retention shift in only in plasma matrix

[HW Mueller]

If your standard is not effected than crud on the column is not the cause (except if it co-elutes). What I find very strange is that the peaks with longer retention time are sharper. Are you sure that those peaks are your analyte?
One other thought: There is really a variable mismatch of sample solvent/matrix and mobile phase.

[Uwe Neue]

It looks to me as if the first part is still in a nearly isocratic section of the C-gram. As you inject more crud, the peak moves more into this section.

[neuger]

HW, what do you mean by variable mismatch? The sample is reconstituted in 0.1% TFA. The MP is dominated by this component near both drug's retention times, we had 10%ACN at one point in our reconst sol'n, but it didn't seem to matter. Please explain what this mismatch means??

Also, I am positive about the peaks. We've done this in neat sol'n 50 times over. The sharper peak is one drug of interest who's retention is NOT affected and the other is, but only in matrix. ... I don't know if this info will mean anything, but the peak that is moving is the phosphorylated version of the same drug at 11 min. They are tricyclic nucleosides. the phosphorylated one at 9min is actually much less hydrophobic than the other at 11min, yet they are retained very close to each other.

Uwe, the gradient is only used to hold the compounds on longer and sweep them off due to endogenous interferences we had. Yes, it does appear that it is coming out in the isocratic part if you just looked at peak shape, but the earlier it comes out, the closer it is actually to still being in the gradient if you review my method. That I simply don't understand. SPE was not too good, zwitterionic drug, we tried the other end, PSA and SAX and it was very difficult, this molecule changes so much at different pHs, but we could try mixed mode in acidic form if all else fails.

[juddc]

Hello again -

Your 9 min peak could actually be in or near the isocratic portion of the chromatogram as Uwe suggested based upon your flow rate and your system volume. Your gradient description only tells us what you programmed the gradient controller to do, not what what the on-column MP composition is at any given time. Given a 2.1x150mm column, I might exect a pretty significant delay if you'r flow rate is slow or your system volume is large (or both).

I'd also suggest that your sharper peak is affected by whatever issue you're having. The reason I say that is if you look at the chromatograms, the retention times are decreasing for both, just much less for the later peak. From highest to lowest retention, the order is purple, green, brown, black for both peaks. I'd predict that if you found intermediate isocratic conditions to separate both peaks, you'd see a greater effect on peak #2 than you do now with the gradient.

Regarding SPE, do you have problems retaining or eluting your compounds? You've said the SPE is difficult on these, but I don't know what you've tried. What I was suggesting with SPE is a setup where your analyte would elute and contaminants, if any, would be retained. Being that your analytical method is RP under which your compounds elute easily, I might suggest RP SPE as a simple cleanup. That way you'd hang any junk on the SPE cartridge and still have all of your analyte in your eluate, which you could then evaporate and reconstitute as you wish.

[neuger]

my bad, I forgot some info...flow is 0.4ml/min. is there really a possible delay of 5 minutes to my column? I start the gradient at 4 min and the peak is at 9, but this still doesn't explain any shifting, does it?

[juddc]

Nope, it doesn't explain the shift itself, but it may explain the differential in the magnitude of the shift between the two peaks. A 5 min delay (or something close to one) might be conceivable at 0.4 mL/min, depending upon your extra column volume. It sounds a bit long, but you're only talking about 2 mL. System volume for a typical Waters Alliance is somewhere around 650 uL, I think. Add your column volume to that and you're up near 2 mL pretty easily, I'd guess.

Do you keep track of system pressure as an acquired channel on your system? That might let you better determing where your peak is eluting in terms of the actual on-column MP composition.

[neuger]

yeah, we can track it, I think I follow what you mean. we could then figure out at least when the change is taking place, right? But where does that lead to?

Also, it is true, both peaks are affected by this in plasma matrix, I think I was the one who actually pointed it out the first time. it is small, but both get shorter retention over time....

in terms of SPE are you saying condition and send plasma through, but then don't wash? We know if we wash with anything it is gone in the wash, we've done it before...with enough volume and force water will take it away...Is it possible in SPE to load the matrix after conditioning and flow this thru to waste and then simply elute with water in the next step and hope to have few components other than the hydrophilic analyte elute?

[Uwe Neue]

Your analyte needs to be a bit longer retained than the protein etc. junk that breaks through an Oasis packing. Then you can take your analyte off in a simple step with a small amount of organic, just a bit more than your initial conditions in the gradient. 20% MeOH will probably do, maybe even less could still be OK. This leaves the gunk that seems to change your elution behind on the cartridge, and proteins break through.

[HW Mueller]

I am not sure, anymore, of what is described here. You said somewhere that a standard injection is normal even after abnormal sample injections. That would rule out that a builtup of crud causes the rt shifts and peak broadening indicated in the chromatograms. That´s why I thought it must be caused by something that is co-eluted. The most likely candidates could be due to a changing matrix that would cause a changing wash in (mismatch of mobile phase and sample solvent), for instance: Evaporating TFA in the sample will cause a change of pH which might cause a steadily increasing pH mismatch, etc., etc.

[juddc]

I am not sure, anymore, of what is described here. You said somewhere that a standard injection is normal even after abnormal sample injections. That would rule out that a builtup of crud causes the rt shifts and peak broadening indicated in the chromatograms. .

Maybe - usually even - but possibly not. I can see a scenario whereby a significant portion (but not all) of the crud is washed off with his gradient and that successive sample injections yield what he's seeing because a good sized slug of crud is hitting the column with each shot and building up, but is not very tightly bound to the column. Running a blank after some sample injections could conceivably clean up the column enough to yield reasonably reliable results for subsequent standards. This might at least be consistent with his second peak being not nearly so affected as the first. It might be instructive to have a look at what's going on at a lower wavelength.

I would think solvent mismatch would be unlikely due to the fact that he's reconstituting in MP, though I agree absolutely that the method would be quite sensitive to MP variations, given his column geometry and injection volume.