GPC and a negative peak

[Aussiemosis]

Hello I am running GPC polystyrene standards using DMF + 0.05M LiBr on a 2695 system and 2414 detector with Styragel HT2, HT3 and HT4 in series.

I get a negative peak around 32mins even in a blank (I always thought that was air, could it be anything else?). The smallest molecule standards (489Da) comes out around 36mins, gradually coming out earlier as the MW increases, but needing something like 19000Da to elute before the negative peak at 32mins.

Is there any way around this problem?

Thank you.

[Uwe Neue]

Considering where the negative peak is coming out it can't be air. Must be a component of the mobile phase, including something that you are extracting from somewhere. Unfortunately, a negative peak in the RI does not mean that this analyte is missing in your injection, so it could come from your injection vial/cap etc. However, considering that the peak is partially excluded, I would think that it is something with a reasonably large molecular weight. I would examine everyhting that is in contact with the sample and the mobile phase and think about possible extractables.

[Aussiemosis]

Hello Uwe and thank you for replying.

I am still getting the negative peak when injecting 0µl of air (from an empty vial) which is the closest I can get to a true non-injection that I know of. I guess that means the only thing that happens is the sample loop valves do their thing and trigger the run, but I still get the negative peak half an hour later.

So just now that led me to think of two possiblities - something from the needle wash frit or something coming from the sample loop valves. But then I don't know if the valves work anything like a rheodyne or what it's made of. Easy enough to investigate those two tomorrow.

But do either of those sound plausible or do you have any other suggestions that I could try?

Thank you.

[HPLCCONSULT]

You did not mention what type of HPLC (brand/model) you are using. It can matter (many different types of A/S and A/I's). Have you tried to do a "dummy injection" ? *Not just a 0 ul injection. Do the same run with the A/S (I assume you have an A/S?) valve being toggled just like in an injection from load to inject, but no sample selection or movement. Only the valve should switch. A second run with no valve switch would be compared to see if the peak disappears. If it does, then you know the injector (or something associated) is the source. Have you PM'd the injector (rotor seal) ? Is everything running correctly ? RI's are super sensitive to disturbances so the peak could be from anything mechanical as well as a contaminate on one of the column frits or MP bottles. Start with the basics to rule sources out.

[Aussiemosis]

Hello HPLCCONSULT it's a Waters 2695 system with 2414 RI detector.

What you speak of, I have always known as a 'Null Injection' when using Agilent 1100 systems from setting the injections command to Off.

This is something I have wanted to try but can't work out how to do it on a 2695 controlled by Empower 2, do you know how to do it?

Thank you.
Daniel.

[AA]

In Empower its called, inject immediate sample (or standard). The method runs without doing an injection.