The Sample Extraction Step is Key

[Jumpshooter]

*I have an SOP that says after you solvent extract the sample, then FILTER it through a PFTE 25mm,0.45um Acrodisk. The acrodisk I have is dimensioned "10mm,0.2um". I have used it but it takes FOREVER to filter. We have some other acrodisks that are "25mm, 0.45um' but they are marked as "GHP Membraned/Glass fiber". Is it OK to use these glass fiber ones instead of the PFTE type? What difference does this sort of matrix make?

[tom jupille]

Is it OK to use these glass fiber ones instead of the PFTE type?

Not without revalidating the method

What difference does this sort of matrix make?

Who knows? Some of your analyte might adsorb on the filter. Or you may leach contaminants from the filter. That's why you have to revalidate.

[Dan]

The approrpiate filter depends on the method. Thus, the SOP should not specify the type of filter.

PTFE is a suitable filter membrane for organic solvents and solvents that may be up to 10-20% aqueous. Trying to push an aqueous (or mostly aqueous) solution through a 0.45 um PTFE filter is hard enough; using a 0.2 um PTFE filter makes the task even harder.

Some suggestions:
- For a solution that is mostly aqueous: use a nylon, a glass fiber or mixed glass fiber/nylon filter.
- For a solution that has 40-50% organic, use a PVDF filter.
- For a solution that is mostly organic, use a PTFE filter.
- For HPLC analyses, generally the filter pore size needs to be 0.45 um (0.2 um is generally used for sterilization requirements).

These are just general suggestions and there is a lot of leeway in them. And there are many more options. Best bet is to check with your filter manufacturer for guidance/ recommendations.

Regards,
Dan

[Mark Tracy]

Maybe, maybe not. The diameter makes a big difference in filtering speed, a 25mm will be 6x faster than a 10mm. The porosity makes a big difference, but you may pass particulate matter if you use a larger pore. The membrane material could be many things, and sometimes it is important, sometimes not. The glass fiber is an integrated pre-filter on top of the membrane, and usually a good thing for heavily loaded samples. I don't know what GHP is, but you can run some compatibility tests: First, run some sample extract through it and check for haze or light scattering compared against the other filter; a laser pointer is good for this. Then put some extraction solvent through it and compare it to a blank; there should not be any extraneous peaks. You then can run your low calibration standard through and see if the peak is the same size before and after filtration.

[Dan]

Jumpshooter,

I was assumiing that your methods and SOPs are different documents.

If by "SOP" you meant a method, then you must use that filter as described.

However, as Tom noted, you can re-validate for a different filter. You would not need a full validation, a method equivalency validation experiment should suffice (that is if your QA group agrees).

Good Luck!

Regards,
Dan

[Dan]

Mark,

Thanks for the catch.

I should have mentioned that my comments/suggestions above were based on keeping the 25mm diameter filter size that Jumpshooter has in his SOP.

Regards,
Dan

[Mark Tracy]

Jumpshooter,
Are you working in a regulated environment, or some other situation? If it is regulated, the SOP is carved in stone and you need permission and a bunch of data before changing anything, and much paperwork afterward. In other settings, you may have more freedom.

[DR]

I would count the use of the wrong filter diameter and poer size as a "minor deviation" from the method as the first 2 filters mentioned use essentially the same membrane (PTFE) - switching to a completely different polymer membrane (GHP) and adding a glass fiber prefilter (GF) is a much bigger deal, requiring a filter study (checking for recoveries, interferences).

Order more of the right filter. They're a LOT cheaper in 1000 count bags (slightly different part number, ThermoFisher & VWR can & will get them even if the PN doesn't pop up in their online catalogs).

[Consumer Products Guy]

Jumpshooter - I'll admit I only glanced over the other responses. In your case, you're using a smaller diameter filter (significantly smaller surface area) and smaller pore size, both which make filtering tougher. Then you use a different membrane type, that's bad (requiring validation in my mind, or documenting that results are equivalent using both), way different than just changing the diameter or pore size (which wouldn't require re-validation at all, in my mind). You could also use small-volume autosampler vials (cone-shaped bottoms or use inserts) so you wouldn't have to filter as much.

[Jumpshooter]

OK, thanks a bunch guys (and girls). I will order some of the PTFE 25mm, 0.45um acrodisks. Looks like a pack of 200 costs $350 from MidlandScientific.

[DR]

I bet they'll get you 1000 of them for <<$1400 if you go for the bag.

[Jumpshooter]

OK, after reading thru all of your comments I have been rendered sufficiently PARANOID to even try to run the samples that were extracted using the "glass fiber filters" (n = 6 samples). I have only run the samples that were extracted with the "PTFE" syringe filters ( n= 3). I guess I will have to wait on the order fulfillment to conduct re-extractions.