Role of Collector Alternating Charged Patches on Transport of

Feb 4, 2013 - Fundamental and Computational Sciences Directorate, Pacific ... Institute for Advanced Science and Technology, University of Illinois at...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/est

Role of Collector Alternating Charged Patches on Transport of Cryptosporidium parvum Oocysts in a Patchwise Charged Heterogeneous Micromodel Yuanyuan Liu,† Changyong Zhang,‡ Dehong Hu,‡ Mark S. Kuhlenschmidt,§ Theresa B. Kuhlenschmidt,§ Steven E. Mylon,∥ Rong Kong,⊥ Rohit Bhargava,⊥ and Thanh H. Nguyen*,† †

Department of Civil and Environmental Engineering, the Center of Advanced Materials for the Purification of Water with Systems, University of Illinois at Urbana−Champaign, Urbana Illinois 61801, United States ‡ Fundamental and Computational Sciences Directorate, Pacific Northwest National Laboratory, Richland Washington 99354, United States § Department of Pathobiology, University of Illinois at Urbana−Champaign, Urbana Illinois 61801, United States ∥ Department of Chemistry, Lafayette College, Easton Pennsylvania 18042, United States ⊥ Department of Bioengineering, Micro and Nanotechnology Laboratory and Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana−Champaign, Urbana Illinois 61801, United States S Supporting Information *

ABSTRACT: The role of collector surface charge heterogeneity on transport of Cryptosporidium parvum oocyst and carboxylate microsphere in 2-dimensional micromodels was studied. The cylindrical silica collectors within the micromodels were coated with 0, 10, 20, 50, and 100% Fe2O3 patches. The experimental values of average removal efficiencies (η) of the Fe2O3 patches and on the entire collectors were determined. In the presence of significant (>3500 kT) Derjaguin−Landau− Verwey−Overbeek (DLVO) energy barrier between the microspheres and the silica collectors at pH 5.8 and 8.1, η determined for Fe2O3 patches on the heterogeneous collectors were significantly less (p < 0.05, t test) than those obtained for collectors coated entirely with Fe2O3. However, η calculated for Fe2O3 patches for microspheres at pH 4.4 and for oocysts at pH 5.8 and 8.1, where the DLVO energy barrier was relatively small (ca. 200−360 kT), were significantly greater (p < 0.05, t test) than those for the collectors coated entirely with Fe2O3. The dependence of η for Fe2O3 patches on the DLVO energy barrier indicated the importance of periodic favorable and unfavorable electrostatic interactions between colloids and collectors with alternating Fe2O3 and silica patches. Differences between experimentally determined overall η for charged heterogeneous collectors and those predicted by a patchwise geochemical heterogeneous model were observed. These differences can be explained by the model’s lack of consideration for the spatial distribution of charge heterogeneity on the collector surface.



INTRODUCTION

heterogeneous distribution of charge on most collectors. The oxides of Fe, Al, and Mn in soils are important sources of geochemical heterogeneities14,15 and their fraction in the total mass of the soil is typically in the range of 0.5−22%14,16,17 and could reach 58% in sediments contaminated by acid mine drainage.14,18 Though Fe (III) oxide crystal sizes are generally less than 1 μm, large discrete Fe (III) oxide crystals can form in locations where Fe concentration exceeded 10%.14,19 This geochemical charge heterogeneity caused by metal oxides has

Cryptosporidium parvum (C. parvum) is identified as a pathogenic microorganism that targets mammals, including humans. Outbreaks of cryptosporidiosis have been reported frequently each year due to the challenges in removing Cryptosporidium oocysts from drinking or recreational water.1−3 Filtration is recommended to control oocysts, and facilities without filtration are required to include Cryptosporidium in their existing watershed control provisions.4 Therefore, investigation of oocyst transport in porous media has been extensively studied in the past decades.5−13 Colloid transport studies generally assume uniform distribution of surface charge for both colloids and collectors. However, in most natural or engineered porous media, there is a © 2013 American Chemical Society

Received: Revised: Accepted: Published: 2670

October 9, 2012 January 3, 2013 February 4, 2013 February 4, 2013 dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

Figure 1. (A) Geometry of micromodel and (B) scanning electron microscopy (SEM) picture of the heterogeneous silica/Fe2O3 collectors at Acc.V = 5 kV. Length of pore network: x = 12 mm, width of pore network: z = 8 mm, height of collector: 22−28 μm, collector diameter: 190 μm, pore space: 130 μm, and pore throat: 36.3 μm. Energy dispersive X-ray spectroscopy (EDS) spectrum of Fe2O3 (C) and silica (D) surfaces at Acc.V = 15 kV.

been found to play an important role in oocyst transport.5,8,10,20 Transport experiments of oocysts in columns packed with mixed quartz sand and ferric oxyhydroxides indicated that oocyst attachment was not proportional with the fraction of ferric oxyhydroxide.8 Colloid filtration theory (CFT) was refined by considering surface interactions to study the influence of charge heterogeneity on colloid transport.15,21 Specifically, the presence of small amounts of charge heterogeneity on the collector, from 2 to 20% by surface area, led to deposition that was orders of magnitude larger.15 However, the attachment mechanism in the presence of collector charge heterogeneity has not been fully understood due to the lack of information about the distribution of charge heterogeneities. Collector charge heterogeneity at both microscopic and macroscopic scales can influence colloid deposition.15,22−26 In the case of microscopic charge heterogeneity, the size of charged heterogeneous patches is much smaller than the colloids.15,22 For these systems, the parameters such as the zeta potential, the interaction energy, and the attachment efficiency are found to be sensitive to the size and number of microscopic-scaled heterogeneities.8,22,24 For macroscopic-

scale charged heterogeneous patches, isolated patches are assumed to be significantly larger than the colloids, and the interactions at patch boundaries can be neglected. For these systems, a linear correlation between attachment efficiency and fraction of charge heterogeneity was found.8,15,16,27,28 It is important to note that the patchwise model breaks down when the size of the collector heterogeneous patches becomes comparable with the size of the colloids.22,24 The interactions between polystyrene latex particles (2.1 μm) and macroscopic-scale charged heterogeneous patches (5 μm) that are comparable in size were studied using a radial stagnation point flow (RSPF) system with glass substrates modified with positively charged amino-silanized strips.22 A “hydrodynamic bump” effect caused by alternating attractive and repulsive electrostatic interactions was proposed to explain a significant overprediction of the patchwise geochemical heterogeneity model that assumes linear correlation between attachment efficiency and the fraction of charge heterogeneity.22 In contrast, simulation results of an Eulerian model (convection-diffusion−migration equation) for 1 μm-sized colloid attachment in a RSPF system with circular alternating charged strips (10 μm)29 or to spherical collectors with 2671

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

alternating charged strips (5 μm)24 found colloid attachment two times higher compared to the prediction by patchwise geochemical heterogeneity model. Thus, this contradiction should be evaluated further through additional well controlled experiments. A better mathematical model is required to describe and predict oocyst transport in geochemical charged heterogeneous porous media. We fabricated silica micromodels containing collectors coated with 0, 10, 20, 50, and 100% Fe2O3 to create macroscopic-scale (9.9 or 49.7 μm) patchwise charge heterogeneity. Collectors coated with 0% and 100% Fe2O3 were used to study deposition entirely under unfavorable and favorable conditions, respectively. Collectors coated with 10− 50% Fe2O3 were used to study the influence of alternating unfavorable and favorable conditions. These micromodels were used to study the influence of surface charge heterogeneity on oocyst and carboxylate microsphere (∼5 μm) transport by direct observations, and for the subsequent comparison of results with a patchwise geochemical heterogeneity model.



MATERIALS AND METHODS Patchwise Charged Heterogeneous Micromodel Fabrication. Materials and reagents used in this study are listed in the Supporting Information (SI). The micromodel with/ without Fe2O3 coating (ca. 100 nm thickness) was fabricated following a photolithography procedure7 as described in the SI. Each micromodel (Figure 1A) includes 1838 uniformly distributed cylindrical collectors with a diameter of 190 μm, a pore-body of 130 μm, a pore-throat of 36 μm, and a porosity of 0.45 (Figure 1B). The height of the collector in different micromodels ranges from 22 to 28 μm. For the patchwise charged heterogeneous collectors, Fe2O3 bands covered the wall of the cylindrical collectors from top to bottom and an area adjacent to the collectors (Figure 1B and and SI Figure 1SG). The width of the Fe2O3 band was 9.9 μm, 9.9 μm, and 49.7 μm for collectors coated with 10, 20, and 50% Fe2O3, respectively (Figure 2SB, 2SC, and 2SD). The coating was characterized by Raman spectroscopy, scanning electron microscopy (SEM), and energy dispersive X-ray spectroscopy (EDS), as described in the SI. Microspheres Preparation and C. parvum Oocyst Preparation and Characterization. Fluoresbrite YG Carboxylate Microspheres (4.5 μm in diameter) were purchased from Polysciences, Inc. C. parvum oocysts (4−5 μm in diameter) were propagated using an infected calf and purified as described in the SI. The microspheres and oocysts were washed twice with deionized water followed by centrifugation (17000 × g for 2 min) before being resuspended in the experimental electrolyte. The oocysts were characterized with Fourier transform infrared spectroscopic (FT-IR) imaging as described in the SI. Surface Potential Measurement and DLVO Energy Profile. A Zetasizer Nano analyzer (Malvern Instruments) was used to measure the electrophoretic mobilities of oocysts, microspheres, pulverized silica, and pulverized Fe2O3 particles suspended in 1 mM NaCl solution over a pH range from 3 to 10. Preparation of pulverized silica particles was described in a previous publication.7 Pulverized Fe2O3 particles were ground from pieces of an Fe2O3 source using an agate mortar. The electrophoretic mobilities were converted to zeta potential using the Smoluchowski equation. DLVO energy profiles at 1 mM ionic strength at pH 4.4, 5.8, and 8.1 for oocyst-silica surface, oocyst-Fe2O3 surface, micro-

Figure 2. (A) A 3-dimensional reconstructed confocal microscopy image of attached microspheres on a silica collector coated with 20% Fe2O3. The geometry outline of the collector was drawn to illustrate the silica (transparent) and Fe2O3 (shadow) surfaces. (B) A side view of the confocal microscopy image of the attached microspheres on the collector wall. (C) The distribution of microsphere fluorescent signals in the collector depth. A micromodel depth of 0 was the heterogeneous silica/Fe2O3 surface and a depth of 24 was the pyrex glass. The green spheres illustrate the size of the microspheres. Experimental solution chemistry: 1 mM NaCl, pH 7.1 ± 0.1, buffered with 0.05 mM NaHCO3.

sphere-silica surface, and microsphere-Fe2O3 surface were calculated from the electrostatic interaction model by Hogg et al.30 and the retarded van der Waals interaction model by Gregory.31 The equations and parameters used in our calculation are reported in previous publications.5,7 The calculation of the Hamaker constants is described in the SI. The Hamaker constants of oocyst−water−silica, oocyst− water−Fe2O3, microsphere−water−silica, and microsphere− water−Fe2O3 systems were 1.2 × 10−21, 2.2 × 10−21, 2.2 × 10−21, and 4.0 × 10−21 J, respectively (SI Table 1S). An extended DLVO model (XDLVO),32 containing the acid−base interaction term, was employed, but the exponential function of this acid−base term increased drastically for separation distance close to the cutoff distance of 0.158 nm. For oocysts, the energy barrier calculated by the XDLVO for separation distance below 5 nm was thousands of kT even at the highest ionic strength where the maximum attachment efficiency was observed in a previous study.6 Micromodel Experiments. Transport experiments were conducted in micromodels coated with 0, 10, 20, 50, or 100% Fe2O3 in 1 mM NaCl solution at pH 4.4, 1 mM NaCl solution at pH 5.8 ± 0.1, and 1 mM NaHCO3 solution at pH 8.1 ± 0.1. Concentrations in the range from 0.5 × 106 to 1.5 × 106 particles/mL of oocysts and microspheres were selected as sufficient to observe colloids attached to the micromodel collectors. A hemacytometer (INCYTO, 22-600 series) was used to quantify colloid concentration. A lower concentration was used in favorable conditions and a higher concentration was used in unfavorable conditions to achieve a detectable number of oocysts attached to the collectors while avoiding 2672

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

inverse Debye length at 1 mM ionic strength (κ = 1.04 × 108 m−1) and the smallest patch size (R = 9.9 × 10−5 m) resulted in a smallest transition of κR (1 × 104). The overall partial attachment efficiency on the charged heterogeneous surface can be described as follows:

particle aggregation. The ionic strength and the colloid concentration used in this study were much lower than the conditions at which colloid concentration may influence attachment.33 The details of micromodel experiment setup (SI Figure 3S) and calculation of colloid average removal efficiencies (η) and attachment efficiencies (α) are described in the SI. Briefly, the micromodel was saturated with colloid free solution. Then a solution with a predetermined colloid concentration was pumped into the micromodel at a constant volumetric flow rate, corresponding to an average linear velocity in the pore network of 1.86 mm/s. For microsphere transport, the microspheres attached on collectors were directly observed using a fluorescent microscope (Leica, DMI5000 M) through a 20 × objective and a FITC filter. Fluorescent images were recorded using a camera (Qimaging Retiga 2000R Fast 1394) and analyzed with Image Pro 7.0 software. For oocyst transport study, after a transport experiment was completed, Crypt-aGlo(Waterborne, Inc.), a fluorescent antibody that specifically binds with oocyst surface protein, was pumped through the micromodel in the dark for 20 min. Images of oocyst attachment were recorded with fluorescent microscopy. Thus current experimental setup did not allow us to determine the attachment rate as a function of time. Oocysts or microspheres that attached to the 1838 collectors were determined by direct counting. Colloids attached to silica or Fe2O3 surfaces were counted separately to determine the average removal efficiency (η) for only Fe2O3 patches (ηFe2O3), only silica patches (ηsilica), and for the entire collectors (ηtotal). The values of η were calculated as the ratios of the number of colloids attached to the patches or collectors over the number of colloids approaching the patches or collectors.34 The average attachment efficiencies (α) of oocysts at pH 5.8 and 8.1 were determined by the ratio between the average single collector removal efficiencies (ηtotal on collectors coated with 0, 10, 20, 50, or 100% Fe2O3) and the maximum average single collector removal efficiencies for favorable conditions (ηtotal on 100% Fe2O3 surface) at the same pH. The average attachment efficiencies (α) of microspheres at pH 4.4 and 5.8 were determined similarly. However, due to the presence of the high energy barrier for the microsphere attachment onto Fe2O3 surfaces at pH 8.1, ηtotal of microspheres onto 100% Fe2O3 surface at pH 5.8 was used as the maximum average single collector removal efficiency to calculate attachment efficiency (α) for microspheres at pH 8.1. This is further discussed in the Results and Discussion section below. A control experiment to study the role of sedimentation in the charged heterogeneous micromodel and the influence of the adjacent Fe2O3 surface at the bottom of the micromodel was conducted using laser scanning confocal microscopy (Zeiss LSM 710 Upright). The vertical distribution of microspheres attached to a charge heterogeneous collector wall at 1 mM NaCl, pH 7.1 ± 0.1, buffered with 0.05 mM NaHCO3, was observed. Details of this experiment are found in the SI. Patchwise Geochemical Heterogeneity Model. The patchwise geochemical heterogeneity model15,35 was based on the assumptions that (1) each of the isolated patches is a homogeneous surface and can be distinguished on the basis of size, and (2) if the transition between two adjacent patches, κR, is greater than 1 (κ is the inverse Debye length and R is the size of the patch), then the interactions at patch boundaries can be neglected. In our charged heterogeneous micromodel, each of the isolated silica and Fe2O3 patches was homogeneous. The

α = λαFe2O3 + (1 − λ)αsilica

(1)

where λ is the fraction of Fe2O3 surface, αFe2O3 is attachment efficiency on Fe2O3 surface and αsilica is attachment efficiency on silica surface.



RESULTS AND DISCUSSION Characterization of the Patchwise Charged Heterogeneous Collector. The SEM image of the charged heterogeneous collectors is shown in Figure 1. As shown in Figure 1B, the silica collector wall was coated with Fe2O3 (lighter area) and the coating area was completely covered with Fe2O3. The Raman spectra of the Fe2O3 layer (SI Figure 4S) are consistent with spectra of hematite reported in a previous publication.36 More than 10 EDS spectra (Figure 1C) were acquired at different locations on the Fe2O3 surface. All spectra revealed the presence of Fe, O, and Si. The X-rays used in EDS can probe about 2 μm in depth;37 therefore, the presence of Si was due to the presence of the silica substrate under the Fe2O3 surface. EDS spectra from silica regions (Figure 1D) showed no presence of Fe2O3. In order to test whether the Fe2O3 coating dissolved during transport experiments, a 1 mM NaNO3 solution at pH 4.4 (the lowest pH of our experiments) was pumped into the micromodel coated with 0 or 50% Fe2O3 for 14 h. Inductive coupled plasma optical emission spectrometry (ICP-OES) was used to measure the free Fe3+ concentration in the effluents. The Fe3+ concentrations in the effluent of micromodels coated with 0 or 50% Fe2O3 were below detection limits (5−10 μg/L) for this method. These results indicated that dissolved Fe2O3 was negligible during the duration of the deposition experiments. Characterization of Oocyst Surface. FT-IR spectra were acquired to characterize oocyst surface functional groups and chemical heterogeneity. Multiple layers of highly concentrated oocysts were placed on a crystal adapter. Two representative spectra of viable oocysts scanned at different locations of the oocyst layers are shown in Figure SI 5S. The FT-IR spectra indicated the presence of amides (1638, 1544, and 1338 cm−1 for amide I, amide II, and amide III, respectively), carboxylate (1400 cm−1 for COO−), phosphate (1238 cm−1 for PO2−), and polysaccharide (1152 and 1078 cm−1 for sugar ring vibration and C−O−C, C−C) functional groups. These functionalities are consistent with the spectra of oocyst surfaces that have been reported previously.38,39 The presence of multifunctional groups on oocyst surface indicated the surface charge heterogeneity. More interestingly, an unknown peak at 1002 cm−1 was shown only in spectra b (SI Figure 5S). This peak was observed in only a few of the many spectra acquired within the 100 × 100 μm. Three samples were scanned and we consistently observed the variation of the peak at 1002 cm−1 at different locations on the oocyst layers, which may also contribute to some degree of oocyst surface chemical heterogeneity. Though current resolution of FT-IR imaging is not sufficient to quantify surface charge heterogeneity on a single oocyst, future improvements in FT-IR imaging resolution 2673

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

will contribute to more robust characterization of surface chemical heterogeneities. Zeta Potential of Colloid and Collector Surfaces. The zeta potentials of colloid and collector surfaces in 1 mM NaCl solution at different pH are shown in SI Figure 6S. Based on the zeta potential data, the isoelectric points of oocysts, microspheres, silica, and Fe2O3 surfaces were determined to be approximately at pH 3.8, 3.2, 4.0, and 6.7, respectively. Commonly reported values for the isoelectric point of SiO2 and synthetic hematite are pH 2.31 - 3.1740 and pH 6.9, respectively.41 Our measurement of the isoelectric point of Fe2O3 surfaces was comparable to the isoelectric points reported in the literature. The measured isoelectric point of SiO2 was slightly higher than that reported in the literature probably due to the difference in colloid size42 or the oxidation method used to create the micromodels. A 3-Dimensional Confocal Imaging for Microsphere Distribution on Charged Heterogeneous Surface. The 3dimensional confocal microscopy imaging results (Figure 2) of microsphere distribution on the vertical wall of a silica collector coated with 20% Fe2O3 in 1 mM NaCl at pH 7.1 show the influence of sedimentation and the adjacent Fe2O3 surface at the bottom of the micromodel. The geometry outline of the collector is shown in Figure 2A. The shadow areas on the collector wall indicate the Fe2O3 coating. The vertical distribution of microspheres attached on the collector wall was determined from the fluorescent signal of 24 images taken every 1 μm from the bottom to the top, as shown in Figure 2B and 2C. The normalized fluorescent signal represented the abundance of microspheres attached on each layer. The sum of the fluorescent signal for the height below 5 μm, that is, 20% of the total depth, was only half of the signal from one microsphere. The fluorescent signal was highest at 8−12 μm from the bottom of the collector. Note that the middle of the collector was at 12 μm from the bottom of the collector. It was concluded that the majority of the microspheres were attached in the middle of the collector wall, and the influence of sedimentation and the adjacent Fe2O3 surface at the bottom of the micromodel was negligible. Oocyst and Carboxylate Microsphere Attachment on Charged Heterogeneous Surface. Results of oocyst and microsphere attachment on charged heterogeneous collectors at different pH are shown in Figure 3 and SI Figure 7S, respectively. The average single collector removal efficiencies (ηtotal) for oocysts and microspheres on the entire ensemble of collectors are shown in SI Figure 8S. The values of DLVO energy barrier between colloid and collector surfaces calculated for different pH values are shown in Table 1. In general, oocyst and microsphere attachment at the same pH increased with the percentage of Fe2O3 on the collector surfaces due to the reduction in the energy barrier for attachment of the colloids onto Fe2O3 surfaces as compared to that of the colloids onto silica surfaces. Oocyst and microsphere attachment on the collectors coated with the same amount of Fe2O3 decreased with increasing pH due to increased energy barrier for attachment of the colloids onto silica surfaces (Table 1). For example, as the energy barrier for the microsphere/silica surface attachment increased from 360 kT at pH 4.4 to 3500 kT at pH 5.8 to 15 000 kT at pH 8.1, the ηtotal for microsphere on collectors coated with 10% Fe2O3 decreased from 3.5 × 10−3 to 2.6 × 10−4 to 3.6 × 10−5 (SI Figure 8S). It can be seen in Figure 3 that on the silica collectors coated with 10% Fe2O3, more than 93% of oocysts were attached onto

Figure 3. Attachment of oocysts on silica collectors coated with 0, 10, and 100% Fe2O3. Experimental solution chemistry: 1 mM NaCl at pH 5.8 or 1 mM NaHCO3 at pH 8.1. Linear velocity = 1.86 mm/s.

Fe2O3 patches at pH 8.1, whereas only 30−60% of oocysts were attached on Fe2O3 patches at pH 5.8. The preference of oocysts attached onto Fe2O3 patches was observed at pH 8.1 due to the larger difference between ηFe2O3 and ηsilica at pH 8.1 (2.7 (±1.3) × 10−2 vs 8.4 (±4.7) × 10−5) than that at pH 5.8 (2.3 (±0.2) × 10−2 vs 3.7 (±1.8) × 10−3) as listed in Table 1. It is also shown in Table 1 and SI Figure 8S that oocyst and microsphere attachment on silica surfaces (ηtotal on entire silica surfaces and ηsilica on silica patches) at pH 4.4 was orders of magnitude greater than that at pH 5.8 and 8.1. Attachment under unfavorable conditions has been observed and has been widely reported and attributed to nanoscale surface chemical/ physical heterogeneity and secondary minimum.5,7,43 In the current study, DLVO energy profiles showed no secondary minimum at ionic strength as low as 1 mM NaCl. Therefore, the attachment under unfavorable conditions probably was attributed to nanoscale surface chemical/physical heterogeneity on the colloid and silica surfaces. The difference of ηtotal or ηsilica at different pH was due to the different energy barriers at different pH. For example, the energy barrier between microsphere and silica surfaces was orders of magnitude lower at pH 4.4 (360 kT) than at pH 8.1 (15 000 kT). In addition, the energy barrier between oocyst and Fe2O3 surfaces was 210 kT at pH 8.1 and there was no energy barrier between oocyst and Fe2O3 surfaces at pH 5.8 (Table 1). However, for collectors coated entirely with Fe2O3, ηtotal for oocysts obtained at pH 8.1 was not significantly different (p = 0.41, t test) with that at pH 5.8 (0.8 (±0.1) × 10−2 vs 0.6 (±0.2) × 10−2). This observation contradicts the estimated higher DLVO energy barrier between oocyst and Fe2O3 2674

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

Table 1. Values of DLVO Energy Barrier, Separation Distance of the Energy Barrier for Oocysts and Microspheres Interacting with Silica or Fe2O3 Surfaces, and Average Single Collector Removal Efficiencies (η) for Oocysts and Microspheres on Entirely Fe2O3 Surface, Fe2O3 Patches, Entirely Silica Surface, or Silica Patches of Charged Heterogeneous Collectorsa ηtotal on entirely Fe2O3 surface pH oocyst oocyst microsphere microsphere microsphere

5.8 8.1 4.4 5.8 8.1

ΦT

H

± 0.1 ± 0.1

210

9

± 0.1 ± 0.1

4200

5

ηFe2O3 on Fe2O3 patches

λ=1 0.6(±0.2) 0.8(±0.1) 0.4(±0.0) 3.2(±1.4) 1.1(±0.2)

λ = 0.1

× × × × ×

−2

10 10−2 10−2 10−3 10−3

2.3(±0.2) 2.7(±1.3) 1.5(±0.2) 1.1(±0.4) 0.2(±0.1)

× × × × ×

10 10−2 10−2 10−3 10−3

ηtotal on entirely silica surface oocyst oocyst microsphere microsphere microsphere

pH

ΦT

H

5.8±0.1 8.1 ± 0.1 4.4 5.8 ± 0.1 8.1 ± 0.1

198 222 360 3500 15000

9 15 7 4 1

× × × × ×

λ = 0.5

1.1(±1.1) × 10−3

1.9(±1.8) × 10−3 0.4(±0.5) × 10−3

ηFe2O3 on silica patches

λ=0 2.3(±0.6) 4.4(±3.8) 2.6(±1.2) 0.3(±0.1) 0.3(±0.2)

λ = 0.2 −2

λ = 0.1 10−3 10−5 10−3 10−4 10−5

3.7(±1.8) 8.4(±4.7) 2.2(±0.3) 1.7(±1.5) 1.9(±1.0)

× × × × ×

10−3 10−5 10−3 10−4 10−5

λ = 0.2

λ = 0.5

3.6(±5.4) × 10−4

8.3(±6.8) × 10−4 1.8(±2.6) × 10−4

ΦT (kT): DLVO interaction energy barrier H (nm): Separation distance of the maximum energy barrier between colloid and the collector surface η: Average single collector removal efficiencies λ: Fraction of Fe2O3 surface.

a

surfaces at pH 8.1 compared to that at pH 5.8. Furthermore, when the energy barrier and separation distance between oocyst and 100% Fe2O3 surfaces at pH 8.1 were similar to those between oocyst and silica surfaces at pH 5.8 (210 kT at 9 nm vs 198 kT at 9 nm), ηtotal for 100% Fe2O3 surface at pH 8.1 was significantly higher (p < 0.05, t test) than ηtotal for entire silica surface at pH 5.8 (0.8 (±0.1) × 10−2 vs 2.3 (±0.6) × 10−3). The difference between experimental observation and DLVO theory could be attributed to surface complexation between oocyst and Fe2O3 surfaces suggested by Gao et al.44 Their results suggested that monodentate complexes between surface carboxylate groups of the oocyst and iron from the mineral surface were the predominant complexes at low pH. At higher pH, however, stronger binuclear bidentate complexes became the predominant mechanism of surface complexation. Table 1 shows that at pH 8.1, above the isoelectric points for all studied surfaces, the energy barriers between microsphere and Fe2O3 surfaces were 4200 kT. Due to the presence of the high energy barrier, ηtotal on 100% Fe2O3 surface at pH 5.8 (3.2 (±1.4) × 10−3) was used as the maximum average single collector removal efficiency (η0) to calculate attachment efficiency (α) of microsphere attachment at pH 8.1. Comparison between Measured and Patchwise Geochemical Heterogeneity Model Predicted Attachment Efficiency. Oocyst and microsphere attachment efficiencies on charged heterogeneous surfaces at different pH were determined from silica micromodels coated with 0, 10, 20, 50, or 100% Fe2O3. The experimental data (symbols) were compared to attachment efficiencies predicted by a patchwise geochemical heterogeneity model (dash lines) as shown in Figure 4. The fraction of Fe2O3 surface (λ) and experimental attachment efficiencies on entirely silica (αsilica) and entirely Fe2O3 (αFe2O3) collectors were used to calculate the theoretical attachment efficiencies on charge heterogeneous collectors using eq 1. The measured overall oocyst attachment efficiency on collectors coated with 10% Fe2O3 at pH 5.8 and 8.1 were 2.1 and 3.4 times higher, respectively, than those predicted by the patchwise geochemical heterogeneity model (Figure 4A). However, the measured overall microsphere attachment efficiencies on charged heterogeneous collectors at pH 5.8 and 8.1, where the DLVO energy barrier between microsphere

Figure 4. Attachment efficiency (α) of oocysts (A) and microspheres (B) on silica collectors coated with 0, 10, 20, 50, and 100% Fe2O3. (C) Illustration of “hydrodynamic bump” effect.22 Experimental solution chemistry: 1 mM NaCl at pH 4.4 (black), 1 mM NaCl at pH 5.8 (red), or 1 mM NaHCO3 at pH 8.1 (blue). Linear velocity = 1.86 mm/s. Shown values are averaged values and standard deviation of at least three measurements.

and silica surfaces was more than 1 order of magnitude greater than that between oocyst and silica surfaces, were 0.3−0.8 times of that predicted by the model (Table 1 and Figure 4B). Interestingly, the measured overall microsphere attachment efficiency on collectors coated with 10% Fe2O3 at pH 4.4, where the DLVO energy barrier between microsphere and silica (360 kT at 7 nm) was comparable with that between oocyst and silica surfaces at pH 5.8 and 8.1 (198 kT at 9 nm and 222 kT at 15 nm, respectively), was 1.2 times higher than that predicted by the model. 2675

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

simulate colloid transport on patchwise charged heterogeneous collectors with a low energy barrier (340 kT at 3 nm). Their model predicted higher colloid attachment on charged heterogeneous collector compared with the patchwise geochemical heterogeneous model. The authors also noticed significantly higher local attachment rates at favorable patches compared to a favorable homogeneous surface due to local colloid concentration accumulating at the leading edge of the favorable patches. In addition, Abudalo et al.8 also reported an under-prediction of the patchwise geochemical heterogeneity model for oocyst attachment in charged heterogeneous column. The difference between colloid attachment on 100% Fe2O3 surface and Fe2O3 patches of patchwise charged heterogeneous collectors implies the importance of periodic electrostatic interactions with charged heterogeneous surfaces. When macroscopic-scale charge heterogeneity within porous media is comparable to colloidal size, it is essential to consider spatial distribution of charged heterogeneous patches in any transport model. Environmental Implications. A small fraction of Fe2O3 on patchwise charged heterogeneous collector significantly enhanced oocyst attachment (Figure 4). The results indicate that with the same amount of positively charged surface, the patchwise charged heterogeneous collectors are expected to remove more pathogens than a mixture of positively and negatively charged collectors. Furthermore, ηFe2O3for oocysts on charged heterogeneous collectors was higher than ηtotal for oocysts on 100% Fe2O3 collectors. The difference between colloid attachment on 100% Fe2O3 surface and Fe2O3 patches of heterogeneous collectors illustrates the importance of periodic electrostatic interactions in colloid transport on charged heterogeneous surface. We recommend considering the spatial distribution of charged heterogeneous patches and colloid-collector surface interaction in a transport model when macroscopic-scale charge heterogeneity is present.

The overprediction of colloid attachment on charged heterogeneous surface by patchwise geochemical heterogeneous model was attributed to significantly greater (p = 0.05, t test) attachment of microspheres on collectors coated entirely with Fe2O3 than attachment on Fe2O3 patches. For example, ηtotal obtained for microsphere attachment on collectors coated entirely with Fe2O3 were 3.2 (±1.4) × 10−3 at pH 5.8 and 1.1 (±0.2) × 10−3 at pH 8.1. However, ηFe2O3 for the microsphere attachment on collectors coated with 10% Fe2O3 were 1.1 (±0.4) × 10−3 at pH 5.8 and 0.2 (±0.1) × 10−3 at pH 8.1. The overprediction of colloid attachment on charged heterogeneous surface by patchwise geochemical heterogeneous model has been observed previously22 and was explained by the “hydrodynamic bump” effect as illustrated in Figure 4C. The hydrodynamic bump effect was related to the periodic favorable and unfavorable interactions between the colloids and the alternating charged heterogeneous patches on the collector surface. Elimelech et al.22 suggested that the repulsive interaction between the colloids and the unfavorable patches repelled the colloids against the collector surface and prevented the colloids from approaching the downstream Fe2O3 patches. As the colloids moved around the collector surface with the solution, the probability of colloid attachment on downstream favorable patches was reduced. We suggest that the “hydrodynamic bump” effect also influences colloid transport in our micromodel with patchwise charged heterogeneous collectors, where there were significantly different unfavorable interactions. For example, when the energy barriers between the microsphere and the silica or Fe2O3 surface at pH 8.1 were 15 000 kT and 4200 kT, respectively, we expect that the presence of a very high energy barrier between the microsphere and silica patches would prevent the colloids from approaching the downstream Fe2O3 patches. As a result, ηtotal obtained for microspheres and collectors coated entirely with Fe2O3 should be higher than ηFe2O3 for the microsphere attachment on Fe2O3 patches of collectors partially coated with Fe2O3. Because the patchwise model ignores the spatial distribution of charged heterogeneous patches, the model overpredicts the attachment on Fe2O3 patches. When the DLVO energy barrier was small, the “hydrodynamic bump” effect was diminished. For example, energy barriers of less than 360 kT were observed for microsphere and silica surfaces at pH 4.4, and for oocyst and silica surfaces at pH 5.8 and 8.1. As shown in Table 1, at pH 5.8 and 8.1 for oocysts and pH 4.4 for microspheres, there was a significantly less (p = 0.05, t test) ηtotal obtained for collectors coated entirely with Fe2O3 than ηFe2O3 for the colloid attachment on collectors coated with 10% Fe2O3 (0.6 (±0.2) × 10−2 vs 2.3 (±0.2) × 10−2 for oocysts at pH 5.8, 0.8 (±0.1) × 10−2 vs 2.7 (±1.3) × 10−2 for oocysts at pH 8.1, and 0.4 (±0.0) × 10−2 vs 1.5 (±0.2) × 10−2 for microspheres at pH 4.4). The increased attachment on Fe2O3 patches compared to 100% Fe2O3 surface led to an under-prediction of attachment by the patchwise model. Bradford and Torkzaban23 suggested that colloids immobilized by secondary minimum on the collector surface would roll and transfer to the low velocity regions and attach to the downstream favorable regions. However, we did not observe microspheres or oocysts rolling on the collector surfaces. For the collectors we tracked during the experiment, those oocysts attached to the backward portion of the collectors as shown in Figure 3 were directly transferred from the solution onto the collector surface. Chatterjee et al.24 used an Eulerian model to



ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in the text. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] . Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by NSF Career Grant No. 0954501. Part of the research was performed at the Environmental Molecular Sciences Laboratory (EMSL), a national scientific user facility sponsored by the DOE, Office of Biological and Environmental Research located at the Pacific Northwest National Laboratory.



REFERENCES

(1) Corso, P. S.; Kramer, M. H.; Blair, K. A.; Addiss, D. G.; Davis, J. P.; Haddix, A. C. Cost of illness in the 1993 waterborne Cryptosporidium outbreak, Milwaukee, Wisconsin. Emerging Infect. Dis. 2003, 9 (4), 426−431.

2676

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

(2) Fong, T. T.; Mansfield, L. S.; Wilson, D. L.; Schwab, D. J.; Molloy, S. L.; Rose, J. B. Massive microbiological groundwater contamination associated with a waterborne outbreak in Lake Erie, South Bass Island, Ohio. Environ. Health Perspect. 2007, 115 (6), 856− 864. (3) Gallay, A.; De Valk, H.; Cournot, M.; Ladeuil, B.; Hemery, C.; Castor, C.; Bon, F.; Megraud, F.; Le Cann, P.; Desenclos, J. C. A large multi-pathogen waterborne community outbreak linked to faecal contamination of a groundwater system, France, 2000. Clin. Microbiol. Infect. 2006, 12 (6), 561−570. (4) Agency, U. S. E. P., National Primary Drinking Water Regulations. In 2009; Vol. EPA 816-F-09−0004. (5) Liu, Y.; Janjaroen, D.; Kuhlenschmidt, M. S.; Kuhlenschmidt, T. B.; Nguyen, T. H. Deposition of Cryptosporidium parvum oocysts on natural organic matter surfaces: Microscopic evidence for secondary minimum deposition in a radial stagnation point flow cell. Langmuir 2009, 25 (3), 1594−1605. (6) Liu, Y.; Kuhlenschmidt, M. S.; Kuhlenschmidt, T. B.; Nguyen, T. H. Composition and conformation of Cryptosporidium parvum oocyst wall surface macromolecules and their effect on adhesion kinetics of oocysts on quartz surface. Biomacromolecules 2010, 11 (8), 2109− 2115. (7) Liu, Y.; Zhang, C.; Hilpert, M.; Kuhlenschmidt, M. S.; Kuhlenschmidt, T. B.; Nguyen, T. H. Transport of Cryptosporidium parvum oocysts in a silicon micromodel. Environ. Sci. Technol. 2012, 46 (3), 1471−1479. (8) Abudalo, R. A.; Bogatsu, Y. G.; Ryan, J. N.; Harvey, R. W.; Metge, D. W.; Elimelech, M. Effect of ferric oxyhydroxide grain coatings on the transport of bacteriophage PRD1 and Cryptosporidium parvum oocysts in saturated porous media. Environ. Sci. Technol. 2005, 39 (17), 6412−6419. (9) Johnson, J. K.; Schmidt, J.; Gelberg, H. B.; Kuhlenschmidt, M. S. Microbial adhesion of Cryptosporidium parvum sporozoites: Purification of an inhibitory lipid from bovine mucosa. J. Parasitol. 2004, 90 (5), 980−990. (10) Hijnen, W. A. M.; Brouwer-Hanzens, A. J.; Charles, K. J.; Medema, G. J. Transport of MS2 phage, Escherichia coli, Clostridium perfringens, Cryptosporidium parvum and Giardia intestinalis in a gravel and a sandy soil. Environ. Sci. Technol. 2005, 39 (20), 7860−7868. (11) Tufenkji, N.; Dixon, D. R.; Considine, R.; Drummond, C. J. Multi-scale Cryptosporidium/sand interactions in water treatment. Water Res. 2006, 40 (18), 3315−3331. (12) Tufenkji, N.; Elimelech, M. Spatial distributions of Cryptosporidium oocysts in porous media: Evidence for dual mode deposition. Environ. Sci. Technol. 2005, 39 (10), 3620−3629. (13) Tufenkji, N.; Miller, G. F.; Ryan, J. N.; Harvey, R. W.; Elimelech, M. Transport of Cryptosporidium oocysts in porous media: Role of straining and physicochemical filtration. Environ. Sci. Technol. 2004, 38 (22), 5932−5938. (14) Ryan, J. N.; Gschwend, P. M. Effect of iron diagenesis on the transport of colloidal clay in an unconfined sand aquifer. Geochim. Cosmochim. Acta 1992, 56 (4), 1507−1521. (15) Song, L.; Johnson, P. R.; Elimelech, M. Kinetics of colloid deposition onto heterogeneously charged surfaces in porous media. Environ. Sci. Technol. 1994, 28 (6), 1164−1171. (16) 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. (17) Coston, J. A.; Fuller, C. C.; Davis, J. A. Pb2+ and Zn2+ adsorption by a natural aluminum- and iron-bearing surface coating on an aquifer sand. Geochim. Cosmochim. Acta 1995, 59 (17), 3535−3547. (18) Nordstrom, D. K.; Alpers, C. N.; Coston, J. A.; Taylor, H. E.; McCleskey, R. B.; Ball, J. W.; Ogle, S.; Cotsifas, J. S.; Davis, J. A. Geochemistry, Toxicity, And Sorption Properties of Contaminated Sediments and Pore Waters from Two Reservoirs Receiving Acid Mine Drainage; United States Geological Survey Staff -- Published Research: March 8−12, 1999.

(19) Fontes, M. P. F. Iron oxide-clay mineral association in Brazilian Oxisols: A magnetic separation study. Clays Clay Miner. 1992, 40 (2), 175−179. (20) Cortis, A.; Harter, T.; Hou, L.; Atwill, E. R.; Packman, A. I.; Green, P. G. Transport of Cryptosporidium parvum in porous media: Long-term elution experiments and continuous time random walk filtration modeling. Water Resour. Res. 2006, 42 (12), W12S13. (21) Song, L.; Elimelech, M. Calculation of particle deposition rate under unfavourable particle-surface interactions. J. Chem. Soc., Faraday Trans. 1993, 89 (18), 3443−3452. (22) 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. (23) Bradford, S. A.; Torkzaban, S. Colloid adhesive parameters for chemically heterogeneous porous media. Langmuir 2012, 28 (38), 13643−13651. (24) Chatterjee, R.; Mitra, S. K.; Bhattacharjee, S. Particle deposition onto janus and patchy spherical collectors. Langmuir 2011, 27 (14), 8787−8797. (25) Duffadar, R. D.; Davis, J. M. Interaction of micrometer-scale particles with nanotextured surfaces in shear flow. J. Colloid Interface Sci. 2007, 308 (1), 20−29. (26) Duffadar, R. D.; Davis, J. M. Dynamic adhesion behavior of micrometer-scale particles flowing over patchy surfaces with nanoscale electrostatic heterogeneity. J. Colloid Interface Sci. 2008, 326 (1), 18− 27. (27) 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. (28) Ryan, J. N.; Elimelech, M.; Ard, R. A.; Harvey, R. W.; Johnson, P. R. Bacteriophage PRD1 and silica colloid transport and recovery in an iron oxide-coated sand aquifer. Environ. Sci. Technol. 1999, 33 (1), 63−73. (29) Nazemifard, N.; Masliyah, J. H.; Bhattacharjee, S. Particle deposition onto charge heterogeneous surfaces: Convection−diffusion−migration model. Langmuir 2006, 22 (24), 9879−9893. (30) Hogg, R.; Healy, T. W.; Fuerstenau, D. W. Mutual coagulation of colloidal dispersions. Trans. Faraday Soc. 1965, 62, 1638−1651. (31) Gregory, J. Approximate expressions for retarded van der waals interaction. J. Colloid Interface Sci. 1981, 83 (1), 138−145. (32) van Oss, C. J. Acidbase interfacial interactions in aqueous media. Colloid Surf., A 1993, 78, 1−49. (33) Haznedaroglu, B. Z.; Kim, H. N.; Bradford, S. A.; Walker, S. L. Relative transport behavior of Escherichia coli O157:H7 and Salmonella enterica serovar pullorum in packed bed column systems: Influence of solution chemistry and cell concentration. Environ. Sci. Technol. 2009, 43 (6), 1838−1844. (34) Tufenkji, N.; Elimelech, M. Correlation equation for predicting single-collector efficiency in physicochemical filtration in saturated porous media. Environ. Sci. Technol. 2004, 38 (2), 529−536. (35) Gibb, A. W. M.; Koopal, L. K. Electrochemistry of a model for patchwise heterogeneous surfaces: The rutile-hematite system. J. Colloid Interface Sci. 1990, 134 (1), 122−138. (36) Jubb, A. M.; Allen, H. C. Vibrational spectroscopic characterization of hematite, maghemite, and magnetite thin films produced by vapor deposition. ACS Appl. Mater. Interfaces 2010, 2 (10), 2804− 2812. (37) Arnoult, C.; Di Martino, J.; Khouchaf, L.; Toniazzo, V.; Ruch, D. Pressure and scattering regime influence on the EDS profile resolution at a composite interface in environmental SEM. Micron 2011, 42 (8), 877−883. (38) Gao, X.; Chorover, J. In-situ monitoring of Cryptosporidium parvum oocyst surface adhesion using ATR-FTIR spectroscopy. Colloids Surf., B 2009, 71 (2), 169−176. (39) Gao, X.; Chorover, J. Amphiphile disruption of pathogen attachment at the hematite (α-Fe2O3)−water interface. Langmuir 2011, 27 (10), 5936−5943. 2677

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678

Environmental Science & Technology

Article

(40) Kosmulski, M. Chemical Properties of Material Surfaces; CRC Press: Boca Raton, FL, 2001. (41) Gutierrez, L.; Li, X.; Wang, J.; Nangmenyi, G.; Economy, J.; Kuhlenschmidt, T. B.; Kuhlenschmidt, M. S.; Nguyen, T. H. Adsorption of rotavirus and bacteriophage MS2 using glass fiber coated with hematite nanoparticles. Water Res. 2009, 43 (20), 5198− 5208. (42) Suttiponparnit, K.; Jiang, J.; Sahu, M.; Suvachittanont, S.; Charinpanitkul, T.; Biswas, P. Role of surface area, primary particle size, and crystal phase on titanium dioxide nanoparticle dispersion properties. Nanoscale Res. Lett. 2011, 6 (1), 27. (43) Bradford, S. A.; Torkzaban, S.; Kim, H.; Simunek, J. Modeling colloid and microorganism transport and release with transients in solution ionic strength. Water Resour. Res. 2012, 48. (44) Gao, X.; Metge, D. W.; Ray, C.; Harvey, R. W.; Chorover, J. Surface complexation of carboxylate adheres Cryptosporidium parvum Oocysts to the hematite−water interface. Environ. Sci. Technol. 2009, 43 (19), 7423−7429.

2678

dx.doi.org/10.1021/es304075j | Environ. Sci. Technol. 2013, 47, 2670−2678