Journal of Environmental Chemical Engineering 9 (2021) 106105

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

journal homepage: www.elsevier.com/locate/jece

Concurrent removal of hardness and fluoride in water by

monopolar electrocoagulation
J.U. Halpegama a, b, e, K.Y. Heenkenda a, Zhiguo Wu a, b, K.G.N. Nanayakkara c, R.M.
G. Rajapakse d, A. Bandara d, Ajith C. Herath e, Xing Chen f, Rohan Weerasooriya a, f, *
a
National Centre for Water Quality Research, National Institute of Fundamental Studies, Hanthana, Kandy 20000, Sri Lanka
b
Postgraduate Institute of Science, University of Peradeniya 20400, Sri Lanka
c
Department of Civil Engineering, Faculty of Engineering, University of Peradeniya 20400, Sri Lanka
d
Department of Chemistry, Faculty of Science, University of Peradeniya 20400, Sri Lanka
e
Department of Chemical Sciences, Faculty of Applied Sciences, Rajarata University of Sri Lanka, Mihintale 50300 Sri Lanka
f
Industry and Equipment Technology Institute, Hefei University of Technology, Hefei 230009, PR China

A R T I C L E I N F O A B S T R A C T

Editor: Despo Kassinos Water treatment based on electrocoagulation (EC) is attractive since required chemicals and colloids are pro­
duced in-situ. However, optimisation of EC operation parameters is necessary to enhance its efficiency. We
Keywords: optimised EC cell parameters by the response surface method (RSM). The optimal removal efficiencies of
Defluoridation hardness (63%) and fluoride (97%) were achieved at 1.98 kW h/m3. With the removal of divalent cations, some
Electrocoagulation
anionic species concurrently remove via an energetically feasible route to adjust the charge balance. When
Fluoride
simulated water is used (450 mg/L TDS, 580 mg/L CaCO3, 10 mg/L fluorides and pH 6.50), 83% hardness and
Hardness
Response surface method 99% fluoride are removed with 0.69 kW h/m2 energy consumption. The chemical species in the solution matrix,
particularly SO2,
4 significantly affect the hardness and fluoride removal efficiencies. The contaminated EC sludge
resulted from feed water is characterised by spectroscopic methods to probe hardness and fluoride removal
mechanisms. In the presence of Mg2+, F- interacts with Al-sludge sites forming ≡ MgF − OH. When Ca2+ and F-
are present, both ≡ CaF − OH and ≡ CaF are formed. In Ca2+, Mg2+ and F- treated Al-sludge dominates – –
CaF− OH and ≡ CaF over ≡ MgF − OH.

1. Introduction activities to restore charge balance via energetically feasible pathways.

Therefore, eliminating major cations in water is required first, and then
Due to intense rock weathering in the tropics, the natural ground­ assess other contaminants in the treated water to determine additional
water enriches lithogenic solutes as fluoride and hardness (viz., divalent treatment, if needed.
cations). The adverse climatic conditions prevailing in these regions Hardness can be removed by reverse osmosis (RO) or nano­
force people to consume large quantities of solutes-enriched water, often membranes (NM), ion exchange resins, lime softening, and electro­
implicating severe health problems [1–3]. Fluoride is an enforceable chemical methods [5,6]. In RO technology, almost all solutes in water
primary water quality standard [4] (maximum contaminants limit, MCL are removed; hence, the treated water becomes unpalatable due to a lack
1.5 mg/L). Hardness is a non-enforceable secondary water quality of solutes. Long-term consumption of solutes-deficient water may also
standard where no MCL is defined. However, hardness (and TDS) ren­ lead to nutritional disorders [7,8]. In lime softening, excess sludge is
ders water unpalatable, restricting its human consumption [2]. There­ generated, and the treated water also requires additional treatment. The
fore, improvement of palatability of the drinking water is essential ion exchange plants generate high TDS water enriched with Na+ that
before treating excess trace solutes as fluoride. When water hardness need different treatment [9,10]. In contrast, the electrochemical
(Ca2+ and Mg2+) is removed, anions in solution may adjust their methods can provide a solution for concurrent removal of fluoride and

* Corresponding author at: National Centre for Water Quality Research, National Institute of Fundamental Studies, Hanthana, Kandy 20000, Sri Lanka.
E-mail addresses: [email protected] (J.U. Halpegama), [email protected] (K.Y. Heenkenda), [email protected] (Z. Wu), [email protected]
(K.G.N. Nanayakkara), [email protected] (R.M.G. Rajapakse), [email protected] (A. Bandara), [email protected] (A.C. Herath),
[email protected] (X. Chen), [email protected] (R. Weerasooriya).

https://doi.org/10.1016/j.jece.2021.106105
Received 2 June 2021; Received in revised form 17 July 2021; Accepted 20 July 2021
Available online 24 July 2021
2213-3437/© 2021 Elsevier Ltd. All rights reserved.
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

hardness. Two types of electrochemical methods, i.e., electrodialysis Table 1

self-reversal (EDR) and electrocoagulation (EC), are used to remove Chemical composition of simulated and groundwater samples.
hardness, fluoride or other solutes in water [11,12]. By EDR technology, Sample type and pH EC TDS Total hardness Fluoride
the hardness in water removes efficiently; however, they show limited location µS/ mg/L mg/L CaCO3 mg/L
success in removing fluoride. The fluoride in water replaces OH- on the cm
membrane, forming specific bonds with the membrane sites, which the Simulated water 6.5 970 450 580 10
EDR self-reversal process cannot regenerate. The disposal of liquid EDR (Ca2+(aq), Mg2+(aq)
concentrates poses an additional issue [13–16]. The electrocoagulation and F-(aq))
L-1 (8◦ 21′ 11.3′′ N, 6.6 970 470 183 0.45
method can be treated as a unification of several water treatment pro­ 80◦ 30′ 07.9′′ E)
cesses, e.g., coagulation, sedimentation, flotation, and oxidation [17]. L-2 (8◦ 19’53.9" N, 7.1 750 380 291 0.48
Unlike membrane technology, in EC, the composition of the inlet water 80◦ 36’04.1" E)
does not require rigorous pretreatment to remove particulates, natural
organic matter, etc. [18,19]. The EC technology can remove inorganic
requires assessment for trace constituents (in this case, fluoride). We
ions, organic compounds, colour, oil, grease, and turbidity in water
used a multi-factorial method based on response surface method (RSM)
[20]. Further, the EC can destruct microbes via cell rupturing under an
to optimise the current flux and reaction time for optimal hardness
electric flux [21]. The sludge generated in EC is compact due to minimal
removal in water [31]. The response variable is only a function of a
enrichment of ions in solution and floats by gas bubbles generated at the
singly varied parameter. Therefore, the desired independent variables
cathode [22]. Electrocoagulation technology is not complex compared
can be combined, and their mutual interactions can also evaluate for
to membrane or EDR methods, and the routine operation is straight­
optimisation [32]. Besides, the RSM method decreases the total number
forward. Hence, compared to membrane technology or EDR, EC
of experiments required to optimise by saving time and experimental
methods are simple, robust, chemical-free, and cost-effective [19,
costs, providing an additional advantage. However, the optimisation
23–25].
process optimises operational conditions without shedding light on
Electrocoagulation does not require external coagulants since they
hardness or fluoride mechanisms [33]. At optimal conditions, the
generate in situ by dissolving metal anodes. Compared to Fe, the Al
Al-derived sludge generated by the EC was analysed by spectroscopic
electrodes used in EC cells is profitable for the following reasons: irre­
methods to elucidate fluoride and hardness removal mechanism.
spective of the current flux, Fe corrodes spontaneously in water; hence
the generation of Fe derived coagulants is uncontrolled. However, the Al
electrodes undergo surface passivation by forming an Al2O3 layer, and 2. Materials and methods
under controlled current flux water [Al3+] can regulate; unlike Fe3+, the
2.1. Materials
Al3+ undergoes polymerisation forming Al(OH)3+ 4+
15 , Al7 (OH)17 ,
3+
Al8 (OH)4+
20 and, Al13 O4 (OH)7+
therefore, the presence of free Al is
24 The aluminium plates (98% purity) required for EC cell design were
limited [19,26]. The required Al derived coagulants can control the obtained from Spectra Industries Lanka (Pvt) Ltd. Analytical grade, A.R.,
demand; hence the sludge production and alum addition can cut off NaF, MgCl2, CaCl2. 2H2O, HNO3 and NaClO4 were obtained from Sigma
substantially. Therefore, we used Al sacrificing anode in EC to examine Aldrich (USA). Milli Q water (Thermo ultra-pure system, Hungary
concurrent hardness and fluoride removal in water. The water hardness 0.055 µS/cm) was used in sample preparation. The required ground­
is removed by precipitating divalent ions on the cathode and adsorption water samples for experiments were collected from two villages within
onto Al(OH)3n flocs produced by the EC cell in near-neutral pH [5,19,27] the same climatic zone in Sri Lanka. (The detailed chemical composition
(e.g. pH 6–8). In determining the effect of Ca2+ or Mg2+ on defluor­ of the water samples is shown in Table 1-S; support documentation). For
idation an optimal value is observed in the presence of Mg2+ by forming optimisation experiments, a simulated water sample was prepared to
an Mg (OH)2 coating on Al-derived colloids that results in low F- match the hardness and fluoride of the natural water using NaF, MgCl2,
retention affinity. The difference in EC defluoridation between Ca2+ and CaCl2. 2H2O and NaClO4 with a controlled matrix. Table 1 shows the
Mg2+ is known to depend on flocs composition. However, the synergy chemical composition of the water samples used.
between Ca2+ and Mg2+ on defluoridation requires a detailed study.
Mostly the EC methods are used to remediate a single contaminant as
fluoride in water. Its versatility in removing multiple pollutants in 2.2. Methods
drinking water is also demonstrated. The EC cells with Fe-Al anode are
used to remove Sb and As [28] or As and F [29]. The substitution of Fe3+ 2.2.1. Lab-scale batch electrocoagulation reactor
ions by Al3+ ions in the solid surface resulted in increased surface sites A laboratory-scale cylindrical batch reactor (Pyrex cell: diameter
for the adsorption of pollutants. However, the concurrent production of 6.4 cm and height 8.00 cm) was used in the experiments. Two plates of
aluminium and iron derived colloids may result in excess wastes that aluminium (dimensions 6.00 cm × 1.60 cm × 0.44 cm; immersion
require additional treatment [28]. Guzman et al. report the efficient depth 5.30 cm) were used as electrodes. The net capacity of the EC
removal of fluoride and arsenate in natural water using EC with Al an­ reactor was 175 mL. The two electrodes were connected in a monopolar
odes where ion exchange and adsorption co-occur for fluoride and configuration. The gap between the electrodes was maintained at
arsenic removal, respectively [30]. In this case, however, the water 1.00 ± 0.01 cm. Regulated current supply was given with high precision
hardness is low (9.8 mg/L). DC power supply (ZHAOXIN PS 305D; resolution voltage-0.1 V [stability
We optimised EC cell parameters for concurrent removal of hardness ≤0.01% + 2 mV], current-0.01 A [stability ≤0.1% + 3 mA]. Before each
(e.g., Ca2+ Mg2+) and fluoride in groundwater. We first optimised the EC experiment, electrodes were rubbed with sandpaper to remove scale and
cell for hardness removal efficiency in the presence of fluoride in water. soaked in 1 M HNO3 and Mill Q water for surface cleaning. The EC cell
The ions removal efficiency depends on water quality (pH, ionic with monopolar electrode configuration, thus designed was used to
strength, temperature, TDS) and EC cell (electrode dimension, applied optimise process time and current density for water hardness removal
current and voltage, processing time) parameters. The performance of according to the method discussed in Section 2.2.2.
EC cell depends markedly on the cell parameters compare to water
quality. The optimisation of EC cell parameters first requires a system­ 2.2.2. Electrochemical cell parameters optimisation
atic approach for major solutes (in this case, hardness) in water. Once Response surface method was used for the statistical design of the
the optimisation is achieved, the composition of the treated water experiments to evaluate the optimum EC cell operational parameters
(process time: X1 and current density: X2) to maximise water hardness

2
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

Table 2 removal efficiency, Y1 according to the following model.

Central composite design observations of independent variables and predicted ∑ ∑ ∑
and experimentally achieved hardness removal efficiency used for the Y = β0 + βi Xi + βij X2i + βij Xi Xj
optimisation. i i i>j

Independent Factor Range and level where Y response function (hardness removal efficiency), β0 is intercept,
factors
X Low High (+1) βi and βij are linear and quadratic coefficients, βij accounts for Xi and Xj ,
(− 1)
and Xi and Xj are independent factors, i.e., current density and process
Time (min) X1 10 45 time. Based on initial experiments, the initial values of independent
Current density X2 0.117 0.353 variables are shown in Table 2. According to the central composite
(A/cm2) design (CCD) developed by Minitab code, a possible combination of
Design variables
Run order Actual levels Observations - hardness
experimental protocols was obtained. Accordingly, thirteen experi­
removal efficiency (%) mental protocols were used to evaluate response function experimen­
Time Current density (mA/cm2) L-1 L-2 tally (Table 2). The model optimisation was carried out with the natural
(min) water collected from the L-1 (Table 1). Afterwards, the model validation
1 3.53
was carried out with the natural water sample collected from L-2 from
52.2 63 64
2 2.35 the same climatic zone. After that the validated model was used to
45 57 59 simulate the chemical composition of the stimulated water.
3 4.71
27.5 48 52 2.2.3. Water quality assessments and sludge characterisation
4 4.71
27.5 48 52
The samples’ pH and electrical conductivity (TDS calculated) were
5 5.89 measured by a multi-parameters analyser (HANNA HI 9811-5, USA).
10 35 41 Cations in solution were measured by inductively coupled plasma op­
6 4.71 tical emission spectrometer (ICP-OES) (ICPA 7000, Thermo, USA) using
27.5 42 51
certified standard solutions for calibration. Fluoride in solution was
7 4.71
2.7 38 31 analysed using ion chromatography (Shimadzu, Japan) equipped with
8 2.35 the conductivity detector (Shimadzu CDD 10A VP, Japan) and auto­
10 38 37 sampler. Quality control data of the measurements made by IC and ICP-
9 4.71 OES as described elsewhere [34–36]. The spike recoveries of fluoride
27.5 51 53
10 4.71
hardness were measured using NIST CRN anion solution (ROTH
27.5 54 54 multi-element IC CAS 2668). In water, 97.83 ± 0.06% recoveries were
11 5.89 obtained for fluoride. For calcium and magnesium detection, ICP CRN
45 62 59 was used (TRaceCERT 54704 Siga Aldrich, USA), and the spike re­
12 7.07
coveries are 100.3 ± 1.51% and 98 ± 1.76% for calcium and magne­
27.5 48 53
13 1.17 sium, respectively. The transmission FTIR method was used to
27.5 48 54 characterise the Al sludge generated by EC experiments. KBr pellets
were prepared at 1:10 sample: KBr ratio. Transmission FTIR spectra

Fig. 1. Variations of hardness removal efficiency in A. groundwater of locations 1 and 2; symbols represent observations. Source water composition, current density
and response time were related to Table 1 and 2. B. Simulated water by EC. Symbols represent various simulated water types. Model calculations were carried out
with an RMS model developed for natural groundwater.

3
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

Fig. 2. Variation of hardness and fluoride removal efficiencies as a function of current density and process time. A. hardness B. fluoride as a function of time and
current density for location 1. Source water composition: TDS 470 mg/L, hardness 183 mg/L CaCO3, fluoride 0.45 mg/L and pH 6.60.

were recorded with an IR spectrometer with a DTGS detector (iS50

Thermo Scientific, USA). All spectra were obtained at 4 cm− 1 resolutions (Y1 ) = 38.4 + 0.170X1 − 0.36X2 + 0.00324X21 − 0.122X22 + 0.0570(X1 X2 )
in the 400 – 4000 cm− 1 spectral range. X-ray diffractograms of (1)
aluminium sludge were obtained for phase identification. An X-ray
diffractometer at 20 kV and 30 mA was operated using Cu-Kα radiation (S = 3.78, R2 = 89.39%, R2adj = 81.81%, R2pre = 71.08%)
at λ = 0.154 nm (Bruker D8 Advance Eco). X-ray photoelectron spec­ The suitability of the RSM model in response function predictions is
troscopy (XPS) measurements were carried out using a Thermo Esca lab embedded in the response variable (S), regression coefficient (R2), R2
250Xi (Thermo Fisher Scientific, Waltham, MA, USA) system using a adjusted, and R2pre predicted values. The high regression coefficient (R2
monochromatic Al Kα source (1320 eV, 140 W). Peak fitting was per­ = 89.39%) and its agreement with adj. R2 complies with a good data fit.
formed according to Gaussian/Lorentzian functions using Origin Pro 9 The S value measures the agreement between data values and response
software. The EC sludge morphology and surface structures were surface. A minimal S value implies a good agreement between mea­
observed by cold field emission scanning electron microscope (Hitachi, surements and modelled data [32,37]. Experimental error of hardness
Japan; SU8020). removal efficiency as shown by lack of fits is statistically insignificant
(Pr(probability > 0.05), and the model adequately predicts observations
3. Results and discussion with 6% uncertainty [37] (Table 3-S). The optimised RSM model was
then validated using an independent dataset from a different location in
3.1. Central composite design model development for hardness removal the same climatic zone (viz. L-2). Accordingly, the hardness removal
efficiency values were calculated, and their agreement with observa­
The efficiency of treating contaminants by EC depends on both feed tions are shown in Fig. 1-A. The residuals, e.g., the difference between
water composition, viz., pH, ionic strength, contaminants load, and EC experimental and predicted responses, could be utilised to investigate
cell parameters, viz. current and process time. Hardness occurs mainly the adequacy of the model [38,39]. Based on a normal distribution,
due to major divalent cations as Ca2+ and Mg2+ in water. To ensure residuals are considered unexplained variations distributed randomly
drinking water palatability, the major species (presently Ca2+ and Mg2+) around zero if the model is a good predictor (Fig. 1-S). Thus, the RMS
require treating first; however, to ensure solution charge balance, an­ model predicts the observations adequately for the aquifer in the dry
ions’ concentration automatically adjusts. We designed experiments to climatic zone (Sri Lanka).
optimise EC cell parameters to maximise hardness removal in natural The production of coagulants by electrocoagulation depends on
groundwater by RSM. The central composite design of independent electrolysis time and current, which determines the size of the gas
design variables, viz. current density and time, are shown in Table 2. The bubbles generated at the cathode [40]. Variations of hardness and
coefficients of the response function for hardness removal efficiency is fluoride removal efficiency as a function of current density and contact
shown in Table 2-S. The ANOVA data of independent design factors time for the water collected from L-1 are shown in Fig. 2-A and B (similar
indicate that the RSM regression model adequately interprets observa­ trends of hardness removal were observed for L-2, hence not shown).
tions (F = 11.80 Table 3-S), and the hardness removal efficiency ranges The contour patterns of hardness and fluoride removals to current
between 35% and 63%. Statistical analysis showed the significance of density and process time differ markedly. At a given process time, the
the linear term (X1, time) in the model with Pr (probability) < 0.05 hardness removal efficiency does not vary significantly with the current
(Table 3-S). However, the coefficients correspond to linear (X1), square density. At a given current density, the hardness removal efficiency
(X21) and cross-term interactions (X1X2) are not significant (P(r)> 0.05). monotonously increases with the processing time. However, the spatial
Accordingly, the optimised RSM model for hardness removal efficiency variation of fluoride removal efficiency shows a variation with the
in water is shown below: current density and process time reaching a hillock morphology [39].
The removal of hardness inducing species from water occurs on the
cathode surface, and the flocs are generated by Al dissolution. Increased

4
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

Fig. 3. A. Water hardness and fluoride removal efficiency using optimised EC. B. Residual aluminium concentration in treated water. Feed water composition is
shown in Table 1 and Table 1-S. Vertical lines depict error bars.

current density resulted in enhanced OH- production at the cathode may

precipitate Ca2+/Mg2+ species on the cathode surface. However, the mAl3+ + nCa2+ + (3m + 2n)H2 O = Alm Can (OH)3m+2n + (3m + 2n)H+
mass balance calculations show that less than 1% of Ca2+ and Mg2+ Mg2+ also use as a co-coagulant with aluminium salts. However,
removed on the cathode surface. However, aluminium flocs provide there is a possibility of forming Mg(OH)2 on the Aly(OH)3x surface.
many reactive sites for Ca2+ and Mg2+ retention [5]. Therefore, hard­ Therefore, Mg(OH)2 blind the Aly(OH)3x flocs [44,45]. Stoichiometric
ness removal efficiency is increased above 50 min process time between coefficients related to the reactions were denoted in m,n,x and y terms.
4.71 and 5.9 mA/cm2 current density. Over 80% of fluoride in water can However, we have noted a negligible mass loss of the cathode at the
also be removed (Fig. 2-B). However, very high current density values end of a treatment cycle. The pH values of both source and treated water
and process time negatively affect hardness (and fluoride) removal ef­ lie between 6.50 and 7.00. The negligible pH variations of the treated
ficiencies resulting in a high activity of Al3+ free ions [19]. In addition, and source waters are ascribed to the high buffer capacity of newly
after the treatment, high density and poor affinity between hydroxide formed polymeric aluminium colloids in neutralising OH- released at the
flocs and gas bubbles create a difficultly in floc separation. Because of cathode [19]. Therefore, additional pH adjustment is not required to
these reasons, elevated current density and electrolysis time conditions obtain the highest hardness and fluoride removal efficiency [43]. The
are not suited for hardness removal. Therefore, 3.53 mA/cm2 current situation is markedly different when hardness removal is accounted for
density and 52.2 min time are adequate for optimal removal of hardness via EC treatment with the groundwater (Fig. 1-B). Typically, a natural
(63%) and fluoride (97%) at 1.98 kW h/m3 energy consumption. The water system consists of coexisting anions, cations and natural organic
operation cost for removing hardness and fluoride in water by EC is also matter. Therefore, removing hardness in natural water is complicated
calculated using the prices of electrode materials (0.76 $/kg) and elec­ compared to simulated water with a controlled matrix. However,
trical energy (0.17 $/kW h) as input parameters according to the simulated water with coexisting ions, viz. NO-3, SO2- - -
4 , Cl and F , hardness
methodology given elsewhere [40]. The hardness and fluoride values of removal efficiency values are somewhat similar to the groundwater.
the treated water are 67.3 mg/L CaCO3 and 0.01 mg/L, respectively. Nitrate does not significantly influence destabilization with
Total operation cost is 0.93 US$/ 0.175 L water. aluminium coagulants due to its weak coordination tendency with metal
ions. However, chloride and sulphate considerably affect coagulation by
3.2. Matrix effects on hardness removal aluminium salts [42]. Hardness removal efficiency for simulated water
with sulphate anion was comparatively low than other coexisting an­
The activities of matrix anions and their chemical nature can change ions. Therefore, according to Fig. 1-B, our data implied a strong coor­
the behaviour of hydrolysed cationic coagulant, directly influencing the dination effect of sulphate to the aluminium hydroxide system [27]. Due
hardness removal efficiency. This interaction depends on the changes of to the surface ionisation and site-specific complex formation reactions,
electronegativity due to the coordination of aluminium and the sulphate anion can destabilise the positively charged aluminium hy­
replacement of hydroxyl ion or changes of positively charged sites on the drolysis products [41].
metal hydroxide precipitates [41,42]. In the absence of SO2- -
4 , Cl , and Further, the formation of CaF2 is also reduced, and therefore hard­
-
NO3 in the matrix, the hardness removal efficiency was closed to 80% for ness removal efficiency decreases with increasing the amount of sul­
simulated water, whereas fluoride removal efficiency was almost above phate [33]. Inhibition of aluminium electrode corrosion and
90% (Fig. 1-B). In each experiment set, a gelatinous deposition layer was competition between sulphate and fluoride ion negatively affect
observed on the anode surface due to the surface coprecipitation defluoridation efficiency. Therefore, reduced fluoride removal effi­
mechanism of defluoridation. ciency can be identified with high sulphate containing feed water.
Formation of CaF2 is possible during the treatment since Ca2+ ion
concentration is above 150 mg/L [43]. Although Aly(OH)3x flocs Al(OH)3− x Fx + ySO2−4 = Al(OH)3− x Fx− 2y (SO4 )y + 2yF−
contribute to F- removal, they also act as a nucleus for the formation of
The negligible effect can be identified with Cl- ions during defluor­
CaF2. Additionally, a positive effect of Ca2+ on fluoride removal was
idation [43]. Research is needed to assess the role of natural organic
probably associated with the coprecipitation of Al3+ and Ca2+.
matter on floc formation and hardness removal efficiency de­
terminations. The matrix of the solution affects hardness removal

5
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

Fig. 4. Scanning electron micrographs obtained for sludge generated after interacting with feed water of different composition. A. F-/Al feed water, B. Ca2+, Mg2+
/Al feedwater C. Ca2+, Mg2+, F-/Al feedwater.

porosity can be identified with a large surface area. Additionally,

amorphous sludge particulates are also present.

3.3.2. XRD characterisation

By the dissolution of Al sacrificing anode in EC, the Al-sludge is
generated. The chemical composition and the structure of the sludge
depend on the feed solution composition. The XRD data of Al derived
sludge generated in the EC cell upon interactions with hard water in the
presence or absence of F- are shown in Fig. 5. The XRD peaks at 2θ0
18.590, 20.440, 26.330, 50.650, and 52.360 confirmed gelatinous
Aly(OH)3x sludge phase believed to form via Al13 kegging in monopolar
EC configuration [48,49]. When Aly(OH)3x derived sludge interacts with
hard water, carbonate and hydroxide phases of calcium and magnesium
also emerge. The peaks at 2θ = 23.880, 29.360, 36.50, 39.50, 48.50, 56.60
correspond to the calcium carbonate phase readily discern in the
Al-sludge generated by fluoride hard water treatment [50]. AlF3.3 H2O,
CaF2, and MgF2 phases can also be identified when Al-derived sludge
interacts with hard water enriched with fluoride. Sharp diffraction peaks
correspond to AlF3.3 H2O are identified at 2θ = 24.730, 27.080, 32.890,
44.620, 53.110, 57.320, 59.430, 67.360, 70.630 [51]. The peaks at 2θ
= 15.150, 30.390, 48.700 confirm the presence of aluminium fluoride
hydroxide [29]. The relative intensity of XRD peaks correspond to
Aly(OH)3x has reduced when Ca2+ and Mg2+ is present in association
with F-. The reduced XRD peak intensity of the sludge confirms the
improved defluoridation in the presence of Ca2+ and Mg2+. Therefore,
the cubic phase of the fluorite (CaF2 space group Fm3m) can be iden­
tified. The XRD peaks at 2θ = 28.320, 47.150, 55.840, 68.730 corre­
Fig. 5. XRD patterns EC sludge after interacting with feed water of different sponded to (111), (202), (311) and (400) planes, respectively [29,52].
composition. B: fluoride feedwater G: hardness feedwater and H: hardness and The diffraction peaks at 2θ = 32.620, 46.650, 54.470 confirm the MgF2
fluoride feedwater. phase in the sludge [53].

efficiency. Therefore simulated water removal efficiency is not matched 3.3.3. XPS characterisation
with the developed model as contaminated water. The removal effi­ The high-resolution x-ray photoelectron spectroscopic spectrums of
ciencies of fluoride and hardness in different water types using opti­ O1s, C1s, Al2p, Ca2p, Mg1s, and F1s received by interacting EC sludge with
mised EC reactor are shown in Fig. 3.A. The residual aluminium feed solution containing Ca2+, Mg2+ and F- is shown in Fig. 6. The Al2p
concentration in treated water is always below the WHO maximum peak at 74.27 eV is ascribed to Aly(OH)3x phase [54]. The O1s peak at
contaminant levels (MCL) as shown in Fig. 3. B [46]. 531.8 eV occurs due to bulk hydroxyls in Aly(OH)3x phase [54]. The
( )
observed Al O
atomic ratio ~ 3 favours Aly(OH)3x stoichiometry
3.3. Sludge characterisation
(Table 4-S). As EC process time evolved, the aggregation of gelatinous
3.3.1. SEM characterisation Aly(OH)3x particulates is discerned, and the formation of γ-AlO(OH)
SEM images were used to investigate the morphology of optimised phase with high crystallinity is favoured, as supported by the reduc­
( )
condition sludge (Fig. 4). The Al sludge generated by fluoride treatment
ed O
ratio. The peak at 75.48 eV due to Al in fluorinated alumina
shows spherical and porous nanoparticles consist of aluminium oxide Al

hydroxide species (black circles in Fig. 4. A) [47]. However, in the (AlOFx) supports fluoride removal on aluminium oxide hydroxide
optimised sludge of Ca2+, Mg2+ /Al and Ca2+, Mg2+, F-/Al superb

6
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

Fig. 6. XPS curve fit of Al2p, O1s, C1s, Ca2p, F1s and Mg1s spectrum of optimised condition Ca2+, Mg2+,F-/Al sludge. The elemental composition data of the sludge
materials is given in Table 4-S.

7
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

Fig. 8. A postulated mechanism for concurrent removal of fluoride and hard­

Fig. 7. FTIR spectra of A: aluminium sludge B: sludge with F- C: sludge with ness in water by EC. Double-headed arrows represent possible bending vibra­
Mg2+ D: sludge with Ca2+ E: sludge with Mg2+ and F- F: sludge with ca2+ and F- tions in proposed species.
G: sludge with Ca2+ and Mg2+ H: simultaneous hardness (Ca2+, Mg2+) and
fluoride sludge. band at 1650 cm− 1 due to H-O-H bending vibrations confirms the
presence of surface-adhered water on ≡ Aly (OH)3x [66]. The bands
substrate [55] (Fig. 6. A). The XPS spectrum of O1 s has resolved into at 3676 cm− 1 and 3448 cm− 1 persist even after the evacuation, can be
three peaks. The peak at 530.74 eV binding energy is ascribed to lattice ascribed to -OH stretching and H-bonded -OH vibrations of ≡ AlOH
oxygen in Al2O3 [56]. The adsorbed water on sludge creates a peak at [67]. Further, the appearance of a band at 1524 cm− 1 suggests the
533.17 eV [54,56] (Fig. 6. B). The spin-orbit doublet of Ca 2p is presence of surface ≡ AlOH bending vibrations. The IR spectrum in
observed at binding energies of 347.14 eV for Ca2p 3/2 and 350.68 eV for Fig. 7. B was measured after interactions of F- with ≡ Aly (OH)3x . A new
Ca2p ½. These peaks are assigned to CaF2, associated with – – Ca-O and
– IR band appeared at 744 cm− 1, and the band at 1020 cm− 1 due to
– Ca-OH sites [57,58]. The difference in binding energy between Ca2p
–
– ≡ Al − O − H deformation vibration shifted to higher wavenumber [68,
3/2 and Ca 2p 1/2 bands is 0.8 eV [59]. This value increases as a function 69], and band at 1375 cm− 1 decreased in intensity while the band at
of O2- and OH- species (Fig. 6. C). The F 1s peak at 685.10 eV indicates 1524 cm− 1 ascribed to the ≡ Al − O − H bending vibrations disappeared
CaF2 in the contaminated sludge [58]. The peculiar behaviour of the due to OH- → F- substitution [29,69]. The emergence of a band at
peak in the vicinity of ~685 eV confirms fluoride removal as CaF2, but 744 cm− 1 can be ascribed to the Al − F vibrations in ≡ AlF(OH) com­
not as AlF3 [59] (Fig. 6. D). plexes [69–71]. When Mg2+ ions were added to the initial sludge, as
The C 1s peak at 289.30 eV confirmed the presence of carbonate in shown in Fig. 7. C, the IR bands around 1000 cm− 1 and 1380 cm− 1
the sludge [60] (Fig. 6. E). In agreement with XRD spectra, the Ca 2p broaden due to weak interactions between Al-sludge and magnesium
peak has spaced spin-orbit components related to ΔC carbonate = 3.54 species forming few ≡ Al − O − Mg − OH. As evidenced from Fig. 7. D,
and interaction of CaF2 species with CaCO3 in the Ca2+, Mg2+, F-/Al compared to Mg2+, the interactions with Ca2+ and Al-sludge show
sludge [61]. Additionally, the strong satellite features further confirm distinct features at 870 cm− 1 and 964 cm− 1 that can be ascribed to
the presence of CaCO3 species in the Ca2+, Mg2+, F-/Al sludge [62]. In Al− O− Ca− OH formation (≡ Al − O − Ca − OH)δ+ species. Further, the
Mg 1s spectrum, the peak at a binding energy of 1304.6 eV implied the appearance of distinct bands at 1792 cm− 1 and 2517 cm− 1 may be due
metal carbonate [63] (Fig. 6. F). Further, O 1s spectrum of Ca2+, Mg2+, to the resonance of the band at 870 cm− 1. When F- ions were introduced
F-/Al sludge indicates carbonate at a binding energy of 533.17 eV with C to the sludge-Mg system, the re-appearance of very small bands around
1s spectra at a binding energy of 289.30 eV [64]. The [O]
atomic per­ 1020 cm− 1 and 1380 cm− 1 indicates the release of some H-bonding
[Mg]
centage for Ca , Mg /Al sludge was 12.10 while 16.28 for Ca2+,
2+ 2+ behaviour of ≡ Al − O − Mg − OH species, probably due to the formation
Mg2+,F-/Al sludge (Table 4-S). The sludge with high oxygen content of ≡ MgF − OH species as evident in Fig. 7. E [72].
implied the presence of oxygen contain debris. Therefore, there is Mg2+ These changes are prominent in the spectrum Fig. 7. E, which is
in MgO, Mg(OH)2.nH2O and MgCO3 forms [65]. obtained with the interaction of F- ions with the sludge-Ca system,
indicating CaF− OH species’ formation together with ≡ CaF. The for­
3.3.4. FTIR characterisation mation of direct ≡ CaF species can be confirmed by observing the
The discrete IR bands are shown in the lattice vibration region, i.e., disappearance of bands in the 900–1100 cm− 1 range of the spectrum
1200–1800 cm− 1 are ascribed to admixture CO2 or H2O adsorption on Fig. 7.D. This has been further supported by the observation of the
Al-derived sludge (Fig. 7. A). The variations in these IR bands provide reduction of H-bonding nature in the spectrum Fig. 7. F where some
important information between surface sites and solute species in­ small distinct features appeared around 3600 cm− 1 freeing OH species
teractions. The IR bands shown in the 900–1100 cm− 1 range is due to indicating the replacement of more OH groups by F-. When Ca2+, Mg2+
the formation of gelatinous ≡ Aly (OH)3x colloids (Fig. 7. A). The for­ interact simultaneously with the sludge (Fig. 7. G), the spectral features
related to Ca2+ interactions are predominant over the Mg2+ interactions;
mation of gelatinous ≡ Aly (OH)3x was observed in experiments. The IR

8
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

when F- is added to the above system formation of more ≡ CaF − OH and Environ. Sci. Technol 53 (2019) 7173–7174, https://doi.org/10.1021/acs.
est.9b02917.
≡ CaF sites compared to ≡ MgF − OH species is evident as the spectra of
[9] L. Birnhack, O. Keller, S.C.N. Tang, N. Fridman-Bishop, O. Lahav, A membrane-
Fig. 7. F and Fig. 7. H are quite similar to each other. The XRD and XPS based recycling process for minimizing environmental effects inflicted by ion-
analyses confirm CaF2, MgF2 and AlF3.3H2O phases in the EC sludge exchange softening applications, Sep. Purif. Technol. 223 (2019) 24–30, https://
resulted by different feed waters. Commitment fluoride and hardness doi.org/10.1016/j.seppur.2019.04.056.
[10] J. Wen, H. Dong, G. Zeng, Application of zeolite in removing salinity/sodicity from
removal mechanism by EC is shown in Fig. 8. wastewater: a review of mechanisms, challenges and opportunities, J. Clean. Prod.
197 (2018) 1435–1446, https://doi.org/10.1016/j.jclepro.2018.06.270.
4. Conclusions [11] M. Sillanpää, M. Shestakova, Electrochemical water treatment methods, 2017.
https://doi.org/10.1016/B978-0-12-811462-9.00002-5.
[12] M.A.C.K. Hansima, M. Makehelwala, K.B.S.N. Jinadasa, Y. Wei, K.G.
The chemical nature of the solution matrix exerts significantly on N. Nanayakkara, A.C. Herath, R. Weerasooriya, Fouling of ion exchange
hardness and fluoride removal efficiency by electrocoagulation. A membranes used in the electrodialysis reversal advanced water treatment: a
review, Chemosphere 263 (2021), 127951, https://doi.org/10.1016/j.
maximum of 83% hardness and 99% fluoride removal efficiencies were chemosphere.2020.127951.
achieved for simulated water within 28 min, consuming a total of [13] S.V. Jadhav, E. Bringas, G.D. Yadav, V.K. Rathod, I. Ortiz, K.V. Marathe, Arsenic
0.69 kW h m− 3 energy. The presence of residual Al3+ in treated water and fluoride contaminated groundwaters: a review of current technologies for
contaminants removal, J. Environ. Manag. 162 (2015) 306–325, https://doi.org/
limits its direct consumption. Further research is necessary to identify 10.1016/j.jenvman.2015.07.020.
aluminium species in water after simultaneous hardness and fluoride [14] D. Clímaco Patrocínio, C.C. Neves Kunrath, M.A. Siqueira Rodrigues, T. Benvenuti,
removal by electrocoagulation. F. Dani Rico Amado, Concentration effect and operational parameters on
electrodialysis reversal efficiency applied for fluoride removal in groundwater,
J. Environ. Chem. Eng. 7 (2019), 103491, https://doi.org/10.1016/j.
CRediT authorship contribution statement jece.2019.103491.
[15] N.B. Goodman, R.J. Taylor, Z. Xie, Y. Gozukara, A. Clements, A feasibility study of
municipal wastewater desalination using electrodialysis reversal to provide
J.U. Halpegama: Methodology, Writing – original draft, K.Y.
recycled water for horticultural irrigation, Desalination 317 (2013) 77–83, https://
Heenkenda: Investigation, Validation, Zhiguo Wu: Visualization, K.G. doi.org/10.1016/j.desal.2013.02.010.
N. Naanaykkara: Resources, R.M.G. Rajapakse: Supervision, A. Ban­ [16] L.J. Banasiak, A.I. Schäfer, Removal of boron, fluoride and nitrate by
dara: Data curation, Ajith C. Herath: Supervision, Validation, Xing electrodialysis in the presence of organic matter, J. Memb. Sci. 334 (2009)
101–109, https://doi.org/10.1016/j.memsci.2009.02.020.
Chen: Conceptualization, Formal analysis, Funding acquisition, Project [17] T. Kim, T.K. Kim, K.D. Zoh, Removal mechanism of heavy metal (Cu, Ni, Zn, and
administration, Resources, Writing – review & editing, Rohan Weer­ Cr) in the presence of cyanide during electrocoagulation using Fe and Al
asooriya: Conceptualization, Funding acquisition, Project administra­ electrodes, J. Water Process Eng. 33 (2020), 101109, https://doi.org/10.1016/j.
jwpe.2019.101109.
tion, Writing – review & editing. [18] D. Mills, New process for electrocoagulation, J./Am. Water Work. Assoc. 92 (2000)
34–43, https://doi.org/10.1002/j.1551-8833.2000.tb08957.x.
Declaration of Competing Interest [19] J.N. Hakizimana, B. Gourich, M. Chafi, Y. Stiriba, C. Vial, P. Drogui, J. Naja,
Electrocoagulation process in water treatment: a review of electrocoagulation
modeling approaches, Desalination 404 (2017) 1–21, https://doi.org/10.1016/j.
The authors declare that they have no known competing financial desal.2016.10.011.
interests or personal relationships that could have appeared to influence [20] K. Padmaja, J. Cherukuri, M. Anji Reddy, A comparative study of the efficiency of
chemical coagulation and electrocoagulation methods in the treatment of
the work reported in this paper. pharmaceutical effluent, J. Water Process Eng. 34 (2020), 101153, https://doi.org/
10.1016/j.jwpe.2020.101153.
Acknowledgements [21] N. Boudjema, N. Drouiche, N. Abdi, H. Grib, H. Lounici, A. Pauss, N. Mameri,
Treatment of Oued El Harrach river water by electrocoagulation noting the effect
of the electric field on microorganisms, J. Taiwan Inst. Chem. Eng. 45 (2014)
The work was supported by the grant received from the National 1564–1570, https://doi.org/10.1016/j.jtice.2013.10.006.
Research Council Sri Lanka (Grant No: NRC-TO-16-015). RW and XC [22] R. Niazmand, M. Jahani, F. Sabbagh, S. Rezania, Optimization of
acknowledge the Program of Distinguished Professor in B & R Countries, electrocoagulation conditions for the purification of table olive debittering
wastewater using response surface methodology, Water 12 (2020) 1687, https://
PR China for abidance in Hefei University of Technology (Grant No. doi.org/10.3390/W12061687.
DL20180052). Reviewers’ comments improved manuscript quality. [23] C.A. Martínez-Huitle, Environment-friendly electrochemical processes, Materials
(2021) 1548, https://doi.org/10.3390/ma14061548.
[24] B. Palahouane, N. Drouiche, S. Aoudj, K. Bensadok, Cost-effective
Appendix A. Supporting information electrocoagulation process for the remediation of fluoride from pretreated
photovoltaic wastewater, J. Ind. Eng. Chem. 22 (2015) 127–131, https://doi.org/
Supplementary data associated with this article can be found in the 10.1016/j.jiec.2014.06.033.
[25] I. Kabdaşlı, I. Arslan-Alaton, T. Ölmez-Hancı, O. Tünay, Electrocoagulation
online version at doi:10.1016/j.jece.2021.106105. applications for industrial wastewaters: a critical review, Environ. Technol. Rev. 1
(2012) 2–45, https://doi.org/10.1080/21622515.2012.715390.
References [26] J. Parthasarathy, N, Buffle, Study of polymeric aluminium (III) hydroxide solutions
for application in waste water treatment. Properties of the polymer and optimal
conditions of preparation, Water Res. 19 (1985) 25–36.
[1] D. Dimkić, Temperature impact on drinking water consumption, Environ. Sci. Proc.
[27] J. Duan, J. Gregory, Coagulation by hydrolysing metal salts, Adv. Colloid Interface
2 (2020) 31, https://doi.org/10.3390/environsciproc2020002031.
Sci. 100–102 (2003) 475–502, https://doi.org/10.1016/S0001-8686(02)00067-2.
[2] S.D. Mengesha, A. Weldetinsae, K. Tesfaye, G. Taye, Organoleptic and palatability
[28] P. Song, Z. Yang, H. Xu, J. Huang, X. Yang, L. Wang, Investigation of influencing
properties of drinking water sources and its health implications in Ethiopia: a
factors and mechanism of antimony and arsenic removal by electrocoagulation
retrospective study during 2010-2016, Environ. Heal. Eng. Manag. 5 (2018)
using Fe-Al electrodes, Ind. Eng. Chem. Res. 53 (2014) 12911–12919, https://doi.
221–229, https://doi.org/10.15171/ehem.2018.30.
org/10.1021/ie501727a.
[3] I.A. Abboud, Geochemistry and quality of groundwater of the Yarmouk basin
[29] S. Aoudj, A. Khelifa, N. Drouiche, R. Belkada, D. Miroud, Simultaneous removal of
aquifer, North Jordan, Environ. Geochem. Health 40 (2017) 1405–1435, https://
chromium(VI) and fluoride by electrocoagulation-electroflotation: application of a
doi.org/10.1007/s10653-017-0064-x.
hybrid Fe-Al anode, Chem. Eng. J. 267 (2015) 153–162, https://doi.org/10.1016/
[4] C. Montero-Ocampo, J.F.M. Villafañe, Effect of dissolved species on the fluoride
j.cej.2014.12.081.
electro-removal from groundwater, ECS Trans. 28 (2010) 57–65.
[30] A. Guzmán, J.L. Nava, O. Coreño, I. Rodríguez, S. Gutiérrez, Arsenic and fluoride
[5] S.L. Zhi, K.Q. Zhang, Hardness removal by a novel electrochemical method,
removal from groundwater by electrocoagulation using a continuous filter-press
Desalination 381 (2016) 8–14, https://doi.org/10.1016/j.desal.2015.12.002.
reactor, Chemosphere 144 (2016) 2113–2120, https://doi.org/10.1016/j.
[6] J.S. Park, J.H. Song, K.H. Yeon, S.H. Moon, Removal of hardness ions from tap
chemosphere.2015.10.108.
water using electromembrane processes, Desalination 202 (2007) 1–8, https://doi.
[31] A.R. Amani-Ghadim, S. Aber, A. Olad, H. Ashassi-Sorkhabi, Optimization of
org/10.1016/j.desal.2005.12.031.
electrocoagulation process for removal of an azo dye using response surface
[7] D.L. Sedlak, The unintended consequences of the reverse osmosis revolution,
methodology and investigation on the occurrence of destructive side reactions,
Environ. Sci. Technol. 53 (2019) 3999–4000, https://doi.org/10.1021/acs.
Chem. Eng. Process. Process. Intensif. 64 (2013) 68–78, https://doi.org/10.1016/j.
est.9b01755.
cep.2012.10.012.
[8] A.M. Naser, M. Shamsudduha, T.F. Clasen, K.M.V. Narayan, Letter to the editor
regarding, ‘the unintended consequences of the reverse osmosis revolution’,

9
J.U. Halpegama et al. Journal of Environmental Chemical Engineering 9 (2021) 106105

[32] M. Ahmadi, F. Vahabzadeh, B. Bonakdarpour, E. Mofarrah, M. Mehranian, [53] Z.Q. Haines, J. Leger, J.M. Gorelli, F. Klug, D.D. Tse, J.S, Li, X-ray diffraction and
Application of the central composite design and response surface methodology to theoretical studies of the high-pressure structures and phase transitions in
the advanced treatment of olive oil processing wastewater using Fenton’s magnesium fluoride, Phys. Rev. B 64 (2001) 1341101–13411010, https://doi.org/
peroxidation, J. Hazard. Mater. 123 (2005) 187–195, https://doi.org/10.1016/j. 10.1103/PhysRevB.64.134110.
jhazmat.2005.03.042. [54] S. Chellam, P. Gamage, T. Tanneru, M.A. Sari, Aluminum electrocoagulation and
[33] U. Tezcan Un, A.S. Koparal, U. Bakir Ogutveren, Fluoride removal from water and electroflotation pretreatment for microfiltration: fouling reduction and
wastewater with a bach cylindrical electrode using electrocoagulation, Chem. Eng. improvements in filtered water quality, in: Desalin. Water Purif. Res. Dev. Progr.,
J. 223 (2013) 110–115, https://doi.org/10.1016/j.cej.2013.02.126. 2014: pp. 1–114. 〈https://www.usbr.gov/research/dwpr/DWPR_Reports.html〉
[34] P. Sarojam, Analysis of wastewater for metals using ICP-OES, 2010. 〈http://www. (viewed on 17-07-2021).
perkinelmer.co.uk/lab-solutions/resources/docs/APP_MetalsinWastewater.pdf〉 [55] R. Ramos, G. Cunge, B. Pelissier, O. Joubert, Cleaning aluminum fluoride coatings
(viewed on 17-07-2021). from plasma reactor walls in SiCl4/Cl2 plasmas, Plasma Sources Sci. Technol. 16
[35] F.S. Stover, R.V. Brill, Statistical quality control applied to ion chromatography (2007) 711–715, https://doi.org/10.1088/0963-0252/16/4/004.
calibrations, J. Chromatogr. A 804 (1998) 37–43, https://doi.org/10.1016/S0021- [56] D. Sun, X. Hong, K. Wu, K.S. Hui, Y. Du, K.N. Hui, Simultaneous removal of
9673(98)00094-6. ammonia and phosphate by electro-oxidation and electrocoagulation using
[36] P. Jackson, Ion chromatography in environmental analysis., 2006. https://doi.org/ RuO2–IrO2/Ti and microscale zero-valent iron composite electrode, Water Res.
10.1002/9780470027318.a0835. 169 (2020), 115239, https://doi.org/10.1016/j.watres.2019.115239.
[37] A. Asfaram, M. Ghaedi, S. Agarwal, I. Tyagi, V.K. Gupta, Removal of basic dye [57] R. Bennewitz, M. Reichling, R.M. Wilson, R.T. Williams, K. Holldack, M. Grunze,
Auramine-O by ZnS:Cu nanoparticles loaded on activated carbon: optimization of E. Matthias, Characterization of Ca aggregates on CaF2 (111)-surfaces by atomic
parameters using response surface methodology with central composite design, force, XPS, and fluorescence microscopy, Nucl. Inst. Methods Phys. Res. B 91
RSC Adv. 5 (2015) 18438–18450, https://doi.org/10.1039/c4ra15637d. (1994) 623–627, https://doi.org/10.1016/0168-583X(94)96298-7.
[38] I.H. Cho, K.D. Zoh, Photocatalytic degradation of azo dye (Reactive Red 120) in [58] A.B. Christie, J. Lee, I. Sutherland, J.M. Walls, An XPS study of ion-induced
TiO2/UV system: optimization and modeling using a response surface compositional changes with group II and group IV compounds, Appl. Surf. Sci. 15
methodology (RSM) based on the central composite design, Dye Pigment 75 (2007) (1983) 224–237, https://doi.org/10.1016/0378-5963(83)90018-1.
533–543, https://doi.org/10.1016/j.dyepig.2006.06.041. [59] C. dos, S. Bezerra, M.E.G. Valerio, Structural and optical study of CaF2
[39] M. Behbahani, M.R.A. Moghaddam, M. Arami, Techno-economical evaluation of nanoparticles produced by a microwave-assisted hydrothermal method, Phys. B
fluoride removal by electrocoagulation process: optimization through response Condens. Matter 501 (2016) 106–112, https://doi.org/10.1016/j.
surface methodology, Desalination 271 (2011) 209–218, https://doi.org/10.1016/ physb.2016.08.025.
j.desal.2010.12.033. [60] D.R. Baer, J.F. Moulder, High resolution XPS spectrum of calcite (CaCO3), Surf. Sci.
[40] M. Kobya, E. Demirbas, F. Ulu, Evaluation of operating parameters with respect to Spectra 2 (1993) 1–7, https://doi.org/10.1116/1.1247719.
charge loading on the removal efficiency of arsenic from potable water by [61] T. Roychowdhury, S. Bahr, P. Dietrich, M. Meyer, A. Thißen, M.R. Linford, Calcite
electrocoagulation, J. Environ. Chem. Eng. 4 (2016) 1484–1494, https://doi.org/ (CaCO3), by near-ambient pressure XPS, Surf. Sci. Spectra 26 (2019), 014025,
10.1016/j.jece.2016.02.016. https://doi.org/10.1116/1.5109266.
[41] R.D. Letterman, S.G. Vanderbrook, Effect of solution chemistry on coagulation with [62] C.S. Gopinath, S.G. Hegde, A.V. Ramaswamy, S. Mahapatra, Photoemission studies
hydrolyzed Al(III). Significance of sulfate ion and pH, Water Res. 17 (1983) of polymorphic CaCO3 materials, Mater. Res. Bull. 37 (2002) 1323–1332, https://
195–204, https://doi.org/10.1016/0043-1354(83)90100-8. doi.org/10.1016/S0025-5408(02)00763-8.
[42] G.P. Hanna Jr., A.J. Rubin, Effect of sulfate and other ions in coagulation with [63] F. Khairallah, A. Glisenti, XPS study of MgO nanopowders obtained by different
aluminum (III), J.-Am. Water Work. Assoc. 62 (1970) 315–321, https://doi.org/ preparation procedures, Surf. Sci. Spectra 13 (2006) 58–71, https://doi.org/
10.1002/j.1551-8833.1970.tb03912.x. 10.1116/11.20060601.
[43] Q. Zuo, X. Chen, W. Li, G. Chen, Combined electrocoagulation and electroflotation [64] M. Santamaria, F. Di Quarto, S. Zanna, P. Marcus, Initial surface film on
for removal of fluoride from drinking water, J. Hazard. Mater. 159 (2008) magnesium metal: a characterization by X-ray photoelectron spectroscopy (XPS)
452–457, https://doi.org/10.1016/j.jhazmat.2008.02.039. and photocurrent spectroscopy (PCS), Electrochim. Acta 53 (2007) 1314–1324,
[44] F. Shen, X. Chen, P. Gao, G. Chen, Electrochemical removal of fluoride ions from https://doi.org/10.1016/j.electacta.2007.03.019.
industrial wastewater, Chem. Eng. Sci. 58 (2003) 987–993, https://doi.org/ [65] A. Hamwi, C. Latouche, V. Marchand, J. Dupuis, R. Benoit, Perfluorofullerenes:
10.1016/S0009-2509(02)00639-5. characterization and structural aspects, J. Phys. Chem. Solids 57 (1996) 991–998.
[45] H. Zhao, B. Zhao, W. Yang, T. Li, Effects of Ca2+ and Mg2+ on defluoridation in [66] E. Balan, M. Lazzeri, G. Morin, F. Mauri, First-principles study of the OH-stretching
the electrocoagulation process, Environ. Sci. Technol. 44 (2010) 9112–9116, modes of gibbsite, Am. Mineral. 91 (2006) 115–119, https://doi.org/10.2138/
https://doi.org/10.1021/es102540t. am.2006.1922.
[46] World Health Organization (WHO), Aluminium in drinking-water: background [67] L. Chen, H.X. Wu, T.J. Wang, Y. Jin, Y. Zhang, X.M. Dou, Granulation of Fe-Al-Ce
document for development of WHO Guidelines for drinking-water quality, 1998. nano-adsorbent for fluoride removal from drinking water by spray coating on sand
〈http://www.who.int/es/〉 (viewed on 17-07-2021). in a fluidized bed, Powder Technol. 193 (2009) 59–64, https://doi.org/10.1016/j.
[47] S. Tanada, M. Kabayama, N. Kawasaki, T. Sakiyama, T. Nakamura, M. Araki, powtec.2009.02.007.
T. Tamura, Removal of phosphate by aluminum oxide hydroxide, J. Colloid [68] C. Morterra, G. Magnacca, A case study: surface chemistry and surface structure of
Interface Sci. 257 (2003) 135–140, https://doi.org/10.1016/S0021-9797(02) catalytic aluminas, as studied by vibrational spectroscopy of adsorbed species,
00008-5. Catal. Today 27 (1996) 497–532, https://doi.org/10.1016/0920-5861(95)00163-
[48] M. Saalfeld, H, Wedde, Refinement of the crystal structure of gibbsite, A1(OH)3, 8.
Zeitschrift Für Krist, Mater 139 (1974) 129–135. [69] R. Liu, W. Gong, H. Lan, Y. Gao, H. Liu, J. Qu, Defluoridation by freshly prepared
[49] H.D. Megaw, The crystal structure of hydrargillite, Аl(OH)3, Z. Kristallogr. Mater. aluminum hydroxides, Chem. Eng. J. 175 (2011) 144–149, https://doi.org/
87 (1934) 185–205. 10.1016/j.cej.2011.09.083.
[50] V. Apalangya, V. Rangari, B. Tiimob, S. Jeelani, T. Samuel, Development of [70] R. Liu, W. Gong, H. Lan, T. Yang, H. Liu, J. Qu, Simultaneous removal of arsenate
antimicrobial water filtration hybrid material from bio source calcium carbonate and fluoride by iron and aluminum binary oxide: competitive adsorption effects,
and silver nanoparticles, Appl. Surf. Sci. 295 (2014) 108–114, https://doi.org/ Sep. Purif. Technol. 92 (2012) 100–105, https://doi.org/10.1016/j.
10.1016/j.apsusc.2014.01.012. seppur.2012.03.020.
[51] E.H.E. Marlene C. Morris, Howard F. McMurdie, D.M.G. Boris Paretzkin, Harry S. [71] U. Gross, S. Rüdiger, E. Kemnitz, K.W. Brzezinka, S. Mukhopadhyay, C. Bailey,
Parker, Winnie Wong-Ng, Standard X-ray diffraction powder patterns section 21 — A. Wander, N. Harrison, Vibrational analysis study of aluminum trifluoride phases,
data for 92 substances, National Bureau of Standards Monograph 25 Washington p. J. Phys. Chem. A 111 (2007) 5813–5819, https://doi.org/10.1021/jp072388r.
148, 1985. [72] S. Wuttke, A. Vimont, J.C. Lavalley, M. Daturi, E. Kemnitz, Infrared investigation of
[52] M. Schreyer, M. Guo, L. Thirunahari, S. Gao, F, Garland, Simultaneous the acid and basic properties of a sol-gel prepared MgF2, J. Phys. Chem. C 114
determination of several crystal structures from powder mixtures: the combination (2010) 5113–5120, https://doi.org/10.1021/jp911584h.
of powder X-ray diffraction, band-target entropy minimization and Rietveld
methods, J. Appl. Crystallogr. 47 (2014) 659–667.

10