Do I need a Guard Column?

[robin]

Uwe
Clearly adding a guard column as you describe would give an improved plate count initially simply by having a longer column. I am sure the suppliers would be very happy to supply guards on such a premise but I would not be an advocate of this. You assume that the negative effects associated with column coupling and increased analysis times through use of an extended bed length are insignificant. Of course with dirty samples the increased plate count would be lost if the guard column rapidly becomes contaminated.
The reality of every day operations means that guard columns are used for convenience rather than as an essential element. The use of guard columns may well be unavoidable for the type of samples you handle. The point I make is that the use of guard columns should be regarded as a method of last resort and in the situations you describe they are the lesser of 2 evils

[Kostas Petritis]

Of course with dirty samples the increased plate count would be lost if the guard column rapidly becomes contaminated.

Robin,
In the above case you would be thankfull to use a guard column otherwise you would have to change your separation column. Furthermore, for analytical columns (i.e. 4.6 and 2.1 ID, 10 cm or more in length) coupling is very easy and increase in analysis time insignificant.

I think that most people agree that guard columns are good for a certain type of matrices. SPE or other sample pre-treatments can help that much...

[tom jupille]

Sometimes use of a guard column can be a conscious choice. Extensive sample cleanup is not free; the additional cost in time, solvent, filters, etc. have to be balanced against the cost of changing guard cartidges.

[HW Mueller]

Uwe,
when your samples require that the column be subjected to a thorough cleaning after about 1-20 injections an experiment that uses highly pure standards has little bearing (except that such an experiment shows that the column might be ok if you can get rid of the dirt).
Also, denaturing does not necessarily mean precipitation. Urea denatures proteins, yet keeps most of them solvated, as an example.
Robin,
do you really have an example where crud on a guard caused "visible" deterioration of the chromatogram, even though no plugging, severe restriction, or a void was involved? Surprisingly, I am not sure wether I have seen this, I only Know that brown-black material accumulates quite rapidly with "dirty" samples. Since I assumed that this is detrimental to a column I use guards for such situations, even though a guard only slows the formation of this discoloration in the main column and that I have no unequivocal evidence that the black crud really effected the chromatogram. This brown-black stuff has not been removed by stringent cleaning.

[robin]

HW
I agree with much of what you say. It depends what we define as dirty samples. Excluding so called “crudâ€

[Black Sunshine]

In my opoinion, there is a give and take when using guard columns. You may extend the life of your column and increase the theoretical plates but you will also increase the tailing effect of your cromatogram. A big consideration as to whether or not to use a guard column, is whether or not the method will be scaled up.

[HW Mueller]

Black Sun... , how do you get an increase in theoretical plates, yet get more tailing? If you mean at some point you will get tailing,...... just replace it (or better ahead of such a time, if you can estimate that).

[Black Sunshine]

HW,
The increase in theoretical plates would simply be the result of a longer bed. The tailing effect I was referring to is more applicable to semi-prep, prep, and process work. This is why I also stated that it is important to know whether or not yhe method would be scaled up.

[HW Mueller]

OK, two separate things. In any case, if your guard causes tailing it appears that you made a mistake, for instance: It´s too large and badly packed, has a wrong stat phase material, has voids.