Chromatography Forum

Hello all,

we have accidentally re-quilibrated an HLPC method at 25% instead of 40%B, which are the initial conditions.

The results were very good however. Might this be a good practice to adapt? It should help initial trapping of the analytes at the head of the column.

Is this being done routinely elsewhere?

Best,

Arne

Hi Arne, this would not be standard practice however all you have done is essentially added an extra step in the gradient program equal to the delay volume of the system.

so if the re-equilibration phase of the gradient is sitting at 25% but the start of the next injection the gradient is at 40% the next injection will not start until the pump reaches 40% but the solvent composition after the pump but before/on the column will still be 25%.

so if your gradient should be eg Tmin(%B)= 0min(40%) 5min(90%) 5,1min(40%) 10min(40%)

your new gradient with the incorrect re-equilibration phase assuming a delay of 1min would be 0min(25%) 1min(40%) 5min(90%) 5.1min(25%) 10min(25%)


it is unlikely your system actually has a delay of 1min this is obviously flow rate dependant and I dont think this would effect your chromatography too much however it is generally bad practice to inject during a gradient change so i would advise if you want to keep the 25% re-equilibration to add it and an initial hold period to the beginning of your gradient then add your old gradient on top of that.

cheers and good luck,
Chromavore

we have accidentally re-quilibrated an HLPC method at 25% instead of 40%B, which are the initial conditions.

Is this being done routinely elsewhere?

I never did that, we equilibrated at initial conditions ONLY.

I made a good living by paying careful attention to such details.

well, I've been making a good living for ten years as well and never did such a thing before...

Obviously, this was an oversight during method development. I had been thinking about an initial step-gradient for better trapping that is programmed in.

So the question would have been better worded as: "Is the practice of an intial step for better trapping used in gradient elution?" rather than "Do people re-equilibrate at the wrong %B?".

Thank you for your insight, though.

Best

Arne

Hi Arne,

Just curious: why do you want to trap things at the head of your column in the first place? Just to reduce the separation of the injected matrix prior to reaching the stationary phase? Or is it to remove any injection solvent interference you may be running across?

I wouldn't do this, but the reason's not obvious.

Trapping is actually quite sensible if you want to inject large sample volumes. It happens automatically on a gradient method. If you injected 50uL of a sample dissolved in running-solvent on an isocratic method at 200uL/min flow, it'd be horrible, because your peak would be 15 seconds wide before it's even started to spread as it passes down the column. So if you start your gradient at a level where the analyte binds tightly, using a sample solvent that's weak, you can inject an enormous volume and the analyte will build up as a tight band at the start of the column, as if it had been injected in a tiny volume.

What's happened in the original poster's situation is an extreme of what gradient developers often do. Gradient people often start with a simple gradient from 5-95% and then realise that all their peaks are eluting between 40 and 50%. They then modify the gradient so it rises pretty fast from 5-40%, then much more slowly to 50%, then faster again to 95%. In the situation the original poster described, the method basically rises instantly from the equilibration percentage (which was a "trapping" percentage, very weak solvent) to something much higher (where maybe the analyte would be moving along the column).

The difficulty is working out when it happens relative to the injection! The previous method ended with the pump at X%, the new one starts at Y%, so at some point the pump must change. But does it change when the new method is loaded? Or when the autosampler declares "time zero"? I'm not 100% sure that every manufacturer will have made the same choice! If it happens when the new method is loaded, then the amount of solvent that's been pumped at Y% before the injection happens will depend on how fast the autosampler is, which is definitely different between manufacturers. If it happens when the autosampler declares "t=0" then whether or not the solvent has reached the sample or column depends on the various dead/delay volumes of the system, which again vary.

So you might end up with a method that works fine on your system, but doesn't work when someone moves it to a system from a different manufacturer, or when you upgrade your system and buy a better autosampler.

And for that reason, I wouldn't do it deliberately, though I'd be quite happy to believe it works!