Page 1 of 1

Waters PorapakRDX SPE / Varian BondElut SPE

Posted: Sat Sep 12, 2009 10:29 pm
by bisnettrj2
I know Uwe will comment on this at some point, probably first....

My lab used to use Waters Porapak RDX SPE cartridges for extraction of explosives in water. We switched to Varian's BondElut cartridge when Waters had a lack of availability of the PorapakRDX for about a month (personally, I think there was a QC issue, as I had a couple poor-performing lots of cartridges at the time, but perhaps I'm wrong). In any case, both SPE cartridges impart interference to the sample extract that inhibit trace analysis of water samples for explosives.

A customer service person I talked to at Phenomenex acknowledged this problem, saying the styrene-divinylbenzene stationary phase was the issue, and that there might be, essentially, 'leftovers' from the manufacturing process of the SDB stationary phase eluting into the sample extract, or possibly the stationary phase itself wasn't stable and some was becoming part of the eluted sample.

My question is - can I eliminate these interferences? Is there a way to condition or clean-up these cartridges to eliminate the intereferences prior to collecting the eluant?

Posted: Sat Sep 12, 2009 11:46 pm
by Uwe Neue
Yes... The standard procedure followed by everybody in the industry is the pre-conditioning of the cartridge with the same solvent that you use for elution of the analytes.

Posted: Sat Sep 12, 2009 11:57 pm
by bisnettrj2
So, if I follow the industry standard, and still get co-extracted interferences, what am I to assume? FYI, I do not have our SOP in front of me, but I believe, when we used Porapak cartridges, we conditioned with 15 mL ACN, then 20 mL H2O, then loaded sample (500 mL to 1 L), then eluted with 5 mL ACN.

Posted: Sun Sep 13, 2009 1:17 am
by Uwe Neue
Well, this won't work for everybody's cartridges with the same efficiency. Both Porapak RDX and Oasis HLB have a lower intrinsic contamination level than competitive products.

Posted: Sun Sep 13, 2009 3:14 am
by bisnettrj2
My experience with the Waters SPE cartridges is that the contamination level is the same or greater than the BondElut cartridges, at a greater cost and lesser reproducibility for the Waters SPEs. I am asking how I can reduce the amount of co-extracted interferencs when using either SPE.

Uwe, nothing you've told me will help me use these SPEs with any more efficacy than I currently use them. Normally, I find your advice insightful and applicable. Here, however, I find it vague and essentially pointless.

Posted: Sun Sep 13, 2009 1:27 pm
by Uwe Neue
OK, but let us look at logic...

You are getting "contamination" even after you wash the contamination from the cartridge with the same solvent with which you elute your analytes. You are getting the same level of "contamination" from the cartridges of different suppliers, while I know that we take special precautions to eliminate intrinsic contamination. I am not convinced that the contamination is coming from the material in the cartridge.

I will nevertheless follow up on this next week and ask about the production records.

For a short-term solution, you may look into using the closely related Oasis packing for the same analysis.

Posted: Sun Sep 13, 2009 2:41 pm
by bisnettrj2
I take it you're intimating that the contamination is in-house, not cartridge related. I'll acquire an outside source of HPLC-grade water (probably through VWR, I'm thinking Aristar Plus HPLC Grade) to extract surrogated blanks. Depending on the results, I either have "dirty" lab water (this contamination has followed my lab, then, across two locations and two DI systems, and two Millipore systems) or I have cartridge contamination issues. I'll post results when I have them.

Posted: Mon Sep 14, 2009 9:11 am
by HW Mueller
Also, this sort of thing has been pointed out many times: If the dirt is only from the cartridge, then it has to diminish with an increase of eluting solvent/solution going through the cartridge.

I am not in any way affiliated with any manufacturers, however, since, as an end-user, I am obviously interested in hearing about real problems with material. This is then an appeal at some of you to at least apply the most basic logic/tests before you knock a product in public.

Posted: Mon Sep 14, 2009 12:36 pm
by bisnettrj2
I figured using different water sources (used to use just DI, now use Millipore, have used two different DI and Millipore systems because our lab moved to a new site) was "at least apply[ing] the most basic logic/tests". I also thought consulting a manufacturer of a third SPE (Phenomenex), and having them say it was the stationary phase contributing to the problem, was basically a good use of logic.

Perhaps more information would help - If I inject an instrument blank of 1:1 lab water:acetonitrile, I don't see any issues in my baseline. If I inject an extracted blank (after using one of the SPEs), I have extra peaks in my baseline. If I inject an extracted environmental sample, I have the same extra peaks in my baseline. The same baseline profile presents itself in extracted blanks and environmental samples. So, logic would tell me my water source would not be contributing to this baseline issue, being that I see the same peaks in lab water extracts and ground water from Kentucky, Nebraska, and Alaska....

Posted: Mon Sep 14, 2009 2:43 pm
by HW Mueller
It is also logical that you contaminate everything before it gets to the SPE. Before claiming that the SPE are the culprit it seems the simple test of checking whether eluent diminishes the undesired material is essential.

Posted: Tue Sep 15, 2009 3:46 am
by bisnettrj2
To evaluate my potential (actually, my extraction staff's potential) contamination of my extracted waters, I did the following today, which will inject tomorrow...

1. Pass 3x5 mL aliquots of Burdick and Jackson (Honeywell) HPLC-Grade Acetonitrile through an unconditioned Varian BondElut cartridge. My hypothesis is that, if HWMueller is correct, the degree of contamination will decrease with each pass of solvent.

2. Collect each 5 mL pass in a separate VWR "Trace Clean" 20 mL vial.

3. Inject only the water I use to cut my injections (normally, all my injections are diluted 1:1 with water, with acetonitrile being the extraction solvent).

4. Next, inject just the acetonitrile I used to extract the SPE.

5. After that, 1:1 water and acetonitrile.

6. After that, each pass of 5 mL, cut with the same volume of water.


This does deviate from the extraction procedure we normally follow, but the SPE conditioning for the Varian SPE involves using lab water, and I wanted to avoid adding any possible 'unknowns' to the SPE test. My boss has been using B&J ACN for years, and has not had issues. I can try to get an outside supplier's ACN, but everyone knows how dicey that can be these days (just had our ordering person tell me of a 'great deal' on ACN, to have it turn out to be a company selling 'stale' ACN, according to his contacts).

HWMueller, does this seem like a good way to address the issue? Precontamination has to be minimal, as I sacrificed some ACN to rinse the flask and volumetric pipet before using them, and rinsed the SPE manifold with ACN before adding the SPE cartridge, and everything else seems clean when analyzing soil samples (vials, water used to do 1:1 dilution, instrument, etc.). The ACN shouldn;t be the source, as we use the same ACN to extract soil samples, with no issues.

Anything else I should consider?

Posted: Tue Sep 15, 2009 8:56 am
by HW Mueller
Yes, the most obvious, namely a test of the water by passing different amounts of it through the cartridge, eluting with the ACN and checking whether the amount of dirt correlates with the amount of water.

Actually, at this point it seems important to summarise what I think you are doing: You condition the cartridge with water then ACN. All samples which are placed on the cartridge contain this same water (are diluted with this same water). Elution is done with ACN, this eluate shows the dirt with all the samples.
If this is the case than your argument, "If I inject an extracted environmental sample, I have the same extra peaks in my baseline. The same baseline profile presents itself in extracted blanks and environmental samples. So, logic would tell me my water source would not be contributing to this baseline issue, being that I see the same peaks in lab water extracts and ground water from Kentucky, Nebraska, and Alaska.... " is negated.

A side remark: I think the systematic approach and logic can not be stressed enough. I am just going through some 700 HILIC chromatograms I obtained in connection with aminoacid modification. I just discovered some things I did which don´t make sense, probably have to go back to the lab on this one.

Posted: Thu Sep 17, 2009 4:18 pm
by Uwe Neue
I followed up on manufacturing and QC of Porapak RDX used for the EPA Method 8330. The material is of the same quality as in the past, and has a very low and controlled level of interferences. Nothing has changed in the preparation process nor in the QC, and you should be able to get high-quality blank chromatograms as for example shown in our catalogue or other relevant literature.