Chromatography Forum

Troubleshooting unacceptably high RSD values for repeat injections on an old method, I have noticed a pattern in the peak areas where odd injections in the sequence will be ~5% greater than even injections. All from same vial. I switched to a simple isocratic MeOH/H20 run to pick up caffeine in water, but with the same autosampler settings, loop size, injection type, etc. and have the same problem. Seems like sample is getting stuck somewhere and then unstuck with each run. I tried using a full loop injection instead of a partial loop injection and it seems to have helped a little - but this changes my method. I am using a Dionex ICS-3000 with an AS-AP autosampler and the default settings (for draw speed, dispense rate, etc.) for my sample loop size (25ul, 10ul injections when doing push partial and 25 ul when push full).

Any ideas on what's happening and what else I should be trying out?
Thanks!

Hi Marie666,

One possibility--if the ICS-3000 uses a Rheodyne injection valve, is the partial-loop volume always being set at a value less than one-half of the volume of the loop? As I recall in older Dionex systems, the needle/loop are isolated from the flow path (it's not a flow-through needle injector), so the cautions regarding loop volumes must be rigorously enforced.

So, for 25-uL inj. vol., use at least a 50-uL loop, or use a 20uL-loop minimum for a 10-uL inj. vol..

And...you likely already know that with overfill you'd need to use a minimum of five times the loop volume's worth of sample...e.g. if you want to inject 10 uL, you'd aspirate a minimum of 50 uL for that injection.

Please, see what you think, and my thanks.

Hi and thanks mattmullaney,
You are correct about the Rheodyne valve. My original method uses the 25ul loop and a 10ul partial injection, I was just trying out different configurations to see if I could see a change. I was using an overfill factor of 5 when in full loop mode - Chromeleon actually makes this a minimum requirement when setting up the method.

I also got a suggestion to try using split septa or removing the septum from the vial in case there's a vacuum being created in the vial, which was a good thought. I usually have a puncture offset line in the program but had forgotten it in the caffeine run. I still see the up-down-up-down peak area pattern with no septum though.

Hi Again, Marie666,

Okay then, that was the easy idea I had. The injection volume is well, and there is no apparent issue with vacuum in the sealed vials, check and check.

I'll think on this a bit further...maybe someone else will have a better idea than mine.

Hang in There, I'll write back in a while.

Hi Marie,

Just a quick suggestion, but have you contacted the service help desk? They may be able to help you over the phone?

Hi Anthony,

Thermo says I need a service call, which may be the case but is not quite the answer I was hoping for!

Hi marie666,

Are you still having problems with this? Here are some additional thoughts:

• Double check the sampler's actual settings vs. the recommended values. For a PushFull injection, you should use the following settings with a 250 µL syringe and 25 µL loop:
  • Loop overfill - 5.0
  • Buffer wash factor - 2.0
  • Wash volume - 250.0
  • Wash speed - 20.0
  • Draw speed - 10.0
  • Draw delay - 2.0
  • Dispense speed - 5.0
  • Dispense delay - 2.0
  • Waste speed - 20.0
  • Wash dispense speed - 20.0
  • Sample height - 2.0
  For a PushPartial injection, use a partial cut volume of 10.0 and the values above (with the exception of draw speed which should be 5.0).
• Go into the command menu (F8 when you are viewing the control panel) and make sure the SmallAirBubbleSize is set to 12 µL.
• PunctureOffset should be set to 3.0 mm. This can be set in the program/instrument method or the command window above.
• Make sure the DC valve waste lines are routed back to the AS-AP waste port. See figure 2-7 on page 18 of the operators manual (http://www.dionex.com/en-us/webdocs/110 ... 361-07.pdf).
• Double-check your transfer line volume calibration to make sure it is correct. This is very important!
• How old is your sampler and when is the last time it had a PM? If it's been over a year, it might be time for some preventive maintenance.