Having ongoing problems (plugging) with frits (metal or plastic) in columns or as precolumn filters it would be instructive to have some imput by people using columns, or precolumn filters,
with sieves (screens, reportedly Bischoff and Micra, now Eichrom, have such columns). Specifically: Is there someone who can relate some experience in comparing frits with sieves in regard to plugging?
We have used sieves (Valco) in line (pre injector) for years with very satisfying results. These take a very long time to plug, and if they do, the whole sieve is covered with visible debris. In comparison, I seldom see anything on plugged frits. Unfortunately, there has been no occasion to use columns with sieves in this laboratory.
Hans

We have used the Upchurch inline filters which I think are some sintered stuff (this is what you are calling a frit?) Never filtered mobile phases and only filtered sample when they were obviously in need. The inline filters plug now and then. Never had a column fail with an Upchurch inline filter in front of it. Always could see crap on filter when pressure across it went up. I consider the Upchurch inline filters the best investment in LC.

PS I own no stock in Upchurch and they never have given me a free tee shirt.

Thanks Bill, do you analyse biological fluids or protein solutions, etc.? If we had an occasional plugging I wouldn´t complain... . actually its frequent flow restrictions which get me. Also, for brevity´s sake I didn´t mention above that the problems were actually with pre-columns (guard columns), we also had few pluggings or restrictions in main columns in the last 10 years or so.
Now, Bill, you have prompted me to launch another complaint: It has not been possible to get small amounts of stationary phase for those newer generation columns. My beloved Upchurch (also no affiliation) pre-columns (can be dry-packed by hand)are mostly sitting around idle. That´s why we have largely also gone over to pre-column filters (sintered frits), hence the problems mentioned above. I am not sure yet, whether the guard columns did a better job, though.
Incidentally, the sieves are very smooth looking wire mesh on enlargement.
Hans

HWM,
nope, never did anything biological, other than a phosphate buffer I let set around too long.

Buy your columns from ES Industries (affiliation...long time friend) and ask Mat for packing for guards. They have equivalents for most columns. You will be supporting about the last of the family column mfg. businesses. Not as fashionable as supporting the "family farm", but still worthy.

Sounds like the frits are nucleating precipitation and the wires aren't, but I wouldn't know why.

Thanks again Bill, checking into the possibility to connect with ES from Europe.
I am one of those preferring "one man" mfg´s, but often found that their packing materials just did not measure up.
If one looks at micrographs of frits (there were some nice pictures in an old Micra catalog)one can imagine how proteins can get hooked up in the many crevices. They probably denature in the process, anyway, other proeins must be associating (a coagulation?) with the immobilized ones, etc. It seems that I mentioned this before: We get some one-way valves this way (reverse flow remains fine). The wire mesh (sieve) can be expected to be too smooth for this to happen. That´s another form of my question: Is it proven that sieves let proteins, etc., pass more efficiently?

There are several big differences between wire mesh screens and frits made from particulates. The screens have a very small open area, in the order of 5% or less of the total area. Frits have an open area of around 35%. Screens are very thin, making the bandspreading very very small. Frits are thick, which may cause some bandspreading.
In the early times of HPLC, we (Waters) used exclusively screens. There were continuing problems with plugging, due to sample constituents and from instrument debris. Since the mid 80's we therefore switched everything over to frits. Even with the larger open area of frits, one can occasionally encounter difficulties with plugging. Based on this, I do not recommend to go to screens. For your problem, it is the wrong direction.
How quickly a frit plugs, depends not only on the open area, but also on the pore size of the frits. It appears that only Mac-Mod-Agilent and Waters have discovered this until now. In some of our columns, we are using different frits at the column inlet and at the column outlet. You may want to explore this improvement with your samples. I found that the problems with plasma samples declined quite drastically with this approach.

Could it be that the wires in the mesh were too thick in "the early days", so that Waters has "burned it´s fingers" and doesn´t want to try again? The 500x micrographs of Micra would tend to indicate that their mesh has more open area than their frit (of course, they could have picked a particularily bad frit).
Uwe, have you seen the phenomenon of one, lets say, 50µL injection of plasma/serum completely plugging a previously completely normal frit?
Also, when you used an inlet frit with larger pores, did plugging occur in the stat. phase?
Did anybody ever see plugging due only to stat. phase and only caused by proteins?

I doubt that the wire mesh has changed drastically. To go from 5% open area to 35 % is a giant step, not easily possible with a given technology. I do not know the pictures that you are referring to, but a picture is not what counts. What does count is what goes through a frit, or a screen, for that matter. I have seen a picture of a screen from Bischoff, which may be the same stuff that Micra showed. This was the same technology that we abandoned in the mid 80's.
My colleagues have been doing routine work with up to 200 microL injections of plasma after 1:1 dilution with an internal standard. The frit has NEVER been the problem in these investigations.
In the case where we made the comparison of frits with different pore sizes, we did not get any increased plugging of the stationary phase. The balance was such that the columns with the larger diameter frits lasted longer. Period. If you see this several times, you are convinced. I have to admit, that I did not investigate, if as a final result the packing was clogged or the column, at least I do not recall any data on this. It would make sense to me that the packing finally clogged.
Plugging of frits can be caused by 50 million things, but the fastes is the injection of protein-containing samples. Plasma, serum, milk are all wonderful examples. However, I bet that you can get similar results by injecting a polymer sample into an incompatible solvent.
Actually, we recently created examples where a sample is known to precipitate, but it does not plug the column. This was done on purpose to improve load in preparative chromatography. Of course, under these circumstances the sample is washed from the column afterwards.

Oh, I forgot: there is another technology around, where a frit is formed in a thin wire mesh. This trick is a very good one, since it reduces the volume of the device while keeping a larger open fraction of the device. Have not seen it used though....

Looks like that kills my grand idea.
Incidentally, we had to quit injecting plasma/serum directly (into Pinkerton columns)in one study, because of the severe plugging. The addition of Na2SO4 and methanol + a micro-filtration prior to injection fixed the problem. The plasma/serum was mostly from highly catabolic patients, who have a strange composition of blood proteins.

Of course, there is nothing better than trying things out yourself. I would suggest to get a larger pore size frit for the inlet of the columns. Maybe some other manufacturers will get the hang of this as well, especially if a customer tells them about the new trick...
Take care
Uwe

Dear Uwe,
Do you have an idea who the producer of this "frit formed in a thin wire mesh" is?
Many years ago I got few pieces from the company Gynkotek, but nobody could remember on the source anymore.
Robert

Hmm - same here. It may have been Mott Metallurgical. I'll see if I can find a web side. - Uwe

Uwe,
If the inlet frit has a larger pore size than the outlet frit, does this mean that the column might shed particles if used in the reverse direction?
I know that manufacturers show a flow direction for use of the column, but some people reverse the column when the inlet end gets contaminated and still manage to get some further good use out of it. Of course the stability of a packed bed when used in the opposite direction in which it was packed is another question, even if the frits at the inlet and outlet end are the same size. Any comments with regard to (ab)use of Waters columns in this way?

I did not find a web site for Mott Metallurgical, and otherwise, I was to busy today to remember to look for an address. Sorry!
I favor the approach of protecting a column with a guard column. Then there is no need to reverse the column...
The bleed from a larger frit is really small. I have seen a blip in the baseline every half minute, when a column was used in the wrong direction. If a blib is a particle, you can run the column in the reverse direction for a long, long time before a significant loss of particles has occured. So, as a last resort, I don't see anything wrong with running a column in the reverse direction, even if the inlet frit is larger. I don't recommend it, but I would not panick about it either.