Reusing SPE Disks

[ravenwork]

I would like to hear from the voice of experience regarding the reuse of SPE disks.

Normally, I extract small batches, usually within the capacity of my 6-station manifold, so I have been using new disks for each extraction. However, I recently had a large batch, and the thought occurred to me that adding an extra elution step, and then another initial organic solvent activation, the membrane should be good to go for another 2-3 extractions, easily.

I went ahead and processed one of the disks with 3 blank runs, and then a spike. The spike turned out beautifully, recovery right in the line with what I saw for a fresh disk.

These disks are not cheap, about $10 each. This seems like a no-brainer to me now that I've done it. After all, we don't throw out HPLC columns after one run, right?

So, anyone out there with additional experience and opinions to relate here?

Evan L. Cooper

[skunked_once]

Were the blanks just reagent blanks or did they include the sample matrix? Same question for the spike sample. The limitation to the reuse of the disks may depend on the complexity of the sample matrix. There is most likely a limit on how many times the disks could be exposed to the sample matrix, cleaned, then reused and still retain their separation efficiency. You could try to determine this by reusing the disk multiple times and documenting the results. If you do this, documentation of a certain number of repeat uses may be critical.

[Mark Tracy]

Depending on the QA and regulatory environment in which you operate, you probably will need to do some form of validation. Also, re-use will be matrix dependent, so well water might be OK, but pond water might not. You will spend about a week doing validations, and another week getting all the approvals. How many extraction disks will this buy? How much labor and material is needed to recover the disk? How many months will it take to recover the validation costs? Expect the laboratory manager to ask you these questions and more. Not such a no-brainer anymore, is it?

[chromatographer1]

Evan,

The big issue here in a regulated environment is proving anything to an auditor.

How do you document how many times a disk has been used and if the cleaning of the disk has been done properly?

Thus how do you know the analysis you did with that used disk is accurate?

How many times will you have to throw away and redo data collection due to these problems.

Personally. I don't want the trouble and expense that such 'cost saving' steps could cost me.

$10 is about 10 to 20 min of labor cost.

It is one thing if you are testing a reactor for a college experiment. It is another if you are making medicine.

Lawsuits are not cheap either.

best wishes,

Rod

[ravenwork]

Thanks to everyone for their thoughtful replies. I am well aware of the regulatory and legal issues that can be involved here. That is why I only performed an experiment while I was extracting the samples with fresh disks as is the usual practice I described.

Personally, I feel that analytical chemistry is being drowned beneath a mountain of unnecessary, costly, and time-consuming regulation. I agree that the documentation and experimental requirements may well exceed any cost savings that would be reaped from my disk savings. I am also sure that the documentation and validation requirements far exceed their usefulness.

I typically perform an extraction blank (reagent water) and a spiked reagent water for each batch. If I chose to save a disk for each batch by first running the extraction blank through and then the spike, I can see no technical reason that would invalidate this procedure.

I suppose that was my curiousity. No one addressed the technical issue I raised. Chromatographically speaking, what is the difference between re-using a column and re-using a disk. Someone did broach the issue of particulates, but as long as the disk can maintain through-put, I don't think this effects it's separation efficiency since that will occur within the disk material not just on the disk surface.

Evan L. Cooper

[Bruce Hamilton]

I suppose that was my curiousity. No one addressed the technical issue I raised. Chromatographically speaking, what is the difference between re-using a column and re-using a disk. Someone did broach the issue of particulates, but as long as the disk can maintain through-put, I don't think this effects it's separation efficiency since that will occur within the disk material not just on the disk surface.

So.... A disk is a small column. It's used in a way that loading capacity and breakthrough volume are additional critical parameters.

In a regulated environment we have to demonstrate, in each batch ( sequence ), that the instrument, including column, are still suitabile for the initended purpose by using a system suitability requirement. Usually, that involves some measurement of separation ability, detection ability, and precision, and comparing those to previously-established limits. Your small column is going to require that test as well...

There is no reason why you could not validate the reuse of SPE disks, and provide data to show that a certain number of reuses does not compromise the analysis. The obvious way is to add an internal standard or a surrogate to some of your yucky samples, and establish recovery limits.

However, given the minor detail that your analysis apparently needs them for some reason, you are going to have to show that the reuse does not compromise the analysis in all cases, and any system suitability test is going to be a lot of nausea, given that each SPE filter is going to have a unique history, just like an HPLC column. Some clean samples, some dirty samples....

My suggestion would be review the use of SPE, can you achieve the same results by some other technique ( centrifuge, solvent ), or by using smaller/cheaper SPE filters?. Maybe you are driving a Rolls Royce, when a Mini would suffice?.

However, you have to review the cost-benefit of all proposed changes, probably excluding the attractive option of killing the regulators and their agents. You may be surprised at the labour v consumable components of proposals.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

Remember that one uses precleaning also to keep things off the column which will ruin it.