Can biochar remove [insert chemical of concern] from my water?
a new tool for estimating the potential of biochar water treatment
We often hear: “I’m really concerned about [chemical pollutant X] getting into our water. Will biochar work to remove it?”
The answer, generally, is “Yes, biochar can probably be made to work…” There are several questions that have to be answered, though, such as:
What kind of biochar (how was it made and from what material)?
What background water characteristics influence removal of the target pollutant(s) such as pH and the type and concentration of dissolved organic matter (DOM)?
How much biochar is needed (i.e., how large of a biochar contactor), and how often does the biochar need to be refurbished? How do these factors figure into the feasibility and economics of biochar water treatment?
In a relative sense, how difficult is biochar adsorption of the specific pollutant(s) of interest relative to select chemical pollutants that we have extensively studied and used as the basis for treatment systems design and operation?
The new open-access tool described in this article addresses the last question. In other words:
“How can we know if biochar adsorption is expected to be a good option for removing your chemical(s) of concern from water?”
Creation of this tool was inspired by our correspondent and friend, Mr. Wilfrid from the island of Guadeloupe. From Wilfrid we learned that Guadeloupe has a major contamination issue from the use of chlordecone insecticide, primarily on banana plantations.
Chlordecone is a serious threat to human and environmental health and has been banned in most countries. All those carbon-chlorine bonds make it an environmentally persistent organic pollutant, so there are major ongoing cleanup challenges associated with chlordecone’s legacy of pollution.
Wilfrid wrote to inquire whether biochar adsorbent would work for removing chlordecone from water sources in Guadeloupe. We hadn’t worked specifically with chlordecone in the past, so I needed a way to answer Wilfrid’s question.
That led to the development of the tool described here - to leverage the large datasets we have collected over the years on biochar adsorption of many different compounds and develop a basis for comparison with chlordecone. In other words:
Compared to chemical pollutants we have studied extensively, would we expect chlordecone to be easier or more difficult to remove by biochar adsorption?
And if we develop a method to answer this question for chlordecone, it could in principle be applied to any organic chemical pollutant of interest.
We will return to chlordecone and Mr. Wilfrid’s question at the end of this article. First I will explain how the tool works and provide a link for you to download Version 1.0 and take it for a test drive inputting your own chemical pollutant(s) of concern.
Adsorbate Comparison Tool
[Definition: the adsorbate is the substance that gets adsorbed to the surface of the adsorbent.]
The objective of this tool is to generate a relative ranking based on the degree of difficulty of removing a variety of organic chemical adsorbates from water by biochar.
With this relative ranking established, a particular chemical pollutant of interest can be placed in that ranking to give a first approximation of its difficulty of removal by biochar adsorption, compared with other chemicals of concern in water sources.
For example, if your pollutant of interest is an herbicide, how difficult would we expect it to be to remove from water using biochar compared with other common herbicides? And especially, compared with herbicides for which we have collected extensive biochar adsorption datasets?
The Adsorbate Comparison Tool was developed primarily from a study we published a few years ago:
Kearns JP, Kennedy AM, Shimabuku KK. Models for predicting organic micropollutant breakthrough in carbon adsorbers based on water quality, adsorbate properties, and rapid small-scale column tests. AWWA Water Science, Vol. 4, Is. 2, April 2022.
I wrote a “Cliff’s Notes” version of the article for this Substack if you’d like to read more:
To create the Adsorbate Comparison Tool, I used one of the biochar adsorption models we developed in that paper to make predictions about the difficulty of removing a wide range of organic chemical water pollutants.
The model utilizes physical-chemical characteristics of target pollutants which can be readily obtained for free from online databases. Specifically, the model uses Abraham Solvation Parameters. Without going into detail, Abraham Solvation Parameters describe a variety of solute-solvent or solute-surface interactions. (If you would like more specifics, see the “Cliff’s Notes” article linked above.)
To use the Adsorbate Comparison Tool, you simply cut-and-paste Abraham Solvation Parameters for your compound of interest from an online source into the MS Excel based calculation tool. You can then view how your compound ranks in its degree of difficulty for removal from water by biochar adsorption.
Caveat/Disclaimer
Predictions about the adsorbability of different organic compounds is based on an adsorption model. Remember: all models are wrong; some models are useful. Predictions made using the Adsorbate Comparison Tool are not meant to be the definitive final word on the treatability of pollutant(s) using biochar. Rather, the Tool provides decision support to aid in the design and operation of biochar water treatment units. Ultimately the operator is responsible for water quality/safety. Predictive models such as the one embedded in the Adsorbate Comparison Tool should not be over-relied upon. Monitoring and evaluation of biochar water treatment units is always recommended to verify water quality.
How to use the Adsorbate Comparison Tool
You will need MS Excel to run this tool. Click here to download:
Note: This is Version 1.0 of the Tool. As we continue to make refinements, and incorporate user feedback, subsequent Versions will be uploaded with a notation log regarding changes.
To illustrate how to use the tool, I have selected the insecticide imidacloprid as an example compound of interest.
Imidacloprid happens to be my favorite pesticide. Not because I am in love with chemical pesticides (quite the opposite). But because while doing pesticide surveys in West Bengal, India, in 2007 I came across this in a rural ag supply shop:
![[]](https://substackcdn.com/image/fetch/$s_!XllN!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F857367d8-5fe6-4d0e-a9b1-01ec38a78594_300x400.jpeg "[]")
Step 1
Search for your compound in the PubChem database.
Step 2
Highlight and copy the SMILES code for your compound.
Step 3
Paste your SMILES code into the AbraLlama pediction app and click “Predict.”
Step 4
Highlight and copy the calculated Abraham Solvation Parameters.
Step 5
Type or paste the compound name, SMILES code, and Abraham Parameters into this table in the Adsorbate Comparison Tool (located on the right side of the worksheet “INSTRUCTIONS” tab).
Step 6
Click the “RESULTS” tab to view your results.
Interpretation of results
Model prediction results appear as a series of “pinwheel” plots on the “RESULTS” worksheet. For our example compound imidacloprid the following plots are shown:
The green pinwheel plot indicates where your test compound’s predicted adsorbability falls relative to a list of 40 common agrichemicals.
These 40 agrichemicals are compounds for which the US EPA has established regulatory values for drinking water, and/or the World Health Organization (WHO) has established Guideline Values for drinking water.
As you go around the pinwheel from the top in a clockwise direction, compounds are predicted to be more strongly adsorbing to biochar (i.e., in order of decreasing difficulty for removal by adsorption). Conversely, as you go around the pinwheel from the top in a counterclockwise direction, compounds are predicted to be less strongly adsorbing to biochar and therefore more difficult to remove by adsorption.
Compounds indicated by bold text/color are select “Sentinel Chemicals” - compounds of particular focus in our biochar adsorption research, and upon which we base much of our design and operation specifications for biochar water treatment systems. Here, the herbicides 2,4-D and simazine are highlighted as Sentinel Chemicals in the agrichemical compound class.
The model predicts that imidacloprid is slightly easier to adsorb on biochar than simazine, and significantly easier to adsorb than 2,4-D. We can infer, then - with due caution as stressed in the disclaimer above regarding over-reliance on model predictions - that a biochar water treatment system design to provide high levels of removal for 2,4-D and simazine would be expected to also provide high levels of removal of imidacloprid.
Note also that a compound’s difficulty of removal by biochar adsorption is relatively proportional to the area shown on the pinwheel plot. The predicted most difficult-to-adsorb agrichemical on the list is glyphosate (RoundUp). It is predicted to be so weakly adsorbing that it takes up a large proportion of the pinwheel plot, compressing the rest of the plot and making it more difficult to interpret. Therefore glyphosate has been omitted from the plot for clarity. Just bear in mind that glyphosate is predicted by this model to be very difficult to remove by biochar adsorption! Also, I anticipate that many users will be interested to make their own predictions for RoundUp as it is one of the most well known (and notorious) agrichemicals. (Perhaps I should dedicate a future post for another day just to discussing “What to do about RoundUp?”…)
The purple pinwheel plot indicates where your test compound’s predicted adsorbability falls relative to a list of 20 widely studied perfluoroalkyl substances (PFAS).
Sentinel PFAS perfluorobutane sulfonate (PFBS) and GenX are shown in bold type/color.
The model predicts imidacloprid to be slightly more strongly adsorbed on biochar than GenX, and substantially more strongly adsorbed on biochar than PFBS. Thus (with appropriate precautions) a biochar water treatment system designed for high levels of removal of PFBS and GenX would also be predicted to achieve high levels of removal of imidacloprid.
The blue pinwheel plot indicates where your test compound’s predicted adsorbability falls relative to a list of 42 industrial compounds, flame retardants, and plasticizers of concern to drinking water quality. The industrial chemicals on this list are compounds for which EPA and WHO have established regulatory or guideline values.
Sentinel flame retardant/plasticizer tris 2-chloroethyl phosphate (TCEP) is barely visible as a dark blue line near the top of the pinwheel plot. The example adsorbate imidicloprid is indicated by the yellow wedge near the top of the plot.
For this list of industrial compounds, the model predicts that 28 compounds are more difficult to remove by biochar adsorption than imidacloprid, and that imidacloprid is predicted to be more difficult to remove than Sentinel compound TCEP.
Thus a treatment system design to achieve acceptable levels of TCEP removal might not be adequate for controlling imidacloprid, along with many other industrial organic chemical pollutants.
It bears noting that many of these weakly adsorbing industrial compounds are very small molecules - this in part explains why they are difficult to remove by adsorption, as adsorption acts by nonspecific attractive forces between matter (van der Waals forces). A bit larger molecule, such as imidacloprid, simply has more mass to be attracted to the biochar surface.
If small and weakly adsorbing molecules like the ones shown in this list are at issue, then it could be necessary to design a biochar water treatment system specifically with these compounds in mind. It is most likely possible to design a biochar treatment unit that would work for these compounds, however it would be larger in size and require more frequent refurbishment of the adsorbent than in our “base model” treatment system designs. This could incur elevated costs and/or logistical challenges that might preclude feasibility.
The gray pinwheel plot indicates where your test compound’s predicted adsorbability falls relative to a list of 32 compounds that have been designated by the Stockholm Convention on Persistent Organic Pollutants (POPs) - which aims to eliminate or heavily restrict the manufacture and use of such chemicals due to environmental and human health concerns.
If the industrial chemicals pinwheel plot represents “bad news” of many substances being harder to adsorb by biochar than our example compound imidacloprid, then this plot reveals some “good news.” Namely, that compared with the Stockholm list of POPs, imidacloprid is predicted by the model to be more difficult to adsorb on biochar than all of the Stockholm listed POPs.
Thus, a biochar water treatment system designed to achieve high levels of imidacloprid removal is predicted to also achieve high levels of removal of the Stockholm Convention POPs.
The orange pinwheel plot indicates where your test compound’s predicted adsorbability falls relative to a list of 41 common pharmaceuticals and personal care products (PPCPs) of concern to water quality globally. These compounds are often detected in wastewater and water bodies impacted by wastewater (from humans and livestock).
Sentinel pharma compounds sulfamethoxazole and carbamazepine are shown in bold text/color.
The model predicts imidacloprid to be more strongly adsorbing than sulfamethoxazole, but more difficult to adsorb on biochar than carbamazepine. Thus a biochar treatment unit designed to achieve high levels of removal of sulfamethoxazole is predicted to also achieve high levels of removal of imidacloprid (and also carbamazepine).
Identifying compounds of potential concern in drinking water sources
You might already have a list of potential water contaminants you want to run through our predictive Adsorbate Modeling Tool.
On the other hand, if you feel lost about where to start in identifying potential contaminants of concern, one starting place could be in this review paper (and its supplemental information) I published a few years ago:
Kearns JP, 2020. The role of chemical exposures in reducing the effectiveness of water-sanitation-hygiene (WASH) interventions in Bangladesh, Kenya, and Zimbabwe. Wiley Interdisciplinary Reviews – Water, Vol. 7, Is. 5.
In this study I provided lists of compounds belonging to several classes - agrichemicals, PFAS, industrial compounds, flame retardants, plasticizers, pharmaceuticals and personal care products, etc. - demonstrated to impact water sources around the world. I also provided some accompanying information on human and environmental toxicity as a framework for helping implementers in the WASH (water-sanitation-hygiene) sector prioritize possible drinking water contaminant threats.
Chlordecone and Wilfrid’s question
Now we’re ready to answer Wilfrid’s question about whether biochar adsorption could be effective for remediating chlordecone impacted waters in Guadeloupe.
To reiterate: our exercise here uses a predictive model, and “all models are wrong…” The purpose of the Adsorbate Comparison Tool is to provide a first-level insight - a “scientific educated guess” - about whether a particular compound (chlordecone in this case) would be expected to be a more difficult challenge or a less difficult challenge to remove from water using biochar than other chemicals for which we have large datasets and comparatively a lot of experience with.
With that disclaimer in mind, following Steps 1-6 outlined above, we get the following results for chlordecone:
What happened?!!?!?!?
The model predicts that chlordecone is significantly more strongly adsorbing to biochar than most of our comparison compounds shown on the pinwheel plots. On all the plots except for the Stockholm POPs, you really have to squint to make out chlordecone’s tiny wedge at to top of the pinwheels.
Chlordecone’s wedge is more obvious on the Stockholm POPs pinwheel plot - in fact, chlordecone is already listed in the Stockholm Convention (number 21 on the list in the gray box above), so it was already present in that plot (although its wedge was not labeled due to space constraints).
So - with all caveats and disclaimers and precautions in mind - biochar water treatment is predicted to be a viable option for chlordecone removal. Chlordecone is predicted to be less difficult to remove from water than our Sentinel Compounds, upon which most of our treatment system designs are based.
Perhaps Wilfrid will welcome this as good news for the people of Guadeloupe?