The Design of a Flow Test Apparatus and Experimental Methods for the Quantification of Viral Adhesion to Granular Activated Carbon

Spring 2005 Senior Design Project

Erica Zerfoss

Dept. of Bioengineering
The Pennsylvania State University, University Park, PA 16802

Sponsor: Dr. Bruce Logan, Kappe Professor of Environmental Engineering

 

Executive Summary

An estimated 3.4 million deaths a year are attributable to waterborne diseases.  Recently, the World Health Organization (WHO) has recognized point-of-use water treatment as an effective means of reducing illness in developing country households [1].  WHO also estimates that over one billion people do not have access to safe drinking water. A possible solution to these problems is to use activated carbon point-of-use filters in the plant and at the tap.

Activated Carbon is a highly porous granular material well known for its adsorption properties and ability to remove organic impurities from drinking water.  Because of its highly porous nature, a small amount of carbon is able to adsorb a large amount of dissolved organic matter [2].  Recently, attention has been given to activated carbon for its ability to adsorb viruses and bacteria; however there has not been much research in this area.

The way adhesion is quantified is by calculation of a sticking coefficient (a), which is a measure of the probability of attachment of colloids to a collector surface.  This mathematical model is based on the classic clean bed filtration theory proposed by Rajagopalan and Tien [3].

Currently, there are methods to calculate the sticking coefficient of bacteria on activated carbon using packed activated carbon columns.  This is accomplished by the radiolabeling of the bacteria and using a scintillation counter to quantify the amount adsorbed.  This method cannot work with the bacteriophage MS-2 because this particular virus does not adequately take up radiolabel.  Also, radiolabel is expensive and requires all lab workers to have the proper safety training.  Therefore, a new column flow experiment has been developed that involves the quantification of adhesion by serial dilution plating of influent and effluent as opposed to radiolabeling. 

Four variations of granular activated carbon (GAC) and silica beads have been tested using the bacteriophage MS-2.  Duplicate experiments were subject to student t-tests to determine the reproducibility of the results.  It was determined that the new flow experiment gives reproducible results with 95% confidence.

Objectives

bullet ·Design and production of a new flow test apparatus and experimental methods for the calculation of the sticking coefficient of bacteriophage  MS-2 to GAC. 
bullet ·Alter the mathematical model of the MARK test to meet the constraints of the new experiment.
bullet ·Run column experiments with MS-2 on four variations of GAC to check that the new experiment gives reproducible results.
bullet  

Design Criteria
  1. The flow test apparatus must provide a constant flow rate of 5 mL/min ± 1mL/min with no leakage.
  2. The design must not require the virus to be radiolabeled.
  3. The new experiment must give results that are reproducible (p>0.05).

Design of Flow Test Apparatus

The new flow experiment column was designed to fit on top of a vacuum box that allowed for filtrate collection.  The apparatus is fairly simple with a metal base that fits into the vacuum box and a circular piece of plastic with a hole for the syringe attachment glued to the top (Figure 1 and 2).  The syringe fits tightly into the plastic piece but is easily removed.  With a vacuum pressure of 1 psi, a flow of 5 mL/min is achieved.  The flow controller is a LuerLok connector.  The final product is shown in Figure 3.

 

 

 

Figure 3.    The flow experiment apparatus.  Adhesion is quantified by collection of  effluent.

 

 

Design of Experimental Methods

In the new two-layer column experiment, the filtrate from the column is collected and the fraction retained in the carbon bed is quantified by serial dilution plating. The test is referred to as a two-layer experiment because the suspension effluent is collected after passing through both a GAC packed bed and a supporting filter.  Because adhesion to the supporting filter must also be accounted for, column experiments are also run with the filter alone (Figure 4).

Figure 4.    Two layer column experiment.  The column on the left has both layers, a filter and a GAC bed.  The column on the right has only the first layer, the GF/D filter.

 

Mathematical Model
The bacteriophage concentration before and after flow through the column is quantified by serial dilution plating.  The fraction of viruses retained in the carbon bed is calculated from the fraction retained in the GF/D filter (Fr2) and the fraction retained in the column with both carbon and filter.  Figure 5 shows the setup with variables.

Figure 5. Two-layer column test setup with variables.  (a) Syringe with carbon supported with GF/D filter.  (b) Syringe with only a GF/D Filter.

 

 

Calculation of Fraction Retained in Carbon Bed
Fr2, the percent removed by the filter is calculated by:

 

N1, the number of bacteria not adhered by the carbon, can be calculated by:
 

 

Fr1 can now be calculated by:

 

 

Calculation of Sticking Coefficient
Once Fr1 is calculated, the sticking coefficient of the GAC, aGAC, can be calculated according to the RT clean-bed model [3]. The sticking coefficient is the probability of attachment of colloids to a collector surface.  Because the porosity of the GAC varied greatly in each experiment, the original RT model equation was altered and uses the dry mass and dry density of GAC instead of porosity.

 

The original sticking coefficient calculation:

 

 

The altered sticking coefficient calculation to account for variations in porosity:
 

where dc is the collector particle diameter, L is the length of the silica bed, θ is the bed porosity, Fr1 is the fraction of cells retained in the carbon layer, ρdry is the dry density of carbon, A is the cross-sectional area of the column, mdry is the dry mass of the carbon bed and η is the collision efficiency which is  the fraction of particles that collide with the collector. η is calculated according to the RT clean-bed model [3] and takes into account interception, sedimentation, diffusion and London van der Waals forces.

Design Performance

To test the reproducibility of the new column experiment, column experiments have been performed with MS2 using the four variations of carbon provided by Proctor and Gamble (Cincinnati, OH; 2004) and silica beads (VWR Scientific Products, Bridgeport, NJ) as a control. The sticking coefficients for duplicate experiments are given in Figure 6.

 

Figure 6. The sticking coefficients for duplicate column experiments calculated using the RT clean-bed model

Design Evaluation and Statistical Analysis

 

To determine if the new experimental methods and altered mathematical model give reproducible results, student t-tests were performed (a = 0.05). The results from duplicate experiments using the same collector were not statistically different (p>0.05).  Therefore the new column-flow experimental design and altered mathematical model is an effective method for the quantification of viral adhesion to granular activated carbon.
 

The Performance of GAC
The sticking coefficients ranged from 0.160 to 0.353 on GAC and 0.03 to 0.07 on silica beads.  The predicted log removals ranged from 0.31 to 0.95 for a 1-cm thick slice of GAC and 0.079 to 0.175 for a 1-cm thick slice of silica beads. GAC is an effective tool for the removal of viruses from drinking water.

Conclusions
The new 2-layer column experiment is an effective method for the quantification of viral adhesion to activated carbon.
GAC is an effective tool for the removal of viruses from drinking water.

Acknowledgements
The Kappe Environmental Engineering Laboratory
Dr. Bruce Logan
Proctor and Gamble
National Science Foundation

References
[1] Souter, P.F., Cruickshank, G.D., Tankerville, M.Z., Keswick, B.H., Ellis, B.D., Langworthy, D.E., Metz, K.A., Appleby, M.R., Hamilton, N., Jones, A.L., Perry, J.D.  (2003)  Evaluation of a new water treatment for point-of-use household applications to remove microorganisms and arsenic from drinking water.  Journal of Water and Health.  1:2, 73-84.
[2] Wagenet, L., Lemley, A.  (1995) Activated Carbon Treatment of Drinking Water.  Water Treatment Notes, Cornell University.
[3] Rajagopalan R. and Tien C. (1976) Trajectoryanalysis of deep-bed filtration with the sphere-in-cell porous media model.  A.I.Ch.E.J.  22, 523-533