pls help on the retention time drifting on LC/MS

[acetone]

Happy new year!
I got an issue that:
My samples having 5 components run on a waters UPLC/MS and column was HSS T3 1.8 um, 2.1 x 100 mm and 5 mM ammonium formate in water (A) and 5 mM ammonium formate in methanol (B).
Results: 4 of them were reproducible in RT and 5-deoxyinosine is drifting from 1.71min to 2.39 min in two runs.
The Pka of the 5-deoxyinsoine is 8.8 so the it was protonated under the PH=5 mobile phase.
I wonder if the interaction of the analyte and the mobile phase were of issue?
Thanks in advance.

[tom jupille]

pH 5 is on the ragged edge of the buffering range for formate. With a pKa of 8.8 on your sample, I would not expect that to matter much, but . . .

The first question: is the drift in retention time related to the number of samples or to the amount of time the column spends in the mobile phase (try running a sample, a couple of blanks, and then another sample). If those were "real" samples, you might be looking at some contaminant from the matrix in which case the blanks should have no effect (i.e., the change in tR will be related to the number of *samples*). If it's a mobile phase pH drift issue, then the change in tR will be related to the number of *injections*). In the latter case, perhaps going up to 10 mM (if your MS parameters will take it!) might help.

Other possibilities:

You didn't specify, but I assume you're running a gradient. You might want to go back and check the proportioning accuracy on the pump.

Temperature is always a possible culprit when retention changes.

[TheProphet]

wich run was not repro? the first injection? any chance that your colomn wasn`t equilibrated enough?

TheProphet

[acetone]

The first question: is the drift in retention time related to the number of samples or to the amount of time the column spends in the mobile phase (try running a sample, a couple of blanks, and then another sample). If those were "real" samples, you might be looking at some contaminant from the matrix in which case the blanks should have no effect (i.e., the change in tR will be related to the number of *samples*).

The RT drift was only for 1 component and other 4 were reproducible

If it's a mobile phase pH drift issue, then the change in tR will be related to the number of *injections*). In the latter case, perhaps going up to 10 mM (if your MS parameters will take it!) might help.

This RT drifting is only for one component and the other 4 have similar Pka and similar structures as the drifting one.

Thanks,

[Alexandre]

5 mM may not be enough for some strong samples. Samples should be also in buffer.
If possible use less injection vol., larger column id., dilute sample. Other words you need more buffer capacity. Samples should be in the same solvent as start of the grad, Say if you start with 20%W and 80%MeOH, sample should be exactly in 20/80.
I prefer even a little less strong if solubility not affected, say 25/75.
And filter well. Equilibrate system very well, a little remains from previous person's/assay's mob. phase in the line can create artifacts, equilibrate longer after each injection too.

[tom jupille]

The RT drift was only for 1 component and other 4 were reproducible

That may be true, but it does not answer the question I asked. Does the drift correlate better with the number of injections or with the total time the column spends in that mobile phase?

This RT drifting is only for one component and the other 4 have similar Pka and similar structures as the drifting one.

Again, this may be true, but it doesn't answer the question. Have you tried increasing the buffer concentration?