Can We Filter PFAS Forever Chemicals from Water?
Recap on PFAS toxicity and can filters successfully remove both long and short-chain PFAS?
This is a blog where I go in-depth into questions and topics, I am personally interested in related to contamination and toxicity. The goal is to break down and interpret the actual data behind many of these compounds in the news for a non-chemist like me. Each blog is based on secondary research, primarily from published articles found through Google Scholar and known academic journals, and experience formed during my work with chemical analysis.
DALL-E generated image of water with foam. Foam on water actually correlates to higher presence of PFAS due to its water repelling properties.
TLDR
PFAS are a family of highly toxic and very long half-life compounds; historically mostly long-chain PFAS were produced, recently there has been a shift to short-chain PFAS;
Evidence shows similar toxicity for some of the short-chain PFAS, meaning there is a need to filter both types from water;
There is no consensus filtration method for the full elimination of both short and long-chain PFAS from water;
Typically tandem techniques are used, such activated carbon + ion exchange or reverse osmosis; none of the techniques are capable to achieve the recently introduced EPA limits of 4 and 10 ng/L concentrations for select few PFAS;
EPA recommends single use ion exchange resin for up to 100% long-chain PFAS elimination and reverse osmosis for up to 90% short-chain PFAS elimination.
Compounds I want to cover in this and future posts
3 different substances are a particular concern for food and water today, as they are not universally regulated, can be highly prevalent and research indicates health risks:
PFAS (Per- and polyfluoroalkyl substances) - found in water repellent materials, such as non-stick pans and DWR clothing, these indestructible ‘forever’ compounds repel water and accumulate in our blood, increasing the risk for cancers and other health issues;
Phthalates - plasticizers found in softer plastics, such as PVC (pipes, flooring) and PET (water bottles and food containers) are ubiquitous and are so called ‘endocrine disruptors’. They can disrupt metabolism, reproductive health and the development of children (more on health impacts);
Micro and nano-plastics - these are found pretty much in everything around us and most parts of our bodies. The evidence on health impacts in isolation from other toxic materials which are often found together with microplastics is not clear-cut, but there are more and more studies of potential health risks. I covered the recent breakthrough on nano-plastic identification and quantification here.
This time I am focusing on PFAS and ways to filter it from water.
PFAS are dangerous to human health
PFAS are a family of compounds which serves the useful function of repelling water, while remaining highly stable and increasing product lifespan. That’s why it has been used in chemical formulations for PTFE / teflon plastic and the derivative applications such as electronics, medical devices, non-stick frying pans, water repellent clothing, pizza boxes, popcorn packaging, stainless proof couches, dental floss, cosmetics and many others.
PFAS compounds can be segmented into two types:
long-chain / big compounds with 8+ carbons and
short-chain / smaller compounds with <8 carbons.
It is usually assumed that long-chain PFAS compounds pose higher risks for health and include specific formulations such as PFOA (Perfluorooctanoic acid) and PFOS (Perfluorooctane sulfonic acid). As an example, PFOA (8 carbons) has the following associated health risks (from my earlier post Water repellent clothing and PFAS - are they a danger to us?):
pregnancy-induced hypertension (including preeclampsia),
kidney cancer,
testicular cancer,
thyroid disease,
ulcerative colitis, and
high cholesterol (hypercholesterolemia).
PFOA, given the strong evidence for negative health impact, was banned in the US in 2014 and globally in 2020 (Source). Both PFOA and PFOS is covered by the recent EPA regulation of drinking water and needs to be under 4 parts per trillion of concentration in water (PPT) (Source).
Given the terrible evidence on health impacts, producers using PFAS over the last decade have been shifting to shorter chain PFAS in their formulations which are less studied and (at least for now) not explicitly covered by regulations.
Some of the examples of short-chain PFAS are perfluorobutane sulfonic acid (PFBS, 4 carbons), GenX (a branded name, 6 carbons) and perfluorohexane sulfonic acid (C6, PFHxS). While there is less research on shorter chain PFAS, the evidence coming in indicates high health risks not dissimilar to those of long-chain PFAS. For instance PFBS was found to correlate with reproductive toxicity (Source) and PFBS and PFHxA, among others, were found to result in muscle cell toxicity (Source).
For a more zoomed in take on the impacts of PFAS, I recommend this recent poignant documentary from Bloomberg:
Some of the points raised are:
Pupils in the high school near a PTFE manufacturing plant experiencing all sorts of rare cancers.
Accumulation of PFAS is found in human bodies, drinking water and animals, especially predators. This leads to looming risks for the whole animal ecosystem.
Questions raised how the accumulated levels of PFAS in mothers pass on to children through breast milk affecting their development.
For those with above average PFAS exposure in blood, liver and thyroid function tests are mentioned.
These PFAS compounds can take thousands of years to break down naturally, and in humans the longitudinal data shows an incredibly long blood clearing timeframe. This clearing time is longer for long-chain PFAS as opposed to short-chain PFAS.
CDC measured PFOS in blood and it took 10 years plus for the levels to come down after exposure declined (CDC). I am really not sure what other compound takes this long to clear from the body after exposure is contained.
Well in fact I checked. Here are some similarly persistent compounds in human body - and they are all banned or heavily controlled:
DDT: half life in humans of 6-10 years (Source) - banned;
PCBs: half life in humans of up to several decades (Source) - banned;
Lead: half life of decades in human bones (Source) - heavily controlled.
PFAS mandated safe limits
Last time I looked at PFAS levels in water (PFAS in tap water - overview of concentrations), I settled on a 70 nanograms per day safe intake limit (70 PPT = 70 ng). That was based on public health recommendations for PFOS exposure.
However, the recent EPA decision goes beyond that, with long-chain PFOA, PFOS mandated at <4 PPT (4ng/L of water) at any time in water, while the PFNA (long-chain), PFHxS (short-chain), GenX (short-chain) are mandated at 10 PPT limit individually and combined (10ng/L of water) (NBC news).
Remarkably, the article claims that up to 10% of US water systems are not compliant today (NBC news). And indeed browsing through EPA collected data map, I can see many spots measuring over these limits (EPA PFAS map).
Given the median combined water intake (food and drink) is 4 litres a day (Wikipedia), the EPA limits imply maximum 16 ng daily exposure of each PFOA and PFOS and additional 40 ng of PFNA, PFHxS and GenX combined (EPA). PFBS is also part of the combined mixture controls.
Together with water limits, EPA also updated the cancer risk estimates. In fact virtually any presence of PFOA leads to elevated cancer risks (Source). The link between skin and PFAS contaminated water was also establish (which previously was widely thought not to be a risk), confirming elevated cancer risk through simply showering with PFAS contaminated water (Source).
The numbers quoted are startling. Even at 0.0027 ng/L of PFOA in water exposure for a million people, one incremental cancer case would be observed (Source).
We really need filters to fully remove PFOA and other PFAS.
PFAS filtration
One of the challenges of filtration is actually the ability to accurately measure the PFAS compounds. At levels of 4 ng/L both the detection sensitivity is an issue and also the capability to separate each of the different PFAS from each other. Such measurements require fully equipped chemistry labs with each test taking up to an hour, if not more.
There is not a single solution to get rid of PFAS and all the filtration options do not destroy the PFAS compounds, but only siphon them away. Also the filtration techniques tend to get less effective as we reduce the number of carbons in PFAS, making removal of short-chain PFAS a much harder task (Source).
Apparently, there is no filtration solution which can readily achieve the recently mandated EPA levels of under 4ng/L (Source).
The long list of filtration options to get rid of PFAS are (Source):
Carbon based materials (activated charcoal, biochar) which adsorb the target compounds;
Metal organic frameworks - crystalline materials with porous structure which bind to the target compounds;
Ion exchange polymers - polymers with an electric charge, interfacing with the target compound through ionic bonds;
Nanofilters and reverse osmosis - tiny membranes retain the target compounds in a water stream through size exclusion and charge effects;
Foam fractionation - through bubbling foam is created on the surface level to draw out compounds which are hydrophobic, such as PFAS;
Destruction through hydrothermal alkaline treatment (HALT) - PFAS is decomposed under high pressure, high temperature and high pH conditions;
Precipitation – a process used in wastewater treatment, where coagulation and flocculation are applied;
Oxidation - oxidation affects even PFAS compounds despite their strong bonds. However, there is a risk that this method will create a lot of unwanted byproducts from other compounds found in the same water.
None are 100% effective for short-chain PFAS removal.
Carbon in particular has been observed to sometimes release PFAS back into the stream when overloaded.
All of these methods need continuous replacement or refreshment of the filtration material.
Additionally there are companies working on new technologies which involve engineering specific surfaces, which selectively bind only with PFAS compounds. However, it is unclear how far these technologies are from being validated and adopted.
Having read a number of papers on different approaches and watched various videos I am yet to find a fully convincing solution. Many different approaches are being trialled but there is no consensus yet.
Most approaches deploy two techniques in tandem - e.g., activated carbon followed by ion exchange. While this increases efficiency, I was not able to confirm such approaches are definite in removal of short-chain PFAS.
For now, the clearest guidance comes from the EPA (Technologies for Reducing PFAS in Drinking Water):
Granulated activated carbon can be up to 100% effective for long-chain PFAS for a period of time - but needs careful timing for replacement as it can release PFAS back (Source)! Worth noting that the pitcher carbon filters are not expected to filter PFAS due to short retention time (Source).
Ion exchange (anion exchange resin) can be up to 100% effective depending on the PFAS for a period of time. The EPA suggests single use resins, which can later be disposed and eventually incinerated to destroy PFAS.
Reverse osmosis is said to be up to 90% effective, also removes short-chain PFAS but tends to retain 20% of water. This can work well for home use, as the 20% filtration byproduct can be diverted to the drain. The byproduct water is corrosive though so poses a risk to certain types of piping. It is also worth noting that reverse osmosis treated water is not far off distilled water and is almost cleaned of any useful minerals. The lack of typical minerals and nutrients otherwise found in drinking water can pose health risks.
Additionally, research on the whole house systems using activated carbon indicates large variations, where in some cases the PFAS concentrations only increase (Source).
That’s my take now. Let me know if you would recommend any method over others in particular?