I’m mixing a medium that tends to form persistent bubbles. I have a liquid class and pipetting behaviors that are working pretty well, but when I do mixing prior to aspiration there will is some air ejected with the final dispense of the mix cycles. This doesn’t seem to be happening with a particular number of cycles, but rather the last cycle. I’ve set transport and blowout volume to zero, so I’m confused as to where this air is coming from. Is there some secret air that isn’t listed in the liquid class that gets moved? Where else should I be looking?
Your issue could be your mix flow rate itself. Depending on the volume and viscosity of the liquid, the speed may be churning inside the tip. Unlike aspiration and dispenses, mixing does not utilize the settling time parameter which means the volume may not fully finish aspirating before the dispense step starts which could start introducing air that is already inside the tip with the liquid. My suggestion would be to try halfing or quartering the mix flow rate to see if that reduces the bubble formation.
Just to clarify. Final dispense of the mix? Or final dispense on the dispense to your target?
How much volume are you using for the pre-mix and how much volume is inside the source? Are you using 2mm submerge depth? Are you using a mix position greater than 0? Finally is liquid following turned on?
I guess the air results from the fact that a minor portion of your liquid sticks to the inside of the tip. Consequently, this liquid is replaced by air during dispense.
Is it possible in your case to use independent aspiration and dispensation (surface mode) for mixing instead of using the mixing as part of the aspiration single step?
Then you could aspirate a little more volume than dispensing to perform the mixing procedure.
For example
This is final dispense of the mix step, prior to transferring any samples to destination. The pre-mix is set at a percent of the tip capacity (calculated by number of wells to fill from the single source), which for the test I ran this morning means 650uL. Pipetting from the bottom plus an offset, so there’s a comfortable head of liquid over the tip when it’s mixing
What is the fixed height from the bottom that you are using? Also for your mix, what is the mix position parameter set to?
Does the bubble form using cLLD with 2mm submerge, liquid following, mix position of 0 and also mixing a maximum of 75% of the total volume in the source?
I’m not aware of any system air that is used for mixing purposes, however the properties of the liquid may not help and a pre-mix may need to be avoided to prevent creation of bubbles.
@BrandonBare_Hamilton Here are the mix parameters. Mix speed is still 100uL/s. I’ll note that even at zero there is a comfortable gap between the real bottom of the source (clear 50mL conical) and the tip.
With cLLD, 2mm submerge, etc I still get the bubble. This happens with 5 cycles on the 5th one, and also with 3 cycles on the third one. The medium is essentially DMEM, for what it’s worth, which is a bit goopy but isn’t what I’d call extreme in viscosity.
No bubbles I could see with these settings, or with the mix speed back up to 500. Which supports the notion of something happening in the tip rather than secret system air. Knowing this can you suggest any other changes?
Ok so this confirms we are fighting liquid properties. Since there are no more fine tuning parameters that can be done with the pre-mixing steps, I would suggest utilizing aspiration and dispense steps within a loop to perform your mix. This way more of the parameters are open to you to control the pipetting. @Martin 's suggestion to utilize reverse pipetting is good one to try as that will use the liquid inside the tip to prevent further air from escaping. The settling time within the aspiration / dispense steps will also ensure the liquid has time to make it up the tip before it moves on to the dispense.
For posterity: I added a pre-wet in media prior to mixing/dispensing the samples. Bubbles still form in the media for the first few asp/disp cycles, which is fine because by and large they migrate to the edge of the reservoir, and then no more bubbles appear in subsequent pipettings.
