Methanol in LC columns

[Mattias]

I wonder if the performance of the column is affected when using acetonitrile as eluent, when the column has been used with methanol before.

How hard is it to remove the "methanol memory effects". Are there any standard procedures to speed up the equilibration? I have heard that you need to flush the columns for up to 24h...

[tom jupille]

If anything, it should take longer going the other way. The rule of thumb that I've always used is to flush with about 10 column volumes -- and then do a couple of replicate runs to check for repeatability,

[Mark Tracy]

Most C18 columns don't seem to have this problem as far as I have seen. Some polar embedded columns do seem to care about the history of methanol versus acetonitrile, and can take hours to stabilize. I hypothesize that this has to do with the hindered rotation about the amide bond needing time to find the equilibrium conformation for different solvents.

I have also noticed that polar embedded phases or phases with significant silanol activity remember the last buffer for some time, and take forever to stabilize when switching from a buffered mobile phase to an unbuffered one. For these columns it is usually better to avoid unbuffered mobile phases as much as possible.

In either case, the effect on retention times is minor, but the effects on efficiency and asymmetry can be quite large.