A systematic study was carried out to investigate the attachment of inorganic particles to fabric media in the presence of divalent cations and the dissolved organics in water, on filter media, or on both. The presence of organics (humic acid, HA) reduced the attachment of inorganic particles on the fabric. Specific attachment trends were observed for the inorganic particles–fibre system. For [CaCl2] < 500 ppm, the particle attachment was lower in the presence of 10 ppm of HA compared to its absence. At [CaCl2] ≥ 500 ppm, the attachment in the presence or absence of HA was similar, suggesting that the attachment was independent of the presence of HA. It was also found that the particle attachment to the fabric was lower when HA was present in water compared to when present on the fabric, suggesting that the attachment behaviour of inorganic particles was dependent on water chemistry (i.e. presence of calcium ions and organics in water), which also altered the surface properties of filter media. The removal trends were explained on the basis of particle aggregation, surface charge and Derjaguin–Landau–Verwey–Overbeek (DLVO) theory.

INTRODUCTION

Organics get easily adsorbed onto surfaces and hence can alter the characteristics of filter media and suspended particles, affecting deposition of the particles. Organic matter tends to compete with particles for attachment sites on the collector (Amirbahman & Olson 1995; Abudalo et al. 2010). Removal efficacy of microspheres reduced from 40 to 13% and for oocysts from 49 to 16% on spherical glass beads in presence of 5 ppm organics (Dai & Hozalski 2003). One of the primary reasons for reduction in removal is the change in the surface charge of particles in presence of organics. After exposure to 5 ppm of organics, the zeta potential became significantly more negative, microspheres being more negative than oocysts. This increased the repulsion between the particle and collector, leading to reduced removal. Similar trends of reduction in oocyst or its surrogates (latex beads or microspheres) attachment efficiency were reported elsewhere (Davis et al. 2002; Franchi & O'Melia 2003; Pelley & Tufenkji 2008; Liu et al. 2009; Abudalo et al. 2010). The other reason was surface conformation of humic substances in presence of electrolytes (Amirbahman & Olson 1995). Positively and negatively charged latex spheres were coated with two types of humic substances and their attachment efficiencies to quartz collectors were determined for varying NaCl concentrations. The deposition behaviour of colloids was found to be highly dependent on the surface conformations of the adsorbed humic acid (HA). The differences in steric interactions brought by the surface conformations were primarily responsible for the observed difference in deposition behaviour of coated colloids.

The separate and combined effects of Na+, K+ or Ca2+ and HA on colloid transport through porous granular media were also reported (Tipping & Cooke 1982; Davis et al. 2002; Dai & Hozalski 2002; Franchi & O'Melia 2003; Pelley & Tufenkji 2008; Foppen et al. 2008; Dagaonkar & Majumdar 2012). The role of HA on decrease in surface charge was more significant than increase of surface charges due to calcium ions. HA adsorption on collector surface increased with Ca2+ concentration, making the collector surface more negative, which resulted in lowering of removal efficiency of latex particles.

Fibrous media are reported to be equally efficient in particle removal compared to granular materials (Dagaonkar & Majumdar 2012). Very little has been reported about the particle attachment in the presence of dissolved organics on collector, water or both. The specific objective of this study is to examine in a systematic way the combined effects of dissolved organics and divalent cations on the attachment of inorganic particles on fibrous media. The experimental observations are explained based on aggregation of particles, surface charge and Derjaguin–Landau–Verwey–Overbeek (DLVO) theory.

MATERIALS AND METHODS

Materials

Non-woven polyester fabric with a GSM (gram per square metre) of 500 and thickness of 2 mm was procured from Travencore Fibres Private Limited, Mumbai, India. The inorganic particles were procured from Powder Technology Inc., USA. The chemical composition of the particles is 68–76% SiO2, 10–15% Al2O3, 2–4% Na2O, 2–5% K2O, 2–5% Fe2O3, 2–5% CaO, 1–2% MgO, and 0.5–1% TiO2. The particle size distribution in de-ionized water was d(0.5) ∼ 2.34 μm, d(0.1) ∼ 1.16 μm, and d(0.9) ∼ 4.93 μm. Calcium chloride dihydrate (CaCl2.2H2O) and sodium salt of HA were supplied by Merck, India and Sigma Aldrich, Germany, respectively.

Methods

The fabric was cut into the required dimensions, rolled manually over an acrylic cylinder to form four layers of wound fabric and sealed across the width by an adhesive. The acrylic cylinder was then removed and the top and bottom side of the fabrics were sealed using plastic plates. The top plate was completely closed and the bottom plate had a throttle of 1.5 mm diameter for water flow. The outer diameter and height of the filter (schematic as shown in Figure 1) were 86 mm and 60 mm respectively. The depth of the filter was 8 mm and filter surface area was 162 cm2. The filter was dipped in water (containing total dissolved salts (TDS) of 100 ppm, pH of 7.0) having 10 ppm HA for 15 days to achieve uniform coating of organics on the fabric surface.

Figure 1

The schematic of filter design.

Figure 1

The schematic of filter design.

Performance parameters

Attachment of inorganic particles. The effect of organics on attachment of inorganic particles was studied as follows. Four types of studies were conducted, the details of which are as shown in Table 1.

Table 1

Details of the experiments to study removal of inorganic particles

Case numberHA on fabricHA in model test water
No No 
No Yes 
Yes No 
Yes Yes 
Case numberHA on fabricHA in model test water
No No 
No Yes 
Yes No 
Yes Yes 

Model test water was prepared by adding calcium chloride (10 to 1,000 ppm) in Milli-Q water (TDS-0 ppm, pH-6.5) with and without 10 ppm of HA. The concentration of inorganic particles was 15 ppm. The filter was inserted into a 10 L chamber that was placed over a stand. Filter flow rate at 17 cm water column was ∼230 mL/min and falling head conditions were maintained in all experiments. Output samples were collected after every litre to analyze turbidity. Turbidity of the water sample was analyzed by Merck Turbiquant 1500 turbidity meter.

The attachment was calculated in terms of % turbidity removal as follows: 
formula
1
Since turbidity of water is directly dependent on the concentration of inorganic particles of this size, 
formula
2
  • pH. Water pH was measured by Control Dynamic pH meter manufactured by CD Instrumentation Private Limited, Bangalore.

  • Total dissolved solids. Total dissolved solids of water was measured by using Equinox TDS meter, model TDS 5031.

  • Surface charge (zeta potential). The estimation of zeta potential was carried out using zeta meter. The suspended particles (fibres or inorganic particles) were added to model test water. The cell was filled with this solution and the electrodes (molybdenum as an anode and platinum as cathode) were fitted to the cell. The electrodes were energized at 150 V and the particle movement was tracked to measure the zeta potential.

  • Particle size distribution. The particle size distribution was obtained using a Malvern particle size analyzer, Model number Mastersizer 2000.

RESULTS AND DISCUSSION

Effect of organics and Ca2+ on attachment of inorganic particles

Figure 2 shows the effect of presence of HA on the attachment of inorganic particles at various [CaCl2]. The particle attachment increased with the increase in [CaCl2] in the presence and absence of HA. In the absence of HA on fabric and in water (Case 1), the C/C0 values reduced from 0.68 to 0.03 with the increase in [CaCl2] from 10 to 1,000 ppm suggesting increase in particle attachment. The attachment was found to be maximum in the absence of HA at all calcium levels.

Figure 2

Effect of HA on particle attachment at various [CaCl2] ([CaCl2] in (a) = 10 ppm, (b) = 100 ppm, (c) = 200 ppm, (d) = 500 ppm, (e) = 1,000 ppm; all experiments are replicates).

Figure 2

Effect of HA on particle attachment at various [CaCl2] ([CaCl2] in (a) = 10 ppm, (b) = 100 ppm, (c) = 200 ppm, (d) = 500 ppm, (e) = 1,000 ppm; all experiments are replicates).

The attachment behaviour of inorganic particles in presence of HA was dependent on [CaCl2] in water. For [CaCl2] < 500 ppm, the presence of 10 ppm HA (either on the filter surface or in model test water or in both) lowered the attachment of inorganic particles compared to its removal in absence of HA. The particle attachment when HA was present in water (Case 2) and on fabric and water (Case 4) was comparable and lower compared to absence of HA (Case 1) or presence of HA on fabric. This suggested that the attachment behaviour of inorganic particles was strongly dependent on water chemistry compared to the surface properties of the filter media. The attachment of particles followed a decreasing trend in the order of Case 1 > Case 3 > Case 2 ∼ Case 4.

For [CaCl2] ≥ 500 ppm, the attachments in all cases were similar, suggesting that at higher calcium levels, the contribution of HA to reducing the attachment got negated and was independent of the presence of HA.

The trends reported above related to the specific role of HA when present in water, fabric and both and the specific particle attachment trends due to calcium ions are new and not reported elsewhere. The explanation for these trends is reported below.

Explanation of results

The possible mechanisms to explain the particle attachment to the collector surface in the presence of divalent cations and organics are:

  • (1)

    change in surface charge of particles and collector;

  • (2)

    change in the particle size (and distribution) of particles;

  • (3)

    change in energy barrier between particle and collector;

  • (4)

    other mechanisms (e.g. steric interactions/repulsion).

Surface charge of inorganic particles and fibres

Zeta potential of inorganic particles in model test water with and without 10 ppm HA was measured and is shown in Table 2. In the absence of HA in water, the average zeta potential of inorganic particles increased from −16 to −8 mV with the increase in [CaCl2] in water from 10 to 500 ppm, and then levelled off at higher calcium loading. In the presence of 10 ppm HA in water, the average zeta potential of the particles increased from −20 to −9 mV with the increase in [CaCl2] in water from 10 to 500 ppm, and then levelled off at higher calcium loading. This explains the observation that attachment of particles improved with the increase in [CaCl2] up to 500 ppm and then levelled off in all the cases. Reduction in zeta potential in presence of HA indicates that organics tend to get adsorbed on the inorganic particles and hence reduce their surface charge. This explains lowering of particle attachment in presence of HA in water and on fabric (Cases 2, 3 and 4) compared to its absence (Case 1), as shown in Figure 2.

Table 2

Zeta potential of inorganic particles and fibres at various [CaCl2]

[CaCl2] (ppm)Zeta potential of inorganic particles (mV)Zeta potential of fibres (mV)
Without HAWith HAWithout HAWith HA
10 −16 −20 −38 −40 
100 −13 −17 −25 −29 
200 −12 −15 −15 −24 
500 −8 −9 −12 −22 
1,000 −8 −9 – – 
[CaCl2] (ppm)Zeta potential of inorganic particles (mV)Zeta potential of fibres (mV)
Without HAWith HAWithout HAWith HA
10 −16 −20 −38 −40 
100 −13 −17 −25 −29 
200 −12 −15 −15 −24 
500 −8 −9 −12 −22 
1,000 −8 −9 – – 

Zeta potential of fibres of the fabric with and without 10 ppm HA was measured and is shown in Table 2. It was found that the average zeta potential of the fibres was −38 mV in absence of HA and −40 mV in presence of 10 ppm HA at 10 ppm [CaCl2]. Addition of calcium chloride increased the surface charge of the fibres in presence and in absence of HA. In the absence of HA, the average zeta potential of fibres got increased from −38 to −12 mV with the increase in [CaCl2] in water from 10 to 500 ppm. In the presence of HA on fabric surface, the average zeta potential of fibres increased from −40 to −24 mV with the increase in [CaCl2] in water from 10 to 200 ppm, and then levelled off at higher calcium loading.

Particle size distribution

The average size data (d50) for the inorganic particles at various [CaCl2] were measured. It was observed that addition of calcium chloride in absence of HA increased the net charge of the particles, which induced flocculation effects, i.e. increased the probability of particles to come close to each other, resulting in formation of aggregates. The average particle size increased from 2.5 to 3.2 μm with an increase in [CaCl2] from 10 to 500 ppm and then levelled off at higher calcium levels. The variation in the size range of the aggregates in presence of HA was found in the range of 2.5 to 2.9 μm. This explains that the presence of dissolved organics hindered the particle aggregation in suspension, more likely due to steric repulsion, which led to less removal.

Energy barrier between particle and collector

Increase in the particle removal with the increase in [CaCl2] (in presence and in absence of organics) can be explained by DLVO theory.

The theoretical potential energy of interaction, EDLVO, between the particle and the collector was calculated by assuming a cylindrical collector as a flat plate and the particle as a sphere, respectively. The energy of interaction is written as 
formula
3
where, Evdw is the interaction potential due to van der Waals forces and EEDL is the interaction potential due to electrical double layer repulsion force. Evdw is determined as 
formula
4
EEDL is given by the Wiese and Healy expression for a sphere–plate system as 
formula
5
 
formula
6
where, H is the Hamaker constant for the system collector–water–particle, rp is the particle radius, φ is the reduced potential of either the collector or the particle, ζ is the surface potential as determined by the streaming potential for the collector or the particle. Equation (5) is valid for systems with minimal electrical double-layer distortion (k rp > >1) where δ << rp and for small surface potentials, ζ.

Presence of calcium chloride increases the surface charge of the particles and the collector hence reducing the energy barrier and increasing the depth of secondary minima for better removal. Figure 3 shows that in all cases, the energy barrier reduced with the increase in the calcium loading, and hence favoured the attachment. The energy barrier in presence of HA in water (Case 2) and in water and fabric (Case 4) was higher than the energy barrier in the absence of HA (Case 1) and HA on fabric alone (Case 3) at all [CaCl2]. The energy barrier was in the order of Case 4 > Case 2 > Case 3 ∼ Case 1, which was in line with the removal data of particles.

Figure 3

Energy of interaction curves for particle-fibre system at various [CaCl2] with and without HA ([CaCl2] in (a) = Case 1, (b) = Case 2, (c) = Case 3, (d) = Case 4).

Figure 3

Energy of interaction curves for particle-fibre system at various [CaCl2] with and without HA ([CaCl2] in (a) = Case 1, (b) = Case 2, (c) = Case 3, (d) = Case 4).

To summarize, in addition to specific action caused by Ca2+ in promoting particle adhesion (Davis et al. 2002) through calcium bridging, the reduction in the absolute value of surface charge (Table 2) led to reduction in the repulsive force between particles (Figure 3), which in turn caused aggregation, thereby improving particle attachment to the fabric.

CONCLUSIONS

The effect of dissolved organics on attachment of inorganic particles through a porous fabric filter was studied in the presence of HA in water, on fabric, and both, at different [CaCl2]. The results were explained in terms of changes in zeta potential, particle size distribution and DLVO theory. The conclusions are summarized as follows:

  • (1)

    A specific removal trend was observed for removal of inorganic particles. For [CaCl2] < 500 ppm, the particle removal was lower in the presence of 10 ppm of HA compared to its absence. At [CaCl2] ≥ 500 ppm, the removal efficiencies were almost similar, suggesting that the removal was independent of presence of HA.

  • (2)

    The particle attachment when HA was present in water (Case 2) and on fabric and water (Case 4) was comparable and lower compared to absence of HA (Case 1) or presence of HA on fabric (Case 3), suggesting that the attachment behaviour of inorganic particles was dependent on water chemistry (i.e. presence of calcium ions and organics in water).

  • (3)

    Addition of organics decreased the zeta potential of the particles and the fibres. Addition of CaCl2 at the lower levels increased the zeta potential of particles and fibres but levelled off at higher [CaCl2] (≥500 ppm for particles and ≥200 ppm for fibres).

  • (4)

    The size of the particles increased with the increase in [CaCl2] and was lower in the presence of HA in presence of CaCl2.

  • (5)

    DLVO theory predicted reduction in energy barrier with increase in [CaCl2] and increase in energy barrier in presence of HA.

REFERENCES

REFERENCES
Amirbahman
A.
Olson
T. M.
1995
The role of surface conformations in the deposition kinetics of humic matter-coated colloids in porous media
.
Colloids Surfaces A: Physicochem. Eng. Aspects
95
,
249
259
.
Dagaonkar
M.
Majumdar
U.
2012
Effect of fluid flow, solution chemistry and surface morphology of fibrous material on colloid filtration
.
J. Eng. Fibres Fabrics
7
,
62
74
.
Davis
C. J.
Eschenazi
E.
Papadopoulos
K. D.
2002
Combined effects of Ca2+ and humic acid on colloid transport through porous media
.
Colloid Polym. Sci.
280
,
52
58
.