Environ. Sci. Technol. 2002, 36, 1510-1515
Displacement Effect of NOM on Atrazine Adsorption by PACs with Different Pore Size Distributions Q I L I N L I , ‡ V E R N O N L . S N O E Y I N K , * ,‡ CARLOS CAMPOS,† AND BENITO J. MARIN ˜ AS‡ Department of Civil and Environmental Engineering, University of Illinois at UrbanasChampaign, Urbana, Illinois 61801, and Technology and Research Center, Ondeo Services, 38 rue du president Wilson, 78230 Le Pecq, France
This study investigated displacement of atrazine by the strongly competing fraction of natural organic matter (NOM) in batch and continuous-flow powdered activated carbon (PAC) adsorption systems. Due to the displacement effect, atrazine adsorption capacity in a continuous flow PAC/ microfiltration (MF) system, where the carbon retention time is greater than the hydraulic retention time, decreased with time or NOM throughput. The capacity was lower than that measured in a batch reactor or predicted by the equivalent background compound-ideal adsorbed solution theory (EBC-IAST) method. A mathematical model previously developed to simulate the adsorption process in the PAC/ MF system was modified to take into account the displacement effect. Two types of PACs were tested using a range of influent atrazine concentrations and carbon doses. The extent of atrazine displacement by NOM was found to depend on the type of PAC, while the rate of displacement was a function of PAC type as well as carbon dose. The PAC lost its adsorption capacity for atrazine faster at a lower carbon dose. PAC B, which has a higher percentage of mesopores, lost more atrazine adsorption capacity but at a slower rate than PAC A.
Introduction Activated carbon adsorption is an efficient process for removing synthetic organic compounds as well as natural organic matter (NOM) in drinking water treatment. Because of the competitive effect of NOM, adsorption capacities of activated carbon for most trace organic compounds are significantly lower in natural water than in organic-free water. Isotherms of trace organic compounds in multi-solute systems can be predicted using the Ideal Adsorbed Solution Theory (IAST) (1) provided that the single-solute isotherm parameters of all the components in the mixture are known. Unfortunately, no techniques are currently available to isolate or identify individual NOM compounds in natural water, and the single-solute isotherm parameters for each compound cannot be obtained. Therefore, NOM is often considered as one equivalent background compound (EBC) that has the same competitive effect on the trace organic compounds as the mixture of NOM compounds (2). The IAST combined with the EBC method has been used successfully * Corresponding author phone: (217)333-4700; fax: (217)333-6968; e-mail:
[email protected]. ‡ University of Illinois at UrbanasChampaign. † Technology and Research Center, Ondeo Services. 1510
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 7, 2002
to predict trace organic compound removal from natural water (2-4). However, these studies only investigated adsorption in batch systems, where the amount of NOM relative to the amount of activated carbon is fixed. In continuous flow systems where the adsorbent is held in place while water containing NOM flows through it, the amount of NOM exposed to the adsorbent increases with time, and therefore displacement of trace organic compounds by NOM can happen. It has been found in both laboratory and pilot plant studies that NOM can displace some trace organic compounds from granular activated carbon (GAC). As a result, column effluent concentrations of trace organic compounds can be higher than those in the influent (5). The displacement was attributed to the high concentration of NOM (mg/L) compared to those of trace organic compounds (µg/L). In fact, NOM is composed of compounds with different adsorbabilities, and the competitive effects of NOM components depend on their molecular sizes (6, 7). In the continuous distributions of molecular sizes and adsorbabilities, there should be a fraction of NOM that can access the adsorption sites where a trace organic compound adsorbs and competes with it for adsorption sites. In a continuous flow system with a long carbon retention time relative to the hydraulic retention time, continuous displacement of the adsorbed target compound by the strongly competing NOM fraction can occur as the strongly competing fraction accumulates on the carbon surface, resulting in reduced adsorption capacity for the target compound. Similar to GAC columns, PAC reactors with a carbon retention time longer than the hydraulic retention time may also be subject to the displacement effect. Examples of such systems are floc-blanket, slurry recirculating, and PAC/membrane reactors. This study was undertaken to characterize the displacement of trace organic compounds by NOM. The objectives were to demonstrate the displacement effect of NOM on atrazine adsorption in continuous flow PAC systems with a carbon retention time larger than the hydraulic retention time, to predict atrazine removal in these systems with the displacement effect of NOM taken into account, and to show the role of pore size distribution parameters (micropore/ mesopore volume and surface area) in displacement of atrazine by NOM. The results achieved from this study would improve our understanding of competitive adsorption in natural water and the capability of predicting trace organic compound removal in PAC/membrane systems as well as in GAC columns.
Materials and Methods Water Sources. Organic-free water (OFW) used in the singlesolute tests was deionized-distilled water containing less than 0.3 mg/L dissolved organic carbon (DOC). Natural water used in most of the experiments was a central Illinois groundwater (referred to as GW) treated with a greensand filter to remove excess dissolved iron and manganese and filtered through a 0.45 µm-pore-size microfiltration cartridge to remove particles. The DOC content of the treated GW was 2.7 ( 0.2 mg/L. Water taken from Lake Decatur, Decatur, IL (referred to as LDW) was used in a continuous-flow displacement test. LDW was also filtered through 0.45 µm-poresize membrane filters prior to use. The DOC concentration of the filtered LDW was 3.7 ( 0.2 mg/L. Natural waters were used immediately after preparation to avoid growth of microorganisms. Adsorbate. Atrazine was chosen as the target compound because it is often found in water supplies at concentrations 10.1021/es010870w CCC: $22.00
2002 American Chemical Society Published on Web 02/28/2002
TABLE 1. Relevant Characteristics of PAC A and B property
PAC A
PAC B
BET surface area (m2/g) micropore (
