I have tried to regenerate some old columns, but I am not very sure about in what extent I have got it. I have read about some tests with groups of substances, but I would like to know more about this. May somebody help me?

The only experience I have had in the regeneration of columns are performing the regeneration in the method prescribed by the manufacturer.
There are available sample solutions containing a mixture of substances that can be used to test the columns performance.If this is what you are referring to the column manufacurers can supply you with these substances.

See the discussion "Column Regeneration" August 23, earlier in the Forum, if you haven't already.

If you have run test mixes on the re-worked columns, and the results are as good as a new column, then you did it right!! Congratulations!

There is usually a test chromatogram with conditions and test mix with every new column; you can buy commercial test mixes, or make them yourself. You might want to consult the catalogs for a mix that would be most appropriate for your work. Usually, we just use standards of what we are familiar with, since we don't always use UV detectors.

In addition, to avoid the problem of having a box of 'old' columns in unknown condition, we have dedicated columns to particular projects until they are worn out. This makes it easier to track their status, and saves time in the long run because the instrument only requires small changes (if any) when a project is re-started, instead of completely re-validating the column. There are just enough difference in the columns from different companies to make it a hassle to substitute another column, unless it is a VERY simple analysis. I've started including the column serial number in the documentation to make sure I have not just the same column type, but the very same column, if possible. Helps keep life in the lab simple, leaving more time to work on the really interesting problems.