In water organic analysis, such as pesticides or disinfection by-products, there is methods use calibration solution prepared directly in solvent from stock standards, while other methods prefer to prepare working standards used in calibration as sample prepared ( procedural standards ) to match conditions of standard with sample. So I want to know which method is more accurate for results?
Please check this link;
Comparison of analytical calibration methods
A set of spreadsheets that perform simulations of widely-used analytical calibration methods. Each spreadsheet simulates an "unknown" sample solution whose concentration is to be measured by an analytical instrument that generates a signal related to concentration. You can specify the "true" concentration of the unknown in the sample solution, then the spreadsheet simulates a measurement of it, using the specified calibration method to establish a calibration based on one or more standard solutions and to convert instrument readings into concentration (including typical experimental errors), reports the calculated result, and compares it to the true value that you specified. To be more realistic, these simulations include typical systematic and random errors in both signal and in volumetric measurements. They demonstrate how non-linearity, interferences, and random errors combine to influence the final result and and they allow you to optimize precision and accuracy of the measurement.
http://terpconnect.umd.edu/~toh/models/Bracket.html
As a general rule it is required to match the matrix of the calibrators with that of the sample. Therefore when we do the dilution from the stock standard, it is required to match the matrix of the sample. When we use the solvent to prepare the standards, very often the results closer to zero are not reliable due to matrix effect. For this reason it is also required to work out the minimum detection level using your assay method and below that level the assay is not valid to produce reliable results. Same as the maximum detection levels also have to be established.
I agree with the answer of Hamachandra Weerawarna. Also it is true that deuterated analogs of samples are used usually in method calibrations.
Hi,
1. Very interesting old question !
2. If you can use a very precise method for the concentration measurements of real-life samples then these data will be the best for the calibration process.
3. Is "2" is impossible then I agree with the answer of Hamachandra Weerawarna.
In residue analysis by GC or LC (particulary with MS detector), you would want your standard be in the environment that is close or similar to the sample to reduce the effect from the sample matrix (GC has signal enhancement, LC has signal suprresion). So you can do matrix matched standard curve e.g. prepare blank matrix with no analyte till the end and spike with different level of analyte to construct the calibration curve. It may be easy if you can find the blank matrix. Sometime you cannot, for instance, I found 10 pesticides in plums from Indonesia, where would you find the same plum with no pesticide in it?. If you cannot find the blank, you can use internal standard that match with your analyte (deuterated) but this IS is expensive. Imagine if I am screening 100+ analytes in my samples, it will be costly to buy all the IS. The IS will behave like your target and will cancel out the matrix effect on your peak response. For multiresidue (10+), if you don't have IS, you can do standard addition by prepare four sample extracts and spike sample 2,3,4 with different amount of analyte (1x, 2x, 3x) and plot the cal curve (y is response and x is the concentration). The sample with no spike will have y value with x = 0. The concentration of analyte in the sample would be the value of x at y = 0 or incept/slop (Google the standard addition for details).
- Badges
- Science topic
- Similar topics
- Chemistry
- Organic Chemistry