Journal of Water and Environment Technology, Vol. 9, No.

3, 2011

Development of a Simple, Effective Ceramic Filter for

Arsenic Removal

Md. SHAFIQUZZAMAN*, Md. Mahmudul HASAN*, Jun NAKAJIMA*, Iori MISHIMA**

*Department of Environmental Systems Engineering, Ritsumeikan University, 1-1-1

Nojihigashi, Kusatsu, Shiga 525-8577, Japan.
**Center for Environmental Science in Saitama, 914 Kamitanadare, Kazo, Saitama 347-0015,
Japan.

ABSTRACT
A simple, low-cost filtration system composed of a ceramic filter, an iron net and iron bacterial
sludge was developed to remove arsenic (As) from groundwater. The ceramic filter, made of an
80% clay soil and 20% rice bran mixture on a weight basis, was combined with the iron net and
the iron bacterial sludge in a reactor, and the assembly (As removal filter) was tested for its
ability to remove As from synthetic groundwater. Synthetic groundwater with a varying
composition of As(III), ferrous iron (Fe(II)) and phosphate phosphorus (P) was filtered on a
daily basis. The results showed that both Fe(II) in groundwater and those released from the iron
net were oxidized biologically and/or physico-chemically and that As and P were effectively
removed by adsorption and/or co-precipitation processes. The concentrations of Fe(II) and P in
groundwater were decisive factors in the removal of As . Groundwater Fe concentrations of ≥ 2
mg/L and P concentrations of ≤ 3 mg/L with an Fe/P molar ratio of ≥ 3.0 were required to
achieve an effluent As concentration of less than 50 µg/L from raw water containing 500 µg/L of
As(III). This simple, inexpensive filter could be used to treat As in contaminated regions.

Keywords: arsenic removal, ceramic filter, groundwater, iron bacterial sludge, iron and
phosphorus; iron net

INTRODUCTION
Arsenic (As) contamination in groundwater has caused a devastating health crisis all
over the world, especially in Bangladesh and West Bengal, India (Dhar et al., 1997). In
Bangladesh and other regions of the Indian subcontinent, 70 million people consume
drinking water with As concentrations in excess of maximum contamination level
(MCL) set by the World Health Organization (WHO, 1996) and US Environmental
Protection Agency (USEPA, 2002) (10 µg/L) and by Bangladesh (50 µg/L). Ingestion
of inorganic As can cause cancer of the skin, lung and bladder as well as noncancerous
health problems (Mazumder et al., 1998; Smith et al., 1998). Arsenic in groundwater is
mainly present in inorganic forms, such as arsenate (As(V)) and arsenite (As(III)).
Arsenate is the predominant species in atmospheric conditions or in more oxidizing
environments in the pH range of 6-9. It exists predominantly as oxyanions, such as
H2AsO4- or H2AsO42- (HAsO42-). Arsenite is thermodynamically stable and exists
predominantly as H3AsO3 or HAsO2 under mildly reducing conditions. Arsenite, the
most common species in anaerobic groundwater and the more toxic of the two forms
(Pontius et al., 1994), is generally removed less efficiently than the oxidized As(V)
(Dixit and Hering 2003). Arsenic release from subsurface minerals and sediments is the
main cause of As contamination in groundwater (Chowdhury et al., 2000).

Different techniques for the removal of As from groundwater have been developed and
tested. These include co-precipitation (e.g., with iron or aluminum salts), ion exchange,

Address correspondence to Jun Nakajima, Department of Environmental Systems Engineering,

Ritsumeikan University, Email: [email protected]
Received May 4, 2011, Accepted Jury 22, 2011.
- 333 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

and adsorption by activated carbon and membrane processes (Kartinen and Martin,
1995). Most of these methods require a pre-oxidation step to transform As(III) into
As(V). Arsenic removal by the application of dissolved Fe(II) and metal Fe(0) has
received special attention due to the high removal efficiency of these treatments. Recent
studies have shown that As(III) is partially oxidized to As(V) during the physical-
chemical oxidation of dissolved Fe(II), resulting in a higher As removal efficiency than
that obtained by direct addition of solid Fe(III) (Hug and Leupin, 2003; Roberts et al.,
2004; Berg et al., 2006). Numerous studies of As removal by the application of Fe(0)
have also been reported. In aerobic water, the slow, continuous release of dissolved
Fe(II) from Fe(0) and the subsequent oxidation to solid Fe(III) effectively remove As
(Su and Puls, 2001a, 2001b and 2003; Leupein and Hug, 2005; Lien and Wilkin, 2005).

Dissolved Fe(II) can be oxidized by both physical-chemical and biological processes,

but the dominant process depends on the physical and chemical characteristics of the
raw water and on the process conditions. Biological oxidation of dissolved Fe(II) in
water is caused by several iron-oxidizing microorganisms such as Gallionella sp. and
Leptothrix ochracea (Czekalla et al.,1985; Mounchet, 1992; Michalakos et al.,1997).
The main product from the biological oxidation of dissolved Fe(II) is usually a poorly
ordered iron oxide containing significant amounts of organic matter. Products from the
intermixing of iron oxides, organic matter and bacteria exhibit unique metal retention
properties (Ferris et al., 2000). Arsenic can directly adsorb onto the biological iron
oxides. In addition, As(III) oxidation during biological oxidation of dissolved Fe(II)
may occur (Katsoyiannis and Zouboulis, 2004; Shafiquzzaman et al., 2008).

Several small-scale As removal technologies have been applied in the field in

Bangladesh (Khan et al., 2000; Hussam and Munir, 2007). Although these technologies
have a high As removal efficiency, they also have certain drawbacks, including high
cost and maintenance difficulties associated with the use of sand beds for filtration
(Shafiquzzaman et al., 2009). In contrast, the ceramic filtration process, which uses
porous ceramics (fired clay) to filter microbes or other contaminants from drinking
water, has many potential advantages. Ceramic filters can be manufactured with locally
available materials. They are low-cost, socially acceptable and can function with
minimal maintenance (Chaudhuri et al., 1994; Oyanedel-Craver and Smith, 2008).
Therefore, biological Fe(II) oxidation, co-precipitation with As and subsequent filtration
through a simple ceramic filter instead of a sand bed could be a viable set of strategies
for the removal of As from groundwater.

A locally made filter, called the soil ceramic candle, has been used in rural area of
Bangladesh for the removal of iron (Fe) from groundwater. The arsenic removal
performance of this filter was evaluated, and a modification to the filter was proposed in
our previous study (Shafiquzzaman et al., 2011). In this study, a simple, low-cost As
removal filter composed of a ceramic filter, an Fe(0) net and iron bacterial sludge was
developed and evaluated for its ability to remove As from synthetic groundwater
containing varying compositions of As(III), Fe(II) and P. The manufacturing procedure
for the ceramic filter is described here in detail.

- 334 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

MATERIALS AND METHODS

Ceramic filter manufacturing and characterization
Ceramic filters were manufactured by mixing 80% clay soil and 20% rice bran on a
weight basis. The soil sample and the rice bran were collected from the local brick field
and rice mill, respectively, of the southwestern region (Khulna) of Bangladesh. The soil
sample was dried, ground with a hammer and passed through a 0.5 mm sieve. The rice
bran was also dried and passed through a 1 mm sieve. A particle size distribution of the
collected soil was measured by the laser diffraction method (Shimadzu SALD 3000,
Japan). The mixing ratio of the ingredients was selected after testing the filtration flux
and flexural strength of several ceramic bars prepared with different ratios of soil and
rice bran. The filtration fluxes of ceramic bars made with 0%, 15%, 20% and 25% rice
bran at 28.5 cm of water head were 0.11 ± 0.02, 0.16 ± 0.01, 0.53 ± 0.08 and 1.92 ±
0.39 mL/cm2/min, respectively. The flexural strengths of the bars were 0.75 ± 0.19, 0.48
± 0.15, 0.27 ± 0.14 and 0.08 ± 0.01 kPa, respectively. The results indicated that fired
ceramic bars with a higher percentage of rice bran (25%) were low in strength and
easily broken, while those with lower percentages (10% and 15%) produced a relatively
low filtration flux. On the basis of these results, a mixture of 80% soil and 20% rice
bran was selected for ceramic filters in this study.

To manufacture the ceramic filters, soil and rice bran were combined and mixed until
homogeneous. This dry mixture (800 g) was combined with water to make dough. The
dough was then molded by hand, placed in a polyvinylchloride cylindrical mold (10 cm
diameter and 10 cm height), and compressed manually. The resulting cylindrical
ceramic filters were hollow with one side open. The final ceramic filters had a height of
10 cm and a thickness of 2 cm (Fig. 1a). The filters were sun-dried for 48 - 72 h and
then fired in a muffle furnace in the laboratory or in a small-scale pottery kiln in the
field at 900°C. The ceramic filter manufacturing cost was estimated to be 0.2 - 0.3 USD
per filter.

The apparent porosity of the ceramic filter was tested according to Yang et al. (2007).
The pore size of the filter was estimated by comparing the particle size distribution of
turbid waters (water with clay mixture) before and after filtration with a ceramic filter.

Iron net (Fe(0))

A commercially available iron net (with wire diameter of 0.75 mm) was used without
treatment as a source of Fe(0). The mean surface area and Fe content of the iron net
were 654 ± 28.4 mm2/g and 99 - 99.5%, respectively. The iron net was cut, and the open
cubic boxes (11 × 11 × 11 cm) were formed manually. These boxes were used to
construct the As removal filter.

Iron bacterial sludge

Iron bacterial sludge was collected from natural river sediments (Ogure River, Gifu
Prefecture, Japan.). Microscopic views of the bacterial sludge showed that the sediment
mainly contained the Leptothrix ochracea iron oxidizing species with iron hydroxide
flocs (image not shown). The collected sludge was cultured in the laboratory on a sand
bed with a continuous tap water flow. Twenty-five milligrams of FeSO4·7H2O diluted in
1 L of deionized water was fed daily onto the bed as the sole source of ferrous iron for
bacterial growth. When required, sludge was taken from the bed, weighed on the basis

- 335 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

26 cm
Reactor

Ceramic filter
Raw water

37.5 cm
Iron net

Iron Bacteria
Sludge

10 cm Effluent bucket

Wooden
stand
8 cm
10 cm

6 cm 22.5 cm
(a) (b)

Fig. 1 - Schematic representation of (a) the ceramic filter and (b) the arsenic removal
filter unit

of the suspended solids concentration (mg/L) and used for filtration experiments.

Filter construction
The As removal filter used to treat As in groundwater was composed of a reactor, a
wooden stand and an effluent bucket (Fig. 1b). The reactor was a plastic bucket (15 L)
with a small hole at the bottom to drain effluent water. The opening of the ceramic filter
was affixed to the bottom of the reactor with acryl glue and covered with a cubic iron
net box. Before starting each filtration experiment, the iron bacterial sludge was added
to the reactor. The reactor was placed on the wooden stand, and the effluent bucket was
positioned under the reactor.

Stock solution
The chemicals, all of which were reagent grade, were from either Wako, Japan, or
Nacalai Tesque Inc., Kyoto, Japan. An As(III) stock solution (100 mg/L) was prepared
by dissolving arsenic oxide (As2O3) in deionized water with 5 ml/L of HCl. An Fe(II)
stock solution was prepared by dissolving FeSO4·7H2O in deionized water immediately
before the start of each experiment to avoid premature Fe(II) oxidation. A phosphorus (P)
(1000 mg/L) stock solution was prepared by mixing reagent grade K2HPO4 with
deionized water.

Synthetic groundwater
Synthetic groundwater, representing the geochemistry of Bangladesh groundwater, was
prepared in the laboratory by dissolving the appropriate chemicals (Table 1) (BGS and

- 336 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

DPHE, 2001). Each liter of synthetic groundwater consisted of 10 mg of NH4C1, 500

mg of MgSO4·7H2O, 1.2 mg of NaCl, 400 mg of CaCl2·2H2O, 2 mg of MnSO4·5H2O
and 500 mg of NaHCO3. Synthetic groundwater with varying compositions of As(III),
Fe(II) and P was prepared by diluting stock solutions prior to each filtration experiment.
The pH value was adjusted to 7.0 - 7.4 for every series of experiments by adding
sodium hydroxide or hydrochloric acid.

Filtration experiments
In the filtration experiments, 14 L of synthetic groundwater containing the compounds
of interest (Table 2) was poured daily into the reactor. The water flowed gravitationally
through the filter for 4 - 6 h. The iron net and the ceramic filter in the reactor were fully
immersed during the filtration period (4 - 6 h) and partially immersed during the
remaining 18 - 20 h of the day. Effluent water was sampled daily from the effluent
bucket. The redox potential (ORP), dissolved oxygen (DO), pH and flow rate were
measured daily. Samples were kept in the refrigerator at 4°C until the As, Fe and P
analyses. A total of 17 series of experiments were conducted (Runs 1 - 17) to assess the
As(III) removal performance (Table 2). All laboratory experiments were focused on the
removal of As(III), because As(III), the most common arsenic species in groundwater, is
generally removed less efficiently than the oxidized As(V) (Dixit and Herring, 2003).

Table 1 - Composition of synthetic groundwater used in this study and Bangladesh

groundwater as obtained from BGS data base (BGS and DPHE, 2001)

Synthetic groundwater Bangladesh groundwater

pH 7.5±0.5 7.0±0.2
Ca (mg/L) 100 77± 56
Mg ( mg/L) 40 31±20
Fe ( mg/L) 0-5 5.4±4.8
As (µg/L) 500 200±167
P ( mg/L) 0.9-10 1.5±1.48
Mn ( mg/L) 0.45 0.6±0.7

Table 2 - Experimental details

Experimental As(III) Fe (II) P Fe(0) net Iron bacterial
Runs (µg /L) (mg/L) (mg /L) (g) sludge (mg)

Run 1 500 5 0.9 87 160

Run 2 500 5 0.9 - 160
Run 3 500 5 5 300 160
Run 4 500 5 0.9 300 750
Run 5 500 5 0.9 300 1500
Run 6 500 5 0.9 300 5000
Run 7 –17 500 0-5 0.9-10 600 5000
Run 1-3 were carried out for 70 days and Run 4 - 17 were carried out for 14 days.

- 337 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

Analytical method
Arsenic was analyzed using a polarized Zeeman GFAAS (atomic adsorption
spectrophotometer) (Hitachi Z2700, Japan) or an ICP-MS (inductively coupled plasma
mass spectrometry) (Yokogawa, HP 4500, Japan). The total Fe concentration was
measured using ICP spectrometry (Seiko SPS 4000, Japan). Phosphate was measured
with the molybdenum blue colorimetric method (JEMAI, 2008). The pH, ORP and DO
were measured with a pH meter (HORIBA-D-54S, Japan), an ORP meter (HORIBA-D-
54S, Japan) and a DO meter (HACH-HQ30d, Japan), respectively.

RESULTS AND DISCUSSION

Soil and filter characteristics
The particle size distribution of the soil sample showed that the soil used in the ceramic
filters contained approximately 40% sand, 52% silt and 8% clay. The porosity and pore
size of the filter were 60% ± 1% and 1 - 5 µm, respectively.

As(III), Fe(II) and P removal

The efficiency of As(III) removal with and without the iron net was investigated in Run 1
and Run 2 (Fig. 2a). As shown in Fig. 2a, the As(III) removal efficiency was clearly
higher in Run 1 (with the iron net) than in Run 2 (without the iron net). In Run 1, As(III)
was removed gradually from 500 µg/L to 50 µg/L over 20 days. After 20 days, As in the
effluent decreased below 50 µg/L. In contrast, removal was insufficient in Run 2, with a
constant effluent As concentration of 120 µg/L after 20 days. The concentration of
effluent Fe is shown in Fig. 2b. According to the figure, less than 0.4 mg/L-1 of total Fe
was present in the effluent on the first day, and this concentration remained constant
throughout both runs. In contrast to As, the P concentration was effectively reduced from
an initial concentration of 0.9 mg/L to less than 0.3 mg/L (Fig. 2c).

The pH of the effluent (7.5 - 7.7) was slightly higher than that of the influent (7.0 - 7.4)
in both runs. This expected pH increase was attributed to water decomposition by the
iron net and to the sorption reaction of arsenic, which releases OH- groups from sorbents
as a result of ligand exchange (Su and Puls, 2001a). The influent ORP values, which
ranged from -50 to -150 mV, indicated that Fe in the synthetics groundwater exists as
dissolved Fe(II). The effluent ORP (100 - 200 mV) was more oxidizing than that of the
influent (-50 to -150 mV). The DO levels of the influent (9 ± 1 mg/L) and effluent (8 ±
1 mg/L) indicated oxidation in the reactor. The flow rate remained nearly constant (3.5
± 0.5 L/h for Run 1 and 2.9 ± 0.5 L/h for Run 2) throughout the operation (70 d),
indicating that the filters were not clogged. Roughly 60 - 70 L/d of water, an adequate
amount to supply the drinking and cooking water needs of a middle-sized family of 5 - 6
people in a developing country (Shafiquzzaman et al., 2009), could be treated at this
flow rate.

Effect of iron bacterial sludge on As(III) removal in early phase

In field applications, As concentrations must be reduced below the Bangladesh drinking
water standard (50 µg/L). Results from Run 1 (Fig. 2a) indicated that roughly 20 days
were required to reach As concentrations below this limit. This may have been a result
of insufficient iron in the reactor during the initial phase of operation. Therefore, As(III)
removal was examined by increasing the initial amount of iron bacterial sludge in Runs

- 338 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

500

Δ Run
Run 11 ▲Run2
Run2 ●Run3
Run3 (a)
400

(μg/L)
As(μｇ/L)
300

EffluentAs
Effluent Bangladesh standard limit of As
200

100

0
(b)
(mg/L)
Fe (mｇ/L)

4.04

3.03
Effluent Fe
Effluent

2.02

1.01

0.0

4.0 (c)
(mg/L)
EffluentPP(mｇ/L)

3.0

2.0
Effluent

1.0

0.0
0 10 20 30 40 50 60 70
Time (day)
Fig. 2 - Removal of (a) A(III), (b) Fe and (c) P by the arsenic removal filter

4, 5 and 6. Initially, 750, 1,500 and 5,000 mg of iron bacterial sludge was added for
Runs 4, 5 and 6, respectively (Table 2). The corresponding effluent As concentrations of
less than 50 µg/L were achieved at 13, 6 and 1 days of operation, for Runs 4, 5 and 6,
respectively (Fig. 3), and these results suggested that iron bacterial sludge added at the
beginning of the experiment played an important role in the early phase of filter
operation. Iron bacterial sludge contains biomass with biological iron hydroxides.
Increasing the amount of this bacterial sludge substantially increases the amount of
biogenic iron hydroxides, resulting in faster reduction of As concentration to less than
50 µg/L in the effluent water. Thus, it was demonstrated that 5,000 mg of iron bacterial
sludge should be added to the filter assembly.

Effect of P
The effect of P on As(III) removal was examined in Run 3 (Fig. 2a). As shown in this
figure, 5 mg/L of P as phosphate caused a significant decrease in As(III) removal. The

- 339 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

150
)
AsL(μg/L)
/L v ■Run
Run4 4 ?ΔRun5
R-4 Run ▲Run
R-5 5 ΩR-6Run6
6
g 125
µ
)

( 100 Bangladesh standard limit of As

-1

s
Eff As (µg

A
tn 75
Effluent

eu
lf 50
f
E 25

0
0 5 10 15
Time
Time (day)
(day)

Fig. 3 - Effects of iron bacterial sludge added at the start of the experiment

average effluent As concentration was 192 ± 55 µg/L. The removal efficiency was
reduced by 25% from Run 1 to Run 3 due to higher P concentrations in the synthetic
groundwater. As expected, P was also reduced to an average of 1.0 ± 0.37 mg/L in Run
3 (Fig. 2c). Arsenate and P have similar chemical properties and are known as inner-
sphere complex-forming anions that are strongly sorbed to mineral surfaces such as Fe
oxides (Su and Puls, 2001a and 2003). In our study, a partial oxidation of As(III) to
As(V) might occur (will be discussed later), therefore the decrease in As(III) removal in
the presence of phosphate was primarily due to competition of oxidized As(V) and P for
sorption sites on the precipitated iron hydroxides surfaces. Moreover, the decrease in As
removal might also have been caused by a decrease in the corrosion of the Fe(0) net,
because phosphate is known to act as a corrosion inhibitor of Fe(0) (Armstrong et al.,
1994). This phenomenon was investigated through batch experiments, and results
showed that corrosion of the iron net decreased as the P concentration in the solution
increased (data not shown).

Effect of Fe(II) and P composition in groundwater

To determine the levels of Fe(II) and P in synthetic groundwater beyond which As
would be reduced to less than the Bangladesh limit (50 µg/L), several series of filtration
experiments (Runs 7 - 17) were conducted. Fig. 4 shows the effluent As concentration
under different Fe and P concentrations in synthetic groundwater. As shown in Fig. 4a,
the effluent As concentration decreased linearly as Fe(II) increased. The As(III)
decreasing rate constant was found to be 34.4 µg/mg with an r2 value of 0.998.
According to Fig. 4a, an effluent As concentration of less than 50 µg/L can be obtained
if the groundwater contains at least 2 mg/L of Fe(II) and 0.9 mg/L of P. In contrast, the
effluent As concentration increased linearly as the P concentration in the raw water
increased (Fig. 4b). The rate constant was found to be 44 µg/mg with an r2 value of
0.991.

- 340 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

400
350 y = -34.461x +114.98 (a) y = 44.164x - 89.197 (b)
2 2
(µg/L)

300 R = 0.9908
Eff As (μg/L)

R = 0.9983
250
200
150
100
50
0
0 1 2 3 40 5 10
Groundwater Fe (mg/L) Groundwater P (mg/L)

Fig. 4 - (a) Effluent As concentration as a function of Fe(II) in synthetic groundwater.

P = 0.9 mg/L.
(b) Effluent As concentration as a function of P in synthetic groundwater.
Fe(II) = 5 mg/L. Error bars show the standard deviation of 14 day samples

Based on the best fit line for the measured effluent As concentrations shown in Fig. 4a
and 4b, the following empirical equation (Eq. (1)) was derived by a multiple regression
analysis to estimate the As removal efficiency from the concentrations of dissolved Fe
and P in synthetic groundwater.

As removal (µg/L) = 33 (µg/mg) × Fe (mg/L) - 45 (µg/mg) × P (mg/L) + 430 (µg/L) (1)

Arsenite removal and the corresponding effluent As concentrations calculated with this
empirical equation for different scenarios (different concentrations of Fe and P in the
synthetic groundwater) agreed well with measured experimental values as shown in Fig.
5. According to this figure, an effluent As concentration of less than 50 µg/L can be
obtained if the initial composition of synthetic groundwater is as follows: 500 µg/L of
As(III) , 2 mg/L or more of Fe(II), 3 mg/L or less of P and an Fe/P molar ratio of at least
3. A comprehensive database of groundwater composition in Bangladesh showed that
the average As, Fe and P concentrations are 199 ± 166 µg/L, 5.3 ± 4.8 mg/L and 1.47 ±
1.48 mg/L, respectively (BGS and DPHE, 2001). According to this database, our As
removal filter would be applicable for this range of groundwater compositions. This
study did not consider the influence of other geochemical compounds in groundwater,
such as silicate, sulfate and dissolved organic matter. We plan to conduct additional
experiments on the influence of such compounds on As removal by our filter. We will
also conduct experiments with actual groundwater for a long-term assessment.
Furthermore, bacterial (pathogen) contamination of the effluent will also be measured
and reported to ensure the microbiological quality of the filtered water.

As(III) removal mechanism

The efficient removal of As by the application of zero valent iron (Fe(0)) has been
reported in many studies. In Fe(0) applications, adsorption, surface precipitation and co-
precipitation of As with the corrosion products have been reported as possible
mechanisms of As removal (Su and Puls, 2001a and 2001b; Manning and Amrhein,
2002; Leupin and Hug, 2005). In addition, the dissolved iron present in groundwater as

- 341 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

P=0.9 mg/L P= 2 mg/L P= 3 mg/L P= 4 mg/L P= 5 mg/L

550
500
removal(µg/L)
(µg/L

450
400
Asremoval

350
300
As

250
200
0 1 2 3 4 5
Gorundwater Fe (mg/L)
Fig. 5 - Measured (symbol) and calculated (solid line) concentrations of As in the
filter effluent at varying concentrations of Fe and P in synthetic groundwater.
Error bars show the standard deviation of 14 day samples

Fe(II) plays a significant role in As removal (Hug and Leupin, 2003; Roberts et al.,
2004; Berg et al., 2006). In our filter experiments, less than 0.4 mg/L of Fe was present
in the effluent of Runs 1 and 2 (Fig. 2b), indicating that most of the Fe(II) in synthetic
groundwater as well as that released by corrosion of the iron net was rapidly oxidized
by biological and/or physical-chemical processes and filtered by the ceramic filter.
Based on our experimental conditions, the mean values of pH and ORP in the influent
water were 7.0 - 7.4 and -50 to -150 mV, respectively, throughout the experiment.
Although these values of pH and ORP in the influent are close to the boundary line of
biological Fe(II) oxidation zone in the Eh-pH diagram, they are also present in the
competition zone between biotic and abiotic Fe(II) oxidation (Mounchet, 1992).
Therefore, the experimental conditions of this study suggested that both biological and/
or physico-chemical oxidation of dissolved Fe(II) in the reactor might occur.
Furthermore, the average influent values for ORP (-50 to -150 mV) and pH (7.0 - 7.4)
indicated ion species (Fe2+/Fe(OH)2) in equilibrium. Increasing the ORP value to 100 -
200 mV and the pH to 7.5 - 7.7 in the effluent water changed the equilibrium towards
oxidizing conditions in which Fe(OH)3 (aq) is the predominant species. Therefore, the
oxidized iron in the reactor would exist as poorly ordered iron oxides such as Fe(OH)3
and hydrous ferric oxides (HFO) (e.g., (Fe2O3, 2 - 3 H2O) ) (Snoeyink and Jenkins,
1980; Katsoyiannis and Zouboulis, 2002 and 2004). It is well known that amorphous
Fe(OH)3, HFO and Goethite bind both arsenite (As(III)) and arsenate (As(V)) (Manning
et al., 1998; Hug and Leupin, 2003). In contrast, the oxidation of As(III) parallel to the
biological and/or physico-chemical oxidation of Fe(II) may occur in the reactor and
further enhance the overall As removal efficiency (Katsoyiannis et al., 2002;
Katsoyiannis and Zouboulis, 2004; Leupin and Hug, 2005; Shafiquzzaman et al., 2008).
Moreover, the gradual decrease of effluent As concentration in Runs 1 and 2 (Fig. 2b)
was caused by the accumulation of Fe hydroxides in the reactor. The results were
consistent with the results of As removal efficiency by changing the amount of iron
bacterial sludges (Fig. 3). The possible physico-chemical and biological reactions
during the filtration process in our filter are shown in Table 3.

- 342 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

Table 3 - Possible physico-chemical and biological reactions during the filtration

process

Phenomena Reactions References

Fe(II) release from Fe(0) Fe(0) + 2H2O + 1/2O2 →
? Fe(II) +H2O + 2OH- Leupin and Hug, 2005
net corrosion
Iron bacteria Katsoyiannis et al., 2002;
Oxidation of released Fe(II) + 2H2O +1/2 O2 →? Fe(III) +1/2 H2O + OH- Katsoyiannis and
Fe(II) and naturally Zouboulis, 2004; Leupin
occurring Fe(II) and Hug, 2005
As(III) + Intermediate ((•O2, •OH, Fe(IV)) →
? As(IV)
Physico-chemical As(IV) + O2 →? As(V) + •O2- Leupin and Hug, 2005
oxidation of As(III)
Iron bacteria Katsoyiannis et al., 2002;
Biological oxidation of As(III) + O2 ?→ As(V) Katsoyiannis and
As(III) Zouboulis, 2004;
Shafiquzzaman et al.,
2008
Katsoyiannis et al., 2002;
Surface complexion and Fe(III) + 3H2O →? Fe(OH) 3 + 3H- Katsoyiannis and
precipitation of As Fe(OH) 3 +As →? Fe-As complex + H2O Zouboulis, 2004; Leupin
and Hug, 2005
Shafiquzzaman et al.,
2008

Filter maintenance and cleaning

Filtration experiment data showed that clogging of the filter did not occur during the 70
days of operation. Therefore, minimal maintenance would be required for the novel
ceramic filter in comparison with sand filtration. For example, scrubbing the ceramic
filter with a soft brush or cloth would restore its filtration capacity when clogging
reduces the flow rate to an unsatisfactory level (Shafiquzzaman et al., 2011). The sludge
collected in the reactor should be removed periodically (once a month) to ensure a
satisfactory flow rate. When the sludge is removed, a portion (5,000 mg) should be
retained and added back to the reactor. The iron net should be replaced at the end of its
useful life span. The life span was estimated with isotherm data for As(III) removal by
the iron net and the average groundwater composition in Bangladesh. The As(III)
adsorption capacity of the iron net was found to be 1630 µg/g (Shafiquzzaman et al.,
2011). Based on the isotherm data, the filter could be used for approximately 1 year at a
flow rate of 60 L/d with influent water containing 400 µg/L of As, 5 mg/L of Fe(II) and
1 mg/L of P before As breakthrough (MCL of 50 µg/L) occurred.

Practical implications
The filter proposed in this study could serve as a simple, low-cost treatment technique.
All filter materials are locally available at little to no cost, and the manufacturing
procedure is simple and easy. The iron bacterial sludge could be found in many iron-
rich areas, or the sludge could be produced in the operational filter unit. The As removal
filter unit (including ceramic filter, iron net, reactor and effluent bucket) cost was
estimated at 7 - 8 USD, an affordable price for rural households in developing countries.
Furthermore, the unit cost could be reduced to 4 - 5 USD if clay pot is used instead of
plastic bucket for the reactor and effluent bucket. The estimated cost is nearly 10 times

- 343 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

less than that of the SONO arsenic filter (Hussam and Munir, 2007). The only operation
and maintenance cost is the annual replacement of the iron net or the ceramic filter,
which is estimated to be between 0.5 and 1 USD. In the absence of manufacturing
defects, the filter will produce potable water for at least 1 year.

Filter sludge management

The release of As from the sludge produced in our filter was examined with TCLP
(toxicity characteristic leaching procedure) leachant containing 0.1 mol/L of acetic acid
and 0.064 mol/L of NaOH (USEPA, 1986). The results showed that leachate
concentrations of As did not exceed either the maximum USEPA or the Australian TCLP
leachate values of 5,000 and 300 µg/L, respectively (data not shown). Therefore, the
sludge produced in the reactor would not be classified as hazardous waste material.
However, important factors, such as pH and redox potential, that were not accounted for
in the TCLP analysis may lead to an inaccurate estimate of As leaching in the solid
wastes (Meng et al., 2001; Shafiquzzaman et al., 2010). To minimize this possible risk,
sludge should be kept away from children and discarded into deep pits encased in
concrete.

CONCLUSIONS
This study introduced a simple, effective filter assembly for As removal composed of a
ceramic filter, an iron net and iron bacterial sludge, and its efficacy was tested in
removing As from synthetic groundwater containing varying compositions of As(III),
Fe(II) and P. The main findings from this study include the following:

1. Arsenite (As(III)) was removed by adsorption and/or co-precipitation with the Fe

precipitate.
2. Significant removal of dissolved Fe and P (over 95%) was achieved simultaneously.
3. Initial seeding with iron bacterial sludge played an important role in the early phase
of treatment and quickened the production of an effluent with less than 50 µg/L of As.
4. The efficiency of As(III) removal depended on the concentrations of Fe and P in
groundwater.
5. An Fe/P molar ratio of at least 3 with an Fe concentration of greater than or equal to 2
mg/L and a P concentration of less than or equal to 3 mg/L were required to achieve
an effluent As concentration of less than 50 µg/L by the filter.
6. This simple, inexpensive filter can be made with locally available materials and can
be used to treat As in contaminated waters.

Field-level experiments with groundwater from Bangladesh will be conducted and

reported to assess the long-term viability of the filter.

ACKNOWLEDGEMENTS
The authors would like to thank the Japan Society for the Promotion Science (JSPS) for
the financial support of this research. The assistance of Fumiaki Kakijuka in laboratory
experiments is also gratefully acknowledged.

- 344 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

REFERENCES
Armstrong R. D., Peggs L. and Walsh A. (1994) Behavior of sodium silicate and sodium
phosphate (tribasic) as corrosion inhibitors for iron, J. Appl. Electrochem., 24, 1244-
1248.
Berg M., Luzi S., Trang P. T. K., Viet P. H., Giger W. and Stuben D. (2006) Arsenic
removal from groundwater by household sand filters: Comparative field study,
model calculations and health benefits, Environ. Sci. Technol., 40, 5567-5573.
BGS, DPHE (2001) Arsenic contamination in groundwater in Bangladesh, Volume 2;
Kinniburgh, D. G., Smedley, P. L., Eds.: Final report, Brtish Geological survey
report Wc/00/19; British Geological Survey: Keyworth, UK.
Chaudhuri M., Verma S. R. and Gupta A. (1994) A performance evaluation of ceramic
filter candles, J. Environ. Eng., 120, 1646-1651.
Chowdhury U. K., Biswas B. K., Chowdhury T. R., Samanta G., Mandal B. K., Basu G.
C., Chanda C. R., Lodh D., Saha K. C., Mukherjee S. K., Roy S., Kabir S.,
Quamruzzaman Q. and Chakraborti D. (2000) Groundwater arsenic contamination
in Bangladesh and West Bengal, India, Environ. Health Prespect., 108, 393-397.
Czekalla C., Mevious W. and Henert H. (1985) Quantitative removal of iron and
manganese by microorganisms in rapid sand filters (in situ investigations), Water
Supply, 3, 111-123.
Dhar R. K., Biswas B. K., Samanta G., Mandal B. K., Chakraborti D., Roy S., Jafar A.,
Islam A., Ara G., Kabir S., Khan A. W., Ahmed S. A. and Hadi S. A. (1997)
Groundwater arsenic calamity in Bangladesh, Current Sci., 73, 8-58.
Dixit S. and Hering J. G. (2003) Comparison of arsenic (V) and arsenic (III) sorption
onto iron oxides minerals: Implications for arsenic mobility, Environ. Sci. Technol.,
37, 4182-4189.
Ferris F. G., Hallberg R. O., Lyven B. and Pederson K. (2000) Retention of stortium,
cesium, lead and uranium by bacterial iron oxides from subterranean environment,
Appl. Geochem., 15, 1035-1042.
Hug S. J. and Leupin O. X. (2003) Iron-catalyze oxidation of arsenic (III) by oxygen
and by hydrogen: pH-dependant formation of oxidants in the fenton reaction,
Environ. Sci. Technol., 37, 2734-2742.
Hussam A. and Munir A. K. M. (2007) A simple and effective arsenic filter based on
composite iron matrix: Development and deployment studies from groundwater of
Bangladesh, J. Environ. Sci. Health A, 42, 1869-1878.
JEMAI (Japan Environmental Management Association for Industry) (2008) JIS K0102
Testing methods for industrial wastewater, Japanese Standard Association, Japan.
Kartinen E. O. and Martin C. J. (1995) An overview of arsenic removal processes,
Desalination, 103, 79-88.
Katsoyiannis I. A. and Zouboulis A. I. (2002) Removal of arsenic from contaminated
water sources by sorption on to iron-oxide-coated polymeric materials, Water Res.,
36, 5141-5155.
Katsoyiannis I. A., Zouboulis A. I. Althoff H. and Bartel H. (2002) As(III) removal from
groundwater using fixed-bed upflow bioreactors, Chemosphere, 47, 325-332.
Katsoyiannis I. A. and Zouboulis A. I. (2004) Application of biological process for the
removal of arsenic from groundwaters, Water Res., 38, 17-26.
Khan A. H., Rasul S. B., Munir A. K. M., Habibuddowla M., Alauddin M., Newaz S. S.
and Hussam A. (2000) Appraisal of a simple arsenic removal method for
groundwater of Bangladesh, J. Environ. Sci. Health A, 35, 1021-1041.

- 345 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

Leupin O. X. and Hug S. J. (2005) Oxidation and removal of arsenic (III) from aerated
groundwater by filtration through sand and zero-valent iron, Water Res., 39, 1729-
1740.
Lien H. L. and Wilkin R. T. (2005) High-level arsenite removal from groundwater by
zero-valent iron, Chemosphere, 59, 377-386.
Manning B. A., Fendorf S. E. and Goldberg S. (1998) Surface structures and stability of
arsenic(III) on Goethite: Spectroscopic evidence for inner-sphere complexes,
Environ. Sci.Technol., 32, 2383-2388.
Manning M. L. and Amrhein H. C. (2002) Arsenic (III) and arsenic (V) reaction with
zerovalent iron corrosion products, Environ. Sci. Technol., 36, 5455-5461.
Mazumdar D. N. G., Haque R., Ghosh N., De B. K., Santra A., Chakraborti D. and
Smith A. H. (1998) Arsenic levels in drinking water and the prevalence of skin
lesions in West Bengal India, Int. J. Epidemiol., 27, 871-887.
Meng X. G., Korfiats G. P., Jing C. Y. and Christodoulatos C. (2001) Redox
transformation of arsenic and iron in water treatment sludge during aging and TCLP
extraction, Environ. Sci. Technol., 35, 3476-3481.
Michalakos G. D., Nieva J. M., Vayenas D. V. and Lyberatos G. (1997) Removal of iron
from potable water using a trickling filter, Water Res., 31, 991-996.
Mounchet P. (1992) From conventional to biological removal of iron and manganese in
France, J. Am. Water Works Assoc., 84, 158-166.
Oyanedel-Craver V. A. and Smith J. A. (2008) Sustainable colloidal-silver-impregnated
ceramic filter for point-of-use water treatment, Environ. Sci. Technol., 42, 927-933.
Pontius F. W., Brown G. K. and Chien J. C. (1994) Health implications of arsenic in
drinking water, J. Am. Water Works Assoc., 86, 52-63.
Roberts L. C., Hug S. J., Ruettimann T., Billah M. M., Khan A. W. and Rahman M. T.
(2004) Arsenic removal with iron (II) and iron (III) in waters with high silicate and
phosphate concentrations, Environ. Sci. Technol., 38, 307-315.
Shafiquzzaman M., Mishima I. and Nakajima J. (2008) Arsenic removal from ground
water by sand filtration during biological iron oxidation, Jpn. J. Water Treat. Biol.,
44, 11-20.
Shafiquzzaman M., Azam M. S., Mishima I. and Nakajima J. (2009) Technical and
social evaluation of arsenic mitigation in rural Bangladesh, J. Health Popul. Nutr.,
27, 674-683.
Shafiquzzaman M., Azam M. S., Nakajima J. and Bari Q. H. (2010) Arsenic leaching
characteristics of the sludges from iron based removal process, Desalination, 261,
41-45.
Shafiquzzaman M., Azam M. S., Nakajima J. and Bari Q. H. (2011) Investigation of
arsenic removal performance by a simple iron removal ceramic filter in rural
households of Bangladesh, Desalination, 265, 60-66.
Smith A. H., Goycolea M., Haque R. and Biggs M. L. (1998) Marked increase in
bladder and lung cancer mortality in a region of northern Chile due to arsenic in
drinking water, Am. J. Epidemiol., 147, 660- 669.
Snoeyink V. L. and Jenkins D. (1980) Water chemistry. John Wiley, New York, USA.
Su C. and Puls R. W. (2001a) Arsenate and arsenite removal by zerovalent iron:
Kinetics, redox transformation, and implication for in situ groundwater remediation,
Environ. Sci. Technol., 35, 1487-1492.

- 346 -
 Journal of Water and Environment Technology, Vol. 9, No.3, 2011

Su C. and Puls R. W. (2001b) Arsenate and arsenite removal by zerovalent iron: Effects
of phosphate, silicate, carbonate, borate, sulphate, chromate, molybdate, and nitrate,
relative to chloride, Environ. Sci. Technol., 35, 4562-4568.
Su C. and Puls R. W. (2003) In situ remediation of arsenic in simulated groundwater
using zerovalent iron: Laboratory column tests on combined effects of phosphate
and silicate, Environ. Sci. Technol., 37, 2582-2587.
USEPA (1986) Hazardous waste management system; land disposal restriction.
Appendix I to part 268: Toxicity characteristics leaching procedure (TCLP). Federal
Register; 51(216) p.40643-40654.
USEPA (2002) Office of ground water and drinking water implementation guidance for
the arsenic rule, EPA Report -816-D-02-005 (I3-I4), USEPA,Cincinnati, USA.
WHO (1996) Guidelines for drinking-water quality, Health criteria and other supporting
information, World Health Organization, 2nd ed., Vol. 2, pp.940-49.
Yang L., Ning X., Chen K. and Zhou H. (2007) Preparation and properties of
hydroxyapatite filters for microbial filtration, Ceramics Int., 33, 483-489.

- 347 -