flow direction on the HPLC column

[Uwe Neue]

On the very first 3-micron column that I used in my life, the backpressure increased 3-fold within 24 hours of injecting standards (this was at some point in the early 1980s). Based on this, I concluded that the technology was not quite there yet. Over time, we have learned how to do this better.

Also, the inversion of a column with a wide frit at the column inlet is not an immediate disaster. If you hook the column to the detector for flushing it (which is not the recommended procedure), you get a sharp air-bubble-type signal every half minute or so. With other words, the packing is not pouring out of the column in a hurry.

[HW Mueller]

There is still the question about whether anybody has injected some difficult matrixes (including sticky proteins) on these "unsymmetric" columns. No plugging in both directions (since a large portion of the bed was crudded up)?

Somehow intuition would tend to indicate that it might be more prudent to do this unsymmetric thing with different frits only on a guard column.

[DR]

to jduan: if changing the flow direction can make the bed shift, then the column was poorly packed, period.

to the column manufacturers: if the recommendation is to use an in-line filter anyway, why not make both the inlet and the outlet frits on the column 0.5 micron?

Because doing so would cut into their guard column sales?

I am taken aback that the inlet & outlet frits are no longer the same. That said, I must confess that I have generally found that reversing a column at either end of its useful life has made little difference in its resolving performance or its back pressure. I just replace them, but then I have 1) fairly clean samples, 2) columns that generally cost <$500, 3) some columns with several thousand injections on them that are still OK. I tend to shy away from brands that exhibit back pressure problems before other types of problems, other things being equal.

[Uwe Neue]

Hans, we routinely inject difficult matrices on these non-symmetrical columns. I am not aware of any problems that could be attributed to the lack of a fine filter on top of the column. I actually just looked at a lifetime study on a 3.5 micron column using a MeCN-precipitated plasma sample without a guard column. The column looked fairly unchanged after 1000 injections, with an only marginal increase in backpressure.

[tom jupille]

Uwe, did you have an in-line filter upstream of the column (I surmise "no" from context, but . . .)?

.. and does the 3.5-micron column also use "asymmetric" frits (different porosity on inlet and outlet)?

[HW Mueller]

Ah yes, Uwe, an ACN precipitated sample! Most of the horrendous problems we had was with restricted access columns, where you are supposed to inject plasma/serum. Later I never injected anything, anywhere, unless at least most of the proteins were precipitated with Na2SO4, org solvents, or "removed" with ultrafilters, or a combination of these. Still with some catabolic patients we had problems even at the third column (in series, commonly and strangely called three dimensional). I mentioned this before: such plasma/serum even plugged 0.45µ microfilters!
When we learned to put the right amount of solid Na2SO4 to each plasma/serum sample the problems almost vanished.

[Uwe Neue]

Tom

In this experiment, no inline filter was used and yes, this was an example of an asymmetric column with the larger filter at the inlet.