problems with peak area

[atram]

Dear all,

Sample: citicoline sodium salt
Column: C18
MP: Phosphate buffer solution - MeOH 95:5
(Phosphate Buffer = 0,1M potassium dihydrogen phosphate and phosphoric acid solution of tetrabutyl ammonium at pH 4.5 (50:50))

In order to evaluate linearity of the method, five different concentrations are injected (three replicates for each concentration). The problem is that each first replicate present a lower area than expected, so that it is impossible to have a good R. However, if first replicate of each concentration is rejected, we get a R=0,9999.

I have checked autoinjector and repeatability when 10 replicates of water are injected is ok (weighting the vial after each injection, the CV is less than 1%).

What could be the problem? What can I do?

Thanks

[dhru]

Dear all,

Sample: citicoline sodium salt
Column: C18
MP: Phosphate buffer solution - MeOH 95:5
(Phosphate Buffer = 0,1M potassium dihydrogen phosphate and phosphoric acid solution of tetrabutyl ammonium at pH 4.5 (50:50))

In order to evaluate linearity of the method, five different concentrations are injected (three replicates for each concentration). The problem is that each first replicate present a lower area than expected, so that it is impossible to have a good R. However, if first replicate of each concentration is rejected, we get a R=0,9999.

I have checked autoinjector and repeatability when 10 replicates of water are injected is ok (weighting the vial after each injection, the CV is less than 1%).

What could be the problem? What can I do?

Thanks

Hi atram

i am not sure but may be this problem is due to carry over.

[ym3142]

very interesting;
how did you prepare the samples?concentration? diluent?

how much you injected?

what is the retention?

what is the pH of your aqueous Buffer?


was truely every first of the three of each level was different from the later two? or only one of them? how much difference?

it seems something to do with your prep.
people will come with ideas after you provide more info.

[Noser222]

Are you injecting the low ones first? It sounds like some kind of adsorption taking place, where some percentage of the analyte gets strongly bound to the column and does not elute with the normal mobile phase. If that is the case, it may help to inject a concentration higher than your standards to "saturate" the system.
Another possibility is that your ion-pair system is not fully equilibrated. I noticed on one of my methods that my initial trials on a new column did not give a good standard curve, but improved over time.

[atram]

Dear all,

Yes, the problem appears every first of different replicates, and at every concentration level: each first replicate presents a lower area. The area of the first replicate differs from the later ones by 4% aprox.
We are injecting the low concentrations first, and the difference of 4% less is maintained during all time (5 hours).

More details:
Samples range is from 0.1 ug/ml to 0.3 ug/ml, using water 100% as diluent.
To prevent carry over, we use water 100% to rinse needle.
We inject 20 ul of sample, and before each injection, we rinse loop with 450 ul of water 100% (maximum injected volume is 100 ul)
Retention time is about 4.5 minutes. This method is from Chinese Pharmacopoeia 2005 and we prepare MP exactly as it is described, for this reason we have not checked pH of final aqueous Buffer.

[ym3142]

Thanks, Atram . for your anawers. so another Q, how do you set up the sample set? what is the difference for the LC status between when you inject the first sample and the 2nd for the same level?

what will happen if you inject:
L1(concen Level 1); L2, L3, L4, L5 ; L1, L2, L3, L4 L5; L1, L2, L3, L4, L5 and so on.

[HW Mueller]

This would be understandable if you rinse only between the different concentration levels, not after each injection: Your first injection after the water "sees" an unequilibrized column and thus chromatographs differently. This should not be surprising in view of the non-buffer and the ion pair reagent.

[Alfred88]

Dear atram:

I am suggesting you to make some changes.
1. Change the wash solution to MeOH 50% (or IPA 50%), but don't use water 100%.
2. Change the vial caps to have the pre-cut septum (we found the the caps with starburst septa from Alltech to be the best). I believe that you have got the 1st penetration problem (with the vial cap!).
3. Washing the sampling needle before and after each injection. The volume can be set ~150-200uL. This is more effective and faster.

Disclaimer: I don't work for Alltech.

Good luck,
Alfred

[atram]

Dear experts,

We have been doing some tests, and maybe we know what the problem is, but not its solution.
We have realised that after a sample injection, if a blanc is injected (only MP), a very little peak with RT of our product appears (so little that its area is not calculated). Could it be a carry over problem?

When wash solution has been changed to MeOH 50%, and needle is washed before and after each injection, that small peak which appears at blanc chromatogram has been reduced a bit: With these new conditions, the area of the first replicate differs from the later ones by 2%.

What more can we do?
Thank you in advance.

[ym3142]

this is not a real carry over though you can call so.
it is more like contamination.
becuase part of sample in each injection is stuck some where along the MP path and gets out of the LC with the sample of the following injection. check all connection, especial the column to see if there is some void space which can hold sample. or injector mal function?
Test this theoy by injcetion from higher concentation to lower before you change anything.
good luck

[HW Mueller]

Carryover, its possible sources, and remedies have been discussed extensively here. In short, any mechanically moving parts (inj. valve, etc.) are quite susceptible. Other carryover types have been discussed a bit controversially, as they can be quite specific to the samples. In any case you have to do a highly systematic search (changing one thing at a time) to locate the source. Once you know where it comes from you may think of a solution. We have had one case where we could get rid of carryover only by dismantling the Rheodne, washing rotor and stator carefully, reassembling. Since one can´t do this after every injection we did these in a sequence such that carryover was negligible or subtractable.

[unmgvar]

i would go first by ym3142 test's that you inject a linearity test from higher to lower concentration.
But alfred88 suggestion regarding the septa is also good reasonning. i would test by doing a repeatbility test at one std concentration. inject one concentration 6 times but make triplicates from 2 vials. if your problem is the septa you should see the problem nicely for the first 2 injections of each vial and get an indecent RSD result with them then without them.

if it is a carryover, what type of instrument do you use? is it a split loop type injection? this information is critical for correct troubleshoot.

[atram]

Sorry, but I do not know what is a split loop type injection. Can somebody explain it to me, please? Anyway, I have a Shimadzu HPLC with a SIL-10AVdp autoinjector.

Tomorrow I am going to do the linearity test from higher to lower concentration, and repeatability test from 2 vials at same concentration. Let’s we see what happens!

[atram]

Dear experts,

We have done repeatability test from 2 vials at same concentration (3 injections of each) and the result is that first injection of each vial shows lower area. So, We suppose, as Alfred said before, there is a problem with vial cap! Now we are going to follow Alfred’s advice about pre-cut septum.

Thank you very much for all your advises!

[Peter Apps]

When did you do needle washes etc ?, before each set of triplicates and not between each injection, or between every injection ?

Peter