Dear all:

I have a few questions regarding sample carryover associated with autosamplers.

1. If the injector needle causes sample carryover, how should we choose washing solvents (pH, buffer type and concentration, organic solvent et al)? If a acid is sticky, would it be better if washing solvent is acidic so the analyte is neutral and won't stick to ions on needle surface? If the analyte is a base, what to do?

2. If the switching valve causes sample carryover and we can NOT switch to switch valve made of different materials, is there any possibility to correct the carryover? What about higher buffer concentrations (if there is more sensitivity than required)? Other alternatives?

Thank you very much for your input.

Jimmy,
For the needle, what options does your autosampler have. For example, with one brand, you could pull solution from three different reagent reservoirs and trick the autosampler into putting the solution each time into waste.
For the injector, can you put a fixed loop on your injection valve and instruct the autosampler to overfill by five volumes?
Jim

Ditto Jimmy, let us know the make and model
of the autosampler and we might be able to
help.

I use a Gilson 234 autoinjector so that's what
I know best, although I have used a Kontron
ancient thing with a carousel belt etc.

We experienced carryover problems with Agilent 1100, PE 200 for LC-MS or LC-MS/MS and Alliance 2690 for HPLC. Unfortunately, we don't have many options with respect to washing programs because of software limitations. Our approach is to select suitable washing solvents. That's why I raised the question.

We work on biological samples and can not tolerate carryover of more than 0.02%.

Thank you all for the response.

With the PE 200, there is the option of washing the needle pre, as well as post, injection. This makes for alot of washes which can be performed. If all of these autosamplers have carryover, it may be time to look at the method as opposed to trying to fix autosamplers which may be performing perfectly. Try to choose a wash solvent which the sample is COMPLETELY soluable in and then wash the heck out of it after each injection. The Waters and Agilent needles and loops are part of the flow path and should be washed pretty well during the run.

The Varian autosampler has a superior programable washing apparatus (IMHO) compared to the competition. I have had excellent success in completely washing samples from syringe and needle below detectable amounts. I feel it is the most versatile unit on the market.

25-year Pharmaceutical analyst

We had major caryover problems with our Waters Alliance system. About a year ago, they re-designed the seal pack and reduced the dead volume. The carryover is much reduced. Our system is under contract so it did not cost us anything. It might be a free upgrade (it really ought to be).
We have had problems off and on with the Agilent 1100 LC. The HP 1090 is still the best for minimizing carryover that I've seen.

Carryover can be caused by the rotor valve, something like,for example, Rheodyne valve. In this case, needle wash does NOT help at all. What can be done if the carryover is caused by ROTOR valve? Probably the slelction of sample solbent and/or mobile phase is the key.

Thanks for the suggestions gigen.

There are different materials available for the rotor. If you replace one with another, the specific adsorption problem will go away.

Two of the most common materials for the Rheodyne type of valves are tefzel and PEEK. We have found that PEEK rotor seals are better when it comes to carryover. They are easy to install and only cost about $75.00.

If the carryover problem is coming from the valve, one solution is to use an autosampler that does not draw the sample all the way into the valve. The Summit ASI-100 autosampler achieves this by loading the sample into a length of tubing between the needle and the valve (rather than a loop beyond the valve), then moving the needle to an injection port and reversing the flow. This design eliminates sample waste and minimizes carryover - all the valve ever sees is mobile phase.

Even with valve carryover eliminated, you need to still consider the potential of adsorption on tubing and needle surfaces. Without knowing the nature of the sample components and your mobile phase, it's difficult to guess whether this is likely to be significant or not. With the injector design described above, the tubing and needle are flushed with mobile phase throughout the run, so the likelihood of carryover due to adsorption on internal surfaces is small. The two remaining locations with the potential to cause carryover would be the outside of the needle and the needle port. Using good septa on the vials typically keeps cross-contamination negligible; a needle wash can further reduce carryover.

I need to clarify a point on my previous posting. In a split-loop injection system like the one used by the ASI-100, the sample does not come into contact with the valve during sample loading, but it actually does pass through the valve when the valve is in the inject position, because the needle port is plumbed through the valve. However, the residence time of the sample in the valve is very brief, and the sample is never in contact with the valve when the rotor is turning, so there is much less opportunity for residual sample to end up in the interstices of the valve with the split-loop design than with conventional injection plumbing configurations.