I'm trying to detect PFAS molecules in air samples...

[lukeskywalker52]

Hi there,

I'm working on a project where my aim is to detect the presence of PFAS (Per- and poly-fluoroalkyl substances) in air samples gathered by a system.

I've done a ton of research, but most solutions are unachievable due to price and the available space (the system should be relatively small).

During my research, I looked up IR spectroscopy and it sounded like an excellent idea. At first, I found that PFAS absorption windows are located in the FIR region. This is a problem because all the technology I would need to build the system is uncommon, expensive and very complicated to assemble together. I saw something related to absorption windows in the NIR region, but I'm not sure if it is possible for PFAS molecules to absorb NIR radiation.

Do you guys have any idea if NIR spectroscopy would be possible for detecting PFAS molecules in air samples? Because it would offer me other solutions to build the system.

Also, can you suggest to me everything you know related to PFAS detection in air samples to help me in the project? Also if you have any other interesting solution, please tell me.

Thank you!

 

[berkeman]

Welcome to PF.

Please post links to the reading you have been doing as part of this project. We require links to show us what effort you have been putting into this, as well as to give us an idea of what you already know and what you are likely not to understand yet. Thank you.

 

[Dullard]

Your biggest issue may be sensitivity (how low a concentration do you need to quantify?).

My prediction: After that exercise, you'll find something else to do.

This is not a purely theoretical speculation on my part (but my experience is more in the water regime).

 

[lukeskywalker52]

Welcome to PF.

Please post links to the reading you have been doing as part of this project. We require links to show us what effort you have been putting into this, as well as to give us an idea of what you already know and what you are likely not to understand yet. Thank you.

These are some of the documents I've been reading:

• https://www.chimica-online.it/download/spettri_IR/interpretazione_spettri_IR.htm
• https://www.osti.gov/pages/servlets/purl/1779052
• https://www.atsdr.cdc.gov/pfas/docs/PFAS_FamilyTree_EnvHealthPro-508.pdf
• https://www.pianetachimica.it/NMR/problemi/Spettroscopia_NMR.pdf
• https://moodle2.units.it/pluginfile.php/192474/mod_resource/content/1/Spettroscopia IR .ppt.pdf
• https://www.sciencedirect.com/topics/chemistry/far-infrared-spectroscopy
• more files I've not saved...

Unfortunately, some of them are in Italian, sorry :)

 

[lukeskywalker52]

Your biggest issue may be sensitivity (how low a concentration do you need to quantify?).

My prediction: After that exercise, you'll find something else to do.

This is not a purely theoretical speculation on my part (but my experience is more in the water regime).

I don't know exactly the concentration of PFAs, because I'll take the samples directly on situs and try to analyze those samples in the same system that took them.
Even though, I think the concentration will be pretty low, so a system with relatively high sensitivity would be better.

 

[lukeskywalker52]

Other things I've been looking at are mass sensors like FID (flame Ionization Detector), PID (photoionization detector) and TED (thermal-electron-conductivity detector). I also found some electrochemical sensors like CWE (cantilever-based sensors) and LWE (laser-based sensors).
Does anyone know if those would be able to detect specifically organic molecules with C-F bonds? Do you know any other sensors?

 

[Baluncore]

The analysis of C-F bond NIR absorption will require the identification of compounds that also absorb in the same part of the spectrum, and so might confound the analysis.

A real-time analyser might blow air through a dark sampling chamber that contains a long, folded optical path between mirrors.

If time averaging is acceptable, there are ways to concentrate pollution from the air biologically. As an example, install a beehive, then regularly capture individual bees for later batch analysis.

 

[lukeskywalker52]

The analysis of C-F bond NIR absorption will require the identification of compounds that also absorb in the same part of the spectrum, and so might confound the analysis.

A real-time analyser might blow air through a dark sampling chamber that contains a long, folded optical path between mirrors.

If time averaging is acceptable, there are ways to concentrate pollution from the air biologically. As an example, install a beehive, then regularly capture individual bees for later batch analysis.

So, you're saying that NIR spectroscopy of C-F bonds is actually possible? Do you have any documentation I could read? Thanks!

 

[Baluncore]

So, you're saying that NIR spectroscopy of C-F bonds is actually possible? Do you have any documentation I could read? Thanks!

No. But stronger bonds with higher energy suggest shorter wavelengths.
I don't see a problem with FIR, IR, or NIR instrumentation.
You must find the spectrum for the C-F bond, or the fingerprint of the specific molecule you are looking for.

 

[lukeskywalker52]

No. But stronger bonds with higher energy suggest shorter wavelengths.
I don't see a problem with FIR, IR, or NIR instrumentation.
You must find the spectrum for the C=F double bond, or the fingerprint of the specific molecule you are looking for.

Ok, thanks! I'll do some research on what you are suggesting and come back later.

 

[TeethWhitener]

You must find the spectrum for the C=F double bond

And if you do, go to Sweden to collect your Nobel Prize for overturning modern chemistry.

 

[TeethWhitener]

On a more relevant note, CF bond stretches are around 1200cm-1, well away from near IR.

But more alarming is that you want to try to detect something without establishing a standard for what you’re detecting. You’ll need a known quantity of a known fluorocarbon to calibrate your measurement technique, regardless of what that technique is.

 

[Baluncore]

And if you do, go to Sweden to collect your Nobel Prize for overturning modern chemistry.

Sorry, I was asleep at the time.
You must find the spectrum for the C-F bond, or the fingerprint of the specific molecule you are looking for.

On a more relevant note, CF bond stretches are around 1200cm-1, well away from near IR.

That puts the "stretch" absorption in the mid-IR. But what other common modes are present and where do they appear in the spectrum.

But more alarming is that you want to try to detect something without establishing a standard for what you’re detecting.

Don't be alarmed. Find a new way to detect something, then establish a standard.

PFAS has many C-F bonds, so one aim could be to assess the population of C-F bonds in an air sample, without reference to any particular PFAS molecule.

Assume it is possible, then search for a solution, or a proof that it is impossible because you have over constrained the solution field. The argument that a new technique cannot be found and developed, is both unproductive and over-conservative.

@TeethWhitener If you have an alternative technique that works today, then how about sharing that information.

You’ll need a known quantity of a known fluorocarbon to calibrate your measurement technique, regardless of what that technique is.

Calibration is essential, but only after a technique has been identified and is being tested.

 

[TeethWhitener]

That puts the "stretch" absorption in the mid-IR. But what other common modes are present and where do they appear in the spectrum.

The only modes of anything that appear in near IR are overtones and combination bands. It’s not a great band for anything where you want any degree of sensitivity.

If you have an alternative technique that works today, then how about sharing that information

Googling “pfas air sampling” would be a good start for the OP. Most methods use GC-MS schemes for detection (example), which makes sense, as they’re sensitive, specific, and can be made relatively cheap and compact. Unfortunately, the links OP supplied seem to have nothing to do with air sampling or trace vapor analysis and really are only broad outlines of a number of spectroscopic techniques.

And flame ionization/atomic absorption likely won’t be able to tell PFAS with any specificity over (say) inorganic F, which is probably an issue if you’re anywhere near the ocean.

 

[lukeskywalker52]

Googling “pfas air sampling” would be a good start for the OP. Most methods use GC-MS schemes for detection (example), which makes sense, as they’re sensitive, specific, and can be made relatively cheap and compact. Unfortunately, the links OP supplied seem to have nothing to do with air sampling or trace vapor analysis and really are only broad outlines of a number of spectroscopic techniques.

The example link you sent me seems quite good. I've been struggling to find such resources online, indeed the only ones I found that talk about sampling are in Italian and I thought they might be unhelpful for the thread. I'll keep up with the research on sampling technics starting from the document you sent me. If you have any other resource like this one, please send it to me, I would appreciate it so much! :)

And flame ionization/atomic absorption likely won’t be able to tell PFAS with any specificity over (say) inorganic F, which is probably an issue if you’re anywhere near the ocean.

I've done some research on these and you are right. I had an idea of using some filters to make only PFAs molecules enter the sample and then use this kind of sensor to establish the presence of organic molecules which would be all PFAS (because they would be the only ones passing through the filter).
But this seems quite unachievable

This opens up another question: does anybody know some filters for PFAS? Like some kind of material which can trap PFAS molecules or just let them pass through but not the other organic molecules.

 

[TeethWhitener]

This opens up another question: does anybody know some filters for PFAS? Like some kind of material which can trap PFAS molecules or just let them pass through but not the other organic molecules.

This is part of what the GC in GC-MS does and it’s one reason why the technique is so powerful. One place to start would therefore be the stationary phase of a GC specialized on fluorocarbons, for example:
https://www.sigmaaldrich.com/US/en/...orocol-on-60-80-carbopack-b-/supelco/713-0705

 

[Baluncore]

This opens up another question: does anybody know some filters for PFAS? Like some kind of material which can trap PFAS molecules or just let them pass through but not the other organic molecules.

The GC-MS stationary column is not a trap filter, but is more like a running track. All runners must finish, or the column will become blocked. Following the starting pistol shot, the runners must separate as they run down the track, until they can be selectively sampled from the flow at the finish line.

For GC-MS, an air sample is inserted into a flow column, the time of exit of the constituents being determined by molecular velocity, a square root function of molecular weight. That makes it possible to classify the species present in the sample by molecular weight. The simpler, naturally occurring organic fluorine compounds can be rejected (by time) by that process.

But for a GC-MS to work, there still remains the need to detect the molecules of interest as they exit the column. I see no alternative than to use the IR absorption spectrum of the C-F bond.

 

[TeethWhitener]

But for a GC-MS to work, there still remains the need to detect the molecules of interest as they exit the column. I see no alternative than to use the IR absorption spectrum of the C-F bond.

Yeah that’s what the MS part of GC-MS is for.

 

[lukeskywalker52]

The GC-MS stationary column is not a trap filter, but is more like a running track. All runners must finish, or the column will become blocked. Following the starting pistol shot, the runners must separate as they run down the track, until they can be selectively sampled from the flow at the finish line.

For GC-MS, an air sample is inserted into a flow column, the time of exit of the constituents being determined by molecular velocity, a square root function of molecular weight. That makes it possible to classify the species present in the sample by molecular weight. The simpler, naturally occurring organic fluorine compounds can be rejected (by time) by that process.

But for a GC-MS to work, there still remains the need to detect the molecules of interest as they exit the column. I see no alternative than to use the IR absorption spectrum of the C-F bond.

Thanks for the great explanation, I needed it. And I agree with you on the fact that seems not to be any other alternative than using the IR absorption spectrum of the C-F bond.

Do you have any suggestions for this? (regarding also the things I said in the previous messages)

 

[lukeskywalker52]

Does anybody know anything about paper chromatography or about electrophoresis paper chromatography? Would these techniques be suitable for discovering PFAS molecules or not?
(Always regarding air samples)

 

[lukeskywalker52]

I'm also doing some research right now on chemicals which could react with PFAS because if I find something interesting, I could make the PFAS in samples react with those chemicals and produce specific products which may be easier to discover through other technics... right? (For example, these new products might have a different absorption spectrum which might be easier to analyze)
But these PFAS molecules seem to be pretty stable and they don't react with almost anything... any suggestions?

 

[Baluncore]

But these PFAS molecules seem to be pretty stable and they don't react with almost anything... any suggestions?

If you find the field too complex, then you need to narrow your interest.
For example, if you are concerned with bioaccumulation, then use a similar bioaccumulation mechanism to concentrate the material of interest.

 

[TeethWhitener]

I really don’t understand why you insist on using infrared when mass spec is the gold standard for trace vapor analysis. IR is a terrible choice for this, especially if you’re looking to do something that is sensitive, cheap, and compact. Air detection of PFAS is going to put you easily in the ppb-ppt range if you’re lucky, and you’ll have a bunch of glop you’ll have to try to sort through like finding a needle in a haystack.

 

[TeethWhitener]

I'm also doing some research right now on chemicals which could react with PFAS because if I find something interesting, I could make the PFAS in samples react with those chemicals and produce specific products which may be easier to discover through other technics... right? (For example, these new products might have a different absorption spectrum which might be easier to analyze)
But these PFAS molecules seem to be pretty stable and they don't react with almost anything... any suggestions?

This is also already being done. You break down the PFAS either electrochemically or with a plasma and you detect the inorganic fluoride with a fluoride ion sensor. Problem is, the fluoride ion sensors are only good down to about 1 ppm or so. It’s great if, e.g., you have a known contaminated site and you’re trying to map out the groundwater plumes, but it’s not nearly sensitive enough for what you want to do.

 

[Tiroid]

One alternative is a Quantum Cascade laser that broadbands a region of midIR (say 950 to 1300 cm^-1). This allows discrimination of the different stretches near 1200 and captures secondary or contaminate information concurrently so an algorithm can "correct" for variances. But it is not an easy task, as likely concentration is needed for ppb or ppt (for sure). But GC has drawbacks, especially for handheld etc.