Chromatography Forum

Hi, first post here, so please forgive me if I miss out on any important information.

I've recently had a problem with carryover, and have narrowed it down to my analyte getting stuck in injector loop/rotor seal. I'm working on an Agilent 1100 with Vespel rotor seals;

This is a gradient method, so my current work-around is the following:
1. Inject- 10 µL sample
2. Carryout gradient to top (95%B)
3. Inject - 100µL (full loop volume) of 100% ACN.
4. Remain at 95% B for 1 minute, then switch to initial conditions and reequilibrate.
5. Repeat steps 1-4, N times.

This appears to be working, but I'm using Empower, so each injection alternates which Instrument Method/Method Set is used (ie: Method A, then Method B, then Method A...and so on).

This is somewhat of an inconvenience, and could be difficult to validate. I also work in a department in Pharma, where Method Transfer is very important. So if I can prevent my analyte from sticking to the system in the first place, I think that would be a bit more ideal.

I've ordered some Tefzel rotor seals, to see if they resolve this issue, but does anyone have any alternative suggestions?

Thanks in advance!

Carryover is often compound dependent but for nonspecific adsorption of sticky hydrophobic (reserpine)
I found that 100% ACN is the best rinse solvent so nice job there. The Tefzel was not any better
than the vespel due to the large amount of carbon in the rotor seal. The best for low carryover was
virgin PEEK.

Thanks for the suggestion. We've got the Tefzel installed on one system, which I'm testing today. Looking at the structure of this compound (and consulting those with about 15-20 years more experience than myself), there's nothing on the structure that indicates this should give such carryover problems. It's not particularly hydrophobic, but we'll see what happens with the first series of test injections. If this doesn't help, maybe the PEEK will, so I appreciate the suggestion and will see if that helps at all.

Are you sure it's sticking on the rotor, and not still present on the outside of the needle? The 1100 has no functionality to rinse the outside of the needle, so maybe (depending on the concentration of what you're injecting, the amount of residue left on the needle, and the total volume in the vial of the next sample you inject) this is what's causing the carryover. You can create an injection program with the 1100 (not sure how with Empower, but you can with Chemstation) where you have the needle do a dip in a vial of solvent after picking up the injection aliquot, before moving to the needle seat.

Yes, I'm sure it's sticking to the rotor (and on my initial Blanks with the brand new Tefzel rotor seal/new Stator Face-- it may be adsorbing to the injector loop as well). I've concluded this by doing a non-injection run through the gradient, and still see the API contamination. And it's not that my Mobile Phases are contaminated, as after giving the system a full flush (by switching the valve from mainpass to bypass and running for 30 mins each) with 100%ACN, I am able to remove the contamination.

Also, I do have the functionality to do a needle wash after each injection-- however the needlewash is after a sample has been injected via the needle seat. (ie: the system does an injection of sample, returns the vial to its position in the tray, then brings the wash vial to the needle, and the needle is dipped into the vial-- This is set up in Empower software). Currently, I've used up to three washes, where it just dips into an HPLC vial. It does not wash the inside of the needle though. Whether I use this type of needlewash or not, no real improvement is observed.

I've been waiting for a 1200 to become available, as I'm aware that the setup is slightly different.

Thanks for your help and suggestions.

Some compounds might be sticking to PEEK tubing and SS parts (chelates, paraquat/diquat). You tubing can trap compounds from previous runs, which are now holding current analytes, for example, if you used IP reagents on the system, they might be trapped in PEEK, and now they contribute to a carryover of the oppositely charged analyte

OK, can you program the valve to flip from load to inject during the 95% acetonitrile portion of the run, to see if you can wash off any retained analyte?

Certainly you would expect Tefzel to be the best material but note that the Tefzel seal is black
just like the Vespel. Both contain about 4% graphite to improve the wear properties of the rotor
seal. My conclusion was that the Tefzel and Vespel had more carryover due to the amount of
carbon in the polymer. The virgin PEEK had better carryover because it didn't have the carbon.

Of course the virgin PEEK doesn't last as long as the carbon loaded rotor seals.

I did also find that using fused silica tubing improved carryover likely due to the very smooth
surfaces. Stainless surfaces are incredibly rough though PEEK is better so it just takes longer
to rinse material out.