Application of Colloidal Filtration Theory to Bacterial Attachment in Textile Fibrous Media Uporaba teorije koloidne fi ltracije za pritrjevanje bakterij na tekstilni vlaknati medij

A mechanism to remove Pseudomonas bacteria from contaminated water using textile fi brous media was proposed in this study. The attachment of Pseudomonas bacteria to nylon fi brous media was studied in a laboratory column experiment. A systematic study was carried out to investigate the attachment of bacteria to fi brous material as a function of media mass at a constant solution chemistry. The single-collector contact effi ciency and collision effi ciency from the interaction between bacteria and fi brous media were calculated applying a semi-empirical approach of clean-bed fi ltration theory. The experimental results indicate that removal effi ciency increases with an increase in media mass up to a certain level. This is due to the change in single-collector contact effi ciency and attachment/collision effi ciency, as observed from experimental data regarding removal effi ciency.


Tekstilec, 2018, 61(3), 171-178
Application of Colloidal Filtration Theory to Bacterial Attachment in Textile Fibrous Media water to form DBPs, many of which are carcinogens [3]. In addition, bacteria have developed chlorine-induced antibiotic resistance, meaning that a high dosage of disinfectant is required, leading to the formation of a signifi cant amount of DBPs [4]. Th ere is thus an urgent need to re-evaluate conventional disinfection methods and to consider some innovative approaches that enhance the reliability and robustness of disinfection, while avoiding the formation of DBPs. In the search for alternatives, some attempts have been made by researchers, resulting in the increased use of physicochemical fi ltration for the removal of bacteria from potable water and wastewater because of its simplicity, high effi ciency and low-costs. Th e attachment of bacteria to a fi lter surface is dictated by the adsorption mechanism, and this process does not produce by-products, such as those found in the chemical disinfection process used for water purifi cation [5]. Th e use of granular fi ltration marked the end of waterborne epidemics in the developed world more than a century ago. However, outbreaks of waterborne diseases continue to occur at unexpectedly high levels. Th e major limitations of granular fi lters are their capacity to retain colloidal particles within pore spaces and a low fi ltration rate [6]. In recent years, textile materials have emerged as a substrate to be used as a fi lter media for the removal of colloidal particles from surface water [7]. Superior performance in the removal of colloidal particles from water can be achieved using textile fi brous media, the fi ltration velocities of which are 10 ten times higher than granular fi lter media [8]. Considerable research has been done on the eff ect of physical and chemical factors that control the adsorption or attachment of bacteria in the physicochemical fi ltration process. Th e attachment of bacteria to diff erent materials in physicochemical fi lters depends on the media-suspension particle interaction [9]. Arnold et al. [10] evaluated the eff ect of cell concentration and fl uid velocity on bacteria attachment in diff erent fabric fi lters. Th ey found that cell concentration and fl uid velocity had no signifi cant eff ect on bacteria attachment, but found that the hydrophobic nature of fi lter media has an eff ect on bacterial attachment. Jewett et al. [11] investigated the eff ect of the ionic strength and pH of a suspending medium on bacteria attachment in a short-column experiment containing silica spheres, and found that the attachment of bacteria was significantly aff ected by the ionic strength of the solution, but that pH had no signifi cant eff ect on the attachment of bacteria. Torkzaban et al. [12] studied bacteria transport through quartz sand in a column experiment. Th ey reported that bacteria attachment not only depends on the solution chemistry, but also on the geometry of the fi lter media. Majumdar et al. [13] examined the eff ect of divalent salt and humic acid on bacteria removal through nonwoven polyester fabric. Th ey observed an increase in the attachment of bacteria at a higher concentration of bivalent salt, but that bacterial attachment decreases in the presence of humic acid. Th e above literature indicates that reported studies on bacteria fi ltration in column experiments found that a longer path bed containing glass bed, quartz, Ottawa sand and silica is required [14][15][16]. On the other hand, textile material is used primarily in the form of a fabric for bacteria fi ltration [17]. Th ere are very few studies where textile fi bres are used as a collector in a column experiment for bacteria fi ltration. Moreover, data regarding bacterial attachment behaviour on textile fi brous media as a function of media mass is limited. Th e specifi c objective of this study was to systematically examine the eff ects of media mass on bacterial attachment to a fi brous media surface at a constant solution chemistry of the suspending medium. Th e experimental observations are explained based on the colloidal fi ltration theory.

Filtration theory
Bacteria removal in a packed bed in a constant state can be described using the following one-dimensional fi ltration equation 1 [18]: where C is the bacteria concentration, L is the thickness of the fi lter bed, (1 -f) is the solid fraction, η is the experimental single-collector contact effi ciency, and d c is the collector diameter.
Integrating over the thickness of the packed bed yield: where F p is the fractional penetration and is an indicator of bathe lance between cell adsorption and desorption. Physical factors that account for particle collisions with a porous media are incorporated into the single-collector contact effi ciency, η. Th e single-collector contact effi ciency of a single media particle or collector (η) is the ratio of the number of bacteria that collide with the collector to the number that approach a collector. A variety of analytical solutions have been used to specify the single-collector contact effi ciency for aquasol. Th e Yao model represents analytical solutions for determining predicted single-collector contact effi ciencies based on spherical collectors that were proposed by Logan et al. [19].
where, η D,I and η G , represent theoretical values for the single-collector contact effi ciency when the sole transport mechanisms are diff usion, interception or sedimentation, respectively. Predicted single-collector contact effi ciency calculated numerically can be approximated by the following analytical expression: Predicted single-collector contact effi ciencies are dimensionless numbers and are developed from correlations involving the following dimensionless numbers: where, Pe represents the Peclet number, R and G represent the interception and gravitational numbers, U 0 and U P represent fi lter approach velocity and particle settling velocity, D represents particle diff usivity, d P represents the particle diameter and d c represents the collector diameter. Th e particle settling velocity is obtained from the formula: where, μ and ϑ represent the dynamic and kinematic viscosity of fl uid, g represents the gravitational constant, and ρ p and ρ f represent the particle and fl uid density. Th e particle diff usivity is obtained from the Stokes-Einstein equation as follows: where, k represents Boltzmann's constant and T represents the absolute temperature. Under conditions relevant to most aquatic systems, the experimental single-collector contact effi ciency (η) is lower than the predicted single-collector contact effi ciency (η 0 ) due to repulsive colloidal interactions between particles and collector grains [20]. Th e quantitative assessment of bacterial attachment to a collector surface is carried out by determining the collision effi ciency (attachment) factor (α), and is oft en expressed as the ratio of experimental single-collector effi ciency to the predicted single-collector effi ciency [21].
Attachment effi ciency represents the fraction of collisions (contacts) between suspended particles and collector grains that result in attachment.

Materials and methods
For the study, 100% polyamide 6 fi bres (Polyventure, Kolkata, India) of linear density 3.3 dtex and length 18 mm were used as packing material in column experiments. Th e microbial culture used in this study was Pseudomonas aeruginosa (gram-negative, rod-shaped) provided by the Department of Biotechnology, NIT Jalandhar (India). Also used were sodium hydroxide (NaOH), hydrochloric acid (HCl), sodium carbonate (Na 2 CO 3 ), sodium chloride (NaCl) and a nutrient broth (Deejay Corporation, Jalandhar, India). All these chemicals were laboratory grade and used as received.
Sample pre-treatment Nylon fi bres were scoured with a 0.5 g/L soda ash (Na 2 CO 3 ) solution at 60 °C for 15 minutes at a liquor ratio of 1:50 in order to remove added oils, lubricants, dust, etc. present on the fi bre surface. hydrochloric acid (HCl) and 0.1 M of sodium hydroxide (NaOH) were used to adjust the pH of the model test water to a value of 7. A fresh Pseudomonas aeruginosa culture was mixed with 200 mL of model test water to produce a fi nal concentration of 8.8 × 10 6 cells/mL. Aft er the packing of fi brous media, the column was fl ushed upward under a saturated condition with tap water for 10 minutes to ensure uniform packing and to release any trapped air bubbles. Th e fl ow was then reversed until the concentration of inlet and outlet equalised. Th e concentration was measured in terms of OD (optical density) using a spectrophotometer (Lambda 365, PerkinElmer). Prior to each experiment, the same pH and ionic strength as the model test water without bacteria was passed through the column to free the effl uent from background contaminants in the packed fi brous media. Th e model test water with bacteria was passed through a fi bre column, and outlet bacteria concentration was measured. Th e fl ow rate was measured as the time required to fi lter 200 mL of input water.

Optical density (OD) measurement
Th e optical density of the bacteria concentration in the inlet and outlet model test water was measured using a spectrophotometer (Lambda 365, Perk-inElmer) at 600 nm.

Calculation of single-collector contact effi ciency and collision effi ciency
Single-collector contact effi ciency and collision efficiency were calculated from the experimental values used for the quantitative analysis of the eff ect of media mass on bacterial attachment. Table 1 shows the parameters used in the calculation of single-collector contact effi ciency and bacterial collision (attachment) effi ciencies for the column experiment. Media porosity was calculated using equations 13-15 [22].

Eff ect of media mass on bacteria removal effi ciency
To study the eff ect of media mass on the bacteria removal effi ciency of textile fi brous media, the model test water was passed through a column packed with fi bres with various masses (i.e. 10 g, 12 g, 14 g, 16 g and 18 g). Th e media masses were chosen on the basis of preliminarily experimental observations. Th e experimental results are shown in Table 2 and Figure 1. It was found that bacteria removal effi ciency increased from 30% to 63% by changing media mass from 10 g to 16 g. Aft er that, there was no appreciable change in removal effi ciency (65% for a mass of 18 g), suggesting that the attachment of bacteria to the media surface depends on fi bre mass. Th e values in the table and fi gure represent the average values of three experiments. Th e incremental change in the removal effi ciency of bacteria by changing the media mass was explained by calculating the single-collector contact effi ciency and the attachment or collision effi ciency, which is discussed in the next paragraph.

Eff ect of media mass on single-collector contact effi ciency and collision/ attachment effi ciency
Th e values of single-collector contact effi ciency (η) and collision (attachment) effi ciency, (α) were calculated to make a quantitative comparison of removal effi ciency with various media masses under identical solution conditions. Collision effi ciency is defi ned as the ratio of the experimental single-collector contact effi ciency (η) and the predicted single-collector contact effi ciency (η 0 ). Th e results are presented in Table 3. Th e value of αwas calculated using equation 12. Th e values of predicted single-collector contact effi ciency were calculated using equation 6 and parameter values from Table 1, while experimental single-collector contact effi ciency (η) was calculated using equation 2. Attachment or collision effi ciency (α) is based on n measurements at a 95% confi dence interval.  It is evident from the Table 3 and Figure 2 that predicted single-collector effi ciency increases from 7.21 × 10 -3 to 8.12 × 10 -3 by changing the media mass from 10 g to 18 g. Th is is due to a decrement in the approach velocity of the fi ltration system from 3.04 × 10 -3 to 1.40 × 10 -3 m/s. Enhanced single-collector contact effi ciency through an increase in the media mass was therefore attributed to a change in the approach velocity of water in the fi ltration system. One potential explanation is that a high media mass may lead to the exposure of a higher surface area for the striking of bacteria as the result of high collector contact effi ciency. When the values of the single-collector contact effi ciency are higher, the probability of bacteria attaching to the surface of the fi bre will be high. Th is was in agreement with previous studies [23], in which it was concluded that the hydrodynamic system of the fi ltration process plays an important role in bacterial attachment.
Collision effi ciency is used for the quantitative determination of bacterial attachment to fi brous media. It is evident from Figure 3 that a change in media mass from 10 g to 16 g resulted in an increase in the collision effi ciency (α) from 0.11 to 0.18, followed by a decrease to 0.16 for a media mass of 18 g.

Figure 3: Collision effi ciency as a function of media mass
Th e value of collision effi ciency at 16 g was 0.18 and resulted in a removal effi ciency of 63%. A further increment in mass did not result in any appreciable change in removal effi ciency. Th is can be attributed to the decrement in collision effi ciency as observed.

Conclusion
It is important to understand the eff ect of physical factors on the performance of fi lters designed to remove microbial pathogens from surface water. An attempt was made to link an important factor of the textile porous media to enhance the removal efficiency of bacterial cells. Experimental evidence demonstrated that media mass can play an important role in bacterial removal and attachment. In this study, a fi lter media with various masses was selected for the experiment at a constant solution chemistry. Bacteria attachment and removal efficiency increased with an increase in media mass up to a certain level. According to colloidal fi ltration theory, this is possibly due to a change in the singlecollector contact effi ciency and collision (attachment) effi ciency of the fi brous media.