Agilent 1100 Injection Volume - Not Accurate

[AThompson]

We have an Agilent 1100 system with G1329 autosampler. The system works well - it gives reproducible results across a given injection volume and is linear across different injection volumes. The problem we noticed initially was that the UV detector gave about 20% of the response of what we got on other 1100 systems we have. I figured it was related to the detector but it passed all diagnostics. Today we decided to test the autosampler... We injected 50uL out of a vial of water 10 times. We hoped the weight loss would be ~500 milligrams. It was 120.8 milligrams. We did this same test on another 1100 system we have, and across 10x50uL injections, the weight loss was 500.2 milligrams. Very good!

I have replaced the sample loop and needle on this problematic G1329, as well as the needle seat. No difference. I have it configured properly in A.10.02 Chemstation with a 100uL syringe and 2.3uL seat capillary. I did an R128 reset to restore to factory settings but this hasn't made a difference, either.

Does this point to an electronic board? If so, which board is the most likely culprit? Thanks for any feedback!!

[tom jupille]

Have you double-checked the actual syringe installed?

[Bull76]

Hi,

Have you tried replacing the stator and rotor seal in the injector? Also have you replaced the sample loop or has someone else cut it? Can you confirm the loop size? You could also carry out a pressure test of the autosampler.

Regards,
Bull76.

[Consumer Products Guy]

Assuming that there are no leaks visible the G1329 autosampler has only a few items involved with amount of sample injected.

autosampler needle/seat (small leak)
autosampler rotor seal and related
autosampler metering valve (same piston and same piston seal as on the 100 pump)

I've also seen the loop just before the needle break its weld at the attachment to the 1/16 inch part, near the needle attachment.

http://www.chem.agilent.com/Library/use ... _ebook.pdf

[LCbob]

As suggested do the pressure test first. If that passes then it maybe a small restriction in the needle. Also check your viscosity settings are the same as your other system.
Also suggested has been the Injection loop, I have seen instances where customers have replaced the loop with different tubing, in general this is not a problem as long as the internal diameter is about the same.

[kathyhu]

Hi Mr Athompson,
I am not sure if you can see my questions. I am actually met the same situations with you for my HPLC 1100. My injection volume is only about 20~25% of my real required volume.

I have also almost replaced most of the part related to injection volume. But no big differences. I wondered how do you solved your problem? I may can use as references for my trouble shooting.

thanks

[duck duck goose]

Here is what you do,
1. Take apart the injection valve and sonicate the parts. Replace the rotor seal if you can.
2. Run a pressure test on the autosampler. Block off the injection valve where it connects to your column compartment.
3. Repeat the pressure test but first switch the valve to mainpass or bypass position, opposite to whatever the position was in step 2.

Has the seal in the metering head been serviced recently, and do the vials that you are using have caps on them? If you have a worn metering head seal and tightly sealed vials you may be creating a vacuum in the vial that is leaking back at the metering head causing an improper volume to be drawn.

[kathyhu]

Hi Duck, I have replaced the rotor and rotor seal recently, I have also replaced the piston seal in both the metering unit and pump. I changed new needle and needle seat. The only thing that I did not do is pressure test. I recently ordered a new sample loop. But I don't think this will work neither since my sample loop is not block at all.

Last thing I am thinking is to change the piston in the metering device. Since my machine is very old, I bought second hand, and the injection volume is not accurate at the beginning. I just noticed this recently due to the large volume injection request.

[Fernando]

Hi kathyhu

I think you proceded well, have you cheked the cell path?

Best luck

Fernando