A question about recovery of solid phase extraction

[brightsun]

Dear bro,

May I ask you a question about the recovery of SPE ?

I use C18 cartridge for lindane (organochloride pesticide) extraction from water sample.

I validate the method to test the recovery of method with concentrations of 1.5, 0.5, 0.1 mg/l . The natural pH is about 5 - 6.

But the recovery is very high : 115 ~ 125 %.

According to AOAC standard, the recovery with this range of concentration should be about 80 -110%.

I would like to ask the reason why the recovery is great like that ? And why the recovery is higher than 100%.

Thank you so much !

[Don Shelly]

You are probably seeing some matrix enhancement. Are you running on GCMS or GC/ECD?

[Peter Apps]

Have you run any blanks ?, do they have peaks at lindane's retention time ?

Peter

[brightsun]

You are probably seeing some matrix enhancement. Are you running on GCMS or GC/ECD?

I'm running GC/ECD

Have you run any blanks ?, do they have peaks at lindane's retention time ?

Peter

I ran the blank sample already. There is a peak at lindane's retention time.

Thanks for your reply !

[Don Shelly]

I think Peter has discovered the problem. Halogenated compounds are ubiquitous in the environment. Your contamination is adding to your peak height/area.

[brightsun]

I think Peter has discovered the problem. Halogenated compounds are ubiquitous in the environment. Your contamination is adding to your peak height/area.

But my sample is simulated sample, these samples are not from environment.

[Peter Apps]

No matter where the samples and reagents come from, if you have a peak in the blank it will add to the lindane peak in recovery tests, making it bigger than it would be if the peak was due only to the quantity of analyte that you spiked for the recovery. The bigger peak will lead to a result for a higher quantity of lindane, which will give you spuriously high recoveries. The preferred solution is to operate clean enough to reduce the interfering peak to neglible levels, or to change chromatographic conditions to separate the lindane from the interference. Second choice is to subtract the blank peak area from the lindane peak area, but you can only do this if the interference is at the same level all the time.

Peter

[brightsun]

No matter where the samples and reagents come from, if you have a peak in the blank it will add to the lindane peak in recovery tests, making it bigger than it would be if the peak was due only to the quantity of analyte that you spiked for the recovery. The bigger peak will lead to a result for a higher quantity of lindane, which will give you spuriously high recoveries. The preferred solution is to operate clean enough to reduce the interfering peak to neglible levels, or to change chromatographic conditions to separate the lindane from the interference. Second choice is to subtract the blank peak area from the lindane peak area, but you can only do this if the interference is at the same level all the time.

Peter

Many thanks for your advice !