peaks eluting within column dead time

[CPR]

Hello everyone,

I am running an RP HPLC method where I am seeing a dead time of approx. 2 min based on blank and standard chromatograms. This holds true for the first injection of sample in a run as well, but every subsequent sample injection has rather sharp peaks eluting at approx. 1 min.

The sample is a gel dissolved in diluent (see below) and has a fair share of placebo peaks as would be expected in this formulation. I have added 10 minutes to my run time to eliminate the possibility of late eluters, but never see anything past 10 min. Also, the first injection of standard following my last sample injection does not exhibit these peaks.

Here's my system:
Thermo TSP HPLC
Agilent Zorbax SB C18 5um 4.6x150mm
95:5 ACN/Water mobile phase, isocratic
70:30 ACN/Water Diluent
Flow rate 1.0mL/min
UV @ 277nm
100uL injection (200uL partial loop)

This does not interfere with my peak of interest as they always seem to elute very early in the chromatogram, but I am curious as to why this is happening. If I had to guess it would have something to do with sample overload on the column but I need a large injection volume to have a satisfactory response from the analyte as the sample concentration is very small.

Has anyone seen this? Thanks!

CJ Rassbach

[tom jupille]

Peaks eluting "before t0" are actually pretty common. In fact, that's how size exclusion chromatography works (GPC or GFC) works. Molecules that are too large to fully permeate the pores of the stationary phase or which have the same charge as the stationary phase are excluded from the pore volume and elute early.

That raises the whole question of how one defines t0 in the first place. There were some extended discussions of that a few years back; for example:
http://tinyurl.com/5gfqy4

[danko]

Hi CPR,

It seems to me you may be experiencing an effect I call “The HILIC Effectâ€

[JGK]

In method validation, I would usually triple the normal cycle time in a late -eluting peak test.

adding 10 minutes may not be enough

[Bruce Hamilton]

I don't know your instrument, so this may be completely irrelevant, but if the injector uses an actuated valve, the peak could be a small amount of your sample trapped in the valve ( ege between the rotor/stator for viscous samples ), and inadvertently injected when the valve switches to the "fill" position.

However, you would normally also expect to see very small retained peaks, as well as the larger early unretained peaks.

You may have to play with different injection volumes to see if the injector is the source of your spurious peaks.

Please keep having fun,

Bruce Hamilton

[danko]

Hi JGK,

In method validation, I would usually triple the normal cycle time in a late -eluting peak test.

If the issue here is just late eluting peaks (i.e. not dependent on the injection of a different liquid than the mobile phase) then the peaks wouldn’t be sharp – on the contrary.

Best Regards

[CPR]

Thank you all for your responses, they all bring up good points.

Tom, if it were a SEC issue wouldn't these small peaks be seen at the beginning of the first injection of sample? The problem that I was having was explaining the inconsistency of multiple injections of the same sample. Also, I may have misled you a bit initially. Looking more closely at my run it seems that these small peaks do in fact elute in my standards and blank injections (at approx the same retention time) and they decrease in area with each injection of something other than sample.

Danko, I have measured the dead time based upon the earliest peak or disturbance in the baseline in both the standard and filtered diluent and have seen it at 2 minutes. This also is true for the first injection of sample. It is the subsequent injections of sample that have peaks eluting well before that. As I said to Tom, the peaks do seem to elute with each injection following samples, so your "HILIC" effect seems to be the best hypothesis so far.

JGK, the peaks elute sharply and at the same time at the beginning of my next injection regardless of run time. I do not believe late eluters are to blame.

Bruce, I believe you have accurately described our injection system. Carryover has been known to be an issue with these systems. The response of my analyte is rather small compared to the rest of the chromatogram so it's possible there could be some carryover problem but I can only detect the unretained portion.

Afer all this discussion it seems this development route may be abandoned due to issues outside of my control (i.e. management), but thank you for indulging my curiousity nonetheless!!

CJ Rassbach

[Bintang]

Hi CPR,

I also think it likely that the peak you see is something retained by a HILIC mechanism and eluted by the (in HILIC terms) stronger eluting diluent.
One thing you should also consider is the sample loop washing solution (since you are using partial loop filling) if this is water it will act as an extremely strong eluent.
This may be why you dont see this early peak in the very first injection since your loop is most likely filled counter current to the eluent flow, ie the strong loop wash is ahead of your sample when going through the cloumn.

[CPR]

Bintang, thanks for your thoughts.

The flush solvent was actually my mobile phase (95:5 ACN/H2O). Normally I would keep the flush solvent consistent with my diluent proportions. However, the goal of this experiment was to evaluate the extraction of the sample from the gel as well as from pad applicators using various diluents. Therefore, I figured it was not practical to switch the flush solvent for each individual analysis and was planning to fine tune this part of the method after I had figured the sample prep out, as this was the most challenging obstacle.

Thanks again for your response!