lot and how the sampling is to be undertaken.
How big should the sample be?
How will it be obtained?
How will it be stored? - 9Approaches to sampling Random samplingSampling must ensure that any portion of the bulk has an equal chance of being sampled.
Normally used when we have little information about the material under test.
Analysts often claim to have undertaken random sampling but… is it?
Best undertaken using some form of random number generation which dictates sampling locations. - 10Sampling: examples
- 11Segregation
- 12Sampling From DrumsDecide how many drums to sample
Consider if the contents are homogeneous or require mixing / sampling from multiple depths.
Consider the effect of storage conditions (esp temperature) on the contents of the drum
- 13Sample collection How much sample do I collect?Liquid: depends on the tests to be undertaken, you may have to collect a series of samples (or subdivide a single sample) and preserve / store them in different ways. The minimum is usually 100ml
Solids: depends on factors such as grain size and homogeneity – the coarser the grain size / poorer the homogeneity the bigger the sample
-Gravel sized material 10-20kg
-Sand sized material 1-2kg
-Fine material 100-250g
Air: usually trapping contaminants on a filter paper or sorbant surface at a constant rate of flow over a given time period.
Vegetation: often restricted by availability - 14Sample collection What do I do with the sample after collection?What do I put it in?
How do I store it?
How long can I store it for?
Depends on what you want to measure….
and you should have made this decision before you collected the sample…. - 15Sample preparationMost sample types must undergo a preparation procedure to transform them into a state which is suitable for introduction into the analytical instrument.
The choice of analytical instrument may influence the prep methodology.
Most prep methodologies use a relatively small amount of sample - 16Case study: Trace Metal analysis, sample preparation for solidsChemical prep methodologies typically use between 0.25g and 10g of sample.
You will have collected more than this….
Precise analytical measurements on a non-representative sample are, at best, a waste of time.
You need to take a representative sub-sample of the material you collected - 17Trace Metal analysis: Sample preparation for solids The typical procedure:Drying
Crushing all (or a representative portion) of the sample
Sub sampling the crushed material
Milling to a fine powder with a mean grain size of <75um, considered to be homogeneous - 18Trace Metal analysis: Sample preparation for solids Drying:Temperature selection is important:
- Air dry (or oven dry) at <30°C if mercury is amongst your list of analytes or if you want speciation / bioavailability data.
-Slow process, dry to constant weight.
Oven dry at 105°C if mercury not required
-Quicker, overnight - 19Trace Metal analysis: Sample preparation for solidsAfter drying soil samples are often sieved (@2mm) to remove stones, large roots etc.
Disadvantages / issues?
Coarse samples are then crushed using a jaw crusher: - 20Trace Metal analysis: Sample preparation for solids MillingJaw crushers reduce grains to <5mm size
To reduce further (to <75 um) we must mill. Because the volume of the mill is fixed we reduce the sample volume in a representative manner by coning & quartering or similar means
Representative sub-sampling is critical because jaw crushing may have preferentially turned some phases of the sample into a fine powder whilst leaving others as large grains. - 21Trace Metal analysis: Sample preparation for solidsSample volume reduction
- Coning and quartering - 22Trace Metal analysis: Sample preparation for solidsRiffling / sample dividing
- 23Trace Metal analysis: Sample preparation for solids Milling Several optionsSwing mill
Ball mill
Mortar and pestle - 24Trace Metal analysis: Sample preparation for solids Swing mill
- 25Trace Metal analysis: Sample preparation for solids ball mill
- 26Trace Metal analysis: Sample preparation for solids Mortar and pestle
- 27Trace Metal analysis: Sample preparation for solidsAt the end of these processes the sample will be reduced to a homogeneous fine powder, fit for processing onwards
But……. what have we done…?
Every process leaves a trace - 28Trace Metal analysis: Sample preparation for solids These physical processes will affect the composition of the sample:
- 29RememberPrecise analytical measurements on a non-representative sample are, at best, a waste of time.
Every process leaves a trace. - 30CalibrationEquipment and analytical instruments invariably require calibration against standards.
In physical measurement the fundamental base units of mass, time, temperature are well defined and supported by internationally recognised standards.
These standards are used to calibrate lesser standards which are then, through a traceable chain of measurements used to calibrate local equipment such as balances, pipettes, volumetric flasks etc. The two key standards for most analytical measurements are mass and temperature. - 31CalibrationBalances, volumetric flasks, thermometers:
-Invariably calibrated during manufacture, verification of calibration must be undertaken on a regular basis.
-Balances – traceability and performance checked once a year, internal weight calibration undertaken at regular intervals; a daily weight used to monitor performance.
-Volumetric flasks checked by weight by filling to the mark with dei water.
-Thermometers – checked against a traceable reference thermometer and / or via an ice point check. - 32Instrument calibration 1) Performance calibration (verification)E.g. optical calibration or mass calibration.
Should be part of the regular (?daily) performance evaluation of the instrument.
In some instances where calibration cannot easily be adjusted by the user (e.g. a balance) the calibration of an instrument is verified by performing a task which confirms its performance. - 33Instrument calibration 2) Analyte calibrationcalibration to facilitate the quantification of unknown samples.
We normally assume that analyte signal at the instrument detector is proportional to analyte concentration in the sample. - 34Standard Methods of Calibration and QuantificationExternal standard calibration
Internal standard calibration
(external) Standard addition calibration - 35Standard Methods of Calibration and Quantification In general terms:
- 36Calibration standardsCalibration standards are made by serially diluting high purity stock solutions of one or more analytes.
stock solutions are normally purchased as either single analyte solutions or prepared blends.
Errors often occur in the simple maths associated with the dilution of standards – a straight calibration line with all points falling on the line does not guarantee accurate results - 37External standard calibration of instrumentsInvariably comparative – they rely on calibration lines for each analyte derived from materials of known concentration.
The quality of standards and calibration lines dictate the accuracy of results - 38External standard calibration of instruments
- 39Types of calibration line
- 40Correlation coefficient rA measure of the correlation (not linearity) between the x & y axes
The closer to +1 the stronger the correlation.
for a good correlation
r = > 0.996 - 41Correlation coefficient r
- 42Note: correlation ≠ causation
- 43Internal Standard CalibrationUsed when there are procedures in your analysis that can introduce significant error.
Sample to sample variation error:
-Sample transfer from one container to another
-Sample storage effect such as degradation
-Sample injection volume (Chromatography)
-Short or medium term instrument instability - 44What is an Internal Standard (IS)?A substance that is added at a constant amount to all samples, blanks and the calibration standards.
The Internal Standard is not the same as the analyte but has similar properties to the analyte.
The IS should not interfere with the detection of the analyte and vice versa.
More than one internal standard may be required for multiple analytes. - 45Internal Standard MethodMost convenient when variations in analytical sample size or matrix limit the precision of a technique.
May correct for certain types of noise.
The resulting measurement will be independent of sample size.
Species A & B must not produce signals that interfere with each other. Usually they are separated by wavelength, time or mass.
Informações
- Descrição
- FlashCards sobre Sampling, calibration, quality control and validation, criado por megan trent em 22-09-2020.
- Criado por
- megan trent
- Visualizações
- 0
- Avaliação