Ecological Engineering 94 (2016) 352–364

Contents lists available at ScienceDirect

Ecological Engineering
journal homepage: www.elsevier.com/locate/ecoleng

Performance of conventional starches as natural coagulants for

turbidity removal
Sook Yan Choya, Krishnamurthy Nagendra Prasada, Ta Yeong Wua,c,
Mavinakere Eshwaraiah Raghunandan b,c , Ramakrishnan Nagasundara Ramanan a,c,∗
a
Chemical Engineering Discipline, School of Engineering, Monash University Malaysia, 47500 Bandar Sunway, Selangor, Malaysia
b
Civil Engineering Discipline, School of Engineering, Monash University Malaysia, 47500 Bandar Sunway, Selangor, Malaysia
c
Advanced Engineering Platform, School of Engineering, Monash University Malaysia, 47500 Bandar Sunway, Selangor, Malaysia

artic l e i nf o abstract

Article history: The efficiency of rice, wheat, corn and potato starches in removing turbidity was performed and compared
Received 5 July 2015 to alum and polyaluminium chloride. Using kaolin suspensions, the effects of pH, dosage and need for
Received in revised form 25 May 2016 starch gelatinization was studied. Autoclaved rice starch with larger polymer chain length outperformed
Accepted 28 May 2016
other starches to yield 50% turbidity removal with an optimized dosage of 120 mg/L and initial pH of 4 at
a sedimentation time of 30 min. The integrated use of autoclaved rice starch and chemical coagulants as
Keywords:
a two step coagulation process has further increased the efficiency of turbidity removal by at least 30%.
Coagulation-flocculation
This process has also reduced the amount of chemical-based sludge by 60% which was highly favourable.
Alum
PACl
SEM images of starch treated flocs revealed distinct bridging of kaolin particles in-line with properties of
Natural coagulant respective coagulation mechanisms specifically adsorption and bridging. This study further conduces to
Starch the understanding of starches as coagulants and puts forward a basis for the characterization of resulting
Floc characteristics flocs.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction ring. With ongoing size increment until a steady state is reached,
gravitational sedimentation of these macroflocs can be enhanced.
Coagulation-flocculation which can be achieved with the addi- The application of chemical coagulants specifically inorganic
tion of coagulants; both of chemical and natural-based remains as coagulants is more widespread and heavily depended on owing
one of the most efficient and simplest method adopted in water to their superiority in treating turbid water, wide availability and
treatment industries. Often characterized by its anionic nature, cost-effectiveness (Duan and Gregory, 2003). Published records on
colloidal matters which include kaolin particles would repel one the adverse impacts on human health have surfaced since the 1960s
another and remain suspended to form a stable suspension (Kim (Simate et al., 2012) with more worrying revelation on the potential
et al., 2001; Sincero and Sincero, 2002). Adsorption and charge development of Alzheimer’s disease (AD) (Flaten, 2001; Gauthier
neutralization on these particles induced by positively charged et al., 2000; McLachlan, 1995; Rondeau et al., 2000; Walton, 2013)
chemical coagulants would result in system destabilization. A sec- and senile dementia (Rondeau et al., 2001) with the use of alu-
ondary mechanism known as adsorption and bridging would then minium salts namely alum.
promote agglomeration of neighbouring particles by linking the Countries such as Malaysia classified aluminium-based sludge
microflocs formed to form macroflocs through the addition of as Scheduled Waste (SW204) wherein 2 million metric tonnes of
polymers such as starches (Caskey and Primus, 1986). The starch sludge per year was generated from a total of 462 water treatment
solution which acts as a layer of netting would entangle multiple plants (SPAN, 2010). This large amount of sludge is currently being
microflocs before binding them together over the course of stir- treated in the country’s only prescribed premise in Bukit Nanas,
Negeri Sembilan which has resulted in excessive strain to the facil-
ity and also additional expenditures up to $ 6 million per day for
sludge disposals only (SPAN, 2010). Thus, there is a need to consider
other potential alternatives to minimize the use of aluminium salts.
∗ Corresponding author at: Chemical Engineering Discipline, School of Engineer-
Natural coagulants such as starches could be the proposed solution.
ing, Monash University Malaysia, 47500 Bandar Sunway, Selangor, Malaysia.
E-mail address: ramanan@monash.edu (R.N. Ramanan).

http://dx.doi.org/10.1016/j.ecoleng.2016.05.082
0925-8574/© 2016 Elsevier B.V. All rights reserved.
 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364 353

Natural coagulants are largely non-toxic, eco-friendly and behavior and the explanation on the different coagulation activities
results in less sludge volume in some instances (Ndabigengesere observed.
et al., 1995). Research studies have proven the efficiency of Moringa
oleifera seeds in the treatment of raw surface and synthetic tur- 2. Materials and methods
bid waters. Established coagulants such as tannins have also been
used in real time applications. Besides fruit wastes (Choy et al., 2.1. Materials
2014) and other various vegetables and legumes (Choy et al., 2015),
biopolymers such as starch is also gaining increasing interests for Kaolin heavy powder (Sigma-Aldrich) was used as a model
applications as potential natural coagulants. Being one of the most suspension for the jar test studies. Four native starch samples of
abundant biopolymer after cellulose with a projected global annual rice, wheat, corn and potato purchased from Sigma-Aldrich were
production of 85 million tonnes by 2015 (ISI, 2014), starch is attrac- studied for their feasibility as natural coagulants. For comparison
tive due to its biodegradability, non-toxicity and is also one of the purposes, aluminium salts were used as chemical coagulants in this
lowest costing biopolymer. study. Aluminium sulfate hydrate (Al2(SO4)3 ·H2O) sourced from
Starch is commonly modified into various types of starch- Sigma-Aldrich and liquid polyaluminium chloride 10% supplied by
grafted polymers notably cationic, anionic and amphoteric (Yang Holy Mate Sdn. Bhd., Malaysia were then prepared as solutions of
et al., 2014). The studies performed on other starch derivatives 3% strength. The characteristics of these aluminium salts were sum-
have also been discussed in a more recent review (Oladoja, 2015). marized in Table S1. Distilled water was used to dilute hydrochloric
Such chemical modifications are achieved via graft copolymeriza- acid (Merck) and to dissolve sodium hydroxide pellets (Merck) for
tion whereby synthetic polymers such as acrylamide are combined the preparation of 1 M solutions.
with starch to enhance the overall molecular weight of poly-
mer (Willett, 2009). Although enhanced turbidity removals can 2.2. Preparation of synthetic turbid water
be achieved, negative impacts on both environmental and human
health would be of great concerns. Instead, unmodified starches Twenty grams of kaolin powder was added into 2 L of distilled
could be used as an aid in tandem with chemical coagulants to water. The solution was mixed thoroughly and allowed to stand
minimize the harmful impacts caused by these proprietary coagu- overnight for complete hydration. The top and bottom 400 mL was
lants. Besides the removal of metal cations through ultrafiltration drained away before collecting the filtered stock kaolin solution
(Baharuddin et al., 2014), the use of unmodified starches has using a strainer into a separate beaker. Model turbid synthetic
also been studied in the treatment of semiconductor wastew- water with medium initial turbidity of 165 ± 5 NTU was then pre-
ater (Mohd Omar et al., 2013) and applied as an aid to ferric pared by diluting approximately 700 mL of kaolin stock solution
chloride (Shahriari et al., 2012) with promising results that were with 3500 mL of distilled water. The turbid water was stirred con-
worth further investigations. However, research areas utilizing tinuously to prevent gravitational settling and a series of analysis
native, heat treated-starches as natural coagulants particularly in − turbidity measurement, pH and zeta potential was performed.
the removal of turbidity remained scarce with the exception on Similar model turbid water was prepared and used in all jar testing
studies conducted using palm oil mill effluent where total sus- experiments.
pended solids removal above 80% have been reported (Teh et al.,
2014a,b). The rheological properties of starches can easily be mod- 2.3. Preparation of starch solutions
ified through gelatinization; a process defined by the rupture of
starch granules under sufficient moisture and supplied heat lead- Starch solutions with 3% concentration were prepared using
ing to increased solution viscosity (Biliaderis, 2009). The release distilled water. To determine the effect of gelatinization on the
of starch fractions to the surroundings would facilitate the link- feasibility of starches as natural coagulants, non-autoclaved and
ing of colloidal particles through the bridging mechanism. As such, autoclaved starches were prepared. These solutions were prepared
autoclaved starch solutions would also present enhanced viscos- fresh when needed to avoid potential biodegradation leading to
ity which is one of the key factors leading to their potential use in inconsistent results. For non-autoclaved starches, the prepared
treating turbid water. Similar to that of the earliest known Chi- solutions were used directly in jar tests according to the desired
nese water clarifying agent, glue made from animal hides were coagulant dosage. As for autoclaved starches, the preparation was
also used in the ancient times and their efficiencies were largely similar to previously published protocol (Teh et al., 2014b).
governed by the degree of viscosity and molecular mass (Jahn,
2001). 2.4. Coagulation-flocculation experiments
Motivated by the positive results reported, this research work
aims to evaluate the performance of various conventional starches The conventional 4 jar apparatus (VELP Scientifica JLT4, Italy)
as low-cost natural coagulants and its suitability as a primary or was utilized in all the coagulation-flocculation experiments. A
coagulant aid to aluminium salts. Moreover, integrated use of starch placebo (without the addition of any coagulants) was also prepared
with alum and polyaluminium chloride (PACl) was accessed by con- at the same starting conditions of the jar test. Any pH adjustments
sidering three different modes of coagulant additions; aid, together required were performed with the addition of 1 M HCl and 1 M
and two step coagulation. The respective coagulation mecha- NaOH solutions to obtain the desired pH values within±0.1 pH
nism involved was then identified. Additionally, characterization units. The beakers filled with 500 mL of synthetic turbid water
of starches used and flocs formed was performed using various were placed on the flocculator and agitated at the pre-selected
microscopy techniques namely inverted routine microscopy and rapid mixing intensity of 100 rpm. Subsequently, varying dosages
Scanning Electron Microscopy (SEM), Fourier Transform infrared of the studied coagulants (0–600 mg/L) were added depending on
spectroscopy (FTIR) and X-Ray Diffraction (XRD). This research the respective mode of addition.
study also aimed to elucidate the importance of starch gelatiniza- Rapid mixing was then continued at an intensity of 100 rpm
tion which resulted in enhanced agglomeration of kaolin particles for 2 min followed by slow mixing at 40 rpm for 20 min
leading to water clarification. Knowledge on the physicochemical (Ndabigengesere et al., 1995). At the last 2 min of slow mixing,
properties of conventional starches would have significant contri- 5 mL of sample was collected for the measurement of zeta poten-
bution towards the understanding of the respective coagulation tial. Upon slow mixing, the beakers were removed and transferred
to a flat, safe environment for sedimentation. After 30 min, 30 mL
354 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364

of water samples were collected at approximately 2 cm from the bidimeter, HACH, USA). The efficiency of coagulants was calculated
bottom of the beakers for the measurement of final turbidity. The based on Eqs. (1) and (2):
pH values after coagulation-flocculation were also determined. All
experimental and sample analysis was performed in triplicates at Initial TurbiditySample − Final TurbiditySample
Turbidity Reductions(%)
=
room temperature of 25 ±1 ◦ C unless specified otherwise. Initial TurbiditySample

The sequence at which the coagulants were added during jar test × 100% (1)
varied depending on the studied mode of addition as summarized:

• Individual coagulant Final TurbidityPlacebo − Final TurbiditySample

Coagulation Activity(%)
=
Final TurbidityPlacebo

Only one type of coagulant (alum/PACl/starch) was added before × 100% (2)
the start of the 2 min rapid mixing.
The initial and final turbidity of the sample corresponded to the
• Combined/Integrated coagulants turbidity values before and after treatment respectively. A turbid
sample without the addition of coagulants was also prepared and
Both the natural and chemical coagulants were utilized in a sin- was used in jar test as a placebo where the residual turbidity mea-
gle jar test routine; either added together before the 2 min of rapid sured after sedimentation was termed as the ‘Final Turbidity Placebo’
mixing or natural coagulant was dosed as an aid to chemical coag- in Eq. (2). The pH of the solutions was obtained using a calibrated pH
ulants after the 2 min of rapid mixing at the specified coagulant probe (Seven Easy pH Meter, Mettler Toledo, USA). Zeta potentials
dosages. were measured by the Laser Doppler Electrophoresis technique at
a constant temperature of 25 ◦C using a light scattering equipment
• Two step coagulation (Zetasizer Nano ZS Series, Malvern, UK). The measurements of COD
were performed using a spectrophotometer (HACH DR 2800, HACH,
Germany) upon digestion based on the Chemical-Reactor Digestion
This mode of coagulant addition required 2 series of coagula-
Method (HACH Method 8000).
tion whereby the 1st coagulation took place with the use of rice
starch alone and then followed by the 2nd coagulation with chem-
2.7. Statistical analysis
ical coagulants. The autoclaved rice starch dosage of 120 mg/L was
applied for the 1st coagulation process following the results from
The significance of means within the groups of experimental
the preliminary studies.
data was assessed using one-way analysis of variance (one-way
After 30 min of sedimentation, 300 mL of the supernatant was
ANOVA). The equal variance assumption made by ANOVA was
then carefully transferred into a new beaker while minimizing dis-
verified by performing Levene statistics test of homogeneity of vari-
ruption to the settled flocs. Another round of jar test commenced
ances. Additionally, Tukey’s honestly significant difference (HSD)
and chemical coagulants (alum or PACl) were then added accord-
was performed to determine the relationships and to distinguish
ingly into the solution with the preset dosage. The general jar test
the significantly different pairs of means based on the study param-
procedures were adhered in both the coagulation processes.
eters. The statistical analysis was performed with a significance
level of P < 0.05 using the program; IBM SPSS Statistics 23.
2.5. Characterization of coagulants and flocs generated
3. Results and discussions
The crystallinity of native starch powders were determined
using the X-Ray Diffraction (XRD) method (Bruker Heimer XRD D8, 3.1. Characterization of dry starch powder
UK) coupled with the Brag Bentano rotating beams at a step size of
0.02◦/s. Upon the preparation of non-autoclaved and autoclaved The dry starch powders were characterized according to their
starch solutions, overall changes in the starch structures were bulk densities, surface morphologies and also crystallinity. The
observed using an inverted routine microscope (Eclipse TS100, kaolin heavy powder used as a base for the preparation of synthetic
Nikon, USA). turbid water was found to be the least dense. The bulk densities
For the estimation of sludge volume, the treated synthetic tur- obtained were in increasing order of rice starch, wheat starch, corn
bid water upon jar tests was gently transferred into sedimentation starch and lastly potato starch (Fig. S1).
cones (Nalgene Imhoff Sedimentation Cones, Thermo Scientific, To establish a thorough understanding on the feasibility of using
USA). The solution was then allowed to stand for 24 h before the starches as natural coagulants in the treatment of turbid water, the
sludge volume was obtained. The wet flocs collected were sub- surface morphologies of these starches were also observed under
sequently freeze-dried for 24 h using a freeze dryer (FDU-2100, the Scanning Electron Microscope (SEM). The images obtained were
Eyela, Japan) and zip-locked prior to other analysis. The surface compared as illustrated in Fig. 1. The starch granules were distin-
morphologies of dry samples were analyzed using Scanning Elec- guishable from their particle size, granule shape and particle size
tron Microscope (SEM Hitachi S–3400 N, Japan) under an applied distribution (Jay-lin, 2003). Comparing at a magnification of 2000
voltage of 10–15 kV. Additionally, the presence of functional groups times, it was obvious that potato starch granules have the largest
in the various coagulants and flocs formed was determined using a particle size while rice starch remained the smallest. Besides that,
FTIR spectrophotometer (Nicolet iS-10, Thermo Scientific, USA). all studied starches have smooth, solid surfaces with the absence of
pores. Most starches exhibited a unimodal size distribution except
2.6. Analytical procedures for wheat starch granules which has a significant contrast in particle
size (Fig. 1b) which is also consistent with the literature (Copeland
Mass measurement was performed using a top loading balance et al., 2009). The images obtained in this present work were found
(TX 423L, Shimadzu, Japan) with sensitivities up to 3 decimal places. to be in close agreement to those reported elsewhere (Jane, 2009).
Turbidity was measured in terms of Nephelometric Turbidity Units Starch which is also a biopolymer would exhibit a certain extent
(NTU) using a calibrated turbidimeter (HACH 2100Q Portable Tur- of particle uniformity and distinct peaks would be observed if the
 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364 355

Fig. 1. SEM images of conventional starches at a magnification of 2000 times: (a) rice, (b) wheat, (c) corn and (d) potato.

1800 1800
a b
1500 1500
1200 1200
Counts

Counts

900 900
600 600
300 300
0 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
2θ Angle, degrees 2θ Angle, degrees

1800 1800
c d
1500 1500
1200 1200
Counts

Counts

900 900
600 600
300 300
0 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70 80 90
2θ Angle, degrees 2θ Angle, degrees

Fig. 2. X-Ray diffraction patterns for conventional starches: (a) rice, (b) wheat, (c) corn and (d) potato.

element is crystalline. Although starches are generally not highly direct influence on its performance in water clarification, the past-
crystalline polymers, sharp crystallinity peaks were visible in all ing properties of starches are linked to the degree of crystallinity
the starches tested across the same 20 region of 10◦–30◦ (Fig. 2). which may affect the efficiency of turbidity removal.
Besides, broad peaks were also noted across the analyzed 20 region
indicating the amorphous portion of the starch structure. The per-
3.2. Sole use of starches
centage of crystallinity obtained was arranged in increasing order
of potato starch (23.9%), wheat starch (28.8%), rice starch (30.8%)
3.2.1. Effects of gelatinization of starches on turbidity removal
and followed by corn starch (31%). From the analysis, it can be
The effect of starch gelatinization on the resulting coagulation
inferred that cereal starches exhibit higher particle uniformities
activities was explored by studying the use of non-autoclaved and
as suggested by the higher percentage of crystallinity as opposed
autoclaved starch solutions at the selected pH of 4–10. A pre-
to tuber starch. While the crystallinity of starches may not have
determined coagulant dosage of 120 mg/L from preliminary studies
356 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364

values with the exception of wheat starch at pH 4 with removals

about 11%. Amylopectin, the branched fraction of starch would
assist in coagulation similar to those present in other natural coag-
ulants such as Maerua subcordata (Mavura et al., 2008). Apart from
the amount of amylopectin present, the polymer chain length is also
equally important. The molecular weight of amylopectin could be
used as a reference to polymer chain length where rice starch was
reported to be at 26.8×108 as opposed to other starches at values
below 5.0 ×108 (Jane, 2009). Longer polymer chains would corre-
spond to larger surface areas available for particle adsorptions due
to extended polymer length and enhanced probability of bridging
with other near-by kaolin aggregates (Caskey and Primus, 1986).
Although both wheat and corn starch have the highest content of
amylopectin (Jay-lin, 2003), only rice starch yielded significant tur-
Fig. 3. Turbidity removal achieved using various non-autoclaved starch solutions bidity removals. Following the autoclave of starches, the number
at pH values ranging from 4 to 10 (rapid-mixing velocity = 100 rpm; rapid-mixing
of average chain length could decrease up to 6–18% with increased
time = 2 min; slow-mixing velocity = 40 rpm; slow-mixing time = 20 min; n = 3).
autoclave duration by 2 folds (Han et al., 2003).
The phosphorus content in potato starch namely phosphate
monoesters of approximately 0.08% which were attached to the
amylopectin fraction (Jay-lin, 2003) would most likely cause repul-
sions with the co-charged kaolin particles. Effective bridging of
the neighbouring kaolin particles would be weakened following
the ionic repulsions experienced and thus, contributing to weaker
coagulation activities as observed. In addition, there is a high prob-
ability that the available surface area for kaolin bridging is reduced
as potato starch has been reported to have increased water binding
capacity due to the presence of phosphate monoesters (Hegenbart,
1996).

3.2.2. Effect of gelatinization of starches on its physicochemical

properties
Additionally, the structural changes of starches after autoclave
were observed under an inverted routine microscope. The non-
autoclaved, suspended starch particles in water maintained the
Fig. 4. Turbidity removal achieved using various autoclaved starch solutions at
pH values ranging from 4 to 10 (rapid-mixing velocity = 100 rpm; rapid-mixing overall granular shape (Fig. 5A); similar to that observed under
time = 2 min; slow-mixing velocity = 40 rpm; slow-mixing time = 20 min; n = 3). SEM (Fig. 1). Starch particles do not dissolve in water and can
easily be separated from water if left undisturbed. Hence, non-
autoclaved starch lack coagulation activity (below 20% as most
was used throughout the coagulation-flocculation process. The values calculated were negative) as the responsible biopolymers
results obtained are shown in Figs. 3 and 4. remained enclosed within the starch granules restricting the occur-
At all studied pH values, both non-autoclaved rice and corn rence of bridging. Nevertheless, starch solutions were found to
starch exhibited negative turbidity removals. A small percentage undergo paste formation process or gelatinization with sufficient
of turbidity removal (< 25%) was obtained for wheat and potato heat and moisture. The peak autoclave temperature of 121 ◦C has
starch at pH 4 (Fig. 3). The particle size for both starch granules are surpassed the pasting temperatures of all the starches (Biliaderis,
found to be among the largest (Fig. 1) and interestingly, the turbid- 2009; Copeland et al., 2009). Consistent with the images obtained,
ity removal efficiency obtained seemed to increase with increasing the visible starch granules before the autoclave process was largely
particle size of starch granules. It is likely that small fraction of broken down and the resulting starch solution became almost
colloidal kaolin particles have adhered onto the surface of much clear (Fig. 5B). The water molecules were adsorbed into the starch
larger-sized starch granules. However, it can be concluded that granules leading to swelling and subsequently, viscosity of the
starch solutions in the absence of gelatinization did not display any starch solution increased. As the granules rupture, the enclosed
commendable coagulation activities independent of pH. starch contents are released to the surrounding which facilitates
On contrary, autoclaved starches displayed significant turbid- the bridging mechanism. The introduction of heat via the autoclave
ity reduction at pH 4 especially rice starch with removals close to process has produced a natural polymer which could be used for
50% (Fig. 4). Rice starch has also previously demonstrated plausible water clarification purposes.
coagulation activities at lower pH values when used for the treat- The physicochemical changes undergone by starch solutions
ment of palm oil mill effluent (POME) with a desirable initial pH after gelatinization would also provide useful information to under-
of 2.6 (Teh et al., 2014a). With increasing pH values, zeta potential line the positive coagulation activities observed (Table S2). Upon
of starches became more negatively charged due to the carboxyl autoclave, pH values of cereal starches shifted towards more alka-
groups of proteins (COO−) which varied with the percentage of line regions whereas the reverse was observed in the case of
protein content present (Wongsagonsup et al., 2005). Therefore, tuber starch. The zeta potential value remained weakly negative
bridging would be hindered if both the starch and kaolin particles charged indicating the “not largely ionic” property of autoclaved
have the same charge which is evident from the sharp decline in starches across the wide pH range measured. As the charge of tur-
efficiency once the pH 4 has been exceeded. Despite that, auto- bid synthetic water at different pH values was also negative, charge
claved corn starch appeared to work at basic regions albeit the neutralization was dismissed as the primary coagulation mecha-
reported removals were less than 10%. Wheat and corn starches nism with the use of starches. The use of autoclave has reduced the
exhibited weak, insignificant coagulation activities across all pH organic loadings in starches albeit in lesser amounts. This could be
 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364 357

Fig. 5. Structure of starch solutions when viewed under an inverted routine microscope at a magnification of 20 times for (A) before and (B) after gelatinization of: (a) rice
starch, (b) wheat starch, (c) corn starch, (d) potato starch.

due to the disintegration of hydrogen bonds to form a more viscous of starch addition prior to start of coagulation, the final turbidity
solution (Jane, 2009). Even after the autoclave process, the liquid introduced by various amount of rice starch was measured over
densities of both solutions remained similar with small, insignifi- a fixed volume of distilled water (Table S3). Although the use of
cant changes. autoclaved starch would enhance the initial turbidity of a water
Autoclaved starch solution has an opaque, whitish− like appear- sample, its influence is minimal for the selected coagulant dosage as
ance which may also contribute to some extent of cloudiness in the observed increment was insignificant (< 5%). Autoclaved starch
the prepared synthetic turbid water. To evaluate the significance samples could thus be considered for the treatment of turbid water.
358 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364

ulants has been found rewarding which could be the solution to

address this matter (Sathiyabama, 2012).
Generally, high turbidity removals of around 80% were obtained
for both alum and PACl even at low coagulant dosages (Fig. 7a).
Governed by the wider range of pH workability (Nansubuga et al.,
2013), higher treatment efficiency was obtained with the use of
PACl as opposed to alum which has a narrower optimum pH range
of 5–6 (Shen and Dempsey, 1998). PACl also outperformed alum
when compared at their respective optimized coagulant dosages
which can be justified by the higher charge density (Table S1). Once
the optimized PACl dosage of 1.2 mg/L has been exceeded, sharp
drop in the removal peak was observed indicating the occurrence
of charge-reversal which caused the kaolin particles to be positively
charged during coagulation-flocculation. Effective charge neutral-
Fig. 6. Effect of pH and coagulant dosage on turbidity removal using autoclaved rice ization could no longer occur as both kaolin and PACl have the
starch (rapid-mixing velocity = 100 rpm; rapid-mixing time = 2 min; slow-mixing same charge leading to strong repulsion forces which hindered the
velocity = 40 rpm; slow-mixing time = 20 min; n = 3). attraction of particles. Instead, the efficiency of alum was main-
tained at the 70–80% region across the applied dosages (Fig. 7b).
Even with increased dosages above the desired coagulant dosage,
3.2.3. Combined effects of pH and coagulant dosage insignificant changes were obtained which was consistent with
Based on the jar test results obtained using conventional the findings reported elsewhere (Duan and Gregory, 2003). Unlike
starches, autoclaved rice starch demonstrated the most potential starches, adsorption and charge neutralization and sweep coagu-
to be used for further coagulation-flocculation studies. To illustrate lation are the two potential mechanisms of chemical coagulants
the effects of pH and dosage on turbidity removals, additional runs which are pH and applied dosage dependent (Duan and Gregory,
were performed. Regardless of the coagulant dosage applied at all 2003). At pH 4, Al3+ remained as the dominant soluble species
studied pH values except pH 4, minimal or insignificant turbidity where the primary coagulation mechanism is likely charge neu-
removals were observed (Fig. 6). Hence, it is further confirmed that tralization.
the coagulation-flocculation using the weakly anionic biopolymer
of rice starch is noticeably enhanced in acidic regions.
The increment of coagulant dosage up to 120 mg/L at pH 4 3.3. Integrated use of autoclaved rice starch and aluminium salts
has increased the efficiency of turbidity removal. With further at pH 4
increment beyond this threshold value, negative removals were
obtained for all other studied pH values. Excessive dosing of rice By comparison, it was apparent that the use of rice starch as an
starch has only increased the residual turbidity as long polymer aid yielded higher turbidity removals (Table 1). The differences in
chain of starches were largely overlapping with one another due performance between the use as an aid and simultaneous addi-
to surface saturation and the overcrowding effect (Sharma et al., tion were more pronounced at low coagulant dosages of alum
2006). Thus, a reverse effect was observed once the optimum (< 12 mg/L) and even greater with PACl (> 1.2 mg/L) as deteriora-
dosage had been exceeded. Likewise, reductions in the bare surface tions up to 25% and 50% were observed respectively. As chemical
area for attachment of segments following the surplus of starch coagulants induce particle agglomerations via charge neutraliza-
addition would lead to particle restabilization and to a certain tion at low coagulant dosage, the addition of rice starch which
extent, some form of steric repulsions (Gregory, 2005b). The addi- is weakly anionic at the start of coagulation might have led to
tion of rice starch beyond 300 mg/L has a more prominent impact the shielding of chemical coagulants. Consequently, charge neu-
(> 5%) on the mean turbidity (Table S3) which increased by at least tralization of these opposite charges would be affected with the
2 folds with the doubling of the applied dosage. This could also simultaneous addition of both types of coagulant. It can also be
account for the decline in performance at high coagulant dosages deduced that during simultaneous addition of starches and chem-
of 180–600 mg/L at pH 4 (Fig. 6). Hence, the applied coagulant ical coagulants, some extent of charge neutralization occurred on
dosage must also be optimized along with pH. Prior to treatment, the starch particles rather than the colloidal kaolin particles; lead-
the COD of turbid water at pH 4 was found to be below the lower ing to reduced turbidity removal. At the similar process conditions,
detectable limits of low range (LR) COD vial of 3 mg/L. Following the measured zeta potential in both cases of alum and PACl were
the use of rice starch at the optimized dosage of 120 mg/L, the COD at least 15% more negatively charged when rice starch was added
after treatment was detected at 95 ±2 mg/L. The results obtained simultaneously. Based on the results obtained (Table 1), it can be
indicated that the concentration of organic matter in treated water certain that the different mode of additions have demonstrated
has been raised significantly which is due to the biological nature variations in terms of coagulation mechanisms which affected the
of rice starch. Similar instances have also been cited where the end zeta potential readings.
content of organic matter was elevated when natural coagulants Using the same combination of coagulant dosages, the two step
were used (Ghebremichael, 2007). For example, the use of crude coagulation was then evaluated. The two step coagulation pro-
extracts of shelled and deshelled Moringa oleifera seeds at inter- cess has resulted in substantial improvements to the use of only
mediate dosages have increased the organic loading of treated rice starch which removed close to 50% of turbidity. With suitable
water by above 95% (Ndabigengesere and Subba Narasiah, 1998). dosage of alum (20 mg/L) or PACl (1 mg/L), the overall turbidity
Additionally, lower COD removals in various sources of wastew- removal was heightened by at least 20% (Table 2). At these coag-
ater following the use of starch (Teh et al., 2014b) and foliage of ulant dosages, deterioration of about 10% and an improvement by
Hylocereus undatus (Idris et al., 2013) have also been accounted 60% was expected for alum and PACl respectively when used alone.
for the additional organic loading introduced compared to chem- PACl which has significantly larger charge density has also demon-
ical coagulants. Granted that natural coagulants are associated to strated greater superiority particularly at lower coagulant dosages.
the potential leaching of organic matters during the extraction and Besides enhancing the performance of rice starch, the amount of
isolation process, purification of active compounds in natural coag- chemical based sludge can also be significantly reduced consider-
 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364 359

100.0 100.0
a b
80.0 80.0

Turbidity Removal, %

Turbidity Removal, %
60.0 60.0 Alum

40.0 40.0 PACl

20.0 20.0

0.0 0.0
0 2 4 6 8 10 10 20 30 40 50 60 70 80 90 100110 120
Coagulant Dosage, mg/L Coagulant Dosage, mg/L

Fig. 7. Effect of coagulant dosage on turbidity removal at pH 4 using aluminium salts of alum and PACl at 3% concentration: (a) dosage of 0–10 mL and (b) dosage of 10–120 mL
(rapid-mixing velocity = 100 rpm; rapid-mixing time = 2 min; slow-mixing velocity = 40 rpm; slow-mixing time = 20 min; n = 3).

Table 1
Turbidity removal achieved at pH 4 for combined use of chemical coagulant and autoclaved rice starch either as an aid or added simultaneously.

Dosage of Chemical Coagulant, mg/L Rice Starch Added as an Aid Rice Starch Added Simultaneously

Alum 3% PACl 3% Alum 3% PACl 3%

Turbidity Removal, % Turbidity Removal, % Turbidity Removal, % Turbidity Removal, %

0.00 7.44 (2.45)

0.06 NA 60.61 (0.30) NA 56.33 (0.33)
0.30 NA 78.57 (0.19) NA 71.88 (0.12)
0.60 73.63 (0.33) 88.13 (0.13) 48.39 (0.18) 74.32 (0.56)
1.20 78.37 (0.12) 84.45 (0.22) 53.80 (0.12) 72.73 (0.13)
12.00 85.50 (0.21) 84.53 (0.21) 79.22 (0.18) 58.95 (0.39)
30.00 86.07 (0.21) 79.76 (0.27) 82.39 (0.18) 33.54 (0.35)

Table 2
Turbidity removal achieved at pH 4 following two step coagulation (1st step using autoclaved rice starch, 2nd step using chemical coagulants).

Dosage of Autoclaved 1st Turbidity Dosage of Chemical 2nd Turbidity Overall Turbidity
Rice Starch, mg/L Removal, % Coagulant 3%, mg/L Removal, % Removal, %

A) Alum
120 47.80 (0.55) 1 17.88 (0.48) 57.13 (0.25)
46.22 (0.21) 2 12.66 (0.30) 53.03 (0.16)
47.39 (0.34) 20 44.66 (0.18) 70.89 (0.09)
47.45 (0.40) 50 43.94 (0.12) 70.55 (0.06)
B) PACl
120 45.13 (0.32) 1 60.94 (0.16) 77.40 (0.09)
44.62 (0.35) 2 61.72 (0.43) 78.41 (0.24)
45.37 (0.51) 20 19.31 (0.33) 54.31 (0.19)
45.15 (0.47) 50 0.70 (0.11) 42.56 (0.06)

ing the splitting of biological sludge with the introduction of two bridging of kaolin particles produced flocs that are not only larger
step coagulation. but also highly porous; leading to enhanced probability for the
entrapment of water particles in the porous surface of flocs. Nev-
ertheless, natural coagulants such as the seeds of Moringa oleifera
3.4. Characterization of wet flocs which also relied on adsorption and charge neutralization has been
reported to produce 5 times less sludge volume than proprietary
3.4.1. Normalized sludge volume coagulants (Ndabigengesere et al., 1995).
The amount of settled flocs at the end of a fixed sedimentation With the integrated use of starch and chemical coagulants, the
period of 24 h was measured at an initial pH of 4 and then normal- overall sludge volume measured has increased by at least 40%.
ized accordingly to yield an overall turbidity removal of 80% (Fig. 8). Similar to the observed trend, alum flocs were at least 10% more
This parameter would be important as it provides additional infor- voluminous than PACl except for two step coagulation. As PACl has
mation on the downstream processing of resultant sludge formed a higher charge density, it is likely that the residual rice starch par-
and serves as a basis for the selection of coagulants. Generally, ticles from the first coagulation process were coagulated along with
alum treated flocs was found to be more voluminous (up to 20%) in kaolin particles. Therefore, the total normalized sludge volume gen-
contrast to PACl treated flocs at their optimized dosages which is erated using PACl was 5% larger than alum with twice the amount of
consistent with existing literature (Gebbie, 2006). The 2 folds lower chemical based sludge. From the above results, it was apparent that
sludge volume observed following the use of aluminium salts over simultaneous addition of these coagulants produced the largest
starches could likely be a direct result of the primary coagulation amount of settled sludge in a single coagulation process owing
mechanism responsible for particle agglomerations. The effective
 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364 361

more kaolin particles yielding to its more compacted outlook. How-

ever, starch generated flocs were at least 75% less dense compared
to those produced using chemical coagulants. Besides the amount
of settled sludge, the underlying coagulation mechanism responsi-
ble for particle agglomerations has also affected the measured bulk
densities.
With the inclusion of rice starch to assist chemical coagulants,
the bulk densities of resultant flocs were noticeably reduced (Fig. 9).
The integration between the coagulation mechanisms of chemical
and natural coagulants has led to modifications in the properties
of flocs formed. The bridging of kaolin particles due to rice starch
would result in loosely attached flocs with greater surface porosi-
ties (Gregory, 2005a). Thus, the reduced compactness in flocs is
likely to result in lower floc bulk densities as evident in this study.
Fig. 8. Normalized sludge volume of treated flocs measured after 24 h of sedimen-
At large, the bulk densities of alum and PACl treated flocs followed a
tation using chemical and natural coagulants through coagulation-flocculation to specific trend; aid > two step coagulation > simultaneous addition.
80% turbidity removal in different modes of addition: sole usage, as an aid or added More effective charge neutralization would result in increased floc
simultaneously (rapid-mixing velocity = 100 rpm; rapid-mixing time = 2 min; slow- compactness and gradually affecting the measured bulk densities.
mixing velocity = 40 rpm; slow-mixing time = 20 min; n = 3). Values annotated with
different letters indicates significant differences (one-way ANOVA, Tukey’s HSD;
PACl treated flocs generally yielded the largest bulk density with
P < 0.05). exception of two step coagulation.

3.5.2. Surface morphology using SEM

From the SEM images, the surface morphologies of coagu-
lants and flocs formed from coagulation-flocculation process were
obtained. Kaolin particles which formed the turbid model suspen-
sion appeared to be very fine with an average particle size of 24 µm
(Fig. 10A, a) and “flake-like” structure when imaged up to 5000
times (Fig. 10B, a) which were similar to those reported previously
(Yavuz and Saka, 2013). On contrary, alum granule was found to be
close to spherical (Fig. 10A, b) with rough surfaces which resem-
bled “rod-like” attachments (Fig. 10B, b). As indicated by visual
observations, flocs treated using aluminium salts were evidently
smaller; mean particle sizes of 57 and 129 µm for alum (Fig. 10A,
c) and PACl (Fig. 10A, d) respectively. Contrastingly, the use of
rice and wheat starches (Fig. 10A, e and f) has led to enhanced
Fig. 9. Bulk density of treated flocs using chemical and natural coagulants through floc growth by at least three times with mean particle sizes of
coagulation-flocculation in different modes of addition: sole usage, as an aid or approximately 478 and 315 µm respectively. Diversity in terms of
added simultaneously (rapid-mixing velocity = 100 rpm; rapid-mixing time = 2 min;
shape and overall structure of flocs generated were also observed
slow-mixing velocity = 40 rpm; slow-mixing time = 20 min; n = 3). Values annotated
with different letters indicates significant differences (one-way ANOVA, Tukey’s as chemical coagulants yielded more compacted and relatively
HSD; P < 0.05). spherical agglomerates (Fig. 10B, c and d). Instead, starches pro-
duced flocs that are highly elongated, fibrous and less compact
(Fig. 10B, e and f). These loosely linked, thread-like flocs exhibited a
to the agglomeration of starch particles instead of the intended
three-dimensional open structure; consistent with characteristics
kaolin particles. The extended contact time during rapid mixing
of adsorption and bridging (Bratby, 2006; Li et al., 2006). Multiple
could have expedited the coagulation of more rice starch parti-
branches were observed on a single floc structure with enhanced
cles resulting in higher sludge volume. On the other hand, the two
hollow interiors; easily distinguishable from flocs produced using
step coagulation has generated the least amount of chemical-based
chemical coagulants.
flocs. Flocs generated using only starches are termed as biological
The type of coagulants used in coagulation-flocculation played
sludge and vice-versa in this study. Although turbidity removals
a vital role in determining the features of the flocs formed. Typ-
obtained were slightly below 80%, the two step coagulation reduced
ically, the mean particle size of flocs ranged from approximately
at least 50% of chemical based sludge with respect to the sole use of
200–250 µm in both aided and simultaneous addition of coagu-
chemical coagulant. By incorporating the use of starches to supple-
lants (Fig. 11A, a–d). However, flocs generated using the latter was
ment chemical coagulants, enhanced settling of flocs was achieved
marginally larger in size; contributed by the increased contact time
as shorter sedimentation time was required for the settling of equal
for coagulation-flocculation. The floc size obtained with integrated
volume of flocs.
use of coagulants is intermediate to those used alone suggesting the
presence of synergistic effects of both coagulants. Overall, the flocs
3.5. Characterization of dry flocs resembled those generated using starches alone (Fig. 11B, a–f). A
similar trend was also observed where the resulting PACl generated
3.5.1. Estimation of bulk density flocs in all scenarios were more compacted and larger which agreed
The bulk density of flocs which is one of the often overlooked closely to findings of literature (Malhotra, 1994). This observation
parameter can be used as supporting evidence to understand could also be supported by the higher charge density of PACl com-
the coagulation mechanism and the accompanying floc structural pared to alum. In two step coagulation, the addition of rice starch in
changes. From the results obtained at pH 4 (Fig. 9), PACl treated flocs the first series of coagulation has promoted particle agglomerations
were evidently bulkier than alum treated flocs. PACl with a higher by adsorption and bridging followed by charge neutralization with
charge density could neutralize and allow the agglomeration of the addition of chemical coagulants. The observed flocs were more
360 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364

Fig. 10. SEM images with (A) 100 and (B) 5000 times magnification for (a) kaolin, (b) alum granule and flocs generated from jar test using: (c) alum, (d) PACl, (e) rice starch,
(f) wheat starch.

rounded and this could be due to linking of multiple “spherical” trast, the presence of a significant broad peak at 3288–3279 cm−1
flocs as observed from the sole use of chemical coagulants. (Fig. 12b) corresponded to O H stretching in starches (Kačuráková
and Mathlouthi, 1996). The CH2 stretching and twisting of carbo-
3.5.3. Functional groups detection using FTIR hydrates were indicated by peaks detected at 2931–2929 cm−1 and
The functional groups present in coagulants (Fig. 12a and b), 1338–1336 cm−1 respectively (Kizil et al., 2002; Theivasanthi et al.,
kaolin particles (Fig. 12c) and respective flocs generated (Fig. 12d–f) 2011). According to Fan et al. (2012), the skeletal mode vibration of
were detected using FTIR; a method for the identification of a a-1,4 glycosidic linkage of starches (C O C) peaks at the region of
pure compound or unknown materials and changes in specific con- 927–926 cm−1. The amylose and amylopectin fractions in starches
formations. The IR spectra of alum granule (Fig. 12a) suggested are known to consist this glycosidic bond (Jay-lin, 2003). The other
the presence and stretching of OH groups with various bonds significant peak located at the wavenumbers of 858–853 cm−1 sug-
as observed by peaks at 2952, 1658 and 1062 cm −1 (Mahadevan gested the presence of CH2 deformation (Yao et al., 2002).
Pillai et al., 2000; Ni et al., 2012; Rong et al., 2013). In con-
362 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364

Fig. 11. SEM images with (A) 100 and (B) 5000 times magnification for flocs generated from different mode of additions: aid (a) alum, (b) PACl, together (c) alum, (d) PACl,
two step coagulation (e) alum, (f) PACl.

Similarity between the spectrum of kaolin (Fig. 12c) and vari- at 755–750 cm−1 suggested the stretching of Al O bond in the
ous treated flocs (Fig. 12d–f) suggested the successful removal of tetrahedral structure of AlO4 in kaolin (Heah et al., 2013). As the
colloidal kaolin particles by the studied coagulants. The two dis- IR spectra observed were identical, it can be deduced that sim-
tinct peaks at 3690–3619 cm−1 corresponded to the stretching of ilar functional groups were present in the treated flocs. Despite
O H bonds in kaolin while the presence of silicate was confirmed that, the transmission intensities at the range of 1115–750 cm−1
from the detection of a weak peak at 1115 cm−1 (Heah et al., 2013; were of marginal difference in the order of kaolin > alum treated
Yavuz and Saka, 2013). The subsequent double peaks observed at flocs > PACl treated flocs > rice starch treated flocs. According to
1031–1005 cm−1 can also be attributed to the presence of Si O Fan et al. (2012), variations in band intensities could be linked to
bonds; typical for kaolin-based materials. Another characteristic changes in specific conformations. PACl which is greatly cationic
feature of kaolin; the vibration of Al OH bonds is often assigned produced flocs that have higher absorbance intensities as governed
to wavenumbers of 950–900 cm−1 which was evident from the by the stronger net charge present on the surface of flocs. The inten-
peaks obtained at 913–911 cm−1 (Fig. 12c–f). Moreover, the band sity of transmission is also affected by the number of bonds present.
 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364 363

4. Conclusions

a) In this paper, the application of conventional starches as poten-

tial natural coagulants in treating turbid water and resulting
1658 properties of the generated flocs were investigated. The following
2952
926 conclusions can be drawn from this study:
1062
b) 1. Autoclaved rice starch outperformed other cereal starches
(wheat and corn) as well as tuber starch (potato) with turbid-
1644
2913 1337 ity removal close to 50% at pH 4 with an optimized coagulant
926
3280
999 857
dosage of 120 mg/L.
1149 2. The gelatinization of starch which has resulted in the rupture
of individual particles and the thickening of starch solution is
c) 1077
necessary to yield positive coagulation activities.
% Transmission

3619 3. Governed by the moderate coagulation activities of rice starch,

3692 753
1115 the use in tandem with chemical coagulants through 3 different
913 modes of additions has yielded varying turbidity removal perfor-
d) 1031 1008 mance in the order of: aid > two step coagulation > simultaneous
addition.
3619
3691
1115
912
753 4. The SEM images of flocs through the use of gelatinized starch
1031 1008 confirmed the adsorption and bridging mechanism as noticeable
bridges of kaolin particles and various points of particle branch-
e) ing were detected. Mean floc sizes ranging from 300 to 500 µm
3619 1115 were obtained when starches were used as primary coagulants.
3692 913753
1031 1101 5. Floc properties such as size, bulk density and sludge volume were
also influenced by the corresponding coagulation mechanisms
leading to particle aggregations. Reduced mean floc sizes ranging
f)
from 50 to 150 µm were obtained following the use of chemi-
1115 cal coagulants with adsorption and charge neutralization as the
1030 911753
1005 dominant coagulation mechanism. Additionally, the estimated
bulk densities were noticeably larger; more than 2 times bulkier
than starch treated flocs which agreed closely to the SEM images
obtained.
3700 3200 2700 2200 1700 1200 700
Wavenumbers (cm-1) Acknowledgements
Fig. 12. FTIR spectra of coagulants and flocs: (a) alum granule, (b) rice starch, (c)
kaolin, (d) treated flocs using alum, (e) treated flocs using PACl, (f) treated flocs
The authors are grateful to Monash University Malaysia for pro-
using rice starch. viding the research scholarship and essential support for this work
Choy Sook Yan is a recipient of the Higher Degree by Research
Scholarship (HDR) awarded by Monash University Malaysia. Spe-
cial thanks also to Holy Mate Sdn. Bhd. for providing the PACl stock
liquid used in this study.

Appendix A. Supplementary data

The bridging mechanism which linked several kaolin aggregates
with rice starch could exhibit a higher number of bonds; leading to
Supplementary data associated with this article can be found, in
lower transmission intensities (Fig. 12f).
the online version, at http://dx.doi.org/10.1016/j.ecoleng.2016.05.
The IR spectra obtained with the combined use of coagulants
082.
(Fig. S2a–c) were similar to that of pure kaolin (Fig. 12c). Although
the trend was similar, the use of rice starch with chemical coag-
ulants has drastically reduced the peak intensities at the range of References
1115–750 cm−1. The peak intensity lowering effect could be due to
Baharuddin, N.H., Sulaiman, N.M.N., Aroua, M.K., 2014. Unmodified starch as
the primary coagulation mechanisms involved. The introduction water-soluble binding polymer for chromium ions removal via polymer
of starch has led to bridging of aggregates and thus, lowering the enhanced ultrafiltration system. J. Environ. Health Sci. Eng. 12, 1–10.
transmission intensities as the number of bonds present increased. Biliaderis, C.G., 2009. Chapter 8 −structural transitions and related physical
properties of starch. In: BeMiller, J., Whistler, R. (Eds.), Starch. , third edition.
Besides that, no substantial differences were observed in terms of Academic Press, San Diego, pp. 293–372.
transmission intensities amongst the studied mode of coagulant Bratby, J., 2006. Coagulation and Flocculation in Water and Wastewater Treatment,
addition in all the IR spectra for both alum (Fig. S2A) and PACl (Fig. 2nd ed. IWA Publishing, UK.
Caskey, J., Primus, R., 1986. The effect of anionic polyacrylamide molecular
S2B). Thus, it can be suggested that the changes in specific confor- conformation and configuration on flocculation effectiveness. Environ. Prog. 5,
mations of flocs were absent. Alum treated flocs with rice starch 98–103.
exhibited higher peak intensities in all cases as compared to PACl Choy, S.Y., Prasad, K.M.N., Wu, T.Y., Raghunandan, M.E., Ramanan, R.N., 2014.
Utilization of plant-based natural coagulants as future alternatives towards
which was consistent to the sole use of chemical coagulants. It can sustainable water clarification. J. Environ. Sci. 26, 2178–2189.
also be deduced from the above analysis that the different mode of Choy, S.Y., Prasad, K.M.N., Wu, T.Y., Ramanan, R.N., 2015. A review on common
coagulant addition studied would eventually result in a similar IR vegetables and legumes as promising plant-based natural coagulants in water
clarification. Int. J. Environ. Sci. Tech. 12, 367–390.
spectra and the overall pattern would be governed by the dominant
Copeland, L., Blazek, J., Salman, H., Tang, M.C., 2009. Form and functionality of
compound in the treated flocs. starch. Food Hydrocolloids 23, 1527–1534.
364 S.Y. Choy et al. / Ecological Engineering 94 (2016) 352–364

Duan, J., Gregory, J., 2003. Coagulation by hydrolysing metal salts. Adv. Colloid Ndabigengesere, A., Subba Narasiah, K., 1998. Quality of water treated by
Interface Sci. 100–102, 475–502. coagulation using Moringa oleifera seeds. Water Res. 32, 781–791.
Fan, D., Ma, W., Wang, L., Huang, J., Zhao, J., Zhang, H., Chen, W., 2012. Ndabigengesere, A., Subba Narasiah, K., Talbot, B.G., 1995. Active agents and
Determination of structural changes in microwaved rice starch using Fourier mechanism of coagulation of turbid waters using Moringa oleifera. Water Res.
transform infrared and Raman spectroscopy. Starch 64, 598–606. 29, 703–710.
Flaten, T.P., 2001. Aluminium as a risk factor in Alzheimer’s disease, with emphasis Ni, F., Peng, X., He, J., Yu, L., Zhao, J., Luan, Z., 2012. Preparation and characterization
on drinking water. Brain Res. Bull. 55, 187–196. of composite bioflocculants in comparison with dual-coagulants for the
Gauthier, E., Fortier, I., Courchesne, F., Pepin, P., Mortimer, J., Gauvreau, D., 2000. treatment of kaolin suspension. Chem. Eng. J. 213, 195–202.
Aluminum forms in drinking water and risk of Alzheimer’s disease. Environ. Oladoja, N.A., 2015. Headway on natural polymeric coagulants in water and
Res. 84, 234–246. wastewater treatment operations. J. Water Process Eng. 6, 174–192.
Gebbie, P., 2006. An operator’s guide to water treatement coagulants, 31 st Annual Rondeau, V., Commenges, D., Jacqmin-Gadda, H., Dartigues, J.-F., 2000. Relation
QLD Water Industry Workshop—Operations Skills. between aluminum concentrations in drinking water and Alzheimer’s disease:
Ghebremichael, K., 2007. Overcoming the drawbacks of natural coagulants for an 8-year follow-up study. Am. J. Epidemiol. 152, 59–66.
drinking water treatment. Water Sci. Technol. Water Supply 7, 87–93. Rondeau, V., Jacqmin-Gadda, H., Commenges, D., Dartigues, J.-F., 2001. Re:
Gregory, J., 2005a. Coagulation and Flocculation Particles in Water: Properties and aluminium in drinking water and cognitive decline in elderly subjects: the
Processes. CRC Press, pp. 121–148. Paquid Cohort. Am. J. Epidemiol. 154, 288–290.
Gregory, J., 2005b. Colloid Interactions and Colloid Stability Particles in Water: Rong, H., Gao, B., Dong, M., Zhao, Y., Sun, S., Yue, Q., Li, Q., 2013. Characterization of
Properties and Processes. CRC Press, pp. 63–92. size, strength and structure of aluminum-polymer dual-coagulant flocs under
Han, J.-A., BeMiller, J., Hamaker, B., Lim, S.-T., 2003. Structural changes of different pH and hydraulic conditions. J. Hazard. Mater. 252, 330–337.
debranched corn starch by aqueous heating and stirring. Cereal Chem. 80, 323– SPAN, 2010. Malaysia Water Service Industry & Water Treatment Sludge Issues
328. (Accessed 19.1.2015) http://www.jwrc-net.or.jp/aswin/projects-activities/
Heah, C.Y., Kamarudin, H., Al Bakri, A.M., Bnhussain, M., Luqman, M., Nizar, I.K., trainee files/20101004-15 CR02.pdf.
Ruzaidi, C.M., Liew, Y., 2013. Kaolin-based geopolymers with various NaOH Sathiyabama, M., 2012. Purification of a coagulant protein from seeds of Moringa
concentrations. Int. J. Miner. Metall. Mater. 20, 313–322. concanensis. Water Sci. Technol. Water Supply 12, 329–333.
Hegenbart, S., 1996. Understanding Starch Functionality (Accessed 10.1.2016) Shahriari, T., Nabi Bidhendi, G., Shahriari, S., 2012. Evaluating the efficiency of
http://www.naturalproductsinsider.com/articles/1996/01/understanding- Plantago ovata and starch in water turbidity removal. Int. J. Environ. Res. 6,
starch-functionality.aspx. 259–264.
ISI, 2014. International Starch: Raw Materials (Accessed 31.12.2014) http://starch. Sharma, B., Dhuldhoya, N., Merchant, U., 2006. Flocculants—an ecofriendly
dk/isi/market/index.asp. approach. J. Polym. Environ. 14, 195–202.
Idris, J., Md Som, A., Musa, M., Ku Hamid, K.H., Husen, R., Muhd Rodhi, M.N., 2013. Shen, Y.-H., Dempsey, B.A., 1998. Synthesis and speciation of polyaluminum
Dragon fruit foliage plant-based coagulant for treatment of concentrated latex chloride for water treatment. Environ. Int. 24, 899–910.
effluent: comparison of treatment with ferric sulfate. J. Chem. 2013, 1–7. Simate, G.S., Iyuke, S.E., Ndlovu, S., Heydenrych, M., Walubita, L.F., 2012. Human
Jahn, S.A.A., 2001. Drinking water from Chinese rivers: challenges of clarification. J. health effects of residual carbon nanotubes and traditional water treatment
Water Supply Res. Tech. 50, 15–27. chemicals in drinking water. Environ. Int. 39, 38–49.
−
Jane, J.-l., 2009. Chapter 6 structural features of starch granules II. In: BeMiller, J., Sincero, G.A., Sincero, A.P., 2002. Coagulation, Physical-Chemical Treatment of
Whistler, R. (Eds.), Starch. , third edition. Academic Press, San Diego, pp. Water and Wastewater. CRC Press.
193–236. Teh, C.Y., Wu, T.Y., Juan, J.C., 2014a. Optimization of agro-industrial wastewater
Jay-lin, J., 2003. Starch, Chemical and Functional Properties of Food Saccharides. treatment using unmodified rice starch as a natural coagulant. Ind. Crop. Prod.
CRC Press. 56, 17–26.
Kačuráková, M., Mathlouthi, M., 1996. FTIR and laser-Raman spectra of Teh, C.Y., Wu, T.Y., Juan, J.C., 2014b. Potential use of rice starch in coagulation–
oligosaccharides in water: characterization of the glycosidic bond. Carbohydr. flocculation process of agro-industrial wastewater: treatment performance and
Res. 284, 145–157. flocs characterization. Ecol. Eng. 71, 509–519.
Kim, S.-H., Moon, B.-H., Lee, H.-I., 2001. Effects of pH and dosage on pollutant Theivasanthi, T., Venkadamanickam, G., Palanivelu, M., Alagar, M., 2011. Nano
removal and floc structure during coagulation. Microchem. J. 68, 197–203. sized powder of jackfruit seed: spectroscopic and anti-microbial investigative
Kizil, R., Irudayaraj, J., Seetharaman, K., 2002. Characterization of irradiated approach. Nano Biomed. Eng. 3, 215–221.
starches by using FT-Raman and FTIR spectroscopy. J. Agric. Food. Chem. 50, Walton, J.R., 2013. Aluminum’s involvement in the progression of Alzheimer’s
3912–3918. disease. J. Alzheimer’s Dis. 35, 7–43.
Li, T., Zhu, Z., Wang, D., Yao, C., Tang, H., 2006. Characterization of floc size, −
Willett, J.L., 2009. Chapter 19 starch in polymer compositions. In: BeMiller, J.,
strength and structure under various coagulation mechanisms. Powder Whistler, R. (Eds.), Starch. , third edition. Academic Press, San Diego, pp.
Technol. 168, 104–110. 715–743.
Mahadevan Pillai, V.P., Nayar, V.U., Jordanovska, V.B., 2000. Infrared and polarized Wongsagonsup, R., Shobsngob, S., Oonkhanond, B., Varavinit, S., 2005. Zeta
Raman spectra of (CH3)2NH2Al(SO4)2 ·6H2O. Spectrochim. Acta Pt. A Mol. potential ($) analysis for the determination of protein content in rice flour.
Biomol. Spectrosc. 56, 887–896. Starch 57, 25–31.
Malhotra, M., 1994. Poly aluminium chloride as an alternative coagulant. In: 20th Yang, Z., Wu, H., Yuan, B., Huang, M., Yang, H., Li, A., Bai, J., Cheng, R., 2014.
WEDC Conference Colombo: Affordable Water Supply and Sanitation, Sri Synthesis of amphoteric starch-based grafting flocculants for flocculation of
Lanka. both positively and negatively charged colloidal contaminants from water.
Mavura, W.J., Chemelil, M.C., Saenyi, W.W., Mavura, H.K., 2008. Investigation of Chem. Eng. J. 244, 209–217.
chemical and biochemical properties of Maerua subcordata plant extract: a Yao, Y., Zhang, J., Ding, X., 2002. Structure-retrogradation relationship of rice starch
local water clarification agent. Bull. Chem. Soc. Ethiop. 22, 143–148. in purified starches and cooked rice grains: a statistical investigation. J. Agric.
McLachlan, D.R., 1995. Aluminium and the risk for Alzheimer’s disease. Food. Chem. 50, 7420–7425.
Environmetrics 6, 233–275. Yavuz, Ö., Saka, C., 2013. Surface modification with cold plasma application on
Mohd Omar, F., Sohrab, H., Tjoon Tow, T., 2013. Semiconductor wastewater kaolin and its effects on the adsorption of methylene blue. Appl. Clay Sci. 85,
treatment using tapioca starch as a natural coagulant. J. Water Resour. Prot. 96–102.
Nansubuga, I., Banadda, N., Babu, M., Verstraete, W., Van de Wiele, T., 2013. Effect
of polyaluminium chloride water treatment sludge on effluent quality of
domestic wastewater treatment. Afr. J. Environ. Sci. Technol. 7, 145–152.