Environ. Sci. Technol. 2002, 36, 184-189
Effect of Molecular Scale Roughness of Glass Beads on Colloidal and Bacterial Deposition KARL SHELLENBERGER AND BRUCE E. LOGAN* Department of Civil and Environmental Engineering, The Pennsylvania State University, University Park, Pennsylvania 16802
Molecular-scale surface roughness and charge heterogeneity have been hypothesized as factors that can affect the deposition rates of colloids during their transport in porous media. To test their relative importance, a single batch of cleaned glass beads was divided in half and chemically treated with acid or base to alter surface roughness. Analysis of the topography of 20 glass beads with an atomic force microscope (AFM) indicated that the chromic acid-treated (rough) beads had a root-meansquare roughness of 38.1 ( 3.9 nm, while the sodium hydroxide-treated (smooth) beads had root-mean-square surface roughness of 15.0 ( 1.9 nm. AFM force volume imaging of glass bead surfaces did not reveal surface charge heterogeneity. Filtration experiments with inorganic colloids (latex microspheres, 1 µm diameter) consistently demonstrated that there was a greater retention of latex microspheres on rough than smooth glass beads suspended in either low (10-5 M) or higher (10-1 M) ionic strength (IS) solutions. Collision efficiencies for rough beads were 3050% larger than for smooth beads. Collision efficiencies of bacteria using rough glass beads were also equal to or greater than those measured for smooth beads. In experiments with the perchlorate-reducing bacterial isolate KJ, collision efficiencies were significantly greater on rough rather than smooth beads for two different ionic strength solutions (IS ) 0.05 or 1 M). In another case (IS ) 0.1 M) for KJ, and in filtration experiments with E. coli, collision efficiencies were not significantly different between the rough and smooth beads. We hypothesize that the consistently greater deposition rates of microspheres, but not bacteria, on rough rather than smooth beads are due in part to the presence of polymers on the surfaces of bacteria.
Introduction An understanding of the chemical and physical factors influencing bacterial adhesion can lead to greater control of the fate and transport of bacteria in subsurface environments. Reducing bacterial adhesion can increase the success of bioremediation of contaminated soils and aquifers. To widely disperse pollutant-degrading bacteria via bioaugmentation, the ionic strength of the water can be lowered (1-3), or nonionic surfactants can be added to the cell suspension (4, 5), or low-adhesion isolates can be obtained from monoclonal * Corresponding author phone: (814)863-7908; fax: (814)863-7304; e-mail:
[email protected]. 184
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 2, 2002
populations (6-8). Alternatively, increasing the adhesion of bacteria can be used to limit the migration of pathogenic bacteria in groundwater aquifers. The chemical properties and morphology of a surface are important factors for understanding bacterial adhesion rates. Metal oxides and clays can increase deposition rates of negatively charged particles such as bacteria, although natural organic matter can partially ameliorate these effects (9-11). Even on cleaned and highly spherical glass beads, however, it has been speculated that charge heterogeneities produce localized, highly favorable sites for attachment (12-15). There is experimental evidence of slowly increasing breakthrough during injection of colloids into a packed medium that support the hypothesis of favorable sites (16), and computer models designed to account for surface heterogeneity support this mechanism (14). Cleaning glass bead and quartz particles can alter deposition rates (5, 16) providing indirect evidence of different types of deposition sites. However, favorable deposition sites for bacteria have never been directly measured on cleaned glass beads. While strong acids can be used to clean glass and quartz surfaces they also can etch the surface and increase surface roughness. The importance of molecular-scale roughness has never been systematically investigated in laboratory filtration experiments, although it is generally accepted that particle irregularity is an important factor in colloid transport (12, 17, 18). For example, under identical colloid and solution chemistry conditions greater collision efficiencies of colloids are produced with irregularly shaped quartz particles than with regularly shaped spherical glass beads (5), and relatively rough composite polyamide reverse osmosis membranes (protrusions of several hundred nm) foul faster than smoother cellulose acetate membranes (protrusions of only a few nm) (17, 19). Models of surface roughness have shown that DLVO interaction energy profiles for rough surfaces (protrusions on the order of 1-5 nm) deviate significantly from those derived assuming smooth surfaces (18). To directly measure whether surface roughness and charge heterogeneity could affect colloid transport in porous media, we measured the overall retention of three different particles (two bacteria and latex microspheres) in minicolumns experiments using rough and smooth glass beads. An atomic force microscope (AFM) was used to probe surface topography (roughness) and charge distribution of glass bead surfaces at molecular scales. The dimensions of the end of some AFM tips (as low as 2-10 nm (20)) make it possible to scan extremely small surface protrusions and adsorbed materials such as humic acids on surfaces (21-23). Measurements of the forces between the tip and the surface obey DLVO model predictions of colloid-surface interactions with respect to pH and ionic strength if the pyramid-shaped tip is defined as a spherical colloid in the range of 100-400 nm (24). Therefore, we used the AFM to probe the surface of glass beads to determine if we could measure charge heterogeneities across the surface of the glass beads.
Methods Glass Beads. A single batch of soda-lime glass beads 500750 µm in diameter (Polysciences, Inc., Warminster, PA) was cleaned with 1 N H2SO4 (25), split into two batches, and further treated to alter surface roughness. To increase roughness, glass beads were successively soaked for 24 h and rinsed with deionized water (Milli-Q, Millipore Corp., Bedford, MA) in the following: 36.5-38% HCl, 10% H2CrO4, 36.5-38% HCl, and dried (16). To make beads smoother they were soaked in 12.5 M NaOH for 30 min and then rinsed with 10.1021/es015515k CCC: $22.00
2002 American Chemical Society Published on Web 12/13/2001
ultrapure water. Electrophoretic mobility measurements of the beads were not made because the beads would need to be crushed to reduce particle sizes, altering the exposed surface area. Analysis of Surface Roughness. A Digital Instruments (DI; Santa Barbera, CA) atomic force microscope mounted on an inverted Olympus IX70 optical microscope (Bioscope; Nanoscope III software, versions 4.23 and 4.32) was used to measure surface roughness and probe surface charge distribution. Glass beads were bonded to glass microscope slides with epoxy (Devcon 2-Ton Crystal Clear Epoxy) and surface roughness measured with contact-mode imaging in air using standard silicon nitride tips (type DNP; Digital Instruments) with a radius of curvature of 20 to 60 nm (20). Tip deflection data was converted to root-mean-square height, RMS, using the DI software, calculated as
RMS )
[(zi - zavg)2]1/2 N
(1)
where zi and zavg are the individual height values and the average (datum) height, respectively, and N is the number of samples. Because the surface of a glass bead is curved, height changes produced by this curvature were eliminated before calculating the RMS. To remove the surface curvature, each individual AFM trace over a surface was adjusted by fitting the arc of the data based on a circle of radius R and then subtracting the height due to the arc of the circle as
z ) z0 + [R2 - (x - x0)2]
(2)
where x and z are the x- and z-coordinates of a point on the arc of a circle, and x0 and z0 are the coordinates of the center of the circle. Adjusting the height data produced a flat trace of surface deflection data that was then quantified in terms of the RMS heights using eq 1. Surface Charge Heterogeneity. The distribution of charge on the surface of a glass bead in Milli-Q water was examined by measuring the interaction force between the negatively charged AFM tip and the glass surface using force volume (FV) imaging (DI software version 4.23). To create a FV image, a section of the surface was imaged in contact mode and divided into an F×F matrix, and a force curve was obtained in the center of each grid box. A force curve is a plot of the location of the tip, while the tip is lowered to the surface as a function of the distance the cantilever is lowered. Attraction or repulsion between the tip and the surface is demonstrated by the tip being pulled down to, or pushed away from, the surface, resulting in a deviation from a linear relationship between the tip position with the cantilever displacement. Once the tip hits the surface, the cantilever continues to be lowered for a short period producing the “constant compliance” region where the tip does not move but the laser is deflected due to the bending of the cantilever. The tip is then retracted, and the cantilever is moved to the next point in the sampling matrix. The location of the surface in a force plot is obtained as the intersection between two straight lines generated from the constant compliance region and the approach region (zero interaction force) in a force plot (26). Tip deflection data were converted to forces (nN) using Hooke’s Law, or F ) -kx, where k the spring constant (N/m), and x is the deflection (nm). The spring constant was calculated for each cantilever using DI software (Nanoscope III v. 4.3.2) based on the Cleveland method (27). All force curves were measured in deionized water (Milli-Q, New Bedford, MA). The origin was manually reset for some force curve data (as indicated) by exporting data to a specialized graphics program and identifying the intersection of the constant-compliance and zero-force lines (26, 28).
When all force curves are combined, a FV image is produced that depicts the location (deflection) of the tip at a given height from the surface (29). When the surface is smooth, these heights can be used to measure the relative attraction or repulsion between the tip and the surface based on the tip location relative to the surface. Using the FV image, we hypothesized that it would be possible to find areas of a surface that held greater attraction or repulsion of the negatively charged AFM tip than other areas, providing a map of more or less favorable deposition sites for negatively charged colloids and bacteria. Bacteria and Microspheres. Two strains of bacteria were used: Escherichia coli HCB137 (30) and Dechlorosoma sp. KJ (31, 32). HCB137 is a nonmotile mutant of a well characterized chemotactic strain that has been used to understand the motion of bacteria near surfaces (30). HCB137 cells are 3 µm × 1 µm (equivalent spherical diameter of 1.8 µm). Cell suspensions were grown using 30 g/L of tryptic soy broth (TSB, Sigma Chemical, Co., St. Louis, MO) and harvested during the mid- to late-exponential growth phase. KJ is a newly isolated Gram negative bacterium that respires using perchlorate (31) and has been examined for bioremediation of perchlorate-contaminated waters (32). KJ cells are 2 µm × 1 µm (equivalent diameter of 1.5 µm). Cell suspensions were grown aerobically on acetate (1 g/L) in a mineral salts medium and harvested during the late exponential growth phase (33). Bacteria retained in column experiments described below were counted using a radiolabeling technique (4, 34, 35). 3H-leucine ((1 miCu/mL); 40 µL/100 mL; ICN Biomedicals) was added to a washed cell suspension of E. coli in 0.1 M KCl and incubated at room temperature for 6 h. For KJ suspensions, radiolabeled leucine was added directly to the growth medium, and the cells were allowed to take up the label during growth. Fluorescent carboxylated latex microspheres 1.0 µm in diameter (Polysciences) were washed three times by centrifugation using Milli-Q water. To correlate radiolabel measurements to actual cell numbers, bacteria were stained with acridine orange, and both bacteria and microspheres were counted on prestained black filters (polycarbonate, 0.4 µm pore diameter; Osmonics Corp.) with a microscope (BH2; Olympus Corp.) equipped for fluorescence under blue light at 1000×. Column Studies. Minicolumn experiments with three different particles (radiolabeled bacteria or fluorescent microspheres) were conducted as described previously (34). Briefly, columns made from the barrels of 3 cm3 syringes (0.8 cm inner diameter) were packed with 1.5 g of glass beads supported by two GF/C filters (1.2 µm nominal pore size; Whatman) and attached to a vacuum box. Each filtration experiment (triplicate columns) consisted of the following: 10 mL of rinse solution, 5 mL of sample containing either radiolabeled bacteria or latex microspheres, and 10 mL of rinse solution. The columns were then vacuumed until they were dry. Filtration experiments were conducted with influent concentrations of either 107 cells/mL or 108 microspheres/ mL. Particles were suspended in deionized water having an ionic strength of ∼10-5 M (Milli-Q water) or solutions set at different ionic strengths (0.05 or 0.1 M) using CaCl2. The number of retained particles were counted by extruding the column packing with a syringe plunger and cutting the top 1 cm of the column into three sections. Microspheres were removed from the glass beads by sonication (Bransonic 1210 Ultrasonic Cleaner) for 3 min in a surfactant solution (Tween 20, 0.1%) that was then filtered through black filters for microscopic counting. Direct examination of the glass beads after sonication revealed a small, but insignificant, number of microspheres remained on the beads. Cell suspensions (2 mL) or column packing from a VOL. 36, NO. 2, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
185
FIGURE 2. Traces used to determine average roughness values for the rough and smooth beads.
FIGURE 1. AFM images of glass bead surfaces: (A) smooth bead and (B) rough bead. filtration experiment were combined with 10 mL of scintillation cocktail (CytoScint, ICN Biomedicals) and analyzed by liquid scintillation counting (RackBeta 1217, LKB-Wallac). The retention of unassimilated radiolabel in the column was accounted for by conducting parallel filtration experiments using filtrate from cell suspensions as previously described (5, 34, 35). Filtration theory was used to model particle retention in the packed minicolumns (36). The particle collision efficiency, R, is defined as the fraction of collisions between a particle and a collector (glass bead) that are successful. By measuring the fraction of particles that are retained in the column, FR, the collision efficiency can be calculated (36, 37) as
R)
-2dc ln(1 - FR) 3(1 - θ)ηL
(3)
where dc ) 500 µm is the diameter of the glass beads, θ ) 0.4 the porosity of the medium, η, the collector efficiency, is calculated using the Rajagopalan and Tien model (36), and L is the length of the column (or slice).
Results Glass Beads Roughness. Glass bead surfaces treated with sodium hydroxide were smoother than those soaked in chromic acid as shown in Figure 1 for 30 µm × 30 µm sections of the bead analyzed with contact mode AFM. Duplicate scans of 20 beads (10 rough, 10 smooth) were transformed to remove surface curvature and analyzed for roughness as shown in Figure 2. Based on data from 20 scan lines for each batch of beads, the RMS heights of the sodium hydroxidetreated beads (smooth beads) was 15.0 ( 1.9 nm and the chromic acid-treated beads (rough beads) was 38.1 ( 3.9 nm. Filtration Experiments with Microspheres. There was a greater retention of microspheres on rough than smooth beads suspended in either low (10-5 M) or higher (10-1 M) ionic strength solutions. Collision efficiencies measured in three sections of the column for rough beads under low ionic 186
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 2, 2002
FIGURE 3. Collision efficiencies for latex microspheres measured in minicolumn experiments using rough or smooth beads as a function of solution ionic strength (IS) under clean bed conditions: (A) C0 ) 108 mL-1, IS ) 10-5 M and (B) C0 ) 107 mL-1, IS ) 10-1 M. (Error bars based on (SD of columns run in triplicate.) strength conditions were on average 30-40% larger at low ionic strength and approximately 50% larger at the higher ionic strength than those on smooth beads (Figure 3). The higher retention of microspheres on the rough rather than the smooth beads at 10-5 M was verified in separate experiments by placing glass beads from the column on a microscope slide and counting microspheres on the upper half of the glass bead using fluorescence microscopy (28). Greater retention on rough than smooth beads only occurred if the microspheres did not occupy a large fraction of the surface area or under “clean bed” conditions (38). As shown in Figure 4, there was no difference between the retention of microspheres on rough and smooth glass beads in filtration experiments under conditions (108 mL-1; IS ) 0.1 M) that resulted in a high surface loading (4.3-5.6% surface area, assuming perfect spheres). Others have shown that surface coverages of only 4-5% (microspheres) and 8-9% (bacteria) are sufficient to completely cover or “block” surfaces (39, 40). Collision efficiencies larger than unity were observed in some of these experiments, but others have also found R > 1 (25, 36, 41, 42) indicating that the Rajagopolan and Tien Model filtration equation underpredicted the frequency of collisions. Filtration Experiments with Bacteria. In all filtration experiments with bacteria, the overall average retention of bacteria was higher on rough rather than smooth glass beads (Figure 5). However, these differences were not always
FIGURE 4. Collision efficiencies of latex microspheres measured in minicolumn experiments using rough or smooth beads producing a high surface loading of 4.3 to 5.6% coverage in the top slice of a column. (C0 ) 108 mL-1, IS ) 0.1 M. Error bars based on (SD of columns run in triplicate.)
FIGURE 6. Force volume image (approach curves) of 50 × 50 µm2 surface of a smooth glass bead.
FIGURE 7. Comparison of individual force curves taken from areas of greatest apparent differences, defined as “light” and “dark” regions in Figure 6.
FIGURE 5. Collision efficiencies for bacteria measured in minicolumn experiments using rough or smooth beads at different ionic strengths: (A) KJ and (B) E. coli. (Error bars based on (SD of columns run in triplicate; symbols without error bars are single measurements.) significant in minicolumn tests due to low values of R and variability between replicate columns. At ionic strengths of 0.05 and 1 M, the collision efficiency for KJ suspensions was 40 and 54% larger on the rough beads than those on the smooth beads (Figure 5A). At an ionic strength of 0.1 M, R was 23% larger on average, but the error bars ((SD between triplicate columns) overlapped. Overall, there appeared to be no trend in R with ionic strength. In filtration experiments with E. coli, a trend of increasing retention with surface roughness was again observed, but results were not significant (Figure 5B). A combination of relatively little uptake of the radiolabel and low collision efficiencies (0.02 e R e 0.05) resulted in large variations in R. At solution ionic strengths of 0.05 and 0.1 M bacterial deposition was larger on average for the rough beads than for the smooth beads, but the error bars overlap indicating the results were not significant. At the lowest ionic strength (IS ) 10-5 M) deposition on the rough beads was too low to be measured. Repeated experiments with these bacteria at all three ionic strengths produced similar results of higher average deposition on the rough than smooth beads and low deposition rates in the 10-5 ionic strength experiment. We were unable to reduce the error in deposition measurements for E. coli prior to the exhaustion of our supply of the two batches of glass beads.
Probing Surfaces using Force Volume Imaging. Force volume imaging of glass bead surfaces initially appeared to suggest the presence of surface heterogeneities, but a thorough examination of force curves did not support the presence of charge heterogeneities on these surfaces. A typical FV image of a 50 µm × 50 µm area of a smooth glass bead is shown in Figure 6. This surface is separated into a 64 × 64 grid of force curves for a total of 4096 individual force curves, with each force curve composed of 64 points (the maximum number possible with this AFM software for this number of force curves). Assuming a tip radius of 20-60 nm for this tip (20), this means that 0.21% to 1.85% of the surface in each grid box was sampled for each force curve. The gray scale of each grid point indicates the magnitude of the interaction force between the AFM tip and the surface during the approach of the tip to the surface. Thus, a light color should indicate a small repulsive force and a dark color should indicate a large repulsive force. While the light/dark variations shown at different grid points in Figure 6 apparently support the presence of large differences in tip-surface forces and therefore surface charge heterogeneity, offline analysis of the FV data does not. Individual force curves from three random “light” and “dark” grid points are shown in Figure 7. The location of the surface of the bead can be identified as the intersection of two lines formed from the force measurements taken at far distances from the surface (right side of figure) and after the tip has contacted the surface (far left of figure). After superimposing the six curves as shown in Figure 7, it can be seen that at the resolution of these scans there was no difference in the forces measured between the tip and the glass bead surface. This indicates that the apparent differences in force by the FV VOL. 36, NO. 2, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
187
image shown in Figure 6 is really due to a comparison of the force curves at different points relative to the surface and not differences in interaction forces. Therefore, we were unable to detect any differences in charge distribution using the FV imaging technique.
Discussion Collision efficiencies measured for latex microspheres in filtration experiments were consistently higher for glass beads with an average roughness of 38.1 ( 3.9 nm than for those with an average roughness of 15.0 ( 1.9 nm. The roughness of the glass beads also contributed in some cases to increased deposition of bacteria. Collision efficiencies were significantly larger for strain KJ on rough rather than smooth beads at ionic strengths of 0.05 and 1 M; in three other experiments, there was no significant difference between rough and smooth beads. These results for microspheres and bacteria demonstrate that molecular-scale surface roughness can affect colloid deposition, but that this effect is colloiddependent. Microspheres do not have long polymers on their surfaces and thus interact with the glass beads in a different manner than bacteria. The size of protrusions from glass beads makes them relevant to sizes of polymers on the surface of bacteria and the thickness of the electrostatic repulsive layer. Protrusions from the rough glass beads averaged 38.1 nm but ranged up to 200 nm. These heights can be compared to the repulsive layer thickness characterized by the inverse of the DebyeHuckel parameter (43). At an ionic strength of 10-5 M at 20 °C, the repulsive layer thickness is 43 nm, but it decreases to 1.4 nm for a 0.1 M solution. In order for approaching colloids not to be repelled by the protrusions themselves, we hypothesize that these protrusions on negatively charged glass bead surfaces carry insufficient overall charge to deflect approaching negatively charged microspheres. These nanoscale glass bead protrusions probably penetrate the repulsive layer around the micron-sized colloids and initiate contact. Thus, we observed (Figure 3) consistently greater overall deposition of microspheres on glass surfaces with larger protrusions (rough beads) than on surfaces with smaller protrusions (smooth beads). However, when the colloidal particle was a bacterium, which has polymers on its surface, there was a less clear relationship between deposition and surface roughness. It is well accepted that the overall properties of the colloids, such as electrophoretic mobility and surface hydrophobicity, are important determinants of the probability of colloid adhesion to a glass surface (44). However, these two properties alone are insufficient to predict R, indicating other factors are necessary to explain bacterial adhesion (45). The interaction of the glass bead protrusions with a bacterium is more difficult to understand than with latex microspheres due to the polymers on the surface of the bacterium. E. coli have well characterized surface lipopolysaccharide (LPS) polymers that extend 3.8 to 42.5 nm into solution (46) and therefore are similar in length to the glass bead surface protrusions. AFM measurements suggest that EPS (extracellular polysaccharides) on other bacteria (Pseudomonas putida KT2442, Burkholderia cepacia G4) can span much larger distances of 230-1040 nm (47). AFM measurements also show that these polymers can repel or attract the negatively charged pyramidshaped AFM tip when the tip approaches the bacterial surface (35, 48, 49). This suggests that polymers could increase or decrease the probability of attachment during an initial collision with a nanosized glass bead protrusion. However, if a collision occurs, there can be a strong adhesion force holding the bacterium at the surface. When an AFM tip contacts a bacterial surface there is a large attractive force that holds the tip at the bacterial surface. Although the AFM tip is not identical to glass bead protrusions in size and 188
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 2, 2002
chemical composition, both are negatively charged pyramidshaped surfaces. AFM experiments have shown that once the AFM tip contacts a bacterial surface the force holding the tip on the cell is greater than attractive or repulsive forces measured during the tip approach. Thus, the presence of these polymers may alter the outcome of a bacterium interaction with a rough surface compared to those occurring with a relatively smooth latex microsphere. Charge heterogeneities did not appear to be a factor in bacterial or microsphere attachment to cleaned glass surfaces since localized areas exhibiting differences in attractive or repulsive forces could not be measured using the AFM. Whereas surface charge heterogeneity cannot be eliminated as a factor in bacterial deposition rates to surfaces, if such heterogeneities exist, they must be present at scales smaller than those of the AFM tip dimensions. Given the scale of surface protrusions relative to the minimum detectable dimension for surface heterogeneities, we conclude that surface roughness is the critical factor in determining favorable, versus less favorable, sites of attachment on cleaned glass surfaces. The importance of surface roughness is also likely manifest at larger dimensions as irregular particles are known to produce greater bacterial deposition than on more spherical particles (5). Including molecular scale roughness and particle surface geometrical effects in filtration models should lead to a more accurate prediction of particle transport and deposition rates in porous media.
Acknowledgments This publication was supported by grant CHE-0089156 from the National Science Foundation (NSF) and by ES-04940 from the National Institute of Environmental Health Science, NIEHS. Additional funding for the AFM was provided in part by NSF/IGERT DGE-9972759. The contents of this paper are solely the responsibility of the authors and do not necessarily represent official views of a funding agency. We thank S. Pardi for assisting with microsphere counts, T. Camesano for assistance with AFM measurements, and M. Elimelech and T. Camesano for comments on an earlier manuscript.
Literature Cited (1) Fontes, D. E.; Mills, A. L.; Hornberger, G. M.; Herman, J. S. Appl. Environ. Microbiol. 1991, 57, 2473-2481. (2) Gannon, J. T.; Manilal, V. B.; Alexander, M. Appl. Environ. Microbiol. 1991, 57, 190-193. (3) Jewett, D. G.; Hilbert, T. A.; Logan, B. E.; Arnold, R. G.; Bales, R. C. Water Res. 1995, 29, 1673-1680. (4) Gross, M. J.; Logan, B. E. Appl. Environ. Microbiol. 1995, 61, 1750-1756. (5) Li, Q.; Logan, B. E. Water Res. 1999, 33, 1090-1100. (6) DeFluan, M. F.; Tanzer, A. S.; McAteer, A. L.; Marshall, B.; Levy, S. B. Appl. Environ. Microbiol. 1990, 56, 112-119. (7) DeFlaun, M. F.; Oppenheimer, S. R.; Streger, S.; Condee, C. W.; Fletcher, M. Appl. Environ. Microbiol. 1999, 65, 759-765. (8) Shea, C.; Nunley, J. W.; Williamson, J. C.; Smith-Somerville, H. E. Appl. Environ. Microbiol. 1991, 57, 3107-3113. (9) Mills, A. L.; Herman, J. S.; Hornberger, G. M.; DeJesds, T. H. Appl. Environ. Microbiol. 1994, 60, 3300-3306. (10) Johnson, W. P.; Logan, B. E. Water Res. 1996, 30, 923-931. (11) Kretzschmar, R.; Sticher, H. Environ. Sci. Technol. 1997, 31, 3497-3504. (12) Elimelech, M.; O’Melia, C. R. Environ. Sci. Technol. 1990, 24, 1528-1536. (13) Song, L.; Johnson, P. R.; Elimelech, M. Environ. Sci. Technol. 1994, 28, 1164-1171. (14) Johnson, W. P.; Blue, K. A.; Logan, B. E.; Arnold, R. G. Water Resour. Res. 1995, 31, 2649-2658. (15) Walz, J. Y. J. Colloid Inter. Sci. 1998, 74, 119-168. (16) Litton, G. M.; Olson, T. M. Environ. Sci. Technol. 1993, 27, 185193. (17) Zhu, X.; Elimelech, M. Environ. Sci. Technol. 1997, 31, 36543662.
(18) Bhattacharjee, S.; Ko, C.-H.; Elimelech, M. Langmuir 1998, 14, 3365-3375. (19) Elimelech, M.; Zhu, X.; Childress, A. E.; Hong, S. J. Membrane Sci. 1997, 127, 101-109. (20) Digital Instruments web site, www.di.com. (21) Balnois, E.; Wilkinson, K. J.; Lead, J. R.; Buffle, J. Environ. Sci. Technol. 1999, 33, 3911-3917. (22) Maurice, P. A.; Namjesnik-Dejanovic, K. Environ. Sci. Technol. 1999, 33, 1538-1541. (23) Liu, A.; Wu, R. C.; Eschenazi, E.; Papadopoulos, K. Colloids Surf. A: Physicochem. Engin. Aspects 2000, 174, 245-252. (24) Drummond, C. J.; Senden, T. J. Colloids Surf. A: Physicochem. Engin. Aspects 1994, 87, 217-234. (25) Martin, M. J.; Logan, B. E.; Johnson, W. P.; Jewett, D. J.; Arnold, R. G. J. Environ. Eng. 1996, 122, 407-415. (26) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Nature 1991, 353, 239-241. (27) Cleveland, J. P.; Manne, S.; Bocek, D.; Hansma, P. K. Rev. Sci. Instrum. 1993, 64, 403-405. (28) Shellenberger, K., M.S. Thesis, Department Civil & Environmental Engineering, The Pennsylvania State University, 2001. (29) Heinz, W. F.; Hoh, J. Biophys. J. 1999, 76, 528-538. (30) Vigeant, M. A. S.; Ford R. A. Appl. Environ. Microbiol. 1997, 63, 3474-3479. (31) Logan, B. E.; Zhang, H.; Mulvaney, P.; Milner, M. G.; Head, I. M.; Unz, R. F. Appl. Environ. Microbiol. 2001, 67, 2499. (32) Kim, K.; Logan, B. E. Water Res. 2001, 35, 3071. (33) Mulvaney, P. M. M.S. Thesis, The Pennsylvania State University, 2000. (34) Gross, M. J.; Albinger, O.; Jewett, D. G.; Logan, B. E.; Bales, R. C.; Arnold, R. G. Water Res. 1995, 29, 1151-1158. (35) Camesano, T. A.; Logan, B. E. Environ. Sci. Technol. 1998, 32, 1699-1708.
(36) Logan, B. E.; Jewett, D. G.; Arnold, R. G.; Bouwer, E. J.; O’Melia, C. R.; J. Environ. Eng. 1995, 121, 869-873. (37) Rajagopalan, R.; Tien, C. AIChE J. 1976, 22, 523-533. (38) Yao, K.-M.; Habibian, M. T.; O’Melia, C. R. Environ. Sci. Technol. 1971, 5, 1105-1112. (39) Ko, C.-H.; Elimelech, M. Environ. Sci. Technol. 2000, 34, 3681. (40) Rijnaarts, H. H.; Norde, W.; Bouwer, E. J.; Lyklema, J.; Zehnder, A. J. B. Environ. Sci. Technol. 1996, 30, 2869. (41) Kinoshita, T.; Bales, R. C.; Yahya, M.; Gerba, C. P. Water Res. 1993, 27, 1295-1301. (42) Smets, B. F.; Grasso, D.; Engwall, M. A.; Machinist, B. J. Colloids Surf. B: Biointerfaces 1999, 14, 121-139. (43) Logan, B. E. Environmental Transport Processes; Wiley: New York, 1999. (44) van Loosdrecht, M. C. M.; Lyklema, J.; Norde, W.; Schraa, G.; Zehnder, A. J. B. Appl. Environ. Microbiol. 1987, 53, 1898-1901. (45) Lahlou, M.; Harms, H.; Springael, D.; Ortega-Calvo, J.-J. Environ. Sci. Technol. 2000, 34, 3649-3656. (46) Amro, N. A.; Kotra, L. P.; Wadu-Mesthrige, K.; Bulychev, A.; Mobashery, S.; Liu, G.-Y. Langmuir 2000, 16, 2789-2796. (47) Camesano, T. A.; Logan, B. E. Environ. Sci. Technol. 2000, 34, 3354-3362. (48) Ong, Y.-L.; Razatos, A.; Georgiou, G.; Sharma, M. M. Langmuir 1999, 15, 2719-2725. (49) Razatos, A.; Ong, Y.-L.; Boulay, F.; Elbert, D. L.; Hubbell, J. A.; Sharma, M. M.; Georgiou, G. Langmuir 2000, 16, 9155-9158.
Received for review May 4, 2001. Revised manuscript received September 18, 2001. Accepted October 17, 2001. ES015515K
VOL. 36, NO. 2, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
189
