Question on : Validation by Recovery ?

[Newchromatographer]

Hi folk ,

I've developed an extraction LLE method for extracting pescticide from ( soil/ water ) media .
now
My idea : want to show the efficieny of the method by spiking some concentration in * distilled water * and see how much i recover ?

My friend said : no dont spike in distilled water , he said i should spike in my test samples ( soil / water media ) and see how much i recovere ?

My question is : I feel spiking my pesticide in these soil water will not represent true recovery because apart from a pesticide will breakdown , and I am thinking the correct way is to spike in distilled water but i have proplem that will not give me a picture because i exclude the ( soil / water ) matrix effect .

plz help I will wait your answer ,, thanks .

[tom jupille]

Actually, you should do both.

[Peter Apps]

Question:

What will a spike into distilled water tell you about the performance of the extraction with real samples ?

Peter

[Newchromatographer]

Thanks Tom , I will do both but i think i will extract after spiking by just two second because the compound will degrade fast in soil media , is that what scientist usually do or they rely on something ?

Thanks Peter , when I spike in distilled water i recover 97% but will this reflect that i am going have same recovery in soil water media .Say i got 80% recovery as example in soil media , is that means 20% from my compound degraded or just the method is not good or have less effecieny , I am trying to search for long days but didnt find something convince me or give at least idea .

[Peter Apps]

If your target pesticide degrades so quickly in soil that degradation will compromise your "recoveries" in the time that it takes you to prepare a sample, then how much degradation will occur in the environment in the hours or days between application and your taking a soil sample ?, and how much pesticide will there be left in the soil for you to analyse ?

Peter

[tom jupille]

Spiking distilled water and spiking the matrix give you different information. Spiking the water will show up any problems in extraction efficiency or kinetics or adsorption to glassware, filters, etc. Spiking the matrix will show up problems with matrix effects (obviously!) such as adsorption to soil particles, formation of emulsions, extraction of interfering compounds, etc.

When all is said and done, however, Peter (as usual!) makes a very good point.

[Newchromatographer]

Good points have been posted here,

but can I say testing the first step 'extraction method efficiency' for particular easy degraded pesticide in soil/water media * does not reflect the extraction efficiency ? *

I mean if I come to your lab and I say to you look I have a very efficient extraction method just spike 10 ppm of this analyte in this soil/ water media and you will get 60% recovery ( how can I proof here that 40% loss was as a result of degradation and not as a result of adsorption ? you may say , look for new peak present but this peaks maybe from the soil/water media extracts or other things and not from my parent compounds which of which in order to say 60% is 100% efficiency .For this reason I am using distilled water instead but as Tom said ' the matrix and distilled water give different information '.

not sure if anyone understand me as I really don't understand whats going on lol.

[JGK]

Good points have been posted here,

but can I say testing the first step 'extraction method efficiency' for particular easy degraded pesticide in soil/water media * does not reflect the extraction efficiency ? *

I mean if I come to your lab and I say to you look I have a very efficient extraction method just spike 10 ppm of this analyte in this soil/ water media and you will get 60% recovery ( how can I proof here that 40% loss was as a result of degradation and not as a result of adsorption ? you may say , look for new peak present but this peaks maybe from the soil/water media extracts or other things and not from my parent compounds which of which in order to say 60% is 100% efficiency .For this reason I am using distilled water instead but as Tom said ' the matrix and distilled water give different information '.

not sure if anyone understand me as I really don't understand whats going on lol.

Comments like that are not going to get folks motivated to help you.

If the initial extraction only gives you 60% extraction, you can check whether this is due to degredation by reconstituting the sample and reextracting 2 or more times if necessary and adding the results of each extraction together. if you have alow extraction efficiency you'll see material in the subsequent extraction. If not either you've extracted everthing or its bound tightly to the soil

[Don_Hilton]

The recovery question relates to what the method is all about. Is the question how much analyte was in the sample at the time you took the sample or is the question how much analyte is left behind after a process, such as your extraction. Both are valid questions and have different implications.

If you say "I can analyze for XYZ in soil and recover between 40 and 60% when I do the extraction and the results for your sample are 10 ppm." There is a good degree of conficence that the soil concentration was not at 100 ppm when the sample was taken. If there is a remediation target or a lawsuit over the presence of XYZ over a particular value, then the strength of the argument that the sample was or was not above some limit is known.

And, the key issue is "when the sample was taken" in such a test. (Your spiking of a soil sample would correspond more to "when the sample was taken" in this case.)

If the question is how much unchanged material remains in the soil after extraction - for consumption by bugs or analysis by another chemist who is testing to see if your remediation is complete, you have an entirely different story - and one that can be very difficult to answer completely. Note JGK's suggestion that you take the extracted sample and extract it again. This is a good procedure to follow if you have low recoveries as part of the troubleshooting process.

One last thing. Years ago I was concerened that an analysis I was running and reporting samples on might not be quite right. My boss looked at me and told me that no method is 100% perfect. I was shocked that this might be acceptable. He continued on: "The method is validated?" I had to acknowledge that. "The known biases and limitations are published in the method?" This was also true. "You report the results as determined by this method?" -- Yes. And with that he shrugged and walked off.

For many of us, that recovery study is part of obtaining that method documentation -- not that we try for less than 100% recovery. And, if we cannot get 100%, we live with what we can get. But, we make sure that people who use our methods and data know the limits of what the numbers can mean.

I've seen QC a method or two which were fairly precice, but inaccurate. Because the method was precise and the bias was constant, the method was quite usable for keeping the process in controll.

[Newchromatographer]

JGK , thanks indeed you gave me lots of knowledg in your short comment .

Don , that's top notch comment , I think I ignored many factors which I have to re-consider them a gain .Experience gain from bad mistakes , as JGK signature .