Printed in Great Britain 0043-1354/01/$ - see front matter PII: S0043-1354(01)00117-8
FILTRATION BEHAVIORS OF GIARDIA AND
CRYPTOSPORIDIUM}IONIC STRENGTH AND PH EFFECTS
BING-MU HSU1, CHIHPIN HUANG* and JILL RUHSING PANInstitute of Environmental Engineering, National Chiao Tung University, Hsinchu, Taiwan, R.O.C.
(First received 30 August 2000; accepted inrevised form 6 February 2001)
Abstract}The laboratory-scale filtration tests of Giardia cysts and Cryptosporidium oocysts in both 2 mm-Fglass beads and 2 mm-F polystyrene beads filters were conducted to investigate their filtration behaviors. The protozoan parasites were used as target particles, while the chemical system altered by changing the electrolyte concentration and pH. The results significantly indicate that ionic strength have a positive effect on the removal efficiencies for Giardia cysts and Cryptosporidium oocysts. The removal efficiency of two filters for Giardia cysts slightly decreased from pH 2.4 to 8.7 and decreased significantly in pH as pH up to 8.7, while that for Cryptosporidium slightly rippled beyond pH 8.7, and with the decrease in pH up to pH 8.7. The experimental collision efficiencies from the interactions between colloids and the filter media were calculated with a semi-empirical approach of the single sphere model and clean-bed filtration theory. The results also indicated that experimental collision efficiencies for (oo)cysts corresponded to the (oo)cysts removal efficiencies in all trials, and oocysts exhibits higher collision efficiencies than cysts. #2001 Elsevier Science Ltd. All rights reserved
Key words}filtration, Giardia cysts, Cryptosporidium oocysts, collision efficiency
INTRODUCTION
The protozoan parasites Giardia and Cryptospori-dium have been recognized as the common patho-genic protozoa of the gastrointestinal tract. Many outbreaks of giardiasis and cryptosporidiosis have been reported in the last few decades (Frost et al., 1996; SoloGabriele and Neumeister, 1996). Human infection with these pathogens is usually through either direct contact or ingestion of contaminated food and/or water (Donnelly and Stentiford, 1997; Pell, 1997). Water is perhaps the major source for massive outbreaks of infection, as a result of contamination of source water. Investigation has shown that commonly used disinfectant in the water treatment plants, such as chlorine, is virtually useless in inactivating such kinds of parasites (Haas et al., 1994). Filtration, on the other hand, is likely to be the most practical treatment technology used for Giardia cysts and Cryptosporidium oocysts removal (Swertfe-ger et al., 1999; Fogel et al., 1993).
Some surface properties of the Giardia cysts and Cryptosporidiumoocysts, i.e. electrophoretic mobility and hydrophobicities, have been investigated by Ongerth and Pecoraro (1996) and Brush et al.
(1998). Adin et al. (1999) have conducted a bench scale filtration column to study the removal efficien-cies of Cryptosporidium oocysts. However, the inter-actions of Giardia cysts and Cryptosporidium oocysts with filter media while changing ionic strength and pH of water samples were not realized. The filtration system of actual water treatment process is hard to simulate for their complex components and mechanisms. Many papers determined the overall (oo)cysts removal efficiency of the filtration by evaluating the number difference of (oo)cysts be-tween the inlet and outlet. However, information about the contribution of simplex mechanism for (oo)cyst removal in the filtration system is very limited. Therefore, it is important to discover the simplex mechanism for (oo)cyst removal in the deep-bed filtration system.
Our group has investigated the differences of collision efficiency between organic (algae) and inorganic (kaoline) particles as well as the effects of ionic strength and pH on them (Huang et al., 1999). This approach showed some interesting results including different filtration behavior between inor-ganic and orinor-ganic particles and the significant effects of their surface properties on the filtration, which deserved to go further. In this study, therefore, we investigated the filtration behaviors of Giardia cysts and Cryptosporidium oocysts at various pH values and ionic strengths. Beside that, two filtration media, glass beads and polystyrene beads, were tested in
1
Present address: Department of Medical Research, Chiayi Christian Hospital, Chiayi, Taiwan, R.O.C.
*Author to whom all correspondence should be addressed. Tel.: +886-3-572-6463; fax: +886-3-572-5958; e-mail: cphuang@green.ev.nctu.edu.tw
parallel to realize the effect of different media on the (oo)cyst filtration. The overall removal percentages of (oo)cysts is not the point of this study, but the (oo)cyst collision efficiency during the deep-bed filtration. In other words, we only investigated the microscopic function of (oo)cysts filtration, not macroscopic. Through the investigation of the (oo)cysts filtration in a lab-scale packed column, the single-collector efficiency (Z) was calculated, and the experimental collision (or attachment) efficiency (aexp) with the observed removal efficiency were determined.
MATERIALS AND METHODS
Preparationand Use of the ColumnFilters
The experimental apparatus of this study is shown in Fig. 1. The column for the filter consisted of a 48 mm i.d. 15 cm acrylic column with two adjustable bed supports. Glass beads and polystyrene beads were the two chosen media. They were first cleaned by soaking in the hot 10 g/l Alconox detergent solution for at least 2 h, followed by rinsing a few times with deionized water. They were then soaked in 5 N nitric acid for at least 12 h, and rinsed with deionized water again. All deionized water used for glassware washing and solution preparation was obtained from a Milli-Q system (Millipore Corp.). The porosity of the glass beads and polystyrene beads were measured to be 0.39 and 0.37, respectively, by a volumetric method.
The tracer study was first performed to determine the duration of complete breakthrough. The column was filled with clean media, and distilled water was pumped into the column to saturate the system. A 10 g NaNO3/l solution was
introduced into the column as a step input, and the nitrate concentration in the effluent was monitored with UV absorption measurements (wavelength was 302 nm) at 1-min interval. The NaNO3 concentration was found to be
directly proportional to UV absorption in the range of interest (0–10 g/l). The breakthrough results were indicative of a plug flow system and complete mix system in series.
Preparation of Protozoa-contained sample and calculation for (Oo)cysts
The cysts and oocysts were obtained from Waterborne, Inc. (Louisiana, USA) and the Pleasant Hill Farms (Idaho, USA), respectively, and were diluted to desired
concentra-tions by 0.1% PBS (phosphate buffered saline) as the stock solution. The numbers of cysts and oocysts in the stock solution were counted before seeded experiment using the indirect fluorescent antibody (IFA) staining procedure.
Five hundred milliliters of the deionized water were collected in a plastic 1-l container. The pHs of water samples were adjusted from 2.4 to 11.2 with 0.1 N HClO4and 0.01 N
NaOH in the presence of 1.0 102M NaClO
4 as
back-ground electrolyte. Water samples with various backback-ground electrolyte (eg, 101.0M, 101.5M, 102.0M, 102.5M, and 103.0M) at pH 5.6 were also prepared. Water samples of given background electrolyte and pH level were then seeded with 1.0 104
(oo)cysts by taking a specific volume from the stock solution. Prior to each experimental run, the entire system was washed with the ionic solution of the designated NaClO4and pH, free of protozoa, with an upflow mode for
at least 20 min. Experiments were carried out at room temperature (24–258C) with downward vertical flow at 10.0 ml/min (approach velocity=2.53 105m/s). Forty
milliliters of filtrate were collected 4 min after the time of complete breakthrough.
To investigate the removal efficiencies for Giardia cysts and Cryptosporidium oocysts, we counted the concentration of cysts and oocysts in the effluent and influent by IFA staining procedure described in the instruction provided by the company (HydrofluorTMCombo Giardia/
Cryptosporidium; Ensys, Inc., NC, USA). The collected suspensions and filtrate were applied to each 25-mm diameter cellulose-acetate membrane, stained with fluores-cent-labeled antibodies (HydrofluorTM-Combo Giardia/
Cryptosporidium; Ensys, Inc., NC, USA) and examined with epifluorescent microscope at 200 magnification (Olympus, Japan).
Theoretical approach
Giardia cysts and Cryptosporidium oocysts were the colloidal particles in our column filtration experiments. The rationale of Z involves the physical aspect of a better understood filtration model, while collision efficiency accounts for the less important factors in the chemical aspect (Elimelech, 1992). With the combination of trajectory and dimensional analysis, the process of filtration can be formulated. The calculation of the single-collector efficiency has been derived by Rajagopalan and Tien (1976) as following:
Z¼ 4A1=3
s P2=3e þ 0:72AsN1:8LONR15:8
þ 0:0024AsNG1:2N0:4R ð1Þ
where As, defined as 2ð1 p5Þ=w, accounts for the effect of
adjacent collectors on the collector efficiency, p¼ ð1 eÞ1=3, w¼ 2 3p þ 3p5 2p6, and e the porosity of the porous
medium. NR, indicating the importance of interception, is
the ratio of the size of suspended particle to the size of collector ðap=acÞ Pe represents the role of Brownian
diffusion (2acU=DBM), where U is the approach velocity to
a collector (m/s). The term DBM, defined as kT=ð6pmapÞ, is
the Brownian diffusion coefficient, where m is the dynamic viscosity of fluid density, k the Boltzmann’s constant, and T the absolute temperature. NL0 reflects van der Waals
interaction (H=9pma2
pU), where H is the Hamaker constant
(J). NG accounts for gravity effect (2a2pðrp rÞg=9mU),
where rpis the density of the suspended particles, r the fluid
density, and g the gravitational constant.
By applying the clean-bed filtration model and a mass balance of particles over a differential packed-bed volume, the experimental collision efficiency (aexp) can be related
to the initial (clean-bed) removal, ðC=C0Þ0, as follows
(Elimelech and O’Melia, 1990)
aexp¼ ln C C0 0 4ac 3 1ð eÞLZ ð2Þ Fig. 1. The experimental apparatus of deep-bed filtration in
where acis the radius of the collector, e the porosity of the
filter bed, L the column depth, and Z the single-collector efficiency. The initial clean bed removal efficiency,ðC=C0Þ0,
is the value taken when a complete breakthrough of an inert tracer occurs.
RESULTS AND DISCUSSION Breakthrough curves
The breakthrough curve of column filtration is presented as C=C0with respect to time. C and C0are the effluent and influent concentrations of NaNO3,
respectively. The duration of complete breakthrough described in Fig. 2 for glass beads and in Fig. 3 for polystyrene beads was about 14 min and 13 min, respectively. The porosity (e) of the glass beads, 0.39, was higher than that of the polystyrene beads, 0.37. Therefore, the duration of complete breakthrough for glass beads was longer than that for polystyrene beads. In this study, 14 min were designed as duration of complete breakthrough.
Removal efficiency
The removal efficiencies, 1 C=C0; for both Giardia cysts and Cryptosporidium oocysts with
respect to different media, NaClO4 concentrations,
and pH levels were presented from Figs 4–7. Results showed that removals of cysts were lower than oocysts in all trials. Because of the size of the filter media used in this study, the removal mechanism of these protozoa with respect to the intercept can be neglected. The removal ratio of (oo)cysts was, therefore, influenced mostly by the collision. The removal ratio was determined by the surface char-acteristics of (oo)cyst, especially the hydrophobicity and surface charge. In a separate study of ours, the hydrophobicity and surface charge of oocysts were larger than those of cysts, which are in agreement with the determined result of (oo)cyst removal efficiencies.
Figures 4 and 5 show the removals of (oo)cysts in two kinds of medium with various background electrolyte at pH 5.6. The removal of Giardia cysts was ranged from 0.21 to 0.80 and that of Cryptos-poridiumoocysts from 0.41 to 0.92, and the variation in removal efficiencies by the polystyrene beads was larger than that by the glass beads for both parasites. Removals of both parasites increased with the NaClO4concentration. This finding agreed with the
DLVO theory, that the total interaction energy and
Fig. 2. Breakthrough curve of an inner tracer NaNO3in the
0.1 m bed depth column packed with 2 mm-Fglass beads. The volumetric flow rate is 10.0 ml/min.
Fig. 3. Breakthrough curve of an inner tracer NaNO3in the
0.1 m bed depth column packed with 2 mm-Fpolystyrene beads. The volumetric flow rate is 10.0 ml/min.
Fig. 4. Removal efficiencies of (oo)cysts in packed glass beads column with various concentrations of NaClO4 at
pH 5.6.
Fig. 5. Removal efficiencies of (oo)cysts in packed poly-styrene beads column with various concentrations of
the height of the energy barrier of the reaction decrease as the concentration of NaClO4 in the
suspension increases, and that the van der Waals interaction is independent of the solution chemistry.
The effects of pH, with 1.0 102M NaClO 4 as
the background electrolyte, on (oo)cysts removals from different filtration media were illustrated in Figs 6 and 7. The variation of parasite removal efficiencies on pH values from different packed column was much greater by glass beads (Fig. 6), than by polystyrene beads (Fig. 7). The removal of Giardia cysts was ranged from 0.07 to 0.76 and that of Cryptosporidium oocysts was ranged from 0.59 to 0.92. The removal of both parasites decreased with the pH value. This can be explained by the pH effect on zeta potentials (Ongerth and Pecoraro, 1996). The glass and polystyrene beads carried negative charge at their surfaces, which increased with the pH. Therefore, the repulsive force between particles and collectors were increased. Because the complicated process of life cycles for both protozoa entail the coordinated production, processing, and transport of wall constituents for assembly into a protective (oo)cyst wall, little is known about this process and the identity of (oo)cyst wall constituents. In this study, we extrapolate the surface contained protein
of (oo)cyst by observing the variations on the (oo)cyst removal efficiency at different pH values. Figures 6 and 7 also showed that the removal efficiencies of two filters for Giardia cysts slightly decreased from pH 2.4 to 8.7 and decreased signifi-cantly in pH as pH up to 8.7, while that for Cryptosporidiumslightly rippled beyond pH 8.7, and with the decreased in pH up to pH 8.7. This phenomenon can be explained by the ionization of amino acids on the (oo)cysts surface. All amino acids have a basic skeleton consisting of a a-carbon linked to an amino group (–NH2), a carboxyl group
(COOH), a hydrocarbon atom (H), and R group. The variations among the amino acids occur at the R group, which is different in each amino acid and imparts the unique characteristics to the molecule and to the proteins that contain it (Talaro and Talaro, 1999). Protein molecules, which consist essentially of a-amino acids linked by peptide (–CONH–) linkages, are the major surface compo-nents of the (oo)cysts. The surface charge of (oo)cysts originates from the ionization of R groups in the amino acids. Most of the R groups in amino acids are not ionized at pH=7, and the pKavalues of amino acids are from 1.71 to 3.0 (Smith and Martell, 1976). However, the R groups for monocarboxy diamino acids } lysine and arginine are positively charged at pH=7, and their pKa values are 10.5 and 12.5, respectively. About 99% of the –NH2groups would
ionize to –NH3+ groups when pH¼ pKa 2 (i.e. pH=8.5–10.5), and thus infer the dominant compo-nents of proteins in (oo)cysts surfaces.
In the study of Adin et al. (1999), the removal efficiencies of Cryptosporidium oocysts were respec-tively 0.09 and 0.10 for fined sand and coarse sand at 2.5 102m/min. Although the removal efficiency was influenced by the approach velocity, the removal of oocysts in their study were still lower than ours. In the study of Huang et al. (1999), the reported removal for algae was also much lower than the removals of cyst and oocyst in this study.
Experimental collision efficiencies
The values used for the parameters in equations (2) and (3) are: fluid density (r) is 998 kg/m3, fluid viscosity (m) 8.94 104N s/m2
,. packed column length (L) 0.1 m, (oo)cysts densities (rp) 1080 kg/m3, the diameters (dp) of cysts and oocysts respectively 10.1 106and 5.0 106m, approach velocity (U) 2.53 105m/s, absolute temperature (T ) 298 K, and gravitational acceleration (g) 9.81 m/s2. The Ha-maker constants (H) were determined to be 1.20 1021and 1.15 1021J for both parasites in
glass beads media and polystyrene beads media, respectively. The Boltzman constant (k) is 1.38 1023J/K. The theoretical porosity (e) was 0.39 for glass beads and was 0.37 for polystyrene beads and their sizes (dc) were both 2 103m.
Fig. 6. Removal efficiencies of (oo)cysts in packed glass beads column with various pH in the presence of 0.01 M
NaClO4.
Fig. 7. Removal efficiencies of (oo)cysts in packed poly-styrene beads column with various pH in the presence of
In this study, experimental collision efficiencies were calculated from the results of column filtration experiments. The logarithm of experimental collision efficiency for (oo)cysts under various background electrolyte at pH 5.6 was presented in Fig. 8. The experimental collision efficiency for (oo)cysts gradu-ally increased with the increasing NaClO4
concentra-tion. As the electrolyte concentration increases, the diffuse layer is compressed, which facilitate the successful attachment of the colloidal particles. Figure 8 also points out that at various background electrolyte, the variation of collision efficiencies between (oo)cyst suspension and polystyrene beads was higher than that between (oo)cyst suspension and glass beads.
Figure 9 shows the experimental stability curves of both parasites at various pHs. The experimental collision efficiency decreased with increasing pH, which is similar to the removal efficiency of parasite filtration in different media, where Fig. 9 also displayed that at various pH values, the variation of collision efficiencies of the glass beads was higher than that of the polystyrene beads for both parasites. When the particle size is smaller than a Brownian particle, 1 mm in diameter, the mechanism of particle transport is dominated by diffusion. Therefore, a particle smaller than a Brownian particle will have better single-collector efficiency in the deep filter. As for non-Brownian particles, transport is controlled by gravity, fluid drag and interception in the deep
filter. In this study, the average diameters of Cryptosporidium oocysts and Giardia cysts were measured to be 5.0 and 10.1 mm. Therefore, Giardia cysts resulted in better single-collector efficiency than the Cryptosporidium oocysts. Both Figs 8 and 9 pointed out that the experimental collision efficien-cies of Cryptosporidium oocysts were higher than those of Giardia cysts. The experimental collision efficiencies of Giardia cysts were ranged from 0.02 to 0.10 and those of Cryptosporidium oocysts were ranged from 0.13 to 0.65. This is because of the higher removal efficiency and the lower single-collector efficiency (Z) of Cryptosporidium oocysts. According to the study of Huang et al. (1999), the experimental collision efficiencies of algae (dp¼ 5:0 106m) in all trials were higher than cysts but lower than oocysts.
CONCLUSIONS
A semi-empirical approach for predicting collision efficiencies of Giardia cysts and Cryptosporidium oocysts for the colloid deposition in porous media was presented. The chemical composition of the suspension was changed by adjusting the electrolyte concentrations and pH. The results revealed that the experimental collision efficiencies and removal effi-ciencies of both protozoan parasites had similar trends in all trials. The experimental collision efficiencies of oocysts were higher than cysts. The experimental collision efficiencies of both parasites appeared to be increased with the electrolyte concentrations, and higher experimental collision efficiency of (oo)cysts filtration was observed at lower operating-pH. By adjusting the electrolyte concentrations, the variation in removal efficiencies and experimental collision efficiencies of both para-sites by the glass beads can be less significant than by the polystyrene beads. However, the contrary results between glass beads and polystyrene beads were obtained by adjusting the pH values.
REFERENCES
Adin A., Biswas P., Dutary G. and Sethi V. (1999) Cryptosporidium parvum as a waterborne particle in granular filtration. Proceedings of International Sympo-sium onWaterborne Pathogens, Milwaukee, Wisconsin, USA. 29 August–1 September.
Brush C. F., Walter M. F., Anguish L. J. and Ghiorse W. C. (1998) Influence of pretreatment and experimental con-ditions on electrophoretic mobility and hydrophobicity of Cryptosporidiumparvum oocysts. Appl. Environ. Micro-biol. 64(11), 4439–4445.
Donnelly J. K. and Stentiford E. I. (1997) The Cryptospor-idium problem in water and food supplies. Food Sci. Technol. 30(2), 111–120.
Elimelech M. and OMelia C. R. (1990) Kinetics of deposition of colloidal particles in porous media. Environ. Sci. Technol. 24(10), 1528–1536.
Elimelech M. (1992) Predicting collision efficiencies of colloidal particles in porous media. Water Res. 26(1), 1–8. Fig. 8. Experimental collision efficiency of (oo)cysts with
various concentrations of NaClO4at pH 5.6.
Fig. 9. Experimental collision efficiency of (oo)cysts with various pH in the presence of 0.01 M NaClO4.
Fogel D., Isaacrenton J., Guasparini R., Moorehead W. and Ongerth J. (1993) Removing Giardia and Cryptospor-idiumby slow sand filtration. J. Am. Water Works Assoc. 85(11), 77–84.
Frost F. J., Craun G. F. and Calderon R. L. (1996) Waterborne disease surveillance. J. Am. Water Works Assoc. 88(9), 66–75.
Haas C. N., Hornberger J. C., Anmangandla U., Heath M. and Jacangelo J. G. (1994) A volumetric method for assessing Giardia inactivation. J. Am. Water Works Assoc. 86(2), 115–120.
Huang C. P., Pan J. R. S. and Huang S. H. (1999) Collision efficiencies of algae and kaolin in depth filter: The effect of surface properties of particles. Water Res. 33(5), 1278–1286. Ongerth J. E. and Pecoraro J. P. (1996) Electrophoretic mobility of Cryptosporidium oocysts and Giardia cysts. J. Environ. Engng.-ASCE. 122(3), 228–231.
Pell A. N. (1997) Manure and microbes: Public and animal health problem?. J. Dairy Sci. 80(10), 2673–2681. Rajagopalan R. and Tien C. (1976) Trajectory analysis of
deep-bed filtration with the sphere-in-cell porous media model. AIChE J. 22, 523–533.
Smith R. M. and Martell A. E. (1976) Critical Stability Contants. Plenum Press, New York.
SoloGabriele H. and Neumeister S. (1996) US outbreaks of cryptosporidiosis. J. Am. Water Works Assoc. 88(9), 76–86.
Swertfeger J., Metz D. H., DeMarco J., Braghetta A. and Jacangelo J. G. (1999) Effect of filter media on cyst and oocyst removal. J. Am. Water Works Assoc. 91(9), 90–100.
Talaro K. P. and Talaro A. (1999) Foundations in Microbiology Basic Principles, 3rd ed. Mcgraw-Hill Companies, Columbus, OH.
