Environ. Sci. Technol. 2006, 40, 6091-6097
Removal of E. coli from Water Using Surface-Modified Activated Carbon Filter Media and Its Performance over an Extended Use SUKDEB PAL,† J. JOARDAR,‡ AND J O O N M Y O N G S O N G * ,† Research Institute of Pharmaceutical Sciences and College of Pharmacy, and School of Materials Science & Engineering, Seoul National University, Seoul 151-742, South Korea
Modification of activated carbon (AC) by aluminum hydroxychloride (AHC), and diatomaceous earth by zinc hydroxide changed the zeta potentials of these filter media from negative to positive. The modification method is amenable to room temperature, and eliminates the essential requirement of strong base treatment for making metal hydroxide coated filter media. Solid-state MAS 27Al NMR spectra suggested the presence of Al13-mer in the AHCtreated AC. AHC-modified AC samples were further treated with silver halide, and two antibacterial compounds to prevent microbial growth on filter media. In situ precipitation of silver bromide on AC resulted in formation of nanosized AgBr crystals. Bacteria removal performances of the modified media were tested in columns. For the first time, we demonstrated that only 30 g of either AHC-treated AC (60 × 200 mesh) or nano AgBr supported AC could provide >6 log E. coli removal over ∼1000 L when the input water had a bacterial load of 107 CFU/mL. The filter media were robust enough to perform even when water was passed at superficial velocities 3-10 times the typical velocity (6 cm/min) of water treatment processes. Metal leaching from the modified media was found to be less than the USEPA specified Maximum Contaminant Level.
Introduction In recent years, concentrated efforts have been made to improve the performance of common negatively charged granular media, like sand, anthracite coal, or diatomaceous earth, for removing microbial contaminants from water by utilizing oxides and hydroxides of iron or aluminum (1, 2). Such modified media are positively charged and, therefore, more effective for removing and retaining the negatively charged microorganisms by electrostatic adsorption (1, 3). High removal efficiencies (90-99%) have been demonstrated using these charge-modified filter media for a wide range of waterborne microbes (4). However, performance of these granular media filters decreases significantly over prolonged * Corresponding author phone: 82-2-880-7841; fax: 82-2-8712238; e-mail:
[email protected]. † Research Institute of Pharmaceutical Sciences and College of Pharmacy, Seoul National University. ‡ School of Materials Science & Engineering, Seoul National University. Present address: Centre for Nanomaterials, International Advanced Research Institute for Powder Metallurgy and New Materials (ARCI), Balapur PO, Andhra Pradesh, Hyderabad 500 005, India. 10.1021/es060708z CCC: $33.50 Published on Web 08/31/2006
2006 American Chemical Society
use and only ∼90% (∼ 1 log) bacteria removal can be achieved after 100-200 L. This is far less than the USEPA recommendation of at least 99.9999% (6 log) bacteria removal from drinking water. Commonly, metallic hydroxides are precipitated on the filter matrix by applying a metallic chloride, such as aluminum chloride, followed by addition of a large excess of a base solution at elevated temperatures. As the nature and species of aluminum hydrolysis are greatly controlled by base addition rate, pH, and hydrolysis ratio, use of strong base for metal hydroxide coating demands sensitive process control; otherwise it may result in uncontrolled modification of media surfaces. Activated carbon has widely been used in point-of-use filtration systems due to its high adsorption rates and capabilities to reduce concentrations of toxic organic compounds. However, the inherent drawback of carbon materials is their excellent biocompatibility with bacteria. Bacteria may breed on carbon during the purification process, thus eventually causing the carbon materials themselves to become pollutants. Some researchers have examined the bactericidal effects of depositing colloidal-sized silver onto carbon (5-7). However, a few problems have been pointed out. The processes of applying silver nanoparticle coating often require expensive and cumbersome high vacuum equipment and high temperatures. At high temperature silver particles coalesce to form larger particles with higher attrition probability which in turn cause antibacterial activities to deteriorate, and could lead to skin discoloration called “Argyria”. Thus, the objective of this work is to develop a simple process for preparing a filter media which does not entail sensitive process control, yet results in a modified media that is highly robust and effective in removing microorganisms even over extended use. In this work, we report the development of activated carbon filter media by modifying the filter matrix with aluminum hydroxychloride, a commercial antiperspirant active. Aluminum hydroxychloride (AHC) contains partially neutralized forms of the metal. Many partially neutralized aluminum products contain substantial proportions of the tridecamer Al13. The relatively high stability of Al13 makes it more readily available for adsorption and charge neutralization at around neutral pH. Because they are already partially neutralized, they have a smaller effect on the pH, and reduce the need for pH correction. These materials are also considerably more effective than traditional coagulants (8) at low temperatures. Another objective of this work is to evaluate a coating with antibacterial activity. We deposited nanosized crystals of silver bromide as a bactericide on activated carbon. Diallyl dimethyl ammonium chloride (DADMAC), triclocarban (TCC), and zinc hydroxide were also tested as antibacterial materials. The bacteria removal performances of the modified materials were evaluated against E. coli, chosen as an indicator of fecal contamination, under flow-through condition. Metal content in output water was compared with the USEPA specified MCL (9).
Materials and Methods Chemicals. The chemicals used in this study and their sources are as follows: aluminum hydroxychloride (% Al2O3 42-44%, basicity 84-86%), Chlorhydrol, Reheis; diallyl dimethyl ammonium chloride (% Active 39-42%), Floquat FL 4540, SNF Inc.; zinc sulfate heptahydrate (99.99%), triclocarban, silver nitrate, and potassium bromide, Aldrich; activated VOL. 40, NO. 19, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6091
carbon (base: coconut shell charcoal), Active Carbon Pvt. Ltd.; diatomaceous earth (DE) (particle size 0-1.0 mm), Seema Minerals and Metals. Modification of AC and DE by Aluminum Hydroxychloride and by in situ Precipitation of Zinc Hydroxide. AC or DE was modified (a) with aluminum hydroxychloride at pH 6.0-6.5, and (b) by in situ precipitation of zinc hydroxide at pH 6.0-7.0. AC materials modified with AHC and zinc hydroxide were denoted as AHC-AC and ZS-AC, respectively, whereas DE samples treated with AHC and zinc hydroxide were labeled as AHC-DE and ZS-DE, respectively. Modification of AC by Silver Bromide. Silver bromide was precipitated on activated carbon by adding a solution of potassium bromide to activated carbon previously soaked to saturation in a solution of silver nitrate. The sample was denoted as Ag-AC. Modification of Silver Bromide-Coated AC by Aluminum Hydroxychloride. Silver halide-modified activated carbon was further modified with 10% (w/v) solution of aluminum hydroxychloride. The sample was denoted as AHC-Ag-AC. Modification of AHC-Treated AC by DADMAC and TCC. AHC-modified AC was also treated with (a) an aqueous solution of DADMAC, and (b) an alcoholic solution of triclocarban, thus producing AHC-DADMAC-AC and AHC-TCC-AC, respectively. (See Supporting Information for experimental details). The amount of AHC (wt %) used is mentioned before the appropriate designations of the samples, and the particle size of AC is mentioned within parentheses immediately after the designations, i.e., 2.5% AHC-AC (60 × 200 mesh) describes an AC sample of 60 × 200 mesh size modified using 2.5% AHC. If particle size is not mentioned separately then it refers to sample of 60 × 200 mesh size. Characterization of the Filter Media. The zeta potentials of the filter media were determined using an Anton Paar Electro kinetic Analyzer (EKA, Anton Paar, Graz, Austria). The cylindrical flow cell had an i.d. of 2.0 cm and a packed length of 4.0 cm. The electrolyte was 1.0 × 10-3 M KCl (pH 7.5). Computation of zeta potential was done as described previously (10). Solid-state MAS 27Al-NMR spectra of AHC modified AC were recorded with a Bruker DRX-500 FT-NMR (12 kHz) instrument. Scanning electron micrographs were taken with a JEOL JSM-6330F (FESEM) and a Hitachi S3500N instrument. Elemental mapping of the media surfaces was conducted with EDX (Oxford Inca 400). X-ray diffraction measurements were carried out for Ag-AC using Cu KR radiation from M18XHF rotating anode (MacScience, Japan) at 50 kV and 200 mA. Step scan mode of XRD was used with a step size of 0.02° and dwell time of 2 s per step. The crystallite size of the sample was estimated by XRD line profile analysis using pseudo-Voigt function applied to single peak profile (11). (See Supporting Information for details). Metal leaching from the filter media was analyzed by ICP-OES (Perkin-Elmer Optima 2000). Briefly, 10 L of deionized water was passed through each column packed with 10 g of unmodified and modified materials at a flow rate of 50 mL/min. Multiple samples of product water were collected after passing 4 and 9 L of water and analyzed for the metal ions by ICP-OES. Test Water. Dechlorinated tap water was manipulated for turbidity (initial 6 >6 >6 >6 >6 >6
>6 >6 >6 >6 >6 >6
Rounded to the nearest integer value
length of mass transfer zone depends primarily on the rate of mass transfer, superficial velocity, and concentration of the solute in the feed. When an experiment is carried out under different superficial velocities while maintaining other operating conditions, the length of mass transfer zone does not remain the same and this is reflected in the performance of the media bed. However, the length of mass transfer zone does not change with the length of column if the operating conditions including superficial velocity remain the same. Therefore, it is expected that if a column of larger dimension is operated under the same operating conditions similar bacteria removal can be achieved. In that case, the length of the mass transfer zone would remain the same as in a column of a smaller dimension operated at the same superficial velocity, but final flow rate would be higher. Bacteria Removal Kinetics. Figure 3 shows the bacteria removal kinetics of differently treated activated carbon materials determined under batch conditions following a method described previously (3). (See Supporting Information for experimental details). Assuming that the removal process could be described by a first-order kinetic rate constant (k) (3) the rate of bacteria removal was quantified from the experimental studies. The rate constant was taken to be invariable with respect to time and uniform among the
FIGURE 3. Bacterial removal kinetics of differently treated activated carbon materials as determined by deposition study carried out under batch conditions. A 10 mL volume of 107 CFU/mL E. coli solution at pH 7.0 (final electrolyte concentration 1 × 10-3 M) was added to a number of tubes each containing 2 g of different materials. Tubes were agitated in a shaking incubator (temperature 25 °C) at a speed of 100 rpm. Bacterial removal was determined by microbiologically analyzing the supernatant sampled at various times using standard plate counting method. Batch experiment performed under the same condition in absence of filter media was taken as control. Nonlinear exponential regression coefficients (R 2): 10% AHC-DADMAC-AC (60 × 200) ) 0.9733; 10% AHC-AC (60 × 200) ) 0.9573; 10% AHC -Ag-AC (60 × 200) ) 0.9447; Ag-AC (60 × 200) ) 0.9808. bacteria population. Thus Ct, the concentration of bacteria remaining in the supernatant at time t would be
Ct ) C0e-kt where C0 is the initial bacterial concentration. The rate constants (k) were estimated by fitting this equation to the experimental data using nonlinear exponential regression VOL. 40, NO. 19, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6095
(R 2 ranging from 0.94 to 0.98). 10%AHC-Ag-AC (60 × 200), and Ag-AC (60 × 200) showed the larger kinetic rate constant values (k ) 0.047 and 0.044 min-1, respectively) and indicated the higher removal efficiency. The k values of 10%AHC-AC (60 × 200) and 10%AHC-DADMAC-AC (60 × 200) were estimated as k ) 0.037 and 0.035 min-1, respectively. Similar kinds of results were observed previously on metal hydroxide coated granular media and bacterial deposition data were adequately modeled as a first-order kinetic process (3). The faster removal kinetics as observed in case of Ag-AC and 10%AHC-Ag-AC (60 × 200) may be explained considering a synergistic effect of bacteria deposition and antibacterial effect of Ag. In the case of ACs with antibacterial materials bacteria may remain in the supernatant solution in a viable non-culture state, and hence would not be counted. The antibacterial effect is relatively less pronounced in the case of DADMAC-treated material. Lower rate constant values of DADMAC-treated AC and 10%AHC-AC (60 × 200) suggested that bacteria removal is mainly transport limited in the case of these two materials. Although bacterial removal kinetics under batch condition gives a preliminary idea about the relative efficiencies of differently treated materials, to quantify the removal efficiencies of the materials in a fixed bed process it would be more appropriate to estimate the maximal accumulative number of bacteria retained per unit area of the filter media. However, at present we are unable to give an estimation of the maximal accumulative number of bacteria retained per m2 of the treated activated carbon surface. Effect of pH and Ionic Strength. A column experiment was conducted using distilled water over a range of pH (5.0-9.0) and ionic strength conditions (10-300 mM KCl) with a cell concentration of 107 CFU/mL. Four materials, 10%AHC-AC (60 × 200), 10%AHC-DADMAC-AC (60 × 200), 10%AHC-Ag-AC (60 × 200), and Ag-AC, were tested by passing 5 L of bacterial suspension through columns each containing 10 g of these materials. None of the materials showed any significant change in bacteria removal efficiencies over the range of pH and ionic strength studied. Usually surface potentials of filter media and microbes change with change in pH and ionic strengths. Presence of metal ions also influences the removal process causing flocculation of bacteria or viruses (4, 13, and references therein). The effect becomes profound in the case of negatively charged filter media. However, in the present study >6 log reduction was achieved for all four materials indicating the materials have the potential to be used over a wider range of water conditions. Extended Lifetime Test of Modified AC. Finally, we evaluated the long-term bacteria removal performance of four modified AC samples. Samples (30 g each) were packed in the columns of the prototype. Contaminated water was forced through the columns vertically upward at a superficial velocity of 24-25 cm/min (final flow rate 120 ((5) mL/min). The average log removals of E. coli after passing certain volumes of contaminated water are reported in Table 3. More than 6 log E. coli removal was achieved using 10% AHC-Ag-AC (60 × 200) even after passage of 900 L of contaminated water. Activated carbon modified with only AHC, 10% AHC-AC (60 × 200), and only silver halide, Ag-AC (60 × 200), also performed equally well. The 10% AHC-DADMACAC sample, which was tested up to 550 L only, also showed >6 log removal of bacteria. These results are in line with the USEPA recommendation for drinking water. The bacteria removal performance of the materials was tested with a very high load of bacterial challenge, which is a rare case in real-life systems. This work offers possible solutions in the development of drinking water purification. However, issues like clogging and consequent pressure drop across the inlet and outlet of the housing, and 6096
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 19, 2006
TABLE 3. Performance of Modified Filter Media over Extended Use average log10 removal volume of water passed (L) 10 100 200 350 450 550 600 650 750 850 900 950 1000 a
10% AHC-AC (60 × 200)
10% AHCAg-AC (60 × 200)
10% AHCDADMAC-AC (60 × 200)
Ag-AC (60 × 200)
>6 >6 >6 >6 >6 >6 NA >6 >6 >6 >6 NA >6
>6 >6 >6 >6 >6 >6 NA >6 >6 >6 >6 NA >6
>6 NAa >6 >6 >6 >6 NA NA NA NA NA NA NA
>6 NA >6 >6 >6 NA >6 NA NA NA >6 NA >6
NA: Not analyzed.
designing the filtering system should be addressed as part of the scaling up process. Further studies on the bacteria removal performances of these materials on other pathogenic bacteria and viruses of different physicochemical properties are also necessary to fully evaluate the possible use as new materials for water treatment.
Acknowledgments This work was supported by the Korea Research Foundation Grant funded by Korea Government (MOEHRD, Basic Research Promotion Fund) (KRF-2005-003-C00123).
Supporting Information Available Methods of modification of filter media, crystallite size analysis of AgBr by integral breadth method, experimental details for bacterial removal kinetics, schematic diagram of the prototype and column, 27Al MAS NMR of AC treated with AHC, XRD patterns of AC and Ag-AC, metal leach-out from filter media, and performance of modified filter media to inhibit growth of E. coli on the used media. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Chaudhuri, M.; Sattar, S. A. Enteric virus removal from water by coal-based sorbents: Development of low-cost water filters. Water Sci. Technol. 1986, 18, 77-82. (2) Prasad, V. S.; Chaudhuri, M. Development of filtration/adsorption media for removal of bacteria and turbidity from water. Water Sci. Technol. 1989, 21 (3), 67-71. (3) Truesdail, S. E.; Lukasik, J.; Farrah, S. R.; Shah, D. O.; Dickinson, R. B. Analysis of bacterial deposition on metal (hydr)oxidecoated sand filter media. J. Colloid Interface Sci. 1998, 203, 369378. (4) Farrah, S. R.; Preston, D. R.; Toranzos, G. A.; Girard, M.; Erdos, G. A.; Vasuhdivan, V. Use of modified diatomaceous earth for removal and recovery of viruses in water. Appl. Environ. Microbiol. 1991, 57, 2502-2506. (5) Oya, A.; Yoshida, S. Antibacterial activated fiber derived from phenolic resin containing sliver nitrate. Carbon 1993, 31, 7173. (6) Li, C. Y.; Wan, Y. Z.; Wang, Y. L.; Jiang, X. Q.; Han, L. M. Preparation and characterization of antibacterial pitch-based activated carbon fiber supporting silver. Carbon 1998, 36 (12), 61-65. (7) Wang, Y. L.; Wan, Y. Z.; Dong, X. H.; Cheng, G. X.; Tao, H. M.; Wen, T. Y. Preparation and characterization of antibacterial viscose-based activated carbon fiber supporting silver. Carbon 1998, 36 (11), 1567-1571. (8) Duan, J.; Gregory, J. Coagulation by hydrolysing metal salts. Adv. Colloid Interface Sci. 2003, 100-102, 475-502.
(9) Secondary Drinking Water Regulations: Guidance for Nuisance Chemicals, EPA 810/K-92-001; U.S. Environmental Protection Agency: Washington, DC, July 1992. (10) Scott, T. M.; Sabo, R. C.; Lukasik, J.; Boice, C.; Shaw, K.; BarrosoGiLSetti, L.; El-Shall, H.; Farrah, S. R.; Park, C.; Moudgil, B.; Koopman, B. Performance and cost-effectiveness of ferric and aluminum hydrous metal oxide coating on filter media to enhance virus removal. KONA 2002, 20, 159-167. (11) De Keijer, T. H.; Langford, J. I.; Mettemeijer, E. J.; Vogels, A. B. P. Use of the Voigt function in a single-line method for the analysis of x-ray diffraction line broadening. J. Appl. Crystallogr. 1982, 15, 308-314. (12) Israelachvili, J. Intermolecular and Surface Forces; Academic Press: San Diego, CA, 1992. (13) Gerba, C.; Goyal, S. M.; Cech, I.; Bogdan, G. F. Quantitative assessment of the adsorptive behavior of viruses to soils. Environ. Sci. Technol. 1981, 15, 940-944. (14) Schafer, A.; Harms, H.; Zehnder, A. J. B. Bacterial accumulation at the air-water interface. Environ. Sci. Technol. 1998, 32, 37043712. (15) Dickinson, R. B. A dynamic model for the attachment of a Brownian particle mediated by discrete macromolecular bonds. J. Colloid Interface Sci. 1997, 190, 142-151. (16) Camesano, T. A.; Logan, B. E. Influence of fluid velocity and cell concentration on the transport of motile and nonmotile bacteria in porous media. Environ. Sci. Technol. 1998, 32, 1699-1708. (17) Elimelech, M.; O’Melia, C. R. Effect of particle size on collision efficiency in the deposition of Brownian particles with electrostatic energy barriers. Langmuir 1990, 6, 1153-1163.
(18) Jucker, B. A.; Zehnder, A. J. B.; Harms, H. Quantification of polymer interactions in bacterial adhesion. Environ. Sci. Technol. 1998, 32, 2909-2915. (19) Elimelech, M.; Gregory, J.; Jia, X.; Williams, R. A. Particle Deposition and Aggregation; Butterworth-Heinemann: Boston, MA, 1995. (20) Kinoshita, T.; Bales, R. C.; Yahya, M.; Gerba, C. P. Bacteria transport in a porous medium: Retention of bacillus and pseudomonas on silica surfaces. Water Res. 1993, 27 (8), 12951301. (21) Bottero, J. Y.; Axelos, M.; Tchoubar, D.; Cases, J. M.; Fripiat, J. J.; Fiessinger, F. Mechanism of formation of aluminum trihydroxide from keggin Al13 polymers. J. Colloid Interface Sci. 1987, 117, 47-57. (22) Hiradate, S.; Yamaguchi, N. U. Chemical species of Al reacting with soil humic acids. J. Inorg. Biochem. 2003, 97, 26-31. (23) Spadaro, J. A.; Berger, T. J.; Barranco, S. D.; Chapin, S. E.; Becker, R. O. Antibacterial effects of silver electrodes with weak direct current. Antimicrob. Agents Chemother. 1974, 6, 637-642. (24) Klaus, T.; Joerger, R.; Olsson, E.; Granqvist, C.-G. Silver-based crystalline nanoparticles, microbially fabricated. Proc. Natl. Acad. Sci., U.S.A. 1999, 96, 13611-13614.
Received for review March 24, 2006. Revised manuscript received August 1, 2006. Accepted August 1, 2006. ES060708Z
VOL. 40, NO. 19, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6097
