Slow decrease in peak area when four times injected

[FW_CH8400]

Hello,

API is dissolved in mobile phase
weak and strong wash solution = mobile phase

System:
UPLC Acquity (BSM/SampleManager/ColumnHeaterCooler/PDA)
Column: ACE 3 C18, 150 x 4.6 mm
Detection: 240 nm
Flow: 1.0 ml/min
Temp: 30°C
InjVol: 5 µL
Mobile phase:
A (26 parts):
2 g sodium octane sulfonate are dissolved in 300 mL water, pH is adjusted with acetic acid to 3.5
B (74 parts):
methanol

the API solution is injected four times:
- Inj 1
- 30 other injections
- Inj 2
- 30 other injections
- Inj 3
- 30 other injections
- Inj 4

Areas of the API solution:
Inj 1: 100'000
Inj 2: 96'000
Inj 3: 94'000
Inj 4: 92'000

If the vials, the column, the wash solutions and the mobile phase components are transferred to an other UPLC system and ran on this second system, the pattern is the same.

What effect might cause this phenomenon?

Thank you very much for your help

Florian

[EmpowersBane]

I think your strong wash should be stronger than your weak wash? That way you are more likely to remove any traces of your API on the injector system like valves, needles, upstream tubing etc. I would recommend firstly changing your strong wash to a 30-40% Organic/Water ratio and also thoroughly priming the needle and injector as part of a start up.

Are you 100% sure the API is fully dissolved? This reduction in area over time suggests an API that is slowly coming out of solution and hence less detected from one injection to the next.

The fact it happened on another UPLC rules out a leak or faulty detector.

I would go back to sample prep and be absolutely sure (vortex, continued sonication) that the API is dissolved. Is there any cloudiness at all in the sample solution? Try adding the individual components of the mobile first, starting with organic, to totally dissolve the API. For example if your mobile is 20% Methanol 80% Water and your flask is 100mls for sample prep, then dissolve the sample firstly in 20mls Methanol, sonicate for 30 seconds or so, check until dissolved then add 80mls Water (scale up or down as appropriate).

If that doesn't work, double check your instrument method, loop connections, characterise your loop to confirm correct amount of volumes. Good luck!

[Poly Styrene]

I agree with EmpowersBane that it's likely sample related. The structure (or if not possible, functional group types present) could be of help. What temperature is the sample manager set at? And what is the runtime of the intermittent injections? Between API sample 1 and 4 there are more than 90 injections.. and if each run is 10 minutes a lot could happen during that time.

[FW_CH8400]

Hello,

thanks a lot for the replies.

The API concentration is 100 µg/mL; the structure could be seen here: https://en.wikipedia.org/wiki/Dequalinium. The omission of sodium octane sulfonate in the weak needle wash solution deteriorates the peak shape. In addition, it is necessary to dissolve the API completely (methanol / water in any ratio is not sufficient).

The problem did not occur over years. But in the last 3 - 5 sequences, the phenomenon could be observed.

Sample manager temperature is 25°C, run time per injection is 12 minutes.

Florian

[Poly Styrene]

If you have a fixed-loop injector and are running partial loop injection mode, the weak wash will be co-injected together with the sample which may account for peak shape issues when omitting the salt. What solvent system are the samples dissolved in? State of protonation/deprotonation of the API would alter RT, and could be unstable over time under the 'right' conditions. If it only occurred in the last 3-5 runs, are we certain that it's simply not a sample preparation issue?

[Peter Apps]

What is the API peak area from the last injection if you load a freshly prepared vial just before it is injected ?

Peter

[FW_CH8400]

Hi Peter,

I will try...

Florian