'Skipped' injections and pressure issues..

[mnrevoy]

Hello,

We are currently running a new method on our Acquity H-class UPLC and we seem to be having some issues..

It seems that, on occasion, the first injection of a certain standard level in our calibration curve is either not taken (skipped), or air is injected. It is hard to tell what is happening, other than there are no standard peaks.. at all. The second injection taken from the same vial, however, always seems to work. All of the components elute at their respective retention times and the baseline is flat. This is quite concerning since the last run we performed, this happened 7 different times. The pressure trace for the 'skipped' injections is much higher than the regular samples.

We are assuming that this is a hardware problem.. potentially the needle needs to be replaced?

I would appreciate any and all comments regarding to this problem as we are stumped..

Thank you

[mattmullaney]

Hi Mnrevoy,

Please, refresh my memory, does the Acquity-H have the "needle-in-needle" design as does the orginal Acquity?

The other thing...are you using pre-slit vial caps?

Please see what you think and thank you.

[mnrevoy]

Hello Matt,

The H-class UPLC has a flow through needle design.
We use 10mm black screw caps, with bonded tan PTFE/white silicon.

[mattmullaney]

Hi Again,

My apology...the Acquity-H autosampler needle, is it made up of an outer steel sheath with a narrower polymer needle inside it that moves up and down for each injection? I seem to recall that this is the design...the metal portion pierces the vial, then the PTFE (?) portion extends down further into the vial and withdraws the solution inside for injection into the flow path of the UPLC system. Agreed, the Acquity-H is of the FTN design...my first thought, if this is the design of the needle, could it be getting caught up somehow inside the injector so as not to properly enter the vial?

The vial caps, are they similar to those such as Waters Part Number 186005666CV? Here's a link that may help, see page 15:

http://www.waters.com/webassets/cms/lib ... 1818en.pdf

See what you think and thank you.

[mnrevoy]

Matt,

The needle itself is in the flow path and draws up the sample using a metering syringe. There is no inner polymer that moves up and down.

The vial caps are not pre-slit. (Canadian Life Sciences p/n c06050bc-10)

I wonder if the needle is entering the vial but not drawing up sample..
It is strange that this only happens sometimes though, and always on the first of two injections from the same vial.

[mattmullaney]

Hi again,

Agreed with your suggestion, and thanks for clearing me up on the needle design!

Seems that this is exactly your Acquity-H's trouble. One thing to try would be to do an injection set of two from a vial where the septum isn't there...but the rest of the cap is. Maybe there is a partial vacuum within the vial that is relieved after the initial entry of the needle such that the solution isn't being aspirated.

Please, see what you think.

[vofelyp]

Hi

At first characterize the needle seal, after run the needle seal readinesss test. If both tests are passed the injector would be OK. It seems the needle has clogged a part of septa. Remove the septa from the cap and try to run the sample.

[HippyLabRat]

Sounds like you might be overfilling the vials. NEVER fill and HPLC vial to the brim. Make sure you're leaving enough headspace in the vial so you're not generating a vacuum when your needle is pulling your sample. Filling a vial so the TOP of the meniscus is at the base of the neck is ideal, in my experience.

[ChromNewbie]

The amount of sample added to a vial should have no effect.. that makes no sense at all. when you are injecting small volumes like this there should not be enough suction to create a vacuum.. We had similar problems in the past and once we replaced the needle it was fine. it seemed as though the needs was partially clogged, but once enough pressure was built up it would inject just fine for a few samples then clog right back up. What type of samples are you injecting?