Elsevier

Water Research

Volume 37, Issue 20, December 2003, Pages 4863-4872
Water Research

Pore blockage effect of NOM on atrazine adsorption kinetics of PAC: the roles of PAC pore size distribution and NOM molecular weight

https://doi.org/10.1016/j.watres.2003.08.018Get rights and content

Abstract

Natural organic matter (NOM) in natural water has been found to have negative effects on the adsorption of various trace organic compounds by activated carbon through two major mechanisms: direct competition for sites and pore blockage. In this study, the pore blockage effect of NOM on atrazine adsorption kinetics was investigated. Two types of powdered activated carbon (PAC) and three natural waters were tested to determine the roles of PAC pore size distribution and NOM molecular weight distribution in the pore blockage mechanism. When PAC was preloaded with natural water, the pore blockage effect of the NOM was found to cause a reduction of up to more than two orders of magnitude in the surface diffusion rate of atrazine compared to simultaneous adsorption of atrazine and NOM with fresh PAC. The surface diffusion coefficient of atrazine for preloaded PAC decreased with a decrease in PAC dose or an increase in NOM surface concentration. Because of the pore blockage effect of NOM, a 30% drop in atrazine removal was observed in a continuous flow PAC/microfiltration (MF) system after 7 days of contact compared to the removal predicted from the batch isotherm test. Large micropores and mesopores were found to play an important role in alleviating the effect of pore blockage. A PAC with a relatively large fraction of large micropore and mesopores was shown to suffer much less from the pore blockage effect compared with a PAC that had a much smaller fraction of large pores. Natural waters with different NOM molecular weight distribution caused different extent of pore blockage. The NOM molecules with molecular weight between 200 and 700 Dalton appeared to be responsible for the pore blockage effect.

Introduction

Activated carbon has been widely used in drinking water treatment to remove dissolved organic compounds, including background natural organic matter (NOM) and a number of synthetic trace organic compounds. However, due to a limited understanding of the mechanisms of competitive adsorption, there is yet the need for the development of an accurate model that will predict trace organic compound adsorption from natural water. Previous studies often use the ideal adsorbed solution theory (IAST) [1] to model the adsorption equilibrium of a multiple solute system, which assumes equal access of all adsorbates to all adsorption sites and interaction of adsorbates through a single mechanism: direct competition for adsorption sites. It has been realized only recently that NOM affects trace organic compound adsorption not only by directly competing for adsorption sites but also by blocking carbon pores [2], [3], [4], [5].

A common example of the pore blockage effect of NOM in continuous flow adsorption systems is the “preloading” phenomena found in fixed bed GAC columns. Due to their slower adsorption kinetics, NOM compounds move down the column faster than trace organic compounds. As a result, the GAC at the effluent end of the bed is preloaded with NOM before it is contacted with the trace organic molecules. In fact, all the adsorption systems in which activated carbon is held in the system while water continuously flows through are vulnerable to the pore blockage effect. In these systems, including granular activated carbon (GAC) adsorbers, floc blanket reactors with powdered activated carbon (PAC) addition, and PAC/membrane systems, the activated carbon is partially loaded with NOM and trace organic compounds before more trace organic compounds enter the system, resulting in pore blockage when the surface concentration of NOM gets high enough. The ‘preloading’ effect in GAC adsorbers is usually attributed to occupation of adsorption sites by NOM or to pore blockage by NOM that makes adsorption sites in small pores practically unavailable to trace organic compounds [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. Therefore, most competitive adsorption research on the preloading effect has focused primarily on its impact on the adsorption equilibrium of trace organic compounds [2], [3], [4], [5], [15], [16]. However, a recent mechanistic study by Li et al. [17] showed that, in addition to practically reducing adsorption capacity, pore blockage by large molecules had a strong effect on the adsorption kinetics of atrazine. It was also found by Lebeau et al. [18] that both adsorption capacity and the surface diffusion coefficient of atrazine decreased as the PAC age in an immersed microfiltration system increased. The result is that higher carbon doses are required to achieve given treatment goals compared to those predicted by batch isotherm and kinetic tests using fresh carbon.

Since adsorption equilibrium is rarely reached in continuous flow adsorption systems such as PAC/membrane reactors and GAC columns, it is critical to fully understand the effect of NOM on adsorption kinetics of a trace organic compound so that meaningful, accurate design tools may be developed. However, NOM in different natural waters varies so much that it can have dramatically different competitive adsorption effects [19]. One of the most important characteristics of NOM that affects adsorption is molecular weight. Moreover, the heterogeneity of activated carbon surface is also an important factor in competitive adsorption. The pore size distribution of activated carbon relative to the molecular sizes of adsorbates (e.g. trace organic compounds and NOM) has been found to determine the dominant mechanism of competitive adsorption [15], [17], [20], [21]. Therefore, an accurate evaluation of the roles of NOM molecular weight distribution (MWD) and PAC properties is necessary for optimizing system design and operation to achieve maximum pollutant removal.

The objectives of this study are to: (1) demonstrate the pore blockage effect of NOM on adsorption kinetics of trace organic compounds in batch as well as in continuous flow systems; (2) determine the effect of NOM surface loading on adsorption kinetics of trace organic compounds; (3) evaluate the roles of NOM molecular weight and carbon pore size distributions in pore blockage. This information is needed to improve our understanding of the competitive effect of NOM, which is necessary for better modeling of competitive adsorption of trace organic compounds in natural water. The results from this study will also help water utilities choose the best adsorbent based on NOM characteristics and the physical/chemical properties of the target compound.

Section snippets

Water

Organic-free water was obtained by passing deionized water through a NANOpure ultrapure water system (Barnstead, Dubuque, Iowa). The dissolved organic carbon (DOC) concentration of the organic-free water is lower than 0.3 mg/L. A central Illinois ground water (referred to as GW) and two surface waters were used to study the effect of NOM. The GW was taken from the source immediately before use. It was treated with a greensand filter to remove dissolved iron and manganese and filtered through a

Effect of NOM preloading on atrazine adsorption kinetics: batch experiments

Batch atrazine adsorption kinetic tests were conducted using PAC preloaded with NOM in natural water to determine the effect of NOM preloading on atrazine adsorption kinetics. Doses of 4, 8, 12 and 16 mg/L of PAC A and 2, 4, 8.3 and 12 mg/L of PAC B were preloaded in 2 L of fresh Lake Decatur water (FLDW) for 4 days to obtain a range of NOM surface loading. Following the same procedure, 4, 8, 12 and 20 mg/L of PAC A was preloaded with the one year old Lake Decatur water (DLDW), and 2 and 4 mg/L of

Conclusions

By studying the effect of preloading PAC with natural water, it was found that atrazine adsorption kinetics of PAC could be severely impeded by NOM molecules through the pore blockage mechanism. When PAC was preloaded in natural water, the surface diffusion coefficient of atrazine decreased with increasing surface loading of NOM on PAC, and was reduced by up to more than two orders of magnitude depending on the PAC dose or NOM surface concentration. The comparison of the two PACs showed that

References (27)

  • C.J. Radke et al.

    Thermodynamics of multisolute adsorption from dilute liquid solutions

    J Amer Inst Chem Eng

    (1972)
  • M.C. Carter et al.

    Modeling adsorption of TCE preloaded by background organic matter

    Environ Sci Technol

    (1994)
  • Zimmer G, Crittenden JC, Sontheimer H, Hand DW. Design considerations for fixed-bed adsorbers that remove synthetic...
  • Cited by (0)

    1

    Current address: Department of Chemical Engineering, Yale University, 9 Hillhouse Avenue, New Haven, Connecticut 06510.

    View full text