Using water matrix for construction of calibration curves

[k.schmidt]

Thank you Peter.

Can I know why is it the case? Is it because the recovery of the analyte might be incomplete or enhanced by the matrix?

[Peter Apps]

It is because the fraction of the analyte that partitions into the headspace, is taken as part of the headspace sample, and ends up at the detector is the same as the fraction of the standard that you add as the standard additions. The "recoveries" for native analyte and added standard are identical - so they cancel out of any equation in which you might want to include them.

One of the great strengths of the method of standard additions is that the matrix for the sample and the standard are identical, because you are creating standards by spiking portions of the sample. Under these conditions there is no need for the corrections for recovery that are required to correct for matrix effects when standards are not made up in sample matrix.

Peter

[James_Ball]

Also if doing Method of Standard Addition Calibration, then you are calibrating out the unknown amount in the matrix to determine a response factor for a known amount of analyte.

If you get a response of 25 for the matrix alone at the detector, then you add 10ppm standard and you now get a response of 50, then it would calculate out that the unknown concentration was 10ppm.

If the final response was 75 then it would calculate that the the unknown concentration was 5ppm where 5ppm+10ppm=15ppm which is proportionate to 25response + 50response=75response.

Calculated as: (10ppm/(75response-25response))=RF=0.2 then RF*matrix Response =concentration in matrix; or 0.2*25=5ppm in the matrix.

That is assuming a linear response factor. If you add several incremental known concentrations(10ppm, 20ppm, 50ppm) and calculate response factors for each you can then plot the curve and see if the response is linear with concentration, if not then you have to solve for the curvature to find the unknown concentration which is a little more complicated.

[k.schmidt]

Thank you for your replies.