carry-over in HPLC

[MestizoJoe]

Hi everyone. Does anyone have any advice for avoiding carry over in a 1100 series agilent system?

Here is my current situation:

I am working with a molecule which is soluble in 50% ACN and 50% water. The system is agilent 1100 series with a DAD.

After some runs which went well, I injected multiple blanks (60%ACN 40% water) which each show my molecule is present. The amount which appears to carry-over increases with each injection. the first is around 2 area counts, then 9 then 20, for three blank injections. I injected from several different vials, some vials were injected more than once. Each vial gives a very different amount of carry-over.

Does this sound like carry-over? I am trying to quantify near the QL so this is trouble for me.

Could it be that the seat is contaminated?

[Peter Skelton]

What injection method are you using? For detection at anything coming close to LOQ you should be using an injection with needle wash. An uncapped vial of either your strong mobile phase or something like IPA is good.

Also, tell us about the stuff you're injecting. What concentration are your solutions at and what order are they being run in?

[MestizoJoe]

my concentration is 0.04 ug/mL

i injected with 100uL and got a good chromatogram. i switched the seat capillary to support a 500uL injection and then things started not working.

my sequence has been
blank
standard
blank
blank
blank
standard
blank
blank
blank etc.

i am using needle wash of diluent.

i recently cleaned the seat with diluent but i dont know if that will help.

[lmh]

I'm going to hazard a complete guess. As I understand it, the 1100 deals with sample by sucking it into the needle through a long piece of tubing, with a metering device (syringe!) at the upstream end. I think the idea is that the syringe itself is probably only ever full of solvent, and is almost a dead-end side-branch; sample should never reach it (it only fills the long bit of tubing).

Is the tubing big enough to contain your new extra-large injection? If not, you would probably get horrible carry-over.

I was puzzled by your comment about changing the seat capillary as the long tube is on the needle side, not the seat side.

It is also very important that your needle wash vial has no lid. It should also contain a greater depth of solvent than the sample.

I may be writing rubbish... please correct, someone!

[MestizoJoe]

it was recommended to us by agilent that we change our seat capillary to increase injection. the syringe still can only pull a maximum of 100uL at a time, but it injects it into the seat capillary where it waits until the injection is complete, i assume, before letting the sample pass to the column.

maybe the seat capillary change is easier than a syringe change or something. i dont really know.

[Peter Skelton]

So exactly how is your injection programmed? What software are you using to do this?

What volume are you trying to inject now, the full 500ul?

Are you using any overlapped injection techniques? As the loop (or pseudo-loop in your case) gets bigger, the necessary flush volumes also grow, so if you're using one of the loop-offline preload techniques this could be responsible.

LMH: you're right, usually the 1100 only injects using the metering head and a sample loop, but you can adapt them to give large volume injections where multiple injections of your loop volume are pushed into the seat capillary before the rheodyne valve is switched to the inject position

[lmh]

ooh, thanks for that information!