dilution injector program on 1290 infinity II multisampler

Hi All,

My goal is to make the autosampler dilute our sample by 30 times (possibly more) before injecting 1ul, preferably without mixing in a separate vial. The reason being that we have very low (~1ng/ml) and very high (~2mg/ml) concentrations of some compounds in our samples. I have tried deoptimizing the high concentration compounds and that didnt work because they were still overloading the column.

Normal injection volume without injection program (to get low concentration compounds) is 3ul. We have a 20ul injection loop installed.

The injection program we are trying to use for 30x dilution is pictured below, idea being that if there is only 1ul of the diluted mixture left in the needle, it will inject that amount at the inject step. Methanol with a preslit septa is in vial 1 position, and an empty vial with no cap is our waste vial in vial 5 position. The injection volume on "multisampler" tab was automatically(?) changed to 20ul. The peak areas of standards run with this injector program are 900% observed to expected ratios consistently compared to the same standards run without the injector program. 

I also tried the program pictured below to dilute by 240x. When I tried changing the injection volume back to 1ul on the "multisampler" tab, the method errored at the second mixing step saying something along the lines of volume of the sample loop was exceeded. setting the injection volume back to 20ul allowed the program to run.

My main question is, what volume is injected when using an injector program? The volume left in the loop? or the volume specified in the "multisampler" tab. Also where does it go wrong with getting a 30x and 240x dilution?

I have seen other solutions that involve diluting and mixing in a separate vial prior to injecting, but we really want to avoid that if possible. 

Also please note: injecting only 1ul of undiluted standards does give 1/3 peak areas of 3ul injection. 

  • Hello  ,

    When using an injector program, the injection volume specified on the sampler tab is used whenever you choose to use the 'default' draw amount.

    If your mulitsampler has the mulitwash option, every time the valve is switched to bypass the metering head moves back for pressure relief. This can make it easier to receive the message allowed volume exceeded since you lose a little volume with each valve switch. You could try adding a step right at the beginning to eject maximum volume to waste. That will make certain to rehome the metering device before any other injector steps.

    From your program steps I'm not certain why you are seeing issues. Other than an initial eject step I would suggest using the eject to waste option instead of specifying a vial. Since you are going into your waste vial with the default offset, eventually your needle will be below the solvent level. You are rinsing afterwards, but depending on what is in that vial it may really dirty up the needle.

  • Hi Howard,

    Thank you for the suggestion. I added an eject step to waste at the beginning and changed the other eject step to go to waste as well (new program pictured below), and then ran 3 standard levels with and without the injection program (results pictured below). We expect the calculated concentration to be the same, but the peak areas should be 1/30th and they are in reality closer to 1/10th. 1/30th because the method without injector program injects 3ul undiluted sample. 

    When the eject is set to waste, where exactly is it ejecting to? Is it ejecting into the needle seat and then supposed to go out the waste line in position4 of the valve? If so, what flow is pushing it through? Additionally, we have a multiwash setup where there is actually a flush head attached to position4. So then where is the waste going? The wash steps in the method are set to flush both the needle and seat. Should I change the injector program wash step to only wash the needle?

    Thank you!


  •  ,

    For the Multisampler when the position is set to waste the needle goes to a position over the leak tray near the washport. That way the ejected waste is flushed out the same line as your needle wash solvent.

    As to your dilution issues, I believe I have an idea about what may be happening.

    Even after you eject the maximum amount from the sampling loop, there is still solvent in the loop. This is similar to how with a manual syringe there is always a small amount of solvent remaining in the needle even when the plunger is fully depressed. With any Agilent autosampler, the volume remaining in the 'needle' is the total volume of the loop. Here is a picture from the Multisampler user manual that shows how as sample is drawn up, it comes into contact with the solvent remaining in the sample loop.

    When you are mixing the 9+1 ul of diluent and sample it can actually be mixing and diffusing into the 20ul of solvent remaining the sample loop region. Then when you eject 9ul you aren't really getting rid of 90% of the diluted sample, but maybe only 60-70%, depending on how far into the loop region the sample migrated. 

    To test this, you could change your final eject step to eject maximum and see how much sample still remains diluted in the loop.

    One way you may be able to dilute in the loop would be to draw up some air first before drawing up your diluent and sample. You don't have much volume to work with, but you could try to do a 5 ul air draw before you do your solvent/sample draw and then mix with just the remaining volume. You could also try dropping to a 4.5 ul and 0.5 ul draw on diluent and sample for a little more gap/mixing volume. I don't believe the small amount of air in the loop at injection will disturb things too much. At system pressure the air gap should just go into solution.

    Let us know what you find. 

Was this helpful?