PFAS in Air, Part 4: OTM-45 Resin Cleaning
2 May 2022In previous blog posts I’ve given a general outline of OTM-45, talked about the chromatography I've been able to achieve, and discussed the calibration requirements laid out by the method and the issues I had with them. In this post we’ll dig into the resin cleaning procedures using the Dionex ASE 350 system.
First, let’s talk a bit about resin cleaning. Restek offers Ultra-Clean Resin, which is equivalent to other manufacturers forms of pre-cleaned XAD-2 resins. If you choose not to buy Ultra-Clean Resin or pre-cleaned resin, OTM-45 outlines in the appendix a procedure needed to remove the salts added to XAD-2 to prevent microbial growth and fully clean it for use. It requires an overnight soak in water, an 8-hour Soxhlet extraction with hot water, a 22-hour Soxhlet extraction with methanol, a 22-hour Soxhlet extraction with methylene chloride, and a final 22-hour Soxhlet extraction with 5% ammonium hydroxide in methanol. Finally, the resin is dried using nitrogen, and the method notes that to dry 500g of XAD-2 it can take a full 160 L Dewar of liquid nitrogen used overnight. That’s almost a full week worth of work, plus solvent and nitrogen use. With all of that, it seems like a no-brainer to start with our Ultra-Clean Resin instead.
That said, our Ultra-Clean Resin and pre-cleaned XAD-2 from other vendors still have some residual PFAS background to them. Fortunately, starting from pre-cleaned resin makes for an easier cleaning process, so you don’t have to do the full cleaning outlined for raw resin as given in OTM-45. My 2021 NEMC presentation talked a bit about this and my use of accelerated solvent extraction (ASE) to both clean and extract the resin as an alternative to the Soxhlet and shakeout extractions. The use of ASE allows for cleaning that is much faster, largely hands free, and uses much less solvent than other methods. The initial ASE parameters used are shown in Table 1. 1.
Dionex ASE 350 parameters
- Pressure - 1500 psi
- Temperature - 120° C
- Heating time - 6 minutes
- Static time - 15 minutes
- Cycles - 2
- Rinse volume - 60%
- Solvent - 4:1 Methanol:Acetonitrile
Table 1 –Initial ASE method for resin cleaning and extraction.
When trying the ASE cleanup/extraction procedure with the Ultra-Clean Resin and pre-cleaned XAD-2, both materials showed a fairly significant background of 3 PFAS compounds, shown in Table 2.
| Compound | ASE Blank (ng/g) | XAD-2 (ng/g) | XAD-2 (ng/g) | XAD-2 (ng/g) | Avg. (ng/g) | Ultra Clean Resin (ng/g) | Ultra Clean Resin (ng/g) | Ultra Clean Resin (ng/g) | Avg. (ng/g) |
| PFBA | 0.05 | 0.54 | 0.51 | 0.60 | 0.55 | 0.43 | 0.35 | 0.43 | 0.40 |
| PF4OPeA | 0.09 | 0.10 | 0.11 | 0.10 | 0.10 | 0.63 | 0.45 | 0.53 | 0.54 |
| ADONA | <0.04 | 0.05 | 0.05 | 0.11 | 0.71 | 0.06 | 0.04 | 0.05 | 0.05 |
Table 2 – Blank results of XAD-2 and Ultra-Clean Resin post ASE cleanup/extraction. “ASE Blank” was an empty ASE cell to determine the instrument background.
After seeing these results, I decided to run the resins through a second ASE extraction to see if the first was sufficient to clean it up. The results of this were a bit unexpected. While the Ultra-Clean Resin cleaned up a great deal for PFBA and PF4OPeA, and were unchanged for ADONA; the XAD-2 showed less improvement for PFBA and ADONA, and an increase in PF4OPeA, as shown in Table 3.
| Compound | XAD-2 (ng/g) | XAD-2 (ng/g) | XAD-2 (ng/g) | Avg. (ng/g) | Ultra Clean Resin (ng/g) | Ultra Clean Resin (ng/g) | Ultra Clean Resin (ng/g) | Avg. (ng/g) |
| PFBA | 0.54 | 0.51 | 0.60 | 0.55 | 0.43 | 0.35 | 0.43 | 0.40 |
| PF4OPeA | 0.10 | 0.11 | 0.10 | 0.10 | 0.63 | 0.45 | 0.53 | 0.54 |
| ADONA | 0.05 | 0.05 | 0.11 | 0.71 | 0.06 | 0.04 | 0.05 | 0.05 |
Table 3 – Results of second ASE extraction of XAD-2 and Ultra-Clean Resin
It seemed like the ASE method was less efficient at extracting from the XAD-2 than it was from the Ultra-Clean Resin, since the XAD-2 had much more PFAS left over for the second extraction. A possible reason for this is the fact that while the Ultra-Clean Resin was dry and free flowing, the XAD-2 was slightly wet and clumpy, as shown in Figure 4. It’s possible the wetness of the resin enabled it to hold on to PFAS more strongly. This also made it a bit more difficult to work with, as seen in the sloppy filling of the ASE cell in Figure 1.
Figure 1 – Comparison of Restek Ultra-Clean Resin and pre-cleaned XAD-2.
After NEMC, I did more work on an improved cleaning only method, hoping to get a simple ASE method that would leave the Ultra-clean Resin PFAS free. I settled on using a mix of methylene chloride and methanol, hoping a more aggressive solvent would be more efficient in cleaning. The ASE settings used are shown in Table 4.
Dionex ASE 350 parameters
- Pressure - 1500 psi
- Temperature - 120° C
- Heating time - 6 minutes
- Static time - 15 minutes
- Cycles - 2
- Rinse volume - 60%
- Solvent - 4:1 Methanol:MeCl2
Table 4 – Final ASE cleaning method for Ultra-Clean Resin
So, was this sufficient to get my resin fully clean? Unfortunately, I ended up with a small amount of PFDA that was slightly above my detection limit (in a later blog I’ll discuss how I came to those limits). Table 5 shows the full results of my cleaning.
| ng PFAS/g resin | ||
|
Compound |
MDL |
Blank Avg. |
|
PFBA |
0.06 |
<MDL |
|
PF4OPeA |
0.02 |
<MDL |
|
3:3 FTCA |
0.05 |
<MDL |
|
PFPeA |
0.03 |
<MDL |
|
PFBS |
0.07 |
<MDL |
|
PF5OHxA |
0.01 |
<MDL |
|
PFEESA |
0.01 |
<MDL |
|
3,6-OPFHpA |
0.02 |
<MDL |
|
4-2 FTS |
0.02 |
<MDL |
|
PFHxA |
0.02 |
<MDL |
|
PFPeS |
0.02 |
<MDL |
|
HFPO-DA |
0.22 |
<MDL |
|
PFHpA |
0.10 |
<MDL |
|
5:3 FTCA |
0.07 |
<MDL |
|
FHUEA |
0.06 |
<MDL |
|
PFHxS |
0.06 |
<MDL |
|
FHEA |
0.09 |
<MDL |
|
ADONA |
0.02 |
<MDL |
|
PFeCHS |
0.03 |
<MDL |
|
6-2 FTS |
0.02 |
<MDL |
|
PFHpS |
0.03 |
<MDL |
|
PFOA |
0.22 |
<MDL |
|
PFOS |
0.04 |
<MDL |
|
FOUEA |
0.07 |
<MDL |
|
PFNA |
0.06 |
<MDL |
|
FHpPA |
0.14 |
<MDL |
|
FOEA |
0.18 |
<MDL |
|
9Cl-PF3ONS |
0.03 |
<MDL |
|
L-PFNS |
0.18 |
<MDL |
|
PFDA |
0.05 |
0.07 |
|
8-2 FTS |
0.07 |
<MDL |
|
L-PFDS |
0.32 |
<MDL |
|
FDEA |
0.07 |
<MDL |
|
PFUnA |
0.08 |
<MDL |
|
11Cl-PF3OUdS |
0.04 |
<MDL |
|
FOSA-I |
0.13 |
<MDL |
|
10:2 FTS |
0.04 |
<MDL |
|
PFDoA |
0.10 |
<MDL |
|
PFDoS |
0.06 |
<MDL |
|
N-MeFOSA-M |
0.07 |
<MDL |
|
N-MeFOSE |
0.07 |
<MDL |
|
N-MeFOSAA |
0.05 |
<MDL |
|
PFHxDA |
0.05 |
<MDL |
|
N-EtFOSA-M |
0.17 |
<MDL |
|
N-EtFOSE-M |
0.05 |
<MDL |
|
N-EtFOSAA |
0.08 |
<MDL |
|
PFODA |
0.05 |
<MDL |
|
PFTrDA |
0.08 |
<MDL |
|
PFTA |
0.05 |
<MDL |
Table 5 – PFAS blank results for Ultra-Clean Resin after cleaning.
Why am I seeing PFDA now when it wasn’t present in my initial cleaning studies? Well, in my experience with PFAS analysis contamination is an unpredictable thing, and talks with other labs seem to confirm this. Solvents and consumables that were PFAS free in previous lots may show PFAS in new lots, so constant monitoring of blank levels is required. Keeping a PFAS free workflow is like a game of laboratory whack-a-mole. One day you’re dealing with PFBA from your bottles, the next day you have to bring the mallet down on PFBS from your solvent. With only one compound out of 49 showing a small background, I was satisfied enough to move on to further testing the extractions. I’ll talk about that in the next post, where I’ll compare the ASE extraction to the shake out called for in OTM-45.