When is a column bad?

[sjmarin]

I am doing a study on column degradation. I need a good measurement that I can apply to many different columns to determine when they have failed.

I was looking at asymmetry and efficiency, and possilby retention time, but I need a value (20% redcution in N, 50% reduction in N?). There doesn't seem to be an industry standard.

I am curious as to what people use in their labs to determine when a column needs to be replaced, or what others who have conducted studies like this think.

Thanks.

[tom jupille]

No simple answer, because the effects of changes in column chemistry are method-dependent. An amount of "damage" to the column that will completely ruin one assay can be hardly noticeable with a different assay.

In practice, most methods have (and any validated regulatory method must have) "system suitability" criteria. These may include retention windows, resolution targets, plate numbers, tailing factors, repeatability, etc. for a check sample. When the column no longer meets system suitability, it must be replaced.

Bottom line is, the "failure" criteria depend on the purpose.

[Uwe Neue]

You also need to check other things such as shifts in retention or peak position and increase in backpressure. These things may very well be independent from plates and tailing.

[DR]

^ what they said - in my experience, a column is deemed "bad" when either it won't meet system suitability, typically due to loss of resolution between the critical peak pair, or because it's back pressure is too high for the user to have any confidence that the system won't quit due to excessive pressure in the middle of a run. Once in a while, a column will not be suitable for a particular analysis because it has been exposed to a lot of triethylamine or something else that has had a lasting impact on its chemistry. A run on that column just doesn't look like it would on a new column, so we tend to tag that column "for analysis of X only" or "Not for analysis of Y".
Unfortunately, there are several ways to render a column useless. In addition to those already mentioned, abuse of a column via exposure to mobile phases well outside rated ranges can damage the packing, they can be "dewetted", if they are whacked hard enough (not as big a problem as it was years ago), plugged with insoluble matter or exposed to a series of immiscible solvents, they can deliver unpredictable results.

[sivaiah]

During HPLC method development it self, u can check the trend of column performance parameters like Plate count, Resolution, Tailing etc. Then include the SST limits in MOA/STP for column peformance checking purpose. If column does not meeting the SST limits, then u can change the column.

[Chris Pohl]

DR,

Problems caused by true "dewetting" are totally reversible by treatment with a high solvent content mobile phase. Have you seen otherwise?