Chromatography Forum

Hello,

I am analyzing 05 standards solutions of creatinine , after 1/20 dilution , i inject it into HPLC (C18 column/DAD detection.Here are my results
Concentration mg/l /// Area
2,5 /// 4062,27
5 /// 8096,72
10 ///13255,34
15 /// 20784,85
20 /// 26949,78

I want to calculate the Limit of Detection (LoD), and Limit of Quantitation , can you tell me how, what data i need to calculate them?

You have the data you need for an estimate (you would have to verify by doing replicate injections at the LOD and LOQ).

Use Excel, do a linear regression and get the standard error of the y intercept (SE).
For LOD, it's 3.3 x SE / slope
For LOQ it's 10 x SE / slope

Thank you;

One last question , my standards solutions had these concentration 2,5//5//10//15 and 20 mg/l, a dilution to 1/20 was peformed before injection and the areas of each peak are :
Concentration /// Area
2,5 after dilution 1/20 => 0.125 mg/l /// 4062,27
5 after dilution 1/20 => 0.25 mg/l /// 8096,72
10 after dilution 1/20 => 0.5 mg/l ///13255,34
15 after dilution 1/20 => 0.75 mg/l /// 20784,85
20 after dilution 1/20 => 1 mg/l /// 26949,78

The limit of detection should be calculated with the concetration after dilution or with the stock standards ?
I found these results (with the concetration after dilution) :
SE: 584,7168989
Slope: 25941
LOD : 0,07438286
LOQ : 0,225402605

The limit of detection should be calculated with the concetration after dilution or with the stock standards ?

Depends on whether you want the LOD / LOQ in the original solution or in the diluted solution. Either one is OK, so long as you specify which one you're dealing with.

That said, I personally prefer to base it on the solution as injected (i.e., after dilution).

... the only thing I'd add is that if it turns out your LOD and LOQ are way, way below your lowest standard (I haven't tried it with your figures) then you should really repeat the work with more realistic values. Otherwise they are an extrapolation and may not be terribly accurate. It's always good practice, anyway, with a LOD, to check that you really can detect a peak at the LOD.