Foreign late SEC peaks

[sturgman]

Hello Everyone,
We are running a Waters 2695 solvent/sample management system with a styragel column bank and toluene as the mobile phase. We have identical columns with THF mobile phase as well. Recently we have noticed carryover peaks from sample to sample. I ran a longer than usual sample (our injections usually run for an hour) to see when this late peaks (magenta) came out and noticed that they come out much later than our standard solvent peaks (red). If we run the same samples but replacing the mobile phase to THF and switching the columns we do not see these very late peaks. Can anyone suggest possible reasons for these?

Some questions/thoughts:
1. Can this be due to defective columns (please NO!)
2. It cannot be due to impure solvent as we have gone through different HPLC grade solvents and we see the same thing.


An easy solution is to wait until they come out for the next injection but we would rather not waste that amount of solvent.

Thanks for your help!

[pelle-plutt]

I have seen a similar thing in the past, although I analyse biological samples (vaccines). In my case the solvent front appeared at ~12 min (which is normal for this SEC column) and then a huge peak at ~15 min and a smaller one at 22 min. The huge peak at ~15 min was the antiseptic (Thiomersal or Thimerosal) used in the vaccine. The extra retention was attrubuted to unspecific (hydrophobic?) interactions to the column matrix. Electrostatic interference was supressed with 150 mM NaCl in the phosphate buffer used in the separation.

In your case, I don't think there's something wrong with you column and probably not an impurity in your solvent. It's more likely something in your sample. Have the samples changed in any way recently?

[sturgman]

Thanks for your reply pp. The problem is that those peaks are present even when I inject blank mobile phase into the system. In fact it is interesting that upon injection of the actual polymer standard the first foreign peak remains the same while the second one is deformed. I know that this system should finish a run in about 60 minutes. What type of substance could come out of the column so late? Im running out of ideas.
For the time being Im just making the runs longer. In fact my standard calibration remains very good despite those peaks, we are just using more solvent per sample.

Thanks for any insight.