flow direction on the HPLC column

[tom jupille]

Cliff, re running in reverse, the article also says that after a short while, the direction didn't matter.

[Uwe Neue]

Tom, if you make the inlet frit a small size, the column, i.e. the frit, tends to clog up faster. If you put a frit in front of the frit, the frit will clog up. See my point?

[tom jupille]

Uwe, I see the point, but let me play "devil's advocate" here (I'm good at that ). If you put, say a 2-micron frit on the inlet, does that mean that a 0.5-micron in-line filter upstream of the column is required (as opposed to, say, just recommended)?

[Victor]

Tom,

I'm not sure what you are getting at here. It seems plain to me from the answers from the column people that the column inlet frit gets blocked more quickly and thus it is usually made with larger porosity. I imagine that the column would better protected if a prefrit (still don't understand exactly what this is) is placed upstream of the column, because the rubbish that blocks the column frit will instead block the pre-frit instead; but then it will have to be changed as Mr Uwe says. A guard column containing packing is I guess a much more efficient filter than a small porosity frit on its own, so I would guess a guard column containing packing but with a high porosity inlet frit and a small porosity outlet frit is good enough for Dr Pangloss-the best of all possible worlds.

Let's ask Mr Uwe and Mr Mark from Dionex to forget the frit business, and advise us whether their columns are asymmetric in terms of their packing, and whether, in the absence of frit problems, whether in theory it would be possible to reverse the column and use it in the opposite direction in which it was packed with no performance loss.

[tom jupille]

Victor, I know the answer to that one. Absent frit differences, flow direction is not a significant variable for most HPLC columns. As has been pointed out earlier in the thread, the majority of columns are packed at much higher pressure/flow than encountered in normal operation.

However, the comment about "absent frit differences" is equivalent to the question "Aside from that, Mrs. Lincoln, how did you enjoy the evening at the theater?"; if the inlet frit has a larger porosity than the outlet frit, then reverse-flushing the column is a bad idea!

*For those unfamiliar with American history: US president Abraham Lincoln was assassinated in 1865 while watching a play.

[HW Mueller]

Very agreeable, Tom! As mentioned above, in my work the reversibility of columns proved to be one of the more important advances in column development.
Some proteins not only tend to block frits, but if these get through they also plug the first few mm of the column bed. Other crud also settles on the initial part of the stat phase. To clean that out by passing it through the whole column, on top of that past an outlet frit which is finer than the inlet, is unthinkable to me (since 10+ yeaars). But, to play the devil himself: Maybe reversal of such a column is just the right thing, you loose part of the filling, then it doesn´t work at all, you have to refill it giving a chance to put the same frits on both sides. Darn, forgot, one doesn´t get the hands on that stuff for refilling.

ananda, its not so much the size of proteins, almost all are much smaller than the holes, pores..... It´s the strange behavior of some, apparently they cling to a crevices, then others cling to that one etc.?? Thus, sieves have much less tendency to get plugged, . . . we have been through this before.

[bert]

Darn, forgot, one doesn´t get the hands on that stuff for refilling.

We use the packing material from columns that have passed away; remove outlet frits and pump ......

[Victor]

Tom,

I really liked your post. For completeness of the discussion, what play was Mr Lincoln watching at the time?

[Russ]

Our American Cousin

[Victor]

Thanks Russ-brilliant!

Tom- I had the impression that columns were packed at pressures up to 10,000 psi- that's about 700 bar. This is only 2-3 times the maximum operating pressure of the column. The part of the packing at the bottom of the column during packing would experience a lower pressure than this maximum which does not build up until the column is completely packed. Reversing the column after packing might expose this part of the column to a pressure of the same order as the pressure it experienced during packing. As I said, I never saw a problem in practice but... I go back to the question of whether the column packing is really symmetric.
Can the column making people answer this ?

Did I get it all wrong?

[Mark Tracy]

Unfortunately, the detailed answer to your question is a trade secret. But we pack is such a way as to reduce the effect you mention. Non-uniform packing from one end to the other will have non-optimal efficiency.

[Uwe Neue]

Tom,

I have never seen stuff from an injection migrating through a column and clogging the outlet frit.

Hans,

You can even turn a column with a larger inlet frit around for a bit. The sand does not come pouring out in a hurry. Just don't do it for days...

[Victor]

Thank you for your answer Mark.

Tom- I think this topic of column asymmetry is not as simple as you are making out. However, the way you told the Lincoln story was brilliant. Is that an American saying, or did you make it up yourself- I mean the "Aside from that Mrs Lincoln.." bit- brilliant!

[tom jupille]

Uwe, that's the point I was trying to make: particles small enough to pass a 2-micron inlet frit will probably lodge at the top of the packing {my geometry is rusty, but the interparticle distance for a 2-micron packing is, what, about 0.2 microns? }.

Victor, I wish I were clever enough to make up jokes like that. It's been around for quite a while.

[JA]

If we know a 3.5 µm packing analytical column has a 5 µm inlet frit and a 2 µm outlet frit then reverse flusing is not possible/recommended. We know it is recommended to use a low porosity precolumn filter (frit, or guard) to prevent clogging of the main column but there might be many users out there who do not employ such a thing.

Assuming one didn't have a precolumn filter, which is it worse to have clogged up - a 2 µm inlet frit (if the column had one) or the actual stationary phase on a column with a 5 µm inlet frit (if the interparticle distance is as small as Tom suggests then presumably this is likely to capture a lot more crud).

It's a two-pronged question,
Which is actually the worse of the two cases in terms of column useability (can either or both problems be rectified) and
Which do the column manufacturers perceive as worse - which one is likely to lead to the most consumer calls to the helpdesk, unhappy customers, bad feeling towards a column brand etc.