Agilent autosampler

[d.telaro]

Sorry, but can I ask you which analyte do you usually use to obtain RSD 0,3%? Can you write me also the concentration? and also the coloumn?
Thanks very much

[d.telaro]

If someone of you really obtain RDS 0,3%:
Could you tell me which analite you use and the concentration ?
Which coloumn do you use?
Thanks
Daniela

[gcguy]

A while ago we had an issue with non-reproducible results from an 1100 injection system. We tried the usual troubleshooting and replacing parts but could not get anywhere with it. An Agilent engineer was in fixing something else and I asked him if he had any ideas. He suggested that we should use genuine Agilent vials rather than our cheapo generics. I was sceptical at first but it worked. Checked to see if it was a fluke by going back to the cheap vials and the problem returned.

We will now use Agilent vials in all our 1100s.

GCguy

[d.telaro]

Hi gcguy.
thanks for your reply.
CAn you tell me the analyte that you use and the concentration to obtain RSD 0,3%?And what about the column?

[unmgvar]

d.telaro

you still do not give enough information regarding your equipment.
you say that it is a 1100 but which model exactly? from what i read it is a pulled loop system and not the tipycal splitt loop sampler.
what are your settings in the method?
in what is your sample dissolved? water, MeOH for example.
what is your mobile phase? bad RSD can be the result of bad tailing and bad integration.
what is the peak height that you get? above 1500 mAU RSD of the UV signal gives bad results for last 1100 models.
How old is the sampler, when was it for the last time serviced?
is your pump equiped with a working degasser? if not do you make sure that you degass the mobile phase correclty?
is the RSD of your retention time stable? almost the same as the area RSD -+? then maybe air in the mobile phase does that, or another reason that affects the flow rate stability.

like gcguy kind of suggested the reason could be the septa used. vials have absolutely no effect on RSD quality. have you tried doing 5 injection without a septa? speta caused problems give generally 2 symptoms:
up and down all over the place results from the same vial
the first injection is out , high or low, and all others are good. again from the same vial.

how do you fill your vials. maybe too much? maybe not enough?
is it a flat bottom vial?
how many injections do you make for your test?

what are the Performance Qualification specs for your sampler model, and what method do they use to achieve it?
it should be a very simple method so as to clear out any type of application related factors in order to trully check the sampler. try that method first to get a low RSD for repeatability

[Bruce Hamilton]

To isolate the root cause of the problem, follow the manufacturer's procedures for checking instrument performance.

Use the IQ/OQ/PV protocols suggested by Agilent for your complete system. Note that the types of injector and detector may change the protocol, as well as the pass/fail criteria, so ask your local Agilent people for the relevant protocols - if you don't have them.

I suspect Agilent may use 25 ug/ml caffeine in water, with water as the mobile phase, and no column, but ask them for the qualification protocols..

However, injector precision test is usually performed after you have demonstrated that critical systems, such as the pumps, are also meeting their specifications. Note that peak height precision may beinferior to peak area, and ensure you set instrument up according to the Agilent test protocols.

Once you have confirmed the instrument is OK, then try your own tests, with/without a column to provide comparative data.

Bruce Hamilton

[d.telaro]

I dissolved naftalene 0,1 mg/ml in water. As mobile phase I use acetonitrile/water (65%/35%). The peak height is about 150.
I don't use a degasser because I use an isocratic pump. Do you suggest me to use the degasser?
The retention time is stable.
Do you suggest me not to use the coloumn?
Thanks
Daniela

[Bruce Hamilton]

No. I recommend obtaining relevant Agilent protocols, and using those.

If you do want to persist with your method ( but I wouldn't ), and remove the column, the sample must be dissolved in the mobile phase.

You really should talk to your local Agilent support people.

Bruce Hamilton

[gouki]

Personal experience with agilent autosamplers using a variety of compounds/mixtures should give a <0.5% RSD for injection volumes >2uL. @ 0.1uL I generally experience about 2% RSD.

[dries vrielink]

Hello,

whenever you want to check the reproducibillity of your autosampler, it's always wise to use the same setup as stated by the manufacturer. Normally a typical reproducibillity is specified at certain conditions. So it's quit possible the autosampler reproducibillity can differ from the specifications when tested using a different test setup. Items to keep in mind are the following:

When using a restriction capillary, are you sure the backpressure of your system is high enough for the pulsedamper (inside your pump) to take effect? Otherwise, it could be quit possible, you are observing a small pulsating flow, which can affect the peak width and therefore the reproducibillity.

When using a column, make sure you use a compound with some retention and even better a typical optimum detection wavelength, so you're sure you are only measuring the peak of interest (without any underlying contaminations).

A typical basic setup we use a lot in our factory is:

Pumpflow @ 1.5mL/min, restriction capillary in between pump and autosampler of appr. 70 bars. (this takes care of any pulsation caused by a pulsedamper at low pressure)

Uracil dissolved in HPLC grade water as a testcompound, appr. 50 ppm

UV detection @ 254nm.
Eluens is HPLC grade water

Peakforming tubing (0.25mm PEEK, 10mtr) between autosampler and detector.

Outlet of the detector is appr. 10bars of restrictive tubing (to prevent any air bubbles from forming, but one should keep the maximum pressure resistance of the detector cell in mind offcourse).

Inject appr. 10uL using partial loopfill (this should be possible for loopsizes from 20uL and up)

After 10 injections, this should give you an accurate image of the reproducibillity of your autosampler.

But as stated before, first take a look at what Agilent subscribes as reproducibillity test and if possible use that.

Good luck with your experiments.