Gradient Elution: Delays Between Runs?

[dkreller]

Hi Forum,

I setting up, based on an established literature procedure, a (non-aqueous) RP-HPLC method. The mobile phase (flowing at 1.0 mL/min) has 3 solvents. At the start, it's 100% MeOH. Then, over 15 min, it linearly changes to 50% MeOH, 28% iPrOH and 22% hexane. It seems to only make sense to, at the end of a completed run, pump some of the initial solvent (100% MeOH) through the column to get it ready for the beginning of the next run. The article doesn't say anything about inter-run delays, but maybe it's obvious to people familiar with running gradient elution - a group that I don't obviously belong to.

Is there an amount of the begining composition solvent that is recommended to re-equilibrate the column back to the initial solvent, perhaps expressed as a number of void volumes?

We are using a 100x4.6 mm C18 5 micron Shimadzu column.

Also, what is a good un-retained solute to use for RP-HPLC, to measure void volume? I use acetone in my aqueous LC work, does acetone also work on a C18 RP column? Seem pretty polar, and might not interact with the stationary phase, but I thought I'd ask the pros...

[HPLCCONSULT]

Yes, if you are doing serial runs then you will want to wash and flush your column down after each run to remove any remaining material from the column before you start the next run. The general rule is to wash the column with at least three column volumes of solvent before equilibrating for the next run. Five column volumes is better and is what I use, but just watch how long it takes for the detector signal to become stable again AFTER three column volumes to establish what will work in your case. Remember to wash the column with the strong solvent and then equilibrate with the starting solvent(s).

Here is a link to typical Column Volumes though yours is about 1.16 mls.[http://www.hplctools.com/columnvolume.htm]

[dkreller]

So, something like this would be a place to start?

Between runs:

1) wash column with the strong (end of gradient) solvent for 4 minutes, (4 mL of solvent, about 3 1/2 void volumes) then
2) switch back to the initial solvent for 4 min.

Now for the next question, do you think I shoud use a gradient when I go back to the starting solvent composition?

Thanks,
David

[Uwe Neue]

You can go back in a simple step.

I do recommend to go to 5 column volumes for reequilibration. It is a safer place to start.

Will you be running the method with many repetitions, even going around the clock? Or are just making a few runs?

[dkreller]

Hi Uwe,

Thanks for your advice r.e. sample equilibration volumes.

I will run at least 30 samples, so I will program in a sample set / use the system's autosampler.

David

[Consumer Products Guy]

I do recommend to go to 5 column volumes for reequilibration. It is a safer place to start.

We use more than 5 column volumes usually; most our runs are overnight, so time is not that important. I'd say only about 10% of our assays are gradient though.

I think what's best is to establish how much re-equilibration/post-time works for you (no change in retention time), then use that plus a minute more.

[Uwe Neue]

Thanks for your comment! I did not want to go into too much depth on this. The "5 column volumes" are an old rule of thumb that has been demonstrated to work in many practical cases. However, one can essentially eliminate ANY additional reequilibration, if one follows the thoughts of Pete Carr et al., who wanted to see how fast one can go. If you are willing to throw away the first gradient run (which many of us do anyway), then you will get away with no reequilibration whatsoever, provided you make the injection at exactly the point when the beginning mobile phase of the gradient reaches the injector. The latter requirement is due to the fact that there is a difference in the time when the gradient starts at the pump and when it reaches the injector/column. However, all of these things are too complicated for a newcomer, so I figured the old rule of thumb will work best for him. Plus, he is under no time pressure: the 30 runs will not even keep the instrument busy for the entire night.

[tom jupille]

In addition to all of the above, don't forget to factor in the time required to get through the "dwell volume" (gradient delay volume).

I'm more conservative than Uwe, and generally suggest 10 times the column volume plus 1 time the dwell volume as a starting point, and then experiment with cutting down from there.

[dkreller]

Tom,

Thanks for your comment. How do you quantify the 'dwell volume'?

David

[lmh]

For what it's worth, I'm also a 10-volume sort of person (which is also why I am keen on using as short a column as I can), but perhaps overcautiously I also regard my first gradient run as an overall equilibration run, to be thrown away.

Thanks, Uwe, for the insight on no-equilibration methods. I've also seen methods with only 1 or 2 min of initial conditions before injection, and they scare me slightly. If new solvent replaces old with 100% efficiency, then we wouldn't need to equilibrate for more than 1 column volume in any case. But we know empirically it often makes a difference, so clearly new solvent doesn't replace old perfectly. In which case, starting a new run as soon as the solvent front has hit the column seems a bit risky. In the methods I've seen where there's no re-equilibration, almost invariably the gradient starts at a much lower % organic than the elution area of the peaks of interest, which adds some safety.

[tom jupille]

How do you quantify the 'dwell volume'?

The procedure is described as part of the "Troubleshooting Wizard" on our web site:

http://www.lcresources.com/resources/TSWiz/hs410.htm

[DR]

Too long to be a rule of thumb, but I'd start with 5x whatever is larger (column volume or system dwell volume) - and WATCH YOUR BASELINE over the course of a couple of blank gradient runs - does it return to what it was at the start of the gradient? If it returns well before you've reached your 5x volume, trim the "initial conditions hold" period at the end of the program to about a minute longer than it the baseline takes to return to zero.

Sorry about the caps, but so many chromatographers forget to do this after reaching a certain level of comfort/experience/complacency...