siphoning effect during injection

[MK]

Hi,

what is siphoning effect during injection with agilent 1100 standard autosampler?

thanks
MK

[Mark Tracy]

Normally, siphoning does not happen with this design. The one time it might happen is if you are doing multiple-draw injections into the extended volume capillary. In this case to prevent siphoning, the waste line from the injector should be positioned so that its tip is level with the needle seat; if it is installed correctly, it will already be in the right position.

Siphoning is usually a problem for manual injectors. Most autosamplers don't siphon unless there is a serious leak or a massive bubble in the syringe.

[MK]

I'm injecting maximum 100ul volumes with 1100 standard set up and occasionally seeing decreasing peak area for repeated injections of the same sample injected from different vials. I've been weighing the vials before and after injection and calculated the weight drawn out from the vial. The injected weight also shows a decreasing trend and correlates with peak area. For example my peak area are:

552.189, 546,023, 546.194, 532.954 mAU*s

and injected weights:

101.62, 101.50, 101.50, 101.45 mg

Agilent says this might be siphoning due to large injection volume. Any ideas?

thanks
mikko

[Bryan Evans]

What is your injection volume? I have seen a drop
of solution come out of the injection port waste line
for injection volumes of 100 uL. We used to wrap
the waste line with a kimwipe and than tape it to the
outside of the instrument to prevent it from leaking
into the leak detector.

Although this line is now lower than the needle (and
not at the optimal level) - it still passed our injection
precision/linearity test just fine.

[DR]

You may be experiencing a vacuum effect as a result of too good a seal. This is generally most pronounced when trying to pull a large amount from an over-filled vial. The cure is pre-split septa lined caps.

[Mark Tracy]

If you are seeing a droplet at the injector waste, it can occur for four reasons. One is when the valve switches to the load position, the fluid in the loop decompresses, and spurts out the waste. This is normal and does not affect the injection. The second reason is that when the syringe resets, the fluid is ejected to waste. This is also normal. The third reason is siphoning out of th seat capillary because of a mis-positioned waste line. This does not affect normal injections, and at worst injects a bit of air. The fourth reason would be that the valve rotor seal is worn and during the switch to the inject position, it leaks to the waste. This is not normal and will affect the precision.

[MK]

Thank you all.

due to 100 ul injection volume there is quite a lot mobile leaking out from the injector waste line. I understand this is normal. Sometimes this leaked to leak sensor and the system shut down. To overcome this I ordered and installed a waste adapter from agilent. The adapter directs the waste away still giving a good position for the waste line.

I still have a problem with the injection precision. On a good day it's RSD 0.6%, normally around 3% and on a bad day it's over 5%. Repeated injections in sys.suit. are often showing decreasing trend in peak area. The peak area usually correlates to the mass drawn out from the vial as we've weighed the vials before and after injections. Due to this correlation I think the precision problems are related to the injection event. Based on your advice and our own experience I'm thinking this is not siphoning. Could it be precipitation somewhere in the injector? Samples are diluted in 300mM NaCl-50mM Hepes -buffer and stored at +4C at thermostatted autosampler. I'm cleaning with solution containing 1M NaOH between sample injections. The external surface of the needle is washed with the dilution buffer.

I've tried:
1) lower injection volume of 50 ul
2)lower draw and eject speed of 50 ul/min (instead of default 100 ul/min)
without succes

Any ideas?

thanks
MK

[Mark Tracy]

Can you adjust the wait time between drawing the sample and raising the needle from the vial? If so, try a longer wait. (I haven't used one of these for a few years, and I forget the details.) You may have a partial obstruction in the needle/loop. Another possibility is the piston seal on the syringe may be worn; these last a long time so you tend to forget that there even is a seal there.

[Patrik Appelblad]

Do you have the same result independent of injected analyte?
You have mentioned the sample diluent and wash buffer, but can you tell us anything more about your samples?

[MK]

i had the autosampler serviced and my RSD% is below <1% again. They changed a lot of parts including the piston seal on the syringe.

many thanks!

mikko

[ntruong]

Hi MK,
Since your injection volume is 100uL, why not switch to a larger loop? 900uL?

[MK]

Hi ntruong,

I tried 900 ul loop but it didn't improve my percision.

I remember seeing a thread saying that it's generally not a good practise to inject maximum volumes with a system. Why is this? I couldn't find the thread anymore.

thanks

[tom jupille]

I remember seeing a thread saying that it's generally not a good practise to inject maximum volumes with a system. Why is this? I couldn't find the thread anymore.

As you get up to a high percentage of the loop volume, laminar flow effects begin to build up. Liquid flows more quickly at the tube center than at the wall, so some portion of the sample washes completely through the loop.