Good morning,
I am currently doing some standards for my HPLC. My interest compound is butyric acid and I am injecting it in four different concentrations in order to build the standard and the relative regression analysis (each concentration is done in triplicate). Now, for 1M, 0.5M and 0.05M I got consistent results in terms of area values, with a low deviation. But for my 0.1M sample, the sd of the area values is very high, therefore I'm not sure of the results.
Why this area values can be so high and different between each other?
Thanks,
Maurizio
The ICH guidelines specified a minimum of 5 concentration levels for linearity
Accuracy criteria for an assay method (FDA) is that the mean recovery will be 100 ± 2% at each concentration over the range of 80-120% of the target concentration.
May be your precision (RSD ) is too high
And check your sample prepatayoion of 0.1M solution
Antonina Kokisheva Thanks for the answer.
Sample preparation is the same for all samples, but only on the 0.1M I got this problem. What I got is that in my triplicate analysis, the first one shows a certain value, the second and the third a lower one. I repeated this test three times and the behaviour was always the same (first area value "high", then the second and the third one always lower, even 25% less).
Dear Mauizio,
Are you performing a manual injection of the sample? Are you sure the injection volume, which your injection port has been taking, is the same? Does your program pre-prepare the needle before each run? What sample concentration did you run prior to the first trial of your 0.1M? If you are sure of all these, Please check if the peak heights are the same, and also check if there is a minute variation in the retention time. Are you using a split ratio during the injection? Are there any visible variations in the triplicate peaks?
I can think of few reasons:
- First i would check the preparation step of this particular standard, maybe the pipette is faulted, or the %bias of the used volume is high.
- Did you calculate the %CV? if its less than 15,0% then no problem, and you can proceed.
- Did you check the LOD of this method? maybe the suggested LLOQ is below the limit of detection, therefore giving unstable signals for the lower part of curve.
- if the LC had problems in terms of injection volume, it would appear on few samples, not only on this one, this may appear in methods where internal rinsing is applied, however the effect on the high concentration samples would be insignificant, while it is significant for low concentration.
- the chromatographic parameters may need optimisation to achieve more rugged results, if the used method is gradient, maybe the gradient program is not good enough in terms of time vs organic content.
- I'd also suggest to re-inject from the same vial 5 times, to check if the system had issues or not.
Dear Navnit Kumar Ramamoorthy , dear Yeghig Armoudjian , I'll try to answer both. The HPLC is automatic, so I assume that injection is always the same ( 1 microliter). There's no preparation of needle before injection. During first trial, before injecting 0.1M sample, 0.5M sample was injected, but in between there is always a 10 minutes washing step. Peak heights go down like the area values (high area high peak and viceversa). Retention times are always the same, extremely close.
----------------------------------------------------
The pipette is brand new, bought exactly for these standard construction method. About the %CV, you mean to claculate it based on my area value obtained? (in this case my %cv is around 10%)
The method is always the same for all my samples, also of other organic acids. I did not have similar issues with other acids(acetate, propionate, etc.). I can think of redo the analysis with 5 instead of 3 samples.
Thanks for all your help!
Dear Maurizio,
Your approach seems to be right. However, once you've resolved the issue, please impart knowledge on why the problem occured. It would be useful for us.
- Badges
- Science method