Use of activated carbon in (drinking) water treatment for the elimination of PFAS

Essen, 07.03.2024 – Per- and polyfluoroalkyl substances, or PFAS for short, are increasingly becoming the focus of attention in water treatment. These chemicals, which are also referred to as ‘forever chemicals’, are synthetic fluoroorganic compounds in which several fluorine atoms are bonded to an alkyl chain. According to the OECD definition, at least one carbon atom is fully fluorinated.

They are characterised by a high chemical and thermal stability due to the strong fluorine-carbon bond. The fluorocarbon chain has a hydrophobic character, but at the same time it can also have a hydrophilic character due to a polar head group. They can thus be water and oil repellent and in some cases also repel dirt particles. Due to these properties, they have been and are still used in many areas. We encounter them in everyday objects such as functional textiles, impregnation agents, Teflon pans and special paper.

In addition to their positive properties, their negative properties have increasingly come to the fore in recent years. They accumulate in the environment, in animal and human tissue. Some PFAS are toxic, suspected of being carcinogenic and contributing to neurodevelopmental disorders. This fact, especially in connection with the fact that many PFAS do not degrade in the environment or only do so over very long periods of time, has led to increased monitoring and the use of certain compounds is increasingly being regulated or banned. The explosive nature of the issue is also reflected in the fact that the European Drinking Water Directive (Directive (EU) 2020/2184) requires the application of maximum levels in the sum of PFAS contents.

Aim of the study

Drinking water is our most important nutrient and for this reason it is one of the goods in Germany that is most closely monitored and strictly controlled. The use of activated carbon is common practice in the treatment of drinking water. It effectively removes various pollutants such as pesticides, pharmaceutical residues and hydrocarbons from the water by adsorption. The Drinking Water Ordinance of 20 June 2023 also set binding limit values for PFAS for the first time. From 12 January 2026, a limit of 0.1 µg/l will apply to substances in the PFAS-20 group and a limit of 0.02 µg/l will apply to substances in the PFAS-4 group. These new requirements pose a challenge for drinking water treatment. Here, too, the use of activated carbon offers a suitable solution. For this reason, CarboTech – one of the world’s leading full-service suppliers of activated carbons, based in Essen – has had the adsorption of PFAS on various activated carbons for drinking water treatment tested at the Technology Centre for Water (TZW) in Karlsruhe.

Description of the test

The experiments were carried out according to the Granular Carbon Selection Test (GCS-Test) developed at TZW. This test allows activated carbons to be compared and evaluated in terms of their adsorption properties within a few weeks, depending on the concentration and adsorbability of the pollutants. In the GCS test stand, four small filter columns, each containing 1.7 litres of activated carbon, were operated in parallel. Karlsruhe tap water was used as the test matrix, to which a mixture of different PFAS was added. The substances examined are perfluorobutanoic acid (PFBA), perfluoropentanoic acid (PFPeA), perfluorohexanoic acid (PFHxA), perfluorooctanoic acid (PFOA), perfluorobutanesulfonic acid (PFBS), perfluorohexanesulfonic acid (PFHxS) and perfluorooctane sulfonic acid (PFOS). All substances belong to the PFAS-20 group, PFOA, PFHxS, PFOS also belong to the particularly critical PFAS-4 group. During the experiments, the spectral absorption coefficient (SAC) at 254 nm and the concentrations of the PFAS were recorded at regular intervals in the inlet and outlet of the small filter columns.

Activated carbons tested

In the experiments, various steam-activated carbons with an 8×30 MESH grain size were tested. The selected carbons are based on different raw materials. The DGF 8×30 GL and DGF BX 8×30/65 are based on hard coal, like the majority of the carbons currently used in drinking water treatment. DGK 8×30/65 (coconut) and DGP 8×30/65 (palm kernel), which are based on renewable raw materials and agricultural by-products, are a more sustainable alternative. All selected activated carbons are approved for drinking water in accordance with DIN EN 12915-1. The properties of the activated carbons examined are shown in Table 1.
Overall, DGF 8×30 GL has the highest porosity of all the activated carbons tested, which is also reflected in the lowest tap density. At the same time, it has the highest proportion of mesopores. Based on its raw material and the significantly lower degree of activation, DGK 8×30/65 is significantly more fine-pored and is almost exclusively microporous.

Results of the tests

The breakthrough curves determined for the SAC 254 nm and for the sum concentration of PFAS-20 and PFAS-4. The SAC 254 nm and the concentrations in the filter effluent are plotted against the specific amount of water passed through in bed volumes.

All the activated carbons tested were able to further reduce the already low SAC 254 nm. This means that the activated carbons can remove substances present in the water by adsorption. In addition to the added PFAS, these can be, for example, hydrocarbons. The DGF 8×30 GL showed the best performance, with the lowest concentrations in the effluent for all bed volumes. This is probably due, on the one hand, to the fact that it has the largest total pore volume and the largest volume available for adsorption. On the other hand, this may also be due to the fact that it has both micropores and larger mesopores and thus good adsorption sites for different sized molecules. The weakest performance was shown by the DGK 8×30/65. This is due to the smallest pore volume and may be influenced by the fact that almost exclusively small micropores are present and thus larger molecules cannot be adsorbed. The adsorption properties of activated carbons DGF BX 8×30/65 and DGP 8×30/65 are comparable, but are significantly lower than those of activated carbon DGF 8×30 GL.

With regard to the adsorption of the sum of all PFAS, the DGK 8×30/65 shows the lowest adsorption capacity, analogous to the SAK value. The other activated carbons show comparable adsorption performance; for low bed volumes, the DGF 8×30 GL shows a slightly better removal performance for the sum of all PFAS examined. Overall, it can be seen that the initial concentration of PFAS was not reached at around 13,500 BV for any activated carbon. This means that the equilibrium loading has not yet been reached and that all activated carbons have further capacity for the adsorption of PFAS. Overall, this also shows that the mass transfer zones in the selected experimental setup are very broad.

When considering only the PFAS-4 substances (PFOA, PFHxS, PFOS), DGF 8×30 GL shows the best adsorption performance. In contrast, the activated carbons DGP 8×30/65 and DGF BX 8×30/65 were found to have slightly higher effluent concentrations of the individual substances overall and, as a result, higher values for the sum parameter. The DGK 8×30/65 showed a significantly lower adsorption capacity compared to the other activated carbons. The difference to the other activated carbons is more pronounced in relative terms than when considering all PFAS examined. This is due to the fact that the PFAS-4 substances examined are molecules with chain lengths ≥ C6. These can form stronger interactions with the surface than short-chain PFAS. Due to the molecule dimensions, they also adsorb in wider pores, of which the DGK 8×30/65 has none or only a small proportion. As with all PFAS examined, equilibrium has not yet been reached for the PFAS-4 substances.
If the shake density/bulk density is included in the consideration of the adsorption performance – as is usual when evaluating activated carbons – the DGF 8×30 GL shows the best cleaning performance due to its low density.

Conclusion

The starting point for the experiments was to assess and compare the adsorption performance of various granulated activated carbons for PFAS in drinking water treatment. The testing was carried out with Karlsruhe tap water on small filter columns. It was found that all the activated carbons tested can reduce the PFAS content in the water tested, with DGF 8×30 GL showing the best removal efficiency. The greatest decrease in SAC 254 nm was also observed for DGF 8×30 GL. The results show that a sustainable alternative for the removal of PFAS can be a palm kernel-based activated carbon (DGP 8×30/65), which shows similarly good performance for the sum parameter of PFAS. In summary, all activated carbons can be used to remove PFAS. If there is another pollutant issue, this should be included in addition to the adsorption performance for PFAS.