Published November 10, 2014
Journal of Environmental Quality
TECHNICAL REPORTS vadose zone processes and chemical transport
Sorption of Tetracycline to Varying-Sized Montmorillonite Fractions Huaizhou Xu, Xiaolei Qu, Hui Li, Cheng Gu, and Dongqiang Zhu*
T
he environmental fate and transport of many organic compounds are inextricably linked to the nature and abundance of clay minerals present in soils (Delle Site, 2001; Kaiser and Guggenberger, 2000; Sassman and Lee, 2005; Sheng et al., 2001). Although often dwarfed by partitioning of neutral organic compounds into soil organic matter, sorption by clay minerals plays a critical role for polar and ionizable compounds that can be sorbed by ion exchange and/or surface complexation (Delle Site, 2001; Figueroa et al., 2004; Kaiser and Guggenberger, 2000; Kulshrestha et al., 2004; Sassman and Lee, 2005; Sheng et al., 2001; Tolls, 2001). Furthermore, the aluminosilicate surfaces of clays have been found to exhibit significant reaction activities that catalyze chemical transformation of sorbed organic compounds (Chen and Huang, 2010; Laszlo, 1987). Therefore, sorption of organic contaminants to clays could be a key determinant for many physicochemical processes occurring in the soil environment. Montmorillonite represents a group of 2:1 phyllosilicate minerals that have high cation exchange capacity (CEC), large specific surface area, and the unique feature of swelling/collapsing interlayer structures. The isomorphic substitution within aluminosilicate sheets results in net permanent negative surface charges, which are commonly compensated by exchangeable cations (Tombacz and Szekeres, 2004; Vaccari, 1998). The siloxane surfaces between exchangeable cations are regarded as nanoscale domains suitable for sorption of hydrophobic organic compounds ( Jaynes and Boyd, 1991; Sheng et al., 2001). The hydroxyl groups from broken edges of aluminosilicates can protonate and/or deprotonate, creating pH-dependent charges associated with the minerals (Tombacz and Szekeres, 2004; Vaccari, 1998). The wide distribution of antibiotics in soil and water has raised concerns regarding human health due to the potential selective pressure exerted on microbial communities to develop antibiotic resistance (Boxall et al., 2003; Graff et al., 2003; Tolls, 2001). Among these antibiotics, tetracyclines are a large class of antibiotics that have been administered to livestock and humans. A large portion of tetracyclines cannot be metabolized
Abstract The influence of particle sizes on sorption of tetracycline by clay minerals is poorly understood. In this study, montmorillonite clay fractions with varying particle sizes were prepared by successive centrifugation, and the effects of particle sizes on sorption of tetracycline were evaluated using an equilibrium dialysis method. Sorption isotherms were nearly overlapped for size fractions ranging from 6.38 to 16.00 mm, except for the finest clay fraction (0.41 mm). The relatively low sorption by the fraction with the smallest particles could be attributed to the colloidal nature and high edge-to-surface ratio, which could lead to reduced accessibility of tetracycline to sorption sites (particularly those at the edges). The impact of solution pH and coexisting Na+ and Ca2+ ions on tetracycline sorption was found to differ between the finest fraction and other clay fractions. The results demonstrated for the first time that clay particle size greatly influenced tetracycline sorption to clay minerals and consequently might affect their transport and bioavailability in the environment.
Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
H. Xu, X. Qu, C. Gu, and D. Zhu, State Key Lab. of Pollution Control and Resource Reuse/School of the Environment, Nanjing Univ., Jiangsu 210093, P.R. China; H. Xu, State Key Lab. Pesticide Risk Assessment and Pollution Control, Nanjing Institution of Environmental Sciences, Ministry of Environmental Protection, Nanjing, 210042, P.R. China; H. Li, Dep. of Plant, Soil and Microbial Sciences, Michigan State Univ., East Lansing, MI 48824. Assigned to Associate Editor Thomas Borch.
J. Environ. Qual. 43:2079–2085 (2014) doi:10.2134/jeq2014.04.0182 Supplemental materials are available online for this article. Received 23 Apr. 2014. *Corresponding author ([email protected]).
Abbreviations: AFM, atomic force microscope; AIOC, aromatic ionizable organic compound; (-)CAHB, negative charge–assisted hydrogen bond; CEC, cation exchange capacity; PZC; point of zero charge; PZNPC, point of zero net proton charge; XRD, X-ray diffraction.
2079
in vivo (Boxall et al., 2003; Tolls, 2001) and inevitably enter soils and aquatic systems (Boxall et al., 2003; Graff et al., 2003). Montmorillonites, an important component in soil minerals, can strongly sorb tetracyclines through cation exchange and complexation with metal cations on clay surfaces (Aristilde et al., 2013; Aristilde et al., 2010; Figueroa et al., 2004; Kulshrestha et al., 2004; Li et al., 2010a; Sassman and Lee, 2005; Sithole and Guy, 1987; Tolls, 2001; Wang et al., 2010; Zhao et al., 2012). Significant research efforts have been dedicated to elucidate how the types and chemical compositions of clays influence tetracycline sorption from aqueous solution (Aristilde et al., 2010; Chang et al., 2009a; Chang et al., 2009b; Chang et al., 2012; Chang et al., 2009c; Figueroa et al., 2004; Li et al., 2010a; Li et al., 2010b; Sithole and Guy, 1987). However, little attention has been paid to the interactions between tetracyclines and colloidal clays with varying particle sizes. Natural mineral colloids could serve as effective carriers for strongly sorbed contaminants, significantly enhancing their transport in soils and aquifers (Boxall et al., 2003; Grolimund et al., 1996). The information of colloid-mediated contaminant transport in porous media has been advanced considerably in past two decades (Kanti Sen and Khilar, 2006; Kretzschmar et al., 1999); however, the effects of particle size on sorption of organic contaminants by clay particles is still unclear, which limits the understanding of colloid-mediated transport in soils. The major objective of the present study was to examine sorption of tetracycline by montmorillonite fractions with varying sizes over a range of pH and ionic strength conditions. The results provide a better understanding of the influence of size-dependent clay geochemistry and microstructure on sorption of aromatic ionizable organic compounds (AIOCs). To the best of our knowledge, this is the first study to illustrate the effects of clay particle size on sorption of organic contaminants.
Materials and Methods Clay Fractionations A montmorillonite (Fenghong Inc.) was fractionated by successive centrifugation into five fractions with varying sizes: Bulk, Rest, 2000 rpm, 4800 rpm, and 11,000 rpm. Briefly, 20 g of montmorillonite was dispersed in 1 L of 0.5 mol L-1 NaCl solution for 24 h. The suspension was centrifuged at 60 g for 5 min. The supernatant was collected and was referred to as the Bulk fraction. The residue clay pellet was resuspended in 1 L of 0.5 mol L-1 NaCl. This procedure was repeated three times, and the collected supernatants were combined. The Bulk fraction was then dialyzed against a dialysis membrane tube (3500 Da) (Union Carbide) to remove the excess salts until a negative chloride test of solution with AgNO3 was obtained. The resultant clay suspension was low in ionic strength. The clay suspension was successively centrifuged at 11,000 rpm (11,363 g), 4800 rpm (2164 g), and 2000 rpm (376 g). The centrifugation step was repeated three times with each suspension for 20 min before the next centrifugation speed was used (Gimbert et al., 2005). After each centrifugation, the supernatant was collected, and the clay pellet was resuspended in deionized water. The collected supernatant samples were referred to as the 11,000 rpm fraction, the 4800 rpm fraction, and the 2000 rpm fraction, respectively. 2080
The clay fraction collected after 2000 rpm centrifugation at 376 g was referred to as the Rest fraction.
Sorbate Tetracycline (99%) was purchased from International Laboratory. The molecular weight, water solubility, and stepwise acid dissociation constant (pKa) are listed in Supplemental Table S1. The molecular structure and the pH-dependent speciation of tetracycline are summarized in Supplemental Fig. S1.
Characterization of Montmorillonite Fractions The CEC of the montmorillonite fractions, including Bulk, Rest, 2000 rpm, 4800 rpm, and 11,000 rpm, was determined using the extraction method by 1 mol L-1 NH4Ac and 0.005 mol L-1 EDTA at pH 7.0 (Sparks et al., 1996). The elemental composition of each component was quantified using a X-ray fluorescence spectrometer (ARL-9800, ARL). The size distribution of each fraction was determined using a laser diffraction particle size analyzer (Mastersizer 3000, Malvern). The point of zero net proton charge (PZNPC) (i.e., the pH value at which the net proton surface charge is equal to zero) of all fractions was determined by potentiometric titration using a microcomputer automatic potentiometric titrator (WDDY2008, Datang). The morphology of the 11,000 rpm fraction and the Rest fraction was examined using a Bruker Multimode 8 atomic force microscope (AFM) (Bruker). The measurements of z-potential were performed using a Zata PALS z-potential analyzer. The basal spacing of clay fractions with or without the sorbed tetracycline was measured by an X-ray diffraction (XRD) spectrometer (X’ TRA; ARL).
Sorption Experiments Batch sorption experiment was performed using an equilibrium dialysis method similar to that reported in a previous study (Gu and Karthikeyan, 2005). The amount of clay present in suspension was determined by drying a known volume of clay suspension in an oven (90°C). All clay fractions were then diluted with deionized water to 2 g L-1, and pH values were adjusted to pH 7.0 using 0.1 mol L-1 HCl. Tetracycline stock solution (0.675 mmol L-1) was prepared in 0.02 mol L-1 NaCl solution with pH adjusted to 7.0 using 0.1 mol L-1 NaOH. A dialysis membrane tube (Union Carbide, 3500 Da) was cut into 9-cm-long sections that were pretreated in boiling solution containing 20 g L-1 NaHCO3 and 10 mmol L-1 EDTA at 100°C for 10 min and washed thoroughly with deionized water. For the dialysis experiment, 20 mL of clay suspension was transferred to a 40-mL amber glass bottle equipped with polytetrafluoroethylene-lined screw cap. Then, a given volume of tetracycline stock solution was spiked into the clay suspension. A sealed dialysis cell filled with 5.0 mL of 0.02 mol L-1 NaCl solution was placed into the bottle. The headspace was filled with 0.02 mol L-1 NaCl solution. The bottles were covered with aluminum foil and tumbled at room temperature for 3 d, which was sufficient to achieve apparent sorption equilibrium as indicated in our preliminary study. To account for the possible solute losses from the processes other than sorption to clay (e.g., sorption to glassware, septum, and/or dialysis cell membrane), a calibration curve was prepared under the same treatment as the Journal of Environmental Quality
sorption experiment but in the absence of clay. The calibration curve consisted of eight concentration levels over the range of tetracycline concentration in aqueous solution. Tetracycline in the dialysis cell was analyzed and assumed to be the equilibrium concentration in solution. Clay-sorbed tetracycline was calculated based on the mass balance between the added tetracycline and that present in solution at equilibrium. Another set of sorption experiment was conducted to evaluate the effects of pH and ionic strength on sorption of tetracycline by clays. In the solution, 0.1 mol L-1 HCl or 0.1 mol L-1 NaOH was used to adjust solution pH to achieve equilibrium pH values ranging from 3 to 10. The initial concentration of tetracycline was 0.19 mmol L-1. The ionic strength of solution was prepared with a range from 0 to 0.1 mol L-1 NaCl or CaCl2. All samples were prepared at least in triplicate.
Tetracycline Analysis Tetracycline in the aqueous solution collected from dialysis cells was analyzed by an Agilent 1200 high-performance liquid chromatograph equipped with an ultraviolet detector with a wavelength of 360 nm and a 4.6 mm × 150 mm SB-C18 column. An isocratic mobile phase was used with a flow rate of 1 mL min-1. The mobile phase consisted of 80% water with 10 mmol L-1 of oxalic acid aqueous solutions, 16% acetonitrile, and 4% methanol (v/v/v). Metal cations Na+, K+, Mg2+, and Ca2+ were determined using Dionex ICS-1000 ion chromatography consisting of an IonPac Cs12A column (4 mm × 250 mm) with mobile phase of 20 mmol L-1 methanesulfonic acid (flow rate, 1 mL min-1). Before analysis, the aqueous solution was filtered through a SPE C18 column (SIC18300, Tianjin Fuji Science & Technology Co., Ltd.) to remove tetraycycline from the solution.
Results and Discussion Characterization of Varying-Sized Montmorillonite Fractions After size fractionation, the Rest fraction was found to be the major component, comprising 74.5 wt% of the Bulk montmorillonite (Supplemental Table S2). The 2000, 4800, and 11,000 rpm fractions made up 18.9, 2.43, and 4.17 wt%, respectively. All size fractions and the Bulk montmorillonite had similar CEC (~95 cmol kg-1), with the exception of the Rest fraction (85 cmol kg-1) (Supplemental Table S2). These results are consistent with a previous study reporting that the CEC values are independent on particle size for montmorillonites (Sparks et al., 1996). The relatively lower CEC value of the Rest fraction could be attributed to the presence of impurities such as quartz in the sample, as indicated by higher Si content (Supplemental Table S2) and stronger quartz peak in the XRD spectrum (Supplemental Fig. S2). The XRD pattern also suggests that a small amount (~3%) of CaCO3 was present in all clay fractions; however, the presence of CaCO3 is expected to have little influence on the comparison of tetracycline sorption by clay fractions because tetracycline sorption by CaCO3 was much less than that by montmorillonite (Clausen et al., 2001; Xu et al., 2009) and because CaCO3 content was relatively low and similar in all clay fractions (Supplemental Table S2).
The PZNPC values for all clay fractions ranged from 5.1 to 5.6 (Supplemental Table S2), which are between the point of zero charge (PZC) values of major functional groups on phyllosilicate mineral surfaces such as ºAl-OH (PZC~8) and ºSi-OH (PZC~4) (Tombacz and Szekeres, 2004; Tombacz et al., 1995). Proton consumption to adjust clay suspension pH to ~7.0 (at which sorption isotherms were measured) differed pronouncedly among the clay fractions and decreased in the following order: 4800 rpm > > Bulk > 2000 rpm > Rest > 11,000 rpm (Supplemental Table S2). The added protons were consumed by neutralizing or protonating hydroxyl groups ºS-O- + H+® ºS-OH or ºS-OH + H+® ºS-OH2+, where S refers to Al or Si on clay surfaces or at the edges. The more protons are consumed, the more abundant exposed hydroxyl groups are in the clay fraction (Michot et al., 2004). Thus, among all clay fractions, the 4800 rpm fraction possessed the highest amount of exposed hydroxyl groups, whereas the 11,000 rpm fraction had the least. This was further supported by the fact that the suspension of the 4800 rpm fraction demonstrated the highest unadjusted pH (8.50), whereas the 11,000 rpm fraction had the lowest value (7.60) (Supplemental Table S2). Over the test pH range (3–10), all clay fractions were negatively charged, although the 11,000 rpm fraction was more negatively charged than the other clay fractions (Supplemental Fig. S3). The average volume equivalent diameters of clay fractions were measured by laser diffraction in 0.01 mol L-1 NaCl or CaCl2 at pH 7.0. The size distribution of the 11,000 rpm fraction demonstrated a narrow symmetric mono-peak centered at 356 nm, which was remarkably different from other fractions, with broad peaks representing the range of diameters from 0.3 to 200 mm (Supplemental Fig. S4). The effective particle size (represented as volume equivalent diameter) of clay fractions followed the order: 4800 rpm (16.00 mm) > 2000 rpm (12.10 mm) > Bulk (9.54 mm) > Rest (6.38 mm) >> 11,000 rpm (0.41 mm) (Supplemental Table S3). The equivalent particle sizes of the 4800 and 2000 rpm fractions were larger than the Bulk montmorillonite despite the higher centrifugation rate used to obtain these two fractions. This discrepancy can be reconciled by the relative abundance of hydroxyl groups in these two clay fractions. The clay fraction with more hydroxyl groups (e.g., 4800 rpm) tends to become more stable in aqueous suspension due to electrostatic repulsion and hydration forces introduced by Si-OH and Al-OH groups, resulting in more resistance to the separation by centrifugation. Among all size fractions, only the 11,000 rpm fraction fell into the typical range of colloids with sizes 6, which facilitates the complexation with metal cations on montmorillonite surfaces, resulting in a local maximum sorption at pH ~7.7. As the pH further increased, strong electrostatic repulsion occurred between anionic tetracycline and negatively charged clays. Possible hydrogen bonding between polar tetracycline groups and acidic groups on the surface of clay was also expected to become weaker, resulting in a continuous decrease of sorption (Pils et al., 2007; Wang et al., 2010; Tolls, 2001; Kulshrestha et al., 2004). The difference in tetracycline sorption between the 11,000 rpm fraction and other size fractions was significant over the pH range from 5 to 10, although this trend diminished at lower pH. At very low pH (
Journal of Environmental Quality
TECHNICAL REPORTS vadose zone processes and chemical transport
Sorption of Tetracycline to Varying-Sized Montmorillonite Fractions Huaizhou Xu, Xiaolei Qu, Hui Li, Cheng Gu, and Dongqiang Zhu*
T
he environmental fate and transport of many organic compounds are inextricably linked to the nature and abundance of clay minerals present in soils (Delle Site, 2001; Kaiser and Guggenberger, 2000; Sassman and Lee, 2005; Sheng et al., 2001). Although often dwarfed by partitioning of neutral organic compounds into soil organic matter, sorption by clay minerals plays a critical role for polar and ionizable compounds that can be sorbed by ion exchange and/or surface complexation (Delle Site, 2001; Figueroa et al., 2004; Kaiser and Guggenberger, 2000; Kulshrestha et al., 2004; Sassman and Lee, 2005; Sheng et al., 2001; Tolls, 2001). Furthermore, the aluminosilicate surfaces of clays have been found to exhibit significant reaction activities that catalyze chemical transformation of sorbed organic compounds (Chen and Huang, 2010; Laszlo, 1987). Therefore, sorption of organic contaminants to clays could be a key determinant for many physicochemical processes occurring in the soil environment. Montmorillonite represents a group of 2:1 phyllosilicate minerals that have high cation exchange capacity (CEC), large specific surface area, and the unique feature of swelling/collapsing interlayer structures. The isomorphic substitution within aluminosilicate sheets results in net permanent negative surface charges, which are commonly compensated by exchangeable cations (Tombacz and Szekeres, 2004; Vaccari, 1998). The siloxane surfaces between exchangeable cations are regarded as nanoscale domains suitable for sorption of hydrophobic organic compounds ( Jaynes and Boyd, 1991; Sheng et al., 2001). The hydroxyl groups from broken edges of aluminosilicates can protonate and/or deprotonate, creating pH-dependent charges associated with the minerals (Tombacz and Szekeres, 2004; Vaccari, 1998). The wide distribution of antibiotics in soil and water has raised concerns regarding human health due to the potential selective pressure exerted on microbial communities to develop antibiotic resistance (Boxall et al., 2003; Graff et al., 2003; Tolls, 2001). Among these antibiotics, tetracyclines are a large class of antibiotics that have been administered to livestock and humans. A large portion of tetracyclines cannot be metabolized
Abstract The influence of particle sizes on sorption of tetracycline by clay minerals is poorly understood. In this study, montmorillonite clay fractions with varying particle sizes were prepared by successive centrifugation, and the effects of particle sizes on sorption of tetracycline were evaluated using an equilibrium dialysis method. Sorption isotherms were nearly overlapped for size fractions ranging from 6.38 to 16.00 mm, except for the finest clay fraction (0.41 mm). The relatively low sorption by the fraction with the smallest particles could be attributed to the colloidal nature and high edge-to-surface ratio, which could lead to reduced accessibility of tetracycline to sorption sites (particularly those at the edges). The impact of solution pH and coexisting Na+ and Ca2+ ions on tetracycline sorption was found to differ between the finest fraction and other clay fractions. The results demonstrated for the first time that clay particle size greatly influenced tetracycline sorption to clay minerals and consequently might affect their transport and bioavailability in the environment.
Copyright © American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America. 5585 Guilford Rd., Madison, WI 53711 USA. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.
H. Xu, X. Qu, C. Gu, and D. Zhu, State Key Lab. of Pollution Control and Resource Reuse/School of the Environment, Nanjing Univ., Jiangsu 210093, P.R. China; H. Xu, State Key Lab. Pesticide Risk Assessment and Pollution Control, Nanjing Institution of Environmental Sciences, Ministry of Environmental Protection, Nanjing, 210042, P.R. China; H. Li, Dep. of Plant, Soil and Microbial Sciences, Michigan State Univ., East Lansing, MI 48824. Assigned to Associate Editor Thomas Borch.
J. Environ. Qual. 43:2079–2085 (2014) doi:10.2134/jeq2014.04.0182 Supplemental materials are available online for this article. Received 23 Apr. 2014. *Corresponding author ([email protected]).
Abbreviations: AFM, atomic force microscope; AIOC, aromatic ionizable organic compound; (-)CAHB, negative charge–assisted hydrogen bond; CEC, cation exchange capacity; PZC; point of zero charge; PZNPC, point of zero net proton charge; XRD, X-ray diffraction.
2079
in vivo (Boxall et al., 2003; Tolls, 2001) and inevitably enter soils and aquatic systems (Boxall et al., 2003; Graff et al., 2003). Montmorillonites, an important component in soil minerals, can strongly sorb tetracyclines through cation exchange and complexation with metal cations on clay surfaces (Aristilde et al., 2013; Aristilde et al., 2010; Figueroa et al., 2004; Kulshrestha et al., 2004; Li et al., 2010a; Sassman and Lee, 2005; Sithole and Guy, 1987; Tolls, 2001; Wang et al., 2010; Zhao et al., 2012). Significant research efforts have been dedicated to elucidate how the types and chemical compositions of clays influence tetracycline sorption from aqueous solution (Aristilde et al., 2010; Chang et al., 2009a; Chang et al., 2009b; Chang et al., 2012; Chang et al., 2009c; Figueroa et al., 2004; Li et al., 2010a; Li et al., 2010b; Sithole and Guy, 1987). However, little attention has been paid to the interactions between tetracyclines and colloidal clays with varying particle sizes. Natural mineral colloids could serve as effective carriers for strongly sorbed contaminants, significantly enhancing their transport in soils and aquifers (Boxall et al., 2003; Grolimund et al., 1996). The information of colloid-mediated contaminant transport in porous media has been advanced considerably in past two decades (Kanti Sen and Khilar, 2006; Kretzschmar et al., 1999); however, the effects of particle size on sorption of organic contaminants by clay particles is still unclear, which limits the understanding of colloid-mediated transport in soils. The major objective of the present study was to examine sorption of tetracycline by montmorillonite fractions with varying sizes over a range of pH and ionic strength conditions. The results provide a better understanding of the influence of size-dependent clay geochemistry and microstructure on sorption of aromatic ionizable organic compounds (AIOCs). To the best of our knowledge, this is the first study to illustrate the effects of clay particle size on sorption of organic contaminants.
Materials and Methods Clay Fractionations A montmorillonite (Fenghong Inc.) was fractionated by successive centrifugation into five fractions with varying sizes: Bulk, Rest, 2000 rpm, 4800 rpm, and 11,000 rpm. Briefly, 20 g of montmorillonite was dispersed in 1 L of 0.5 mol L-1 NaCl solution for 24 h. The suspension was centrifuged at 60 g for 5 min. The supernatant was collected and was referred to as the Bulk fraction. The residue clay pellet was resuspended in 1 L of 0.5 mol L-1 NaCl. This procedure was repeated three times, and the collected supernatants were combined. The Bulk fraction was then dialyzed against a dialysis membrane tube (3500 Da) (Union Carbide) to remove the excess salts until a negative chloride test of solution with AgNO3 was obtained. The resultant clay suspension was low in ionic strength. The clay suspension was successively centrifuged at 11,000 rpm (11,363 g), 4800 rpm (2164 g), and 2000 rpm (376 g). The centrifugation step was repeated three times with each suspension for 20 min before the next centrifugation speed was used (Gimbert et al., 2005). After each centrifugation, the supernatant was collected, and the clay pellet was resuspended in deionized water. The collected supernatant samples were referred to as the 11,000 rpm fraction, the 4800 rpm fraction, and the 2000 rpm fraction, respectively. 2080
The clay fraction collected after 2000 rpm centrifugation at 376 g was referred to as the Rest fraction.
Sorbate Tetracycline (99%) was purchased from International Laboratory. The molecular weight, water solubility, and stepwise acid dissociation constant (pKa) are listed in Supplemental Table S1. The molecular structure and the pH-dependent speciation of tetracycline are summarized in Supplemental Fig. S1.
Characterization of Montmorillonite Fractions The CEC of the montmorillonite fractions, including Bulk, Rest, 2000 rpm, 4800 rpm, and 11,000 rpm, was determined using the extraction method by 1 mol L-1 NH4Ac and 0.005 mol L-1 EDTA at pH 7.0 (Sparks et al., 1996). The elemental composition of each component was quantified using a X-ray fluorescence spectrometer (ARL-9800, ARL). The size distribution of each fraction was determined using a laser diffraction particle size analyzer (Mastersizer 3000, Malvern). The point of zero net proton charge (PZNPC) (i.e., the pH value at which the net proton surface charge is equal to zero) of all fractions was determined by potentiometric titration using a microcomputer automatic potentiometric titrator (WDDY2008, Datang). The morphology of the 11,000 rpm fraction and the Rest fraction was examined using a Bruker Multimode 8 atomic force microscope (AFM) (Bruker). The measurements of z-potential were performed using a Zata PALS z-potential analyzer. The basal spacing of clay fractions with or without the sorbed tetracycline was measured by an X-ray diffraction (XRD) spectrometer (X’ TRA; ARL).
Sorption Experiments Batch sorption experiment was performed using an equilibrium dialysis method similar to that reported in a previous study (Gu and Karthikeyan, 2005). The amount of clay present in suspension was determined by drying a known volume of clay suspension in an oven (90°C). All clay fractions were then diluted with deionized water to 2 g L-1, and pH values were adjusted to pH 7.0 using 0.1 mol L-1 HCl. Tetracycline stock solution (0.675 mmol L-1) was prepared in 0.02 mol L-1 NaCl solution with pH adjusted to 7.0 using 0.1 mol L-1 NaOH. A dialysis membrane tube (Union Carbide, 3500 Da) was cut into 9-cm-long sections that were pretreated in boiling solution containing 20 g L-1 NaHCO3 and 10 mmol L-1 EDTA at 100°C for 10 min and washed thoroughly with deionized water. For the dialysis experiment, 20 mL of clay suspension was transferred to a 40-mL amber glass bottle equipped with polytetrafluoroethylene-lined screw cap. Then, a given volume of tetracycline stock solution was spiked into the clay suspension. A sealed dialysis cell filled with 5.0 mL of 0.02 mol L-1 NaCl solution was placed into the bottle. The headspace was filled with 0.02 mol L-1 NaCl solution. The bottles were covered with aluminum foil and tumbled at room temperature for 3 d, which was sufficient to achieve apparent sorption equilibrium as indicated in our preliminary study. To account for the possible solute losses from the processes other than sorption to clay (e.g., sorption to glassware, septum, and/or dialysis cell membrane), a calibration curve was prepared under the same treatment as the Journal of Environmental Quality
sorption experiment but in the absence of clay. The calibration curve consisted of eight concentration levels over the range of tetracycline concentration in aqueous solution. Tetracycline in the dialysis cell was analyzed and assumed to be the equilibrium concentration in solution. Clay-sorbed tetracycline was calculated based on the mass balance between the added tetracycline and that present in solution at equilibrium. Another set of sorption experiment was conducted to evaluate the effects of pH and ionic strength on sorption of tetracycline by clays. In the solution, 0.1 mol L-1 HCl or 0.1 mol L-1 NaOH was used to adjust solution pH to achieve equilibrium pH values ranging from 3 to 10. The initial concentration of tetracycline was 0.19 mmol L-1. The ionic strength of solution was prepared with a range from 0 to 0.1 mol L-1 NaCl or CaCl2. All samples were prepared at least in triplicate.
Tetracycline Analysis Tetracycline in the aqueous solution collected from dialysis cells was analyzed by an Agilent 1200 high-performance liquid chromatograph equipped with an ultraviolet detector with a wavelength of 360 nm and a 4.6 mm × 150 mm SB-C18 column. An isocratic mobile phase was used with a flow rate of 1 mL min-1. The mobile phase consisted of 80% water with 10 mmol L-1 of oxalic acid aqueous solutions, 16% acetonitrile, and 4% methanol (v/v/v). Metal cations Na+, K+, Mg2+, and Ca2+ were determined using Dionex ICS-1000 ion chromatography consisting of an IonPac Cs12A column (4 mm × 250 mm) with mobile phase of 20 mmol L-1 methanesulfonic acid (flow rate, 1 mL min-1). Before analysis, the aqueous solution was filtered through a SPE C18 column (SIC18300, Tianjin Fuji Science & Technology Co., Ltd.) to remove tetraycycline from the solution.
Results and Discussion Characterization of Varying-Sized Montmorillonite Fractions After size fractionation, the Rest fraction was found to be the major component, comprising 74.5 wt% of the Bulk montmorillonite (Supplemental Table S2). The 2000, 4800, and 11,000 rpm fractions made up 18.9, 2.43, and 4.17 wt%, respectively. All size fractions and the Bulk montmorillonite had similar CEC (~95 cmol kg-1), with the exception of the Rest fraction (85 cmol kg-1) (Supplemental Table S2). These results are consistent with a previous study reporting that the CEC values are independent on particle size for montmorillonites (Sparks et al., 1996). The relatively lower CEC value of the Rest fraction could be attributed to the presence of impurities such as quartz in the sample, as indicated by higher Si content (Supplemental Table S2) and stronger quartz peak in the XRD spectrum (Supplemental Fig. S2). The XRD pattern also suggests that a small amount (~3%) of CaCO3 was present in all clay fractions; however, the presence of CaCO3 is expected to have little influence on the comparison of tetracycline sorption by clay fractions because tetracycline sorption by CaCO3 was much less than that by montmorillonite (Clausen et al., 2001; Xu et al., 2009) and because CaCO3 content was relatively low and similar in all clay fractions (Supplemental Table S2).
The PZNPC values for all clay fractions ranged from 5.1 to 5.6 (Supplemental Table S2), which are between the point of zero charge (PZC) values of major functional groups on phyllosilicate mineral surfaces such as ºAl-OH (PZC~8) and ºSi-OH (PZC~4) (Tombacz and Szekeres, 2004; Tombacz et al., 1995). Proton consumption to adjust clay suspension pH to ~7.0 (at which sorption isotherms were measured) differed pronouncedly among the clay fractions and decreased in the following order: 4800 rpm > > Bulk > 2000 rpm > Rest > 11,000 rpm (Supplemental Table S2). The added protons were consumed by neutralizing or protonating hydroxyl groups ºS-O- + H+® ºS-OH or ºS-OH + H+® ºS-OH2+, where S refers to Al or Si on clay surfaces or at the edges. The more protons are consumed, the more abundant exposed hydroxyl groups are in the clay fraction (Michot et al., 2004). Thus, among all clay fractions, the 4800 rpm fraction possessed the highest amount of exposed hydroxyl groups, whereas the 11,000 rpm fraction had the least. This was further supported by the fact that the suspension of the 4800 rpm fraction demonstrated the highest unadjusted pH (8.50), whereas the 11,000 rpm fraction had the lowest value (7.60) (Supplemental Table S2). Over the test pH range (3–10), all clay fractions were negatively charged, although the 11,000 rpm fraction was more negatively charged than the other clay fractions (Supplemental Fig. S3). The average volume equivalent diameters of clay fractions were measured by laser diffraction in 0.01 mol L-1 NaCl or CaCl2 at pH 7.0. The size distribution of the 11,000 rpm fraction demonstrated a narrow symmetric mono-peak centered at 356 nm, which was remarkably different from other fractions, with broad peaks representing the range of diameters from 0.3 to 200 mm (Supplemental Fig. S4). The effective particle size (represented as volume equivalent diameter) of clay fractions followed the order: 4800 rpm (16.00 mm) > 2000 rpm (12.10 mm) > Bulk (9.54 mm) > Rest (6.38 mm) >> 11,000 rpm (0.41 mm) (Supplemental Table S3). The equivalent particle sizes of the 4800 and 2000 rpm fractions were larger than the Bulk montmorillonite despite the higher centrifugation rate used to obtain these two fractions. This discrepancy can be reconciled by the relative abundance of hydroxyl groups in these two clay fractions. The clay fraction with more hydroxyl groups (e.g., 4800 rpm) tends to become more stable in aqueous suspension due to electrostatic repulsion and hydration forces introduced by Si-OH and Al-OH groups, resulting in more resistance to the separation by centrifugation. Among all size fractions, only the 11,000 rpm fraction fell into the typical range of colloids with sizes 6, which facilitates the complexation with metal cations on montmorillonite surfaces, resulting in a local maximum sorption at pH ~7.7. As the pH further increased, strong electrostatic repulsion occurred between anionic tetracycline and negatively charged clays. Possible hydrogen bonding between polar tetracycline groups and acidic groups on the surface of clay was also expected to become weaker, resulting in a continuous decrease of sorption (Pils et al., 2007; Wang et al., 2010; Tolls, 2001; Kulshrestha et al., 2004). The difference in tetracycline sorption between the 11,000 rpm fraction and other size fractions was significant over the pH range from 5 to 10, although this trend diminished at lower pH. At very low pH (
