equilibration time

[gerapas]

is it correct to say that after a gradient, the system has been equilibrated when the pressure is back to its sarting level?

[HPLCaddict]

No. Of course, the pressure will return to its original level, but watching the pressure alone is not sensitive enough.
The system is reequillibrated when the baseline reaches its original level (usually zero when you zero'ed at the start or use autozero).

[tom jupille]

I find it useful to make a distinction between "washout" time and "equilibration" time.

The former is the time required to *physically* replace one solvent with another in the system. As a good rule of thumb, it takes about 10 column volumes to completely switch over from one solvent to another. You have to add to that the dwell time (the time required for the solvent to get from the pump to the column).

The latter is the time required for the system to come to *chemical* equilibrium. That can range from negligible (when you simply change the aqueous/organic ratio) to ridiculous (long-chain ion-pair reagents).

Note that it is possible to do very good, very reproducible chromatography with an incompletely equilibrated system; the catch is that it has to be incompletely equilibrated *exactly the same way* every time.

The only way to know for sure is to do the experiment: keep cutting the equilibration time until you see an effect on the separation -- and then make sure you allow for dwell volume differences if you transfer the method to another instrument.

[lmh]

(1) Personally I wouldn't trust the pressure-back-to-normal approach; what you're relying on is using the back-pressure (i.e. viscosity) as a measure of the solvent mix in the column. It just doesn't feel very reliable to me.
(2) Nor would I trust baseline-back-to-normal unless using a detector that I know gives wildly different reading with one component of the gradient to the other (which is not generally a very happy situation as it means looking for peaks on a wildly changing baseline).

I tend to use Tom's wash-out volume, but appreciate that this is wasteful for high-throughput methods as Tom's right: you can work with a system that's not fully equilibrated.

In terms of short-term equilibration, I throw away my first run in any sequence, and reason that all subsequent runs began with the instrument in the same condition (whatever that was). In reverse phase chromatography, if equilibration is too short, one early symptom can be loss of resolution/peak-shape of early-eluting peaks, when the samples is injected in too strong a sample solvent, or too great a volume. I.e. if your method is only just retaining these peaks (because the column still contains solvent that's almost strong enough to elute), then it is teetering on the brink of failure, and a tiny extra push from a little bit of extra methanol/acetonitrile in the sample will bring it to disaster. Main point: a fully-equilibrated method may be more robust to slightly-incorrect samples than a partially-equilibrated method.

Long-term equilibration is a far more serious problem. If your column has been used with long-chain ion-pair reagents that are gradually washing out, then the symptom is gradually changing retention time over tens or hundreds of runs. Unfortunately this isn't something that can be overcome by altering the re-equilibration time between individual runs.