[Prim]

Hi everyone 

I have a question if that's ok:

Can standard addition be used when the instrument response for the sample (unknown) is zero? Or when the response is less than the LOD?

I have been reading around the subject and have yet to find a definite answer. I have read some papers which do the above, and yet I can't quite sit comfortabably with the method.

Any help, insight etc much appreciated

Prim

[Archer]

Consider the question, there is no such thing as a zero instrument response (unless the detector has been diconnected!).

What is the definition of an LOD

[Prim]

Agreed, but if there is no signal that is stastically a signal or one which can be determined a signal can standard addition be used? The idea I've got from a paper is that they have spiked a known quantity into a sample which normally is below the LOD ! I can't make this sit quite right with myself, nor statsically. I could be missing something though.

LOD is the instrument limit of detection, a signal that is statistically different from the baseline or background noise. 3 x S:N ratio (3x std of the noise).

I have read through 9 different cases of standard addition (Bader 1980 - A systematic approach to standard addition methods) most of the equations use the function Rx (Instrument response of unknow) if the unknow response is a base line value (zero when you subtract the noise) then the unknown concentration will continue to equal zero.

I'm just not sure if you can indeed use standard addition in such examples.  not 100% sure though

Thanks for reply

[marquis]

Standard addition is used to make sure the method and the instrument are behaving.

The standard added should be well above the LOD.  If the standard addition comes
back with normal instrument response, it helps validate that the sample is truly below
 the LOD. 

Does this help answer your question?  Maybe I'm on the wrong track?

Good luck!

[Archer]

LOD is the instrument limit of detection, a signal that is statistically different from the baseline or background noise. 3 x S:N ratio (3x std of the noise).

Why do we set a LOD at 3 x S:N ratio?

What do you think the Standard Addition will do to the S:N ratio of your analyte?

You need to consider the precision of the analytical method to determine  whether anything below the LOD would give a meaningless result. Essentially if the LOD is relatively high but the precision is high then, although a sample with zero material in it will still provide a zero value, a value which would have been below the LOD may give a result.

The Standard Addition is essentially pushing your analyte above the noise, if it's there you will see it.

[Archer]

I have read through 9 different cases of standard addition (Bader 1980 - A systematic approach to standard addition methods) most of the equations use the function Rx (Instrument response of unknow) if the unknow response is a base line value (zero when you subtract the noise) then the unknown concentration will continue to equal zero.

I'm just not sure if you can indeed use standard addition in such examples.  not 100% sure though

Just to be clear that we are reading from the same page when describing "Standard Addition", I am describing adding the calibration standard to your analytical sample to increase the signal intensity.

[Prim]

Thanks both - muchly.

Yes we are talking about the same thing

Ultimately I have a sample in a very messy matrix (urine) I'm using mass spec, so the ion suppression is huge. We have yet to find a SPE method that will retain the compound. From literature it appears that someone used standard addition (added a known quantity) to the urine as the compound is below the LOD without the addition, if detectable at all. They then subtracted the know amount signal from the spiked sample signal.

I just see so much room for error - and I'm struggling to find an actually theoretical basis - though I've read a tonne on standard addition -

Maybe I'm all read out and I'm not seeing the wood for the trees - shall sleep on it.

Thanks much appreciated

[Prim]

Oh gosh I've also just realised I've posted in "students" sorry about that

[Archer]

What type of MS are you using is it a Triple Quad? Are you using it with LC or direct injection?

Have you considered a Molecularly Imprinted polymer (MIP)?

[Hexaflex]

The 3 X S:N is a standard usage in analytical separation techniques to set the LOD. It is the minimum significant signal of a specific analyte that can be measured in that specific system under certain conditions. It is independent from your addition and it should be determined by making standard solutions of your pure analyte in a solvent and measuring these in a simular system as will be used in the eventual method.

But in your case Prim I asume you already did that ? And that the problem you have is the LOD of the analyte in the matrix ? Which depends on the preroccessing method ?

Have you done a ionsupression test first ? To determine how much ionsupression comes from the matrix ? By measuring spiked matrix solutions with respect to standard solutions with simular concentrations ? Also it's best to do this in different concentrations (together with your calibration line) cause the supression can change up the line.

If your supression is known and stable you can do addition in your sample. But if you're talking about a signal lower than the LOD in solvent:analyte solution, the signal you will get extra on your spiked peak is probably lower then the error of your method. You should ofcource define that error first for your spiked matrix solutions ass well as standard solutions.

I hope this helps you a bit but if you want more specific information it's best if you tell wat kind of system you work with and how you determined your LOD.