UPLC Injection linearity issues (1 uL?)

[Rob Burgess]

Almost all of a sudden we have a series of issues with our injector showing poor injection linearity on our Acquity (see table below as an example) i.e. the plot of injection volume versus peak area does not cross the origin (0,0). This has happened on multiple method types (3). We are using PLNO injection with a 10uL loop.

We originally saw it when we replaced our PEEK needle for a stainless steel one for compatibility with red rubber self sealing HPLC vials. Our on-site engineer through a iterative process of changing various parts on the injector module but all to no avail e.g. changing the whole motorised pod, changing the needle 4 times, changing the 10 uL loop, changing the air sensor etc. The engineer is going to do a PM of the whole system to see if anything else is amiss but I cant think of anything obvious. If it doesnt work soon we will have to escalate.

Inj volume Peak areas
1 1413
1.5 31732
2 64365
3 126116
4 187692

It appears that it just isn't injecting 1uL volumes to any significant extent. We didnt use to have an issue with that low a volume on a 10uL loop with our methods. Anyone got any ideas of what is causing this issue?

[Arne]

Rob,

could you inject 0.5 uL? It would be interesting to see what peak area you get then. There are two possibilities: (1) a negative peak or (2) a very small peak. The linearity of you r curve is excellent, which suggests to me that the injector is fine. I would check for detector problems immediately before working on the HPLC any further.
It appears that you simply have an offset of -60,800 on your detector. What type of detector are you using?

http://i.imgur.com/e9AdUPY.jpg <-- picture of curve

[Hollow]

as Arne said, the linearity is fine for me too, but the offset is the problem.
Looks like a missalignment of placing the sample in the loop.

just to have it asked:
after the changes and servicing, all the settings and sensor calibrations in the Acquity Console were done correctly (loop, air sensor)? The Air sensor can be a bit tricky.
Also is the Needle Seat (O-ring) working well (maybe it was set ok for the PEEK needle but not for the SS)?

Other ideas:
what if you do the linearity check from a vial withouth any cap?
How do the PLNO values compare with the ones of a Partial loop linearity?

good luck

[Rob Burgess]

Hi All,

Thanks for your detailed reply, much appreciated.

I am using a DAD. havent tried 0.5µL but will try - not even sure you can do this on the acquity. It would be outside the "working recommendations" of PLNO the 10µL loop

We did some weighings in PLNO mode but not sure they made sense (it was very variable for the 1µL injection 30-50% RSD) and still coming out extremely low wrt peak areas.

1 µL
2 µL
4 µL

mean
12.94 mg
15.5975 mg
17.42667 mg

%rsd
44.93484
6.989255
10.70138

The needle has been replaced several times by our on-site engineer (backwards and forwards between PEEK and SS) with the same issue, so suspect absorption isn't the issue here.

Have changed mobile phase, whole entire injector pod and needle washes. All seal seats and air sensor calibrations have been checked by engineer and double checked by me.

Have tried with and without vial caps and same issue being seen constantly (see below). If you plot the means out it never goes through (0,0).

Inj vol Peak area
1 5929.333333
2 66254.33333
4 190015.3333
6 314183.6667

Have tried in partial loop, but this just gives a distorted peak shape/ extreme split for this particular method but does seem to be "injecting more".

Engineer is going to a full strip down PM on entire kit, and if its still not fixed escalate direct with Waters.

Will also run a 1290 next week to see if there is any sort of similar issue.

[Hollow]

Hi All,
Have tried in partial loop, but this just gives a distorted peak shape/ extreme split for this particular method but does seem to be "injecting more".

actually I don'r have experiences with partial loop, but maybe you can do the test with a method which gives good peaks, even in partial loop? (Does the weak wash compare with the initial mobile phase, maybe this causes the peak split?)
I'm thinking of something like caffeine in water as sample.
With a mobile phase of water/meoh 75/25 it should elute at a k of about 2-3 on most C18-columns, 35°C. use water or meoh <20% as weak needle wash. Detection at 270nm)

(tested on HSS T3,1.8, 2.1x50mm, 0.4ml/min; caffeine at 0.92min)

[Gaetan Glauser]

Hi Rob,

Hope you see this message...we have the same kind of issue now. May I ask you if you found a solution for your injector issue in PLNO mode?

Thanks,
Gaetan