ARTICLE pubs.acs.org/est
Deposition of Silver Nanoparticles in Geochemically Heterogeneous Porous Media: Predicting Affinity from Surface Composition Analysis Shihong Lin,†,|| Yingwen Cheng,‡,|| Yohan Bobcombe,§,|| Kimberly L. Jones,§ Jie Liu,‡ and Mark R. Wiesner*,†,|| †
)
Department of Civil and Environmental Engineering, Pratt School of Engineering, Duke University, Durham, North Carolina 27708, United States ‡ Department of Chemistry, Duke University, North Carolina 27708, United States § Department of Chemical Engineering, Howard University, Washington, DC 20059, United States Center for the Environmental Implications of NanoTechnology (CEINT)
bS Supporting Information ABSTRACT: The transport of uncoated silver nanoparticles (AgNPs) in a porous medium composed of silica glass beads modified with a partial coverage of iron oxide (hematite) was studied and compared to that in a porous medium composed of unmodified glass beads (GB). At a pH lower than the point of zero charge (PZC) of hematite, the affinity of AgNPs for a hematite-coated glass bead (FeO-GB) surface was significantly higher than that for an uncoated surface. There was a linear correlation between the average nanoparticle affinity for media composed of mixtures of FeO-GB and GB collectors and the relative composition of those media as quantified by the attachment efficiency over a range of mixing mass ratios of the two types of collectors, so that the average AgNPs affinity for these media is readily predicted from the mass (or surface) weighted average of affinities for each of the surface types. X-ray photoelectron spectroscopy (XPS) was used to quantify the composition of the collector surface as a basis for predicting the affinity between the nanoparticles for a heterogeneous collector surface. A correlation was also observed between the local abundances of AgNPs and FeO on the collector surface.
’ INTRODUCTION Silver nanoparticles (AgNPs) are finding their way into a growing number of commercial applications.16 Silver in its bulk form has long been known to be bactericidal and has thus been used for disinfection purposes long before the invention of antibiotics. In a nanoparticle format silver can serve as a versatile bactericide with a wide range of uses such as in fabrics,1 water filters,3 filtration membranes,6 and surgical instruments.5 In addition to their bactericidal properties, AgNPs exhibit properties that are attractive for use in highly sensitive optical sensors,7 conductive inks for electronic applications,8 and as colloidal catalysts for organic oxidation.9 The expanding use of AgNPs suggests that these particles may enter wastewater streams and other aqueous media where they may interact with a wide range of other particles and immobile surfaces. The affinity of engineered nanoparticles (ENPs) for relevant surfaces, ranging from sand grains as in water treatment filters and aquifers, to biofilms and biological membranes, is of crucial importance in assessing the fate and eventually the environmental impact of these nanoparticles in natural and engineered systems. The affinity of ENPs for various surfaces provides information necessary to evaluate the mobility, bioavailability, and stability of these nanoparticles in the environment.10 The affinity of nanoparticles for a given surface can be quantified by r 2011 American Chemical Society
the attachment efficiency between the nanoparticles and the surfaces under specified conditions of solution chemistry.11 In this paper, we consider the relative affinity of AgNPs for two types of surfaces that are frequently encountered in natural and engineered systems: silica and iron oxide (mainly hematite). The surface charge of these inorganic oxides is strongly dependent on pH. Silica is usually negatively charged over the range of environmental values of pH, while values for the point of zero charge (PZC) for iron oxides are close to the pH of many aquatic environments.12 This implies that the affinity of negatively charged nanoparticles for an iron oxide surface might be expected to be more sensitive to pH than for a silica surface. In addition, environmental surfaces are typically heterogeneous. It is thus important to understand the effect of surface heterogeneity on the affinity of nanoparticles for the heterogeneous surfaces. The goal of the current study is to assess the affinity of AgNPs for a pure silica surface and a heterogeneous surface coated with hematite and to test the correlation, if there is any, between the affinity of nanoparticle for the collector surface and the Received: January 19, 2011 Accepted: April 27, 2011 Revised: March 30, 2011 Published: May 18, 2011 5209
dx.doi.org/10.1021/es2002327 | Environ. Sci. Technol. 2011, 45, 5209–5215
Environmental Science & Technology surface composition of the collector. The PZC of hematite is typically reported to be in the range of 6.66.9.13 Previous investigators have considered the dynamics of deposition of silica colloid to porous media composed of a mixture of negatively charged uncoated quartz sands and positively charged collectors that are fully coated with either iron oxyhydroxide14 or aminosilane.15 A linear relationship between the attachment efficiency of deposition (RDep) of the colloids for the media, and ratio of the mixture of the positively charged and negatively charged collectors, was observed.15 The current study examines the deposition of AgNPs and considers the effect of heterogeneity at the level of the collector surface as well as heterogeneity at the level of mixtures of collectors of different types. By examining the relationship between particlesurface affinity and surface heterogeneity, we test the hypothesis that the affinity of AgNPs for a mixture of surfaces can be predicted from the individual affinities of these particles for each of the mineral phases.
’ MATERIALS AND METHOD Synthesis and Characterization of the Silver Nanoparticles. Suspensions of charged-stabilized AgNPs were prepared
using the borohydride-reduction method16 which is also briefly restated in the Supporting Information (SI). The resulting AgNPs were stable as an aqueous suspension and were used as is without further purification to prevent aggregation that otherwise occurs following the washing-resuspension or dialysis processes. The volume of background electrolyte used in the subsequent experiments was much larger than that of the stock AgNPs suspension and thus the contribution to ionic strength of the AgNPs stock from residual ions was negligible throughout the range of ionic strengths tested. The UVvis spectrum of the AgNPs was scanned using a spectrophotometer (UVvis 2810, Hitachi, Pleasanton, CA) to confirm the existence of surface plasmon resonance (SPR) characteristic of AgNPs. Both transmission electron microscopy (FEI Tecnai G2 Twin, Hillsboro, OR) and dynamic light scattering using a goniometer (ALV/CGS 3, Germany) were conducted to obtain size information of the AgNPs. The electrophoretic mobility (EPM) of the AgNPs was measured using a Zeta-Sizer (Nanosizer ZS, Malvern Instruments, Worcestershire, U.K.). The EPM measurements were conducted at a fixed ionic strength of 102 M (adjusted by NaNO3) and at a variety of pHs ranging from 5.0 to 8.3. Preparation of Porous Media. Spherical silicate glass beads (Potters Industries Inc., Berwyn, PA) were washed following a protocol by Espinasse17 to remove the organic and inorganic impurities on the glass beads’ surfaces. The procedure for preparing iron oxide (FeO) coated glass beads (FeO-GB) largely follows that of Benjamin18 for coating silicate sand surface for arsenic removal. Two types of FeO-GBs were prepared—one with a higher surface coverage (FeO-GB-HC) and the other with a lower surface coverage (FeO-GB-LC). The detailed procedures for synthesis and purification can be found in SI. In both samples, the iron oxide formed on the collector surface following this procedure is predominantly hematite.12 The FeO-GBs prepared using these two methods were visually different in that the FeOGB-LC had a lighter overall color than the FeO-GB-HC. It should be noted that even though FeO-GB-HC was subsequently confirmed to have a higher FeO surface coverage than the FeO-GB-LC, in both cases, the underlying silica surface was
ARTICLE
only partially covered with FeO as revealed by both electron microscopic imaging and surface composition characterization. Characterization of Glass Beads and the Iron Oxide Coated Glass Beads. The glass beads (GB) and FeO-GBs were characterized with respect to the surface chemistry and morphology of these collectors. First, the surface potential (as approximated by the ζ potential) of the collectors was calculated from streaming potential measurements (ZetaCad, CAD Instruments, France) conducted at the same solution chemistry as for the EPM measurements mentioned above. X-ray photoelectron spectroscopy (Axis Ultra, Kratos Analytical, Chestnut Ridge, NY) was used to determine the surface elemental composition of the FeO-GBs prepared using two different methods. Because the size of the projection area of a collector was similar to the size of the sampling area (0.3 mm 0.7 mm), at least 30 measurements were conducted to generate representative results. The spectra data collected using XPS were processed by CasaXPS (Casa Software Ltd.) to calculate the relative concentrations of Si and Fe on the surface. Finally, scanning electron microscopy (FEI XL30 FEG-SEM, Hillsboro, OR) was employed to observe the morphology of iron oxide coated surface. Column Experiments. The affinities between AgNPs and different surfaces were evaluated by measuring the removal of the AgNPs by a well-defined porous media comprising the collectors of interest. The spherical collectors were packed into a chromatography column (C10/10, GE Healthcare, NJ) with an inner diameter of 10 mm and a height of 6 cm with the assistance of a height adaptor (GE Healthcare, NJ). Mild sonication using a sonication bath was applied to ensure that the column was tightly packed to attain minimum porosity which was measured to be 0.378. The background electrolyte solution was delivered by a single magnetic-drive gear pump (Cole-Parmer Instrument Company, Chicago, IL) at a flow rate of 0.94 mL/min which corresponded to a Darcy velocity of 0.02 cm/s in this system. The addition of AgNPs suspension was achieved by a syringe pump (Harvard Apparatus, Holliston, MA) at a flow rate of 0.04 mL/min that was negligible compared to the electrolyte flow rate. The mass concentration of AgNP in the resulting mixture was 1.35 mg/L. The electrolyte solution and the nanoparticle suspension were mixed in a Y-connector (BioChem Fludics, UK) before the mixture entered the packed column. The chamber of the Y-connector was no more than several microliters in volume, ensuring that aggregation before the column would be negligible due to the very short hydraulic retention time. At least ten pore volumes (PV) of background electrolyte were passed through the column to equilibrate the surface of the porous media before conducting the deposition experiment. The real time relative concentration of AgNPs was quantified by a UVvis spectrophotometer (UVvis 2810, Hitachi, Pleasanton, CA) with a flow module. The primary peak of the UVvis spectra for the AgNPs was at the wavelength of 390 nm. It was confirmed that the absorbance at that wavelength was proportional to the concentration of AgNPs. The GB and the FeO-GB-HC were respectively used as collectors for column tests using background electrolyte solutions of pH 5.0 and pH 8.3 with a variety of ionic strengths to assess and compare the affinity of AgNPs for these two different porous media under various conditions of solution chemistry. To study the effect of mineral heterogeneity of porous media on attachment efficiency and to evaluate the hypothesized linear relationship between the attachment efficiency and surface composition, the FeO-GB-HC was mixed with uncoated GB at different mass 5210
dx.doi.org/10.1021/es2002327 |Environ. Sci. Technol. 2011, 45, 5209–5215
Environmental Science & Technology
Figure 1. (A) Evolution of hydrodynamic radius of AgNPs under different concentration of NaNO3 (pH = 5.0). (B) Change of normalized aggregation rate as NaNO3 concentration increases. The CCC, as the minimal concentration for the diffusion-controlled regime, was determined to be around 300 mM. The error bars represent standard deviations.
ratios for column tests with a background electrolyte solution of pH 5.0 and an ionic strength of 0.01 M. With the same solution chemistry, the affinity of uncoated AgNPs in a porous medium composed of FeO-GB-LC was also measured. The attachment efficiency thereby obtained was coupled with the XPS data to evaluate the validity of using surface composition analysis for conditional prediction of attachment efficiency.
’ RESULT AND DISCUSSION Properties of the Silver Nanoparticles. The stock AgNPs suspension had a clear bright yellow color consistent with a UVvis spectrum that peaked at 390 nm; The hydrodynamic radius of the AgNPs was measured to be 12.09 ( 1.01 nm by dynamic light scattering, which was around twice the average particle size as given by TEM image analysis. The hydrodynamic diameter rather than the diameter determined by TEM was used in calculating the single collector contact efficiency η0 because it is more relevant to Brownian transport of the AgNPs. The critical concentration of coagulation (CCC) for these AgNPs at pH 5.0 was found to be approximately 300 mM (Figure 1), much higher than even the highest ionic strength used in column experiments,
ARTICLE
thereby further ensuring that aggregation did not play a role in these experiments. Morphology and Composition of Collector Surface. The FeO coating on the surface of the FeO-GB-HC was patchy and inhomogeneous as observed by scanning electron microscope (SEM) (Figure S4). Two representative XPS spectra, one for FeO-GB-HC and the other for FeO-GB-LC, are presented, respectively, on which shows an apparent difference in the height of the Fe 2P peaks of the spectra. The ratio Fe/(Fe þ Si) as a measure of surface abundance of Fe compared to Si was 25.34 ( 3.04% for FeO-GB-HC and 6.84 ( 1.87% for the FeO-GB-LC based on more than 30 XPS measurements for each type of sample. Although it can be observed from the SEM images (Figure S4) that the iron oxide coatings were visually inhomogeneous and the surface coverage appears to vary from one bead to another, the Fe/(Fe þ Si) ratios given by XPS measurement were relatively consistent for all the collectors examined. It should also be noted that the relative surface concentration of Fe as measured by XPS was higher than suggested by the patchy coverage observed by SEM, which implies that a portion of the “uncovered” areas might in fact be partially associated with Fe. Surface Chemistry of AgNPs and Collectors. The zeta (ζ) potentials calculated for the AgNPs, GB, and FeO-GB-HC were all negative in the range of pHs tested (Figure S6). (Note that ζ potentials are reported rather than the primary electrokinetic measurements to allow for comparison of measurements of EPM and streaming potential on the same plot.) Comparatively speaking, the ζ potentials of AgNPs were lower (less negative) than those of the collectors. However, such a comparison should be made with caution because the ζ potentials of the nanoparticles were calculated from electrophoresis measurements, whereas those of the collectors were calculated from streaming potential measurements. Nonetheless, theory and experimental evidence suggest that these two techniques yield the same result for large particles.19 It should also be noted that the FeO-GB-HC was still of relatively high negative charge even at a pH value of 5.0—a value that is lower than the reported PZC for hematite,13 which implies that the negatively charged silica patches contribute significantly in determining the overall electrokinetic properties of the collectors in these conditions. Effect of pH and ionic Strength on Particle Deposition. Figure 2A, B, C, and D shows the representative breakthrough curves (BTCs) for AgNPs at low and high pH values with a range of ionic strengths in porous media of GB and FeO-GB-HC, respectively. For the case when the pH of the background solution was adjusted to 5.0, the pH of effluent mixture was measured to be 6.2—still lower than the PZC of the hematite; for the case when the pH of the background solution was adjusted to 8.3, the pH of the effluent was measured to be 8.2. In both porous media, attachment of the AgNPs increased with increasing NaNO3 concentration due to electrical double layer (EDL) compression. Note that the hydrodynamic size of the particles was monitored using DLS at both the influent and effluent and was found to be the same. The lack of aggregation across the column was consistent with the ionic strength being below the CCC and the large volume fraction of porous medium collectors relative to the suspended particle volume fraction. An obvious difference was observed between the shapes of the breakthrough curves obtained using different porous media when pH was lower than PZC of FeO. With the uncoated GB as the porous medium, the effluent concentration of AgNPs reached a plateau soon after injection and remained constant until 5211
dx.doi.org/10.1021/es2002327 |Environ. Sci. Technol. 2011, 45, 5209–5215
Environmental Science & Technology
ARTICLE
Figure 2. AD: Breakthrough curves (BTC) for AgNPs in the following scenarios: (A) GB at high pH; (B) FeO-GB-HC at high pH; (C) GB at low pH; and (D) FeO-GB-HC at low pH. E: Attachment efficiencies calculated from the BTCs in AD. The error bars represent standard deviations.
the termination of injection, which was in accordance with the traditional filtration theory. The breakthrough curves in this case were similar to that of the tracer (NO3). However, for the porous medium of FeO-GB-HC, a flat plateau was not observed even after 3 PVs of injection. The effluent concentration continued to increase gradually as more and more AgNPs deposited onto the collector surface. This phenomenon is explained by the dynamics of deposition in which the negatively charged AgNP block the positively (or less negatively) charged sites so that local favorable deposition may be attenuated and eventually replaced by electrostatic repulsion over longer timeframes due to surface charge reversal, which can be described using the random sequential adsorption (RSA) model.14,20 For this case of dynamic deposition where steady state would not be reached at the early stage, theoretically an attachment efficiency unique to the solution chemistry does not exist because the surface chemistry of the collector is under constant change. The evolution of the BTCs has been successfully described by the RSA model and is not the focus of the current study. In all cases, the average of relative effluent concentration measured between
second to third PV was used to estimate the clean bed attachment efficiency R that represents the relative affinity of particles for the porous medium. The attachment efficiency R calculated from the Ceff/Cinf obtained from experimental breakthrough curves in different scenarios are shown in Figure 2E. Qualitatively consistent with standard DerjaguinLandauVerweyOverbeek (DLVO) theory, the attachment efficiency increased with increasing ionic strength as the diffuse layer was compressed and the EDL interaction was weakened. At high pH, both the GB and iron oxide portion of FeO-GB-HC were negatively charged, as this pH is greater than PZCs of both silica and hematite. Theoretically, the GB is more negatively charged than the FeO-GB-HC since the PZC for GB is further away from the solution pH than that of the FeO. Therefore, at the same ionic strength the attachment efficiency between AgNPs and FeO portion of the FeO-GB-HC should be higher than that between AgNPs and GB. However, the differences in attachment efficiencies between two porous media at ionic strengths up to 0.02 M were not statistically significant. A large enhancement in AgNPs removal 5212
dx.doi.org/10.1021/es2002327 |Environ. Sci. Technol. 2011, 45, 5209–5215
Environmental Science & Technology
Figure 3. Log Rexp vs Φ/kBT for certain deposition experiments. The Rexp is attachment efficiency obtained from column experiments. The energy barrier is calculated using the classic DLVO theory with the measured ζ potentials of the AgNPs and the collectors in the solution chemistry of corresponding experiments. χ is the mass ratio of FeO-GBHC in a medium composed of a mixture of GB and FeO-GB-HC. All data points were collected with ionic strength fixed at 0.01 M.
by the FeO-GB was observed when the ionic strength was increased to 0.03 M, which might be attributed to the elimination of the energy barrier between the AgNPs and the FeO fraction of the collector surface. At low pH, the affinity of AgNPs for FeOGB was noticeably greater than that for GB even at relatively low ionic strength, possibly because this pH was lower than the PZC of hematite so that FeO fraction would bare positive charge leading to favorable deposition of the negatively charge AgNPs. However, the measured overall attachment efficiencies at low ionic strengths were significantly lower than that predicted by assuming that 25% of the surface was covered by hematite and that the hematite fraction would lead to favorable deposition. One possible explanation would be that the AgNPs approaching the positively charged hematite patches might experience a “hydrodynamic pump” effect that may occur during particle deposition onto a surface with a micropatterned charge heterogeneity.21,22 Another possibility is that although the XPS measurement gave a value of 25.34 ( 3.04% for Fe/(Fe þ Si) of the FeO-GB-HC, the fraction of surface area covered by large hematite patches might be significantly smaller than 25%. Part of the total surface Fe concentration detected by XPS might be contributed by the ]Fe located either in the uncovered area as =SiOFe, or underneath the uncovered area as a result of isomorphous substitution. Therefore, the effective FeO patches leading to favorable deposition might be in fact smaller than that suggested by XPS measurements. Relative Insignificance of Zeta Potential ζ. It is tempting to simply apply the classic DVLO theory using measurable global parameters such as ionic strengths and the surface potentials of the particles and the collectors to predict the attachment efficiency R between the particle and the collector surface. Notwithstanding the intrinsic inadequacies of DLVO theory in modeling particle deposition onto a homogeneous surface,23 using the DLVO theory with a heterogeneous surface where surface potential (approximated by ζ potential) varies spatially,
ARTICLE
Figure 4. Attachment efficiencies of AgNPs for bicomponent porous media composed of GB and FeO-GB-HC mixed at different ratios and for a monocomponent porous medium composed of FeO-GB-LC. The background electrolyte solution was adjusted to IS = 10 mM and pH = 5.0. The error bars represent standard deviations.
introduces further problems. For example, Elimelech et al.15 concluded that the deposition kinetics in a similar case was determined primarily by the extent of heterogeneity and that using ζ of the collector surface for prediction would lead to erroneous results. Our experimental observation of AgNPs deposition onto collectors of local heterogeneity (i.e., patchwise heterogeneity on a single collector as compared to column scale heterogeneity) also supports the aforementioned argument. Even without conducting a detailed calculation employing the interaction force boundary layer approximation,24 it can be easily shown that R µ eφ/kBt 23 does not hold if the total interaction energy j is evaluated using the measured ζ potential. (We used expression by Gregory25 for calculating vdW interaction potential and the linear-superposition-approximation-based improved analytical expression for calculating sphere-plate EDL interaction.26) Figure 3 presents Log Rexp vs barrier of the total interaction energy (Φ/kBT) based on column test and zeta potential measurement results. If classic DLVO theory with ζ potentials were a valid predictor of attachment efficiency, all of the data in Figure 3 should fall on a straight line, which was not observed. The starred data points refer to the measurement on GB only (assumingly homogeneous surface). Theoretically, the dashed line (reference line) connecting the starred data points is where valid DLVO predictions are favored. When pH was lower than the PZC of FeO, all the data points are above the reference line, meaning that the particle-collector affinity was higher than that predicted using DLVO theory with the apparent ζ potential of the heterogeneous collectors. The higher the degree of heterogeneity, the higher the extent to which the data deviate from the reference line because a higher fraction of surface was available for favorable deposition. Correlating Attachment Efficiency to Surface Composition in Heterogeneous Porous Media. To evaluate the affinity between AgNPs and other heterogeneous surfaces of known composition (i.e., the percentages of FeO and silica on surface in our case) in a porous medium, it is necessary to understand the effect of the surface composition of the porous medium. The simplest possibility would be a linear relationship in which the affinity of a given type of nanoparticle in a heterogeneous porous 5213
dx.doi.org/10.1021/es2002327 |Environ. Sci. Technol. 2011, 45, 5209–5215
Environmental Science & Technology
Figure 5. Correlation between the local relative concentrations of Fe and of Ag based on XPS analysis. FeO-GB-HC sample collectors were taken carefully from the same longitudinal position in a column through which 1 mL of AgNPs passed.
medium composed of multiple types of collector surfaces is expressed as a composition-weighted average of the affinities of that type of nanoparticle for each of the individual types of collector surfaces comprising the heterogeneous medium.14,27 To test such a linear relationship for the current system, a series of heterogeneous media of varying surface compositions was prepared by mixing the GB and FeO-GB-HC at different mass ratios to evaluate the affinity of the AgNPs for such porous media while the background electrolyte solution was adjusted to pH 5.0 and an ionic strength of 10 mM. As shown in Figure 4, there was indeed a strong correlation (R = 0.992) between the affinity of AGNPs for the heterogeneous media and the mass percentage of the FeO-GB-HC. Also measured was the AgNPs affinity for the FeO-GB-LC, which has a lower FeO fraction and in practical terms is heterogeneous only on the scale of the collector. Comparing the Fe/Si compositions from XPS measurement between FeOGB-HC and FeO-GB-LC, it was possible to assign an equivalent mass ratio of FeO-GB-HC to the FeO-GB-LC, viewing the FeOGB-LC as a mixture of FeO-GB-HC and GB, to quantify the surface composition of FeO-GB-LC for the purpose of predicting the affinity of AgNPs for these media. The compositional data from XPS and the attachment efficiencies from column test are also presented in Figure 4. Both the proportion of Fe and the attachment efficiency for FeO-GB-LC was between that of GB and FeO-GB-HC. Quantifying the surface composition of FeO-GB-LC using the mass ratio of FeO-GB-HC, it was calculated that the Fe proportion of the FeO-GB-LC surface was equivalent to a FeO-GB-HC mass ratio of 0.27(0.08 in a FeO-GB-HC/GB bicomponent porous medium. The fitting equation relating the overall attachment efficiency and the mass ratio of FeO-GB-HC in a heterogeneous column for this case was found to be R = 0.1289χ þ 0.055, where R was the attachment efficiency between AgNPs and χ is the FeO-GB-HC mass ratio of the bicomponent medium. Using this equation and an equivalent χ of 0.27(0.08 for FeO-GB-LC, the attachment efficiency R for the FeO-GB-LC was predicted to be 0.089(0.01 which was different from the experimentally measured attachment efficiency Rexp by only 11%. Considering the fact that the sampling area for XPS with our instrument was only 0.7 mm 0.3 mm, which is insufficient to cover two collectors in a single measurement, the uncertainty associated with surface composition
ARTICLE
measurement is appreciable and it thus can be concluded that the experimentally determined Rexp agreed with predicted value satisfactorily. In addition to correlating the surface composition and the global attachment efficiency obtained from column tests, local evidence was also observed to further support the hypothesized linear correlation between the composition of a surface fraction and the affinity of the AgNPs for that surface fraction. FeO-GBHCs were used as collectors in columns with a background electrolyte of 0.01 M and pH 5.0 and a flow rate identical to that used in all other column tests conducted. One milliliter of AgNPs suspension was allowed to flow through the column before the FeO-GB-HC beads were collected for XPS measurement. Due to the longitudinal distribution of the AgNPs concentration in a column, only a thin slice of the porous medium was sampled to ensure that the FeO-GB-HCs for XPS measurement were all located at the same longitudinal position so as to avoid systematic error. Not only the Fe and Si elements but also the Ag element were analyzed to obtain a Fe/Si/Ag profile for each measurement. Because the FeO coating on the collectors was inhomogeneous and each measurement sampled only one to two collectors, a relatively wide spectrum of relative Fe concentration was obtained ranging from 15.26% to 30.63% for 22 measurements in which the relative Ag concentration ranged from 0.18% to 0.55%. The observed linear correlation between attachment efficiencies is reflected in a linear mapping between the relative Fe concentration and the relative Ag concentration on the collector surface as shown in Figure 5. The amount of the AgNPs retained on the collector surface increases linearly with the abundance of FeO on the collector surface. Such an observation therefore further supports the hypothesis that the affinity of AgNPs for a mixture of surfaces can be predicted from the individual affinities of these particles for each of the mineral phases.
’ ASSOCIATED CONTENT
bS
Supporting Information. Procedures for preparing the heterogeneous collectors (a); TEM image (Figure S1), size distribution (Figure S2) and UVvis spectrum (Figure S3) of AgNPs; SEM image of FeO-GB-HC (Figure S4); XPS spectra of FeO-GB-HC and FeO-GB-LC surfaces (Figure S5); Zeta potential of AgNPs and the porous media (Figure S6); determination of attachment efficiency (b). This material is free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected]; phone: 919-660-5292; fax: 919660-5219.
’ ACKNOWLEDGMENT This material is based upon work supported by the National Science Foundation (NSF) and the Environmental Protection Agency (EPA) under NSF Cooperative Agreement EF-0830093, Center for the Environmental Implications of NanoTechnology (CEINT). Any opinions, findings, conclusions, or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF or the EPA. This work has not been subjected to EPA review and no official 5214
dx.doi.org/10.1021/es2002327 |Environ. Sci. Technol. 2011, 45, 5209–5215
Environmental Science & Technology endorsement should be inferred. We thank CEREGE for kindly offering us the access to their streaming potential instruments and we also thank Dr. Melanie Auffan and Dr. Natalia Solovitch for their help with streaming potential measurements.
’ REFERENCES (1) Lee, H.; Yeo, S.; Jeong, S. Antibacterial effect of nanosized silver colloidal solution on textile fabrics. J. Mater. Sci. 2003, 38 (10), 2199–2204. (2) Baker, C.; Pradhan, A.; Pakstis, L.; Ismat, S. Synthesis and antibacterial properties of silver nanoparticles. J. Nanosci. Nanotechnol. 2005, 5 (2), 244–249. (3) Jain, P.; Pradeep, T. Potential of silver nanoparticle-coated polyurethane foam as an antibacterial water filter. Biotechnol. Bioeng. 2005, 90 (1), 59–63. (4) Li, Q.; Mahendra, S.; Lyon, D.; Brunet, L.; Liga, M.; Li, D.; Alvarez, P. Antimicrobial nanomaterials for water disinfection and microbial control: potential applications and implications. Water Res. 2008, 42 (18), 4591–4602. (5) Bosetti, M.; Masse, A.; Tobin, E.; Cannas, M. Silver coated materials for external fixation devices: in vitro biocompatibility and genotoxicity. Biomaterials 2002, 23 (3), 887–892. (6) Chou, W.; Yu, D.; Yang, M. The preparation and characterization of silver-loading cellulose acetate hollow fiber membrane for water treatment. Polym. Adv. Technol. 2005, 16 (8), 600–607. (7) McFarland, A.; Van Duyne, R. Single silver nanoparticles as realtime optical sensors with zeptomole sensitivity. Nano Lett. 2003, 3 (8), 1057–1062. (8) Lee, H.; Chou, K.; Huang, K. Inkjet printing of nanosized silver colloids. Nanotechnology 2005, 16, 2436–2441. (9) Shiraishi, Y.; Toshima, N. Colloidal silver catalysts for oxidation of ethylene. J. Mol. Catal. A: Chem. 1999, 141 (13), 187–192. (10) Wiesner, M. R.; Lowry, G. V.; Alvarez, P.; Dionysiou, D.; Biswas, P. Assessing the risks of manufactured nanomaterials. Environ. Sci. Technol. 2006, 40 (14), 4336–4345. (11) Lecoanet, H.; Bottero, J.; Wiesner, M. Laboratory assessment of the mobility of nanomaterials in porous media. Environ. Sci. Technol. 2004, 38 (19), 5164–5169. (12) Cornell, R.; Schwertmann, U. The Iron Oxides; VCH Weinheim, 1996. (13) Parks, G. A. The isoelectric points of solid oxides, solid hydroxides, and aqueous hydroxo complex systems. Chem. Rev. 1965, 65 (2), 177–198. (14) Johnson, P. R.; Sun, N.; Elimelech, M. Colloid transport in geochemically heterogeneous porous media: Modeling and measurements. Environ. Sci. Technol. 1996, 30 (11), 3284–3293. (15) Elimelech, M.; Nagai, M.; Ko, C. H.; Ryan, J. N. Relative insignificance of mineral grain zeta potential to colloid transport in geochemically heterogeneous porous media. Environ. Sci. Technol. 2000, 34 (11), 2143–2148. (16) Creighton, J.; Blatchford, C.; Albrecht, M. Plasma resonance enhancement of Raman scattering by pyridine adsorbed on silver or gold sol particles of size comparable to the excitation wavelength. J. Chem. Soc., Faraday Trans. 2 1979, 75, 790–798. (17) Espinasse, B.; Hotze, E. M.; Wiesner, M. R. Transport and retention of colloidal aggregates of C-60 in porous media: Effects of organic macromolecules, ionic composition, and preparation method. Environ. Sci. Technol. 2007, 41, 7396–7402. (18) Benjamin, M.; Sletten, R.; Bailey, R.; Bennett, T. Sorption and filtration of metals using iron-oxide-coated sand. Water Res. 1996, 30 (11), 2609–2620. (19) Silvert, P.; Herrera-Urbina, R.; Tekaia-Elhsissen, K. Preparation of colloidal silver dispersions by the polyol process. J. Mater. Chem. 1997, 7 (2), 293–299. (20) Johnson, P. R.; Elimelech, M. Dynamics of colloid deposition in porous media: Blocking based on random sequential adsorption. Langmuir 1995, 11 (3), 801–812.
ARTICLE
(21) Elimelech, M.; Chen, J. Y.; Kuznar, Z. A. Particle deposition onto solid surfaces with micropatterned charge heterogeneity: The “hydrodynamic bump” effect. Langmuir 2003, 19 (17), 6594–6597. (22) Nazemifard, N.; Masliyah, J.; Bhattacharjee, S. Particle deposition onto micropatterned charge heterogeneous substrates: Trajectory analysis. J. Colloid Interface Sci. 2006, 293 (1), 1–15. (23) Elimelech, M.; Omelia, C. R. Kinetics of deposition of colloidal particles in porous-media. Environ. Sci. Technol. 1990, 24 (10), 1528–1536. (24) Spielman, L.; Friedlander, S. Role of the electrical double layer in particle deposition by convective diffusion. J. Colloid Interface Sci. 1974, 46 (1), 22–31. (25) Gregory, J. Approximate expressions for retarded van der Waals interaction. J. Colloid Interface Sci. 1981, 83 (1), 138–145. (26) Lin, S.; Wiesner, M. R. Exact analytical expression for the potential of electric double layer interaction for a sphere-plate system of constant surface potentials. Langmuir 2010, 26 (22), 16638–16641. (27) Song, L.; Johnson, P.; Elimelech, M. Kinetics of colloid deposition onto heterogeneously charged surfaces in porous media. Environ. Sci. Technol. 1994, 28 (6), 1164–1171.
5215
dx.doi.org/10.1021/es2002327 |Environ. Sci. Technol. 2011, 45, 5209–5215
