Perfluorinated cpds via 537

[Bigbear]

I've been tasked to determine our capability to do 537.
I have a question regarding standards. Who has the most reliable ones at the least expensive price?

Thanks
Bear

[James_Ball]

During UCMR3 we looked into that method and there wasn't any I would call cheap, but the worst thing to find was the labeled components. The one they required was only made by a university professor somewhere and not commercially available. Things might have changed since then.

Just remember to replace all of your teflon lines for mobile phases with PEEK lines.

[Steve Reimer]

We are currently involved in a multi-lab study of a PFAS method (closer to ASTM D7979-16) and the standards were supplied by Wellington.

[Rndirk]

I'm from a European lab so we don't follow EPA methods, but I'm developing the same analysis.

I haven't started yet but ordered the standards from 2 suppliers (it is required for us to check the recovery of one supplier to another one as a QC):

- Single component unlabeled: Cambridge isotopes
- Mixtures unlabeled components: Wellington
- Mixtures labeled components: Wellington

The total sum for everything was about 3500€. The good news is that these compounds are extremely stable and should last us several years.

[Bigbear]

Another question.
Did, or do you extract the calibration standards, or dilute them?

[Rndirk]

Anyone have tips to share on the sample preparation for this method?Chromatography and MS wise, this method runs well. Nice peaks, good sensitivity, good calibration curves. We took precautions in terms of contamination and carry-over, no problems there so far.

However, in our first tests runs we have troubles with recovery (both spiked analytes and internal standards) for higher molecular weight compounds, especially the sulfonic acids. Recoveries are in the low % range for these compounds, while they're fine (60-100%) for the compounds that elute earlier.

Any particular tricks for the SPE protocol that we might be missing? Our calibration standards do not go through the procedure. With these recoveries, we're at the limits of sensitivity so we have to get it better before starting the validation.

[James_Ball]

Anyone have tips to share on the sample preparation for this method?Chromatography and MS wise, this method runs well. Nice peaks, good sensitivity, good calibration curves. We took precautions in terms of contamination and carry-over, no problems there so far.

However, in our first tests runs we have troubles with recovery (both spiked analytes and internal standards) for higher molecular weight compounds, especially the sulfonic acids. Recoveries are in the low % range for these compounds, while they're fine (60-100%) for the compounds that elute earlier.

Any particular tricks for the SPE protocol that we might be missing? Our calibration standards do not go through the procedure. With these recoveries, we're at the limits of sensitivity so we have to get it better before starting the validation.

I haven't extracted the 537 compounds but I am finding that with the other methods like 544, you must do the step where you wash the bottle with solvent and use that solvent to elute the cartridge. Seems a lot of the analytes like to stick to the glass sample bottle and also in the transfer lines. I add the solvent, slowly rock the bottle on its side while I roll it so that the solvent touches every part of the glass, then let it settle to the bottom then use the same transfer line to pull it over into the cartridge to elute. I gained quite a bit of recovery with 544, 525.3 and 530 analytes this way. Makes for a longer extraction process but fixes the low recovery.

[Rndirk]

Thanks James.

Our samples are in polypropylene recipients. I'm not sure if that makes them stick more or less compared to glass.

What you describe was actually tested yesterday. I was hoping that this would fix the problem but unfortunately there's no significant different in recovery...

[James_Ball]

Thanks James.

Our samples are in polypropylene recipients. I'm not sure if that makes them stick more or less compared to glass.

What you describe was actually tested yesterday. I was hoping that this would fix the problem but unfortunately there's no significant different in recovery...

Another possibility I have hear of especially with PAH compounds is to lower the gas flow in the evaporation step and increase the temperature a few degrees. Seems higher gas flows are worse in recovery than higher temperatures.

[Steve Reimer]

With regards to sample containers, they must be rinsed with methanol to remove target analytes. Some of the targets are very prone to sticking on surfaces unless the solvent is 50% methanol or higher. During the development of 537 they went with a high organic final solvent to keep things in solution.
That is why the peak symmetry check.

[Rndirk]

Interesting remarks. We'll be running more tests this week.

We use 50/50 water/MeOH as a final solvent in agreement with the procedure of the EPA counterpart of our country. A higher %MeOH would be allowed, but i'm afraid peak shape of early eluters will suffer even more. Our analysis includes 2 analytes eluting earlier than PFBS (first analyte in 537).

Another question. It is instructed to force the calibration through (0,0), which is odd. The explanation is to better estimate background PFCs. I noticed our calibration is better with 1/x weighting, which the software doesn't allow if forced through zero (divided by 0 i guess?). Is this an issue of our software which tries to weigh the 0 point? I might be missing something here.

Edit: "force" 0 + weighting does work, it was "include" 0 + weighting that wasn't possible.

By default we run 5-6 point LCMS calibration curves without 0, quadratic and 1/x weighting.

[Steve Reimer]

Forcing through zero is an attempt to account for background. If you use linear or quadratic there may be a significant Y intercept. For some compounds the low standard response may not be much larger than the Y intercept. In these cases the potential error at the low standard concentration level can be much larger than the calculated value.
Including zero when there is no blank included in the calibration is not allowed under EPA methods. Forcing zero is allowed and may be recommended or even required for some analytes and methods.

[Rndirk]

I believe we found the fix for our low recovery.

We performed a series of tests with the solvent exchange step separately from the SPE procedure. Different gas flows in the evaporation step didn't gave significant differences in recovery (evaporation @ 60°C). Neither did filtration after the exchange. Glass tests tubes were tested against polypropylene tubes and again no big losses there.

What did the trick was using 1% ammonia/MeOH to elute instead of 0.1% ammonia/MeOH in the SPE procedure. We use Oasis WAX cartridges and for these compounds, ammonia is required to release them (after fixing them with acetate buffer pH 4 to wash). 0.1% ammonia was in the original protocol, but for us increasing this worked. Perhaps our stock solution ammonia has a lower concentration than what's on the bottle...