Project 4

Uncovering Mechanisms of PFAS Adsorption by Granular Activated Carbon to Support PFAS Remediation

Principal Investigator: Detlef Knappe
Co-Investigator: Morton Barlaz

More than 6 million US residents are estimated to consume drinking water that exceeds EPA’s May 2016 health advisory level of 70 ng/L for the summed concentration of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS). This number would increase greatly if the list of PFAS were expanded and/or acceptable PFAS concentrations were lowered. Sources of water contamination by PFAS include storage and use of aqueous film-forming foams (AFFF), industrial and municipal wastewater treatment plants, landfill leachate, land applied sewage sludge, and air emissions from fluorochemical production facilities. Thus, effective treatment technologies for PFAS removal are needed for a wide range of applications.

Granular activated carbon (GAC) adsorption is the most widely employed remediation technology to reduce PFAS exposure through drinking water. Although GAC is a material that has been studied for decades, we lack models that can predict its effectiveness from fundamental adsorbent and adsorbate properties or from bench-scale experiments, especially when there is a need to remove organic pollutants from complex aqueous matrices containing dissolved organic matter (DOM). Here, we argue that current adsorption models incorrectly treat GAC as a homogeneous material with uniformly accessible adsorption sites along its radial dimension. Instead, we hypothesize that intraparticle distributions of PFAS and DOM are non-uniform along the radial dimension at adsorption equilibrium. We propose that PFAS and DOM adsorb primarily in a shell region near the external GAC surface. This hypothesis is supported by limited direct observations of intraparticle adsorbate distributions and by preliminary data illustrating that PFAS adsorption capacity varies with GAC particle size. Furthermore, when our data are interpreted with current models, intraparticle diffusion coefficients describing PFAS sorption kinetics vary with GAC particle size, which is difficult to rationalize when GAC properties are similar among the different size fractions.

In this project, we will identify factors that control PFAS sorption equilibria and kinetics and interpret data with models that take into account the possibility of non-uniform intraparticle PFAS and/or DOM distributions at sorption equilibrium and diffusion length scales that differ from the commonly assumed GAC particle radius. We anticipate that resulting model parameters will be physically meaningful and can be more readily predicted from first principles than parameters for current models. This project is tightly integrated into the proposed Center and will rely on information generated in Projects 1-3 to prioritize PFAS that are especially health-relevant and for which remediation data are lacking.

Microscope icon

To achieve this, Project 4 has two specific aims:

Specific Aim 1. Observe and describe intraparticle adsorbate distributions at sorption equilibrium.

We will conduct batch adsorption isotherm experiments to systematically evaluate factors controlling the intraparticle distribution of PFAS and polystyrene sulfonate (PSS), a model for pore-blocking DOM, at sorption equilibrium. Intraparticle adsorbate distributions will be observed directly (isotope microscopy, scanning electron microscopy coupled with energy-dispersive X-ray spectrometry, and scanning tunneling microscopy coupled with electron energy loss spectroscopy). In addition, we will deduce PFAS penetration depth into GAC particles from adsorption isotherm experiments conducted with different GAC particle sizes and application of a shell adsorption model.

Data points icon


Specific Aim 2. Describe PFAS sorption kinetics.


We will conduct batch tests to identify factors controlling PFAS sorption kinetics. We will systematically study systems of increasing complexity to describe the effects of GAC characteristics, PFAS properties, and DOM type and concentration on intraparticle diffusion coefficients. Intraparticle diffusion coefficients will be determined with a new model that considers the possibility of non-uniform intraparticle PFAS distributions along the radial dimension at sorption equilibrium. The model is expected to yield intraparticle diffusion coefficients that are independent of GAC particle size and that can be linked to collision cross sections of PFAS that we will determine by ion mobility-mass spectrometry.

Together, the proposed specific aims will fundamentally advance our understanding of GAC adsorption processes. By developing novel mechanistic data, will provide a foundation for designing and selecting effective sorbents for PFAS removal. Results of this project will contribute to lower remediation costs, safer management of spent GAC, and cleaner drinking water, which will ultimately reduce human exposure.


Progress Summary

A critical barrier to develop mechanistic adsorption models is that accessibility of adsorption sites inside of GAC particles is not known. An important assumption of current models is that contaminants are uniformly distributed inside of GAC particles at sorption equilibrium. However, recent data obtained by isotope microscopy for relatively small, neutral organic contaminants suggest that this assumption is likely incorrect. An important goal of this project is to determine the distribution of PFAS and competing constituents, such as humic and fulvic acids that naturally occur in water, inside of GAC particles prepared from different base materials. In year 1, isotope microscopy experiments were conducted to visualize the distribution of 18O-labeled perfluorohexane sulfonic acid (PFHxS) inside of GAC particles prepared from coconut shells and sub-bituminous coal. Granular activated carbon (GAC) samples loaded with 18O-labeled perfluorohexane sulfonic acid (PFHxS) were analyzed along with PFAS-free GAC controls. Isotope micrographs showed that we could readily detect the intraparticle distribution of 18O-labeled PFHxS in the analyzed GAC particles, and that intraparticle PFHxS was non-uniformly distributed. Results of experiments conducted in year 1 are informing the experimental design of planned experiments for year 2.