To eliminate organic micropollutants (OMPs) from (surface) water, activated carbon adsorption is a cost-effective technology to remove a broad range of OMPs without producing any byproducts. However, co-existing dissolved organic matter (DOM), at much higher concentrations (mg C/L) than OMPs (ng/L-μg/L), inducing adsorption competition, can interfere with OMP removal. Direct site competition and pore blocking are two DOM competition mechanisms, and low molecular weight (LMW) DOM has been recognized as the major competitor in the site competition against OMPs. However, the insights into DOM molecular properties are limited with regard to DOM competition. Therefore, the objective of this research was to relate (LMW) DOM properties to the competitiveness against OMPs, clarify the mechanism of direct site competition, and explore a useful DOM surrogate to predict DOM competitiveness.
Model DOM compounds (mDOMs) could be described individually and more accurately with molecular properties than a complex, real DOM matrix in water. To elucidate the impact of LMW DOM characteristics (hydrophobicity/polarity and aromaticity) on DOM competitiveness, fifteen model compounds (mDOM), differed in functional groups (hydroxyl, phenol, carboxyl groups, etc.), were used to represent several elemental structures of LMW DOM. By temporarily occupying adsorption sites prior to OMPs, LMW mDOM was found to be more competitive to inhibit OMP adsorption kinetics than OMP adsorption equilibrium. Although OMPs were more preferentially adsorbed onto activated carbon than mDOM, the large concentration asymmetry (~500 μg DOC/μg OMP) made a few mDOM compounds exert strong competition against OMPs. The mDOM competitiveness increased when compounds were more hydrophobic and more aromatic, whereas π-π interactions were more important to determine mDOM competitiveness than hydrophobic interaction for LMW mDOM compounds. As an integrated indicator, mDOM adsorbability, defined by mDOM adsorption capacity, was considered better to associate with mDOM competitiveness than hydrophobicity and aromaticity individually. The competition was found to be strong between strongly adsorbable mDOM and weakly adsorbable OMPs, where weakly adsorbable mDOM could even co-adsorb with strongly adsorbable OMPs with little to no competition.
To relate DOM adsorbability to competitiveness in natural waters, a two-stage adsorption procedure was designed to differentiate the adsorption of DOM fractions and OMPs by removing variously adsorbable DOM fractions with activated carbon pretreatment and analyzing the competitiveness of the remaining DOM fractions. Our results demonstrated that adsorbable (LMW) DOM was not necessarily competitive against OMPs. In addition, an increasing amount of DOM competitors was observed against the weaker adsorbable OMPs, compared to their stronger adsorbable counterparts. Similarly, more DOM competitors were identified at high initial OMP concentrations, due to the increased loading of OMPs on activated carbon, highlighting the variable roles (varying competitiveness/ complementary adsorption) of differently adsorbable DOM fractions in competition.
To elucidate the role of molecular weight (MW), polarity and aromaticity in DOM competition from a natural DOM with a complex molecular composition, activated carbon and anion exchange resin (AER) pretreatment served for differentiating competitive DOM from natural DOM. Ultrahigh-resolution Fourier transform mass spectrometry was employed for the DOM analysis at the molecular level. A large percentage of molecular formulas in untreated DOM was PAC-absorbable (97.8% for 40 mg PAC/L), while ~75% of PAC-absorbable formulas were considered poorly competitive, since these molecular formulas were not detected in DOM remaining after AER pretreatment that was highly competitive. The semi-quantitative analysis revealed that aromaticity was the dominant factor for LMW DOM adsorbability and competitiveness. In contrast, with higher MW, the competitiveness of an increasing number of aromatic DOM compounds was diminished due to strong dissociation induced by relatively high polarity.
Finally, the interference of ozone-modified NOM with the adsorption of 2-methylisoborneol (MIB, an odorous OMP) was studied in three natural waters and one standard humics solution in order to study how ozonation influences the competitiveness of DOM with different MW distributions. In the three natural waters, it was found that reducing NOM competition against MIB was found to coincide with increasing ozone consumption. The cleavage of the macromolecules in a standard humics solution, with larger molecular weight and higher aromaticity than the humics in natural waters, only induced a slightly stronger competition under low/moderate ozone consumptions. Overall, the declined aromaticity outweighed the produced LMW DOM in the competitiveness of DOM against MIB in ozonated natural waters. The UV absorbance of the LMW DOM was better correlated with the competitiveness of ozonated/non-ozonated waters than the LMW DOM concentration itself, underlining the role of LMW hydrophobic aromatics in competitive adsorption prediction.
From this thesis, it could thus be concluded that the DOM competition against OMPs is not ubiquitous for all (adsorbable, LMW) DOM fractions. The amount of DOM competitors, as well as their competitiveness, strongly varies with to OMP adsorbability and the initial OMP concentrations (i.e., the concentration asymmetry). For LMW DOM, aromaticity was a key characteristic to promote DOM competitiveness, while the high polarity reduced the DOM competitiveness by DOM dissociation (and thus high hydrophilicity/polarity). To project the competitiveness of ozonated DOM of which the hydrophobicity and aromaticity were simultaneously diminished, LMWUV can then be a handy DOM surrogate instead of LMW DOM concentration itself.