Fromic acid and periodic retention time shifting

[Zydrunas Stanius]

Dear Colleagues,

I do have a problem with one forensic method I am implementing. That's amphetamine family drugs. There are problems with peak tailing as comounds are bases, thats old an 100 times sloved problem ind different ways. Anyway the main prolem what ruines my life now is strange phenomena. I am not sure what generates it.

Phenomena: Even using isocratic run, Retention times of all the peaks with each following injection are shifting by getting shorter (equally for each compound). And every shift in time is equal to previous shift. Column has small internal volume, so it couldn't be just stupid lack of column equilibration with used mobile phase composition. It's fluched with tens of column volume before injections. The same problems occurs with the gradient. Just after half a day retention timne sof peaks are becoming more or less stable. I was playing around a lot.

Could it be something with formic acid interaction with stationary phase? Maybe there are some procceses going behind that are slow and needs time an a lot of flushing.

Maybe you have experienced something similar? Anything similar with retention times shifting? I would be very greatful for any hints.


The conditions:

2.2 µm particle column (suppose to be fast LC) 3 mmx75 mm;
Mobile phase: A: 0.5 % formic acid in water
B: ACN

All chromatogram is about 4.5-5 mins (4 compounds)
Starting ratio of eliuents is more or less 95% (A) / 5% (B)

Tried different gradient shapes (with short isocratic start or with sudden gradient form very start).




Thankyou in advance

Zydrunas

[Kostas Petritis]

Are you able to recover the retention times of your compounds after full column regeneration, reequilibration and analysis of your compounds? I do not believe that it has to do anything with the mobile phase composition, more likely build up from matrix compounds or loss of stationary phase bonding...

[Zydrunas Stanius]

Dear Kostas,

The column is brand new. Maybe the 0,5% Formic acid is quite cruel for the column, having in mind that ACN conc. in full flow is only 8 %. There was no real matrix applied on column, as no real samples were injected in this column lifetime. Has done just about 50 injections with standard solutions.

formic acid was used here as a low pH generator for keepeing silanol groups inactive as possible, because amphetamine and its "relatives" are very basic. But the significant tailing, what I was trying to avoid was still here.

What I have done now is changed from Formic acid to amonium acetate buffer with pH 4 . Now it's not too bad. The RT fluctuates around the average RT very bit, and there is no RT shifting trend as before. Even the tailing become less "in action". will check this in next days to see this is stable.


will see what happends

best wishes

[HW Mueller]

Are the "not bad" results, using NH4OAc, obtained with the column that you think was handled "quite cruel" with formic? So the formic acid was not cruel, after all?
That retention time shift in the original post indicates a flow rate change, but the increments of the shift can not stay constant over long. That indicates that either the observation was done once, over a relatively short time, or is based on wrong information. Also, since you did not answer Kostas´ question I would tend to interpret the whole thing to be a fluke, nothing to be learned, especially not about formic acid.

[Bruce Hamilton]

My suspicion would not be the adverse effect of formic acid on the column, but your sample solvent or injector. I would have the formic acid in both components of the mobile phase to keep the concentration constant.

I would also be injecting the sample in the initial mobile phase or a similar composition buffer. I suspect your CH3CO2NH4 buffer is improving the pH control, and thus peak shape. I'd also want to ensure a partial-fill injector was filled with initial mobile phase by performing some large volume blank injections at the start.

If you run isocratically, you will get a broad peak on most columns, unless you move the mobile phase pH up whilst using a suitable column.

When looking for impurities in amphetamine-like compounds, I've found ( 250 x 4.6 x 5um C18 ) that 0.1% TFA in H2O:CH3CN and a gradient of 90:10 to 30:70 in 25mins gave a reasonable main peak at 11 mins.

Please keep having fun,

Bruce Hamilton

[Uwe Neue]

You did not tell us what kind of column you are using. Some columns are less stable than others.