Hindawi
Advances in Materials Science and Engineering
Volume 2021, Article ID 6679422, 12 pages
https://doi.org/10.1155/2021/6679422
Research Article
Synthesis and Characterization of Modified Kaolin-Bentonite
Composites for Enhanced Fluoride Removal fromDrinkingWater
Ebenezer Annan,1 Emmanuel Nyankson,1 Benjamin Agyei-Tuffour ,1
Stephen Kofi Armah,1 George Nkrumah-Buandoh,2 Joanna Aba Modupeh Hodasi,2
and Michael Oteng-Peprah3
1Department of Materials Science and Engineering, School of Engineering Sciences, College of Basic and Applied Sciences,
University of Ghana, Accra, Ghana
2Department of Physics, School of Physical and Mathematical Sciences, College of Basic and Applied Sciences,
University of Ghana, Accra, Ghana
3Department of Water and Sanitation, University of Cape Coast, Cape Coast, Ghana
Correspondence should be addressed to Benjamin Agyei-Tuffour; bagyei-tuffour@ug.edu.gh
Received 10 October 2020; Revised 9 December 2020; Accepted 5 January 2021; Published 16 January 2021
Academic Editor: Lingxue Kong
Copyright © 2021 Ebenezer Annan et al. .is is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.
Fluoride-contaminated drinking waters are known to cause severe health hazards such as fluorosis and arthritis. .is paper
presents the encapsulation of iron oxide nanoparticles in kaolin-bentonite composites adsorbents (KBNPs) for the removal of
fluoride from drinking water by adsorption compared with kaolin-bentonite composite (KB). Adsorbents with an average weight
of ∼200mg and ∼7mm diameter (granules) were prepared in the ratio of 10 :10 : 0.1 for kaolinite, bentonite, and magnetite
nanoparticles, respectively..e granules were air-dried and calcined at 750°C and contacted with 2mg/L sodium fluoride solution
at varying time periods. .e adsorbents were characterized using Fourier transform infrared spectroscopy (FTIR), scanning
electron microscopy (SEM) formulation, and Brunauer–Emmett–Teller (BET), whereas the adsorption mechanism and the
kinetics were explained using the Langmuir isotherm, Freundlich models, and pseudo-first-order and pseudo-second-order
models. .e results showed that the BET surface areas for the granules were 10m2/g and 3m2/g for KBNPs and KB, respectively.
.e SEM images for the adsorbents before and after adsorption confirm the plate-like morphology of kaolin and bentonite. .e
FTIR analyses of bentonite (3550 cm−1–4000 cm−1) and kaolin (400–1200 cm−1) correspond to the structural hydroxyl groups and
water molecules in the interlayer space of bentonites and the vibrational modes of SiO4 tetrahedron of kaolin, respectively. .e
KBNPs composites also recorded a fluoride removal efficiency of ∼91% after 120 minutes compared with 64% for KB composites
without Fe3O4 nanoparticles. .e adsorptions of fluoride by the KBNPs and KB granules were found to agree with the Freundlich
isotherm and a pseudo-second-order kinetic model, respectively. .e results clearly show that the impregnation of clays with
magnetite nanoparticles has significant effect in the removal of fluoride, and the implication of the results has been discussed to
show the impact of clay-magnetite nanoparticles composites in the removal of fluoride from contaminated water.
1. Introduction Organization (WHO) recommends a fluoride intake of less
than 1.5mg/L [1]. Small amounts, less than 1.0mg/L, of
Fluoride contamination in drinking water is commonly fluoride in ingested water are usually considered to have
associated with groundwater sources due to demineraliza- beneficial effects on the teeth and skeletal systems and have
tion..e effects of the contamination depend on the fluoride been shown to decrease the rate of occurrence of dental
concentration and the duration of continuous uptake. cavities, particularly among children [2]. However, excess
Fluoride content in drinking water can be beneficial or intake of fluoride (>1.5mg/L) leads to various diseases such
detrimental to human health. .e World Health as fluorosis, arthritis, and brittle bones [3, 4]. Dental
2 Advances in Materials Science and Engineering
fluorosis is a common symptom of high fluoride ingestion with iron oxide nanoparticles for the removal of fluorides
which is classified by water with fluoride concentration from contaminated groundwater.
between 1.5 and 4mg/L. .e mottling of teeth in mild cases Hence, in this research, kaolinite (K) and bentonite (B)
and embrittlement of bones are signs of dental and skeletal composites impregnated with iron oxide nanoparticles were
fluorosis for water ingested with fluoride concentrations investigated for their potential to remove fluoride ions from
between 4 and 10mg/L [5]. groundwater. Magnetite nanoparticles were mixed with the
Fluoride is widely distributed in geological environment clay minerals into granules and subjected to adsorption
[5] and generally released into the groundwater by slow studies. Scanning electron microscopy (SEM), Fourier
dissolution of fluorine-containing rocks [6]. Various min- transform infrared spectroscopy (FTIR), and Bru-
erals, for example, fluorite, biotite, and topaz and their nauer–Emmett–Teller (BET) analyses performed on the
corresponding host rocks such as granite, basalt, syenite, and composites and the adsorption kinetics were explained using
shale, contain fluoride that can be released into the the Langmuir, Freundlich, and pseudo-first-order and
groundwater [7, 8]. .us, groundwater is a major source of pseudo-second-order models.
human intake of fluoride. Besides the natural geological
sources for fluoride enrichment in groundwater, various 2. Materials and Methods
fluorochemical industries such as aluminum smelting are
also contributors to fluoride pollution [9]. All chemicals used were of analytical grade and were ob-
Fluoride concentration levels in groundwater in many tained from Sigma-Aldrich, United Kingdom. Magnetite
African countries have been reported by Kut et al. [10] and nanoparticles, Fe3O4 with particle size range of 50–100 nm,
Ghana is reported to have concentrations greater than 4mg/ bentonite clay (item number 682659), and kaolinite (item
L in the Bongo and Bolgatanga districts in the northern number 03584), were acquired from Sigma-Aldrich.
regions [11, 12]. Indeed, some municipalities within the
above-mentioned districts have recorded fluoride concen-
trations in groundwater to be as high as 12mg/L [12]. It is 2.1. Preparation of Adsorbents. .e adsorbents were pre-
therefore critical to consider the treatment of fluoride- pared in the ratio of 10 :10 : 0.1 by weight for kaolinite,
contaminated water in these locations. .e use of clay bentonite, and magnetite nanoparticles, respectively. .e
materials and coagulation and precipitation, membrane compositions were thoroughly mixed and granules with
filtration, electrochemical processes, and adsorption tech- diameter of 0.5–0.7 cm range and an average weight of
niques are among the methods used in the removal of 250mg were formulated. .e green granules were air-dried
fluoride from potable groundwater sources. Adsorption and subsequently calcined at 750°C. During the calcination,
processes are economical especially in cases where adsor- an initial ramp of 5°C/min was used for 30 minutes before it
bents are made from indigenous materials and a previous was increased to 10°C/min for the remaining period until the
study by Puka [13] reports fluoride adsorption on clays maximum temperature. .e temperature was then main-
which contains oxides of iron, aluminum, and silicon. tained at 750°C for 30 minutes before self-cooling to ambient
Other adsorbents that have been reported for the re- temperature. .e granules shown in Figure 1 are the green
moval of fluoride from water include activated and im- bodies (Figures 1(a) and 1(c)) and the calcined granules
pregnated alumina [14–16], rare Earth oxides [17], activated (Figures 1(b) and 1(d)). .e calcined granules’ brown-red
clay [13, 18], impregnated silica [19], carbonaceous materials coloration is due to the presence of iron.
[20], solid industrial wastes like redmud, spent catalysts, and
fly ash [21, 22], zeolites and related ion exchangers [23], and 2.2. Characterization of Adsorbents. .e calcined granule
biosorbents [24]. Iron oxide, usually in red mud, clays, and weighing ∼200mg was subjected to surface area and pore
other deposits, has been reported as better adsorbent in size analysis with nitrogen gas physisorption at 77K using a
comparison with other adsorbents for removal of fluoride Micromeritics ASAP 2020 (Micrometrics Instrument Cor-
[25]. .is is due to the fact that the high affinity for ion poration, Norcross, USA). .e surface area was calculated
exchange is such situations when fluoride is attached to using BET (Brunauer–Emmett–Teller) method for relative
other support structures. Tomar and Kumar have also in- pressures between (P/P ) � 0.05–0.3. .e pore size was
dicated that a mixture of metallic oxides enhances fluoride 0estimated with the BJH (Barrett–Joyner–Halenda) method
adsorption [26]. Composites chromium (III)–zirconia bi- using the desorption data. .e samples were degassed at
metallic oxide and zirconium iron oxide/clay are also 393K for 2 hours before the nitrogen gas adsorption in-
documented to show good removal rate for fluorides [27]. trusion analysis. .e FTIR analysis was performed with
.e removal of fluoride with kaolinite (∼18.2%) and ben- PerkinElmer Frontier (Perkin Elmer, Ohio, USA) and
tonite (∼46%), respectively, has been reported by Mohapatra TESCAN MIRA3 FEG-Scanning Electron Microscope
et al. [28] and Kau et al. [29] who reported bentonite to (TESCAN, United Kingdom) was used to obtain
possess higher fluoride adsorption capacity than kaolinite. microimages.
.e maximum adsorption efficiencies for iron oxide coated-
kaolinite and iron oxide coated-bentonite clays are known to
be 61% and 80%, respectively, according to Puka [13]. 2.3. Batch Adsorption Experiments. Fluoride solutions were
However, the authors are unware of any published study that prepared by diluting the prepared stock solution (100mg/L
has formed kaolin and bentonite composites impregnated fluoride) to 2mg/L. .e adsorption experiments were
Advances in Materials Science and Engineering 3
(a) (b)
(c) (d)
Figure 1: Adsorbents used in the adsorption experiments. (a) Green KB adsorbent. (b) Calcined KB adsorbent. (c) Green KBNPs adsorbent.
(d) Calcined KBNPs adsorbent.
carried out in 100mL beakers with 2 g (equivalent to 10 temperature of 25± 1°C. All other parameters of the exper-
granules) of adsorbent granules and a liquid volume of iment with regard to batch adsorption are the same, except for
50mL with an initial fluoride concentration of 2mg/L. .e contact times which were 120 minutes and 140 minutes for
beakers were placed on an orbital shaker (STUART, Staf- KB and KBNPs, respectively..ese times are whenmaximum
fordshire, UK) at constant speed of 200 rpm at 25± 1°C. removal efficiency was obtained for the respective granules.
After an adsorption time of 20 minutes, the granules were .e basic Freundlich and Langmuir isotherms modeling was
filtrated with 0.45 μm cellulose acetate membranes and 5ml then undertaken and key parameters were computed.
of TISAB 1 (Sigma-Aldrich, UK) was added to the filtered .e amount of fluoride adsorbed was calculated using
solution to maintain the ionic strength. .e fluoride con- the following equation:
centration was measured using a fluoride electrode H14110 V
(HANNA Instruments, Inc., Woonsocket, USA). .e ad- q � ( C0 − Ct􏼁 , (1)
sorption experiments were repeated for different adsorption m
times and the average removal efficiencies recorded. .e where q is the fluoride adsorbed (mg/g), C0 is the initial
percentage removal of adsorbents and kinetic modeling concentration of fluoride (mg/L), Ct is the concentration of
parameters were determined to explain the adsorbents’ fluoride in solution at given time (mg/L), V is the solution
fluoride removal potential and mechanism. For each set of volume (L), and m is the adsorbent dosage (g). .e per-
granules, the experiment was repeated thrice, and the av- centage removal is given by the following equation:
erage recorded.
Furthermore, the dosage of adsorbent for both granules C0 − Ct
was increased from 2 grams to 10 grams and the removal A% � 􏼠 􏼡 × 100, (2)C
efficiency evaluated. .e initial pH for experiments under- 0
taken was within 5.5–7.5 pH range and equilibrium reached at whereA% is the percentage of fluoride ions adsorbed and the
a pH of 6.5.0. .e experiment was undertaken at constant remaining parameters are defined as in equation (1).
4 Advances in Materials Science and Engineering
3. Results and Discussion After the adsorption experiments, the bands observed
between 500 and 1030 cm−1 are characteristic vibrations of
3.1. Characterization of the Clay Composite Adsorbent metal oxides (Al/Fe). .e peaks observed at 3417–3660 cm−1
Materials. .e BET surface area of the different composite may be assigned to the -OH stretching frequencies of
materials was determined using nitrogen physisorption as gibbsite or comparable metal oxide such as iron oxide [30].
shown in Table 1. .e surface areas for the granules are After fluoride adsorption, intensity of the peaks was found to
10m2/g and 3m2/g for KBNPs and KB, respectively. .e be decreased and shifted slightly to higher wavelength.
magnitude of the surface area gives an idea of possible Minor peaks observed from 1620 to 1652 cm−1, which could
available particles for adsorption. In general, the hysteresis is be assigned to the H–O–H bond stretching, disappeared or
closely related to features of pore structure and underlying reduced drastically after the fluoride adsorption.
adsorption mechanism. .ere are two distinctive features of
the type H3 loop: (i) the adsorption branch resembles a type
II isotherm and (ii) the lower limit of the desorption branch 3.2. Effect of Contact Time. .e effect of contact time is
is normally located at the cavitation-induced p/p0. Loops of crucial in the understanding of the binding processes of
this type are given by nonrigid aggregates of plate-like fluoride ions and the time of equilibrium which strongly
particles. .e type II isotherm results in the shape which is depends on factors such as pore structure of adsorbent,
the result of unrestricted monolayer-multilayer adsorption adsorbent particle size or surface area, and adsorbent
up to high p/p0. .e scanning microscopic images for the concentration. .e results obtained for percentage of fluo-
adsorbents before and after adsorption are shown in ride removal as a function of contact time for the composite
Figures 2(a)–2(d). Both images confirm the plate-like be- granules are presented in Figure 4. Kaolin-bentonite with
havior according to IUPAC description of the hysteresis of nanoparticles (KBNPs) granules was found to have average
the BET. If the knee is sharp, Point B, the beginning of the percentage removal of 87% compared with the average value
middle almost linear section, usually corresponds to the of 62% for kaolin-bentonite without nanoparticles (KB)
completion of monolayer coverage. A more gradual cur- representing 25% increase in fluoride removal when mag-
vature is an indication of a significant amount of overlap of netic nanoparticles were combined with KB. .e BET
monolayer coverage and the onset of multilayer adsorption. hysteresis classification for both granules was described as
.e thickness of the adsorbed multilayer generally appears type-II H3 which implies that the pores are slit-shaped pores,
to increase without limit when (p/p0) � 1. which appear to be more pronounced in the KBNPs. .is
.e FTIR spectra of bentonite (B) and kaolinite clays (K) could be responsible for the increase in adsorption observed
and the composites (KB and KBNPs) are shown in in KBNPs. .e maximum percentage removal for KBNPs
Figures 3(a)–3(c). .e bands between 3550 cm−1 and and KB was 91% at 120 minutes and 68% at 140 minutes
4000 cm−1 corresponded to the structural hydroxyl groups contact time, respectively. Both KB and KBNPs were found
and the water molecules in the interlayer space of the raw to have minimum percentage fluoride removal of 64% and
bentonite. FTIR spectrum of bentonite in the lower region 48%, respectively, at 20 minutes’ contact time..is preempts
shows bands at 1104, 1032, 976, 797, 695, 538, 470, and that adsorption of fluoride in the case of adsorbent under
433 cm−1. .ese bands are due to the vibrational modes of this study is therefore time-dependent. .e sharp increase
SiO4 tetrahedron. from 20 minutes’ contact time to 120 minutes for both
.e band at 3615 cm−1 was due to O-H stretching, and a granules is due to the fact that initially all adsorbent sites
broad band centered on 3404 cm−1 was due to the interlayer were vacant and the solute concentration gradient was high
and intralayer H-bonded O-H stretching. .e band at [31, 32]. Nevertheless, increase in contact time beyond
1637 cm−1 represented the H-O-H bending vibration of 120min did not increase the adsorption efficiency, which
water, while the band at 1634 cm−1 might be attributed to the might be due to the presence of fewer adsorption sites and a
siloxane (-Si-O-Si-) group stretching. .is indicates the lower fluoride ion concentration and/or as a result of
possibility of the hydroxyl linkage between octahedral and competition for adsorption sites between fluoride and hy-
tetrahedral layers. A very sharp and intense band is observed droxyl ions [33, 34]. .e asymptotic value obtained for both
at 1634 cm−1 due to the asymmetric OH stretch (defor- granules confirming equilibrium being established can be
mation mode) of water and is a structural part of the explained by minimal diminishing available sites for ad-
mineral. .e band at 976 cm−1 is due to Al-OH and sorption [35].
662 cm−1. .e phenomenon of adsorption is primarily dependent
.e band observed at around 3620 cm−1 in the case of on particle size and/or active surface area. .is implies that
Figure 4 has been ascribed to the inner hydroxyls (crystalline both powder and granules can be used for adsorption of
hydroxyl), and the bands observed around the other three fluoride ions. Although we adopted the use of granular
characteristic bands (3684, 3645, and 3620 cm−1) are gen- ceramics, our studies compare well with reported studies
erally ascribed to vibrations of the external hydroxyls. .e where powders where employed. Kebede et al. [36] reported
bands in the 1000 cm−1 to 500 cm−1 region are dominated by the use of iron ore (particle size less than 0.075mm) for
functional groups Si-O and Al-OH..e bands 788 cm−1 and fluoride adsorption with a percentage removal of 86% at pH
750 cm−1 are attributed to Si- quartz (as also in 1113 cm−1 Si- 6 and after 120 minutes’ contact time resulting in an ad-
O quartz), whereas 996 cm−1 and 910 cm−1are the vibrations sorptive capacity of ∼1.72mg/g [36]. Also, Puka found
due to OH deformation usually linked to 2Al3+. percentage removal of fluoride for kaolinite and bentonite
Advances in Materials Science and Engineering 5
Table 1: Adsorption isotherms’ parameters and IUPAC classification for adsorbents granules.
Clay composites BET surface area (m2) Pore size, desorption (nm) Isotherm, hysteresis classification
KB 3 12 Type II; H3
KBNPs 10 10 Type II; H3
2 μm 5 μm
(a) (b)
2 μm 5 μm
(c) (d)
Figure 2: .e SEM images of KB and KBNPs: before adsorption ((a) and (c)) and after adsorption ((b) and (d)).
coated with iron oxide to be 61% and 80%, respectively. (reduced to ∼81%) was observed when the adsorbent dosage
Furthermore, the maximum efficiency of fluoride removal was increased to 10 g. Similar trend is observed for KBNPs,
has been documented as 92% at 120 minutes’ contact time where it increases from 91% (2 g of adsorbent dose) to 96%
and equilibrium time for researchers which used powdered (6 g of adsorbent dose) and then decreases to 91% (10 g of
pyrolusite ore [37]. Longer contact time of 180 minutes has adsorbent dosage) from the maximum value.
been reported by [38, 39], while contact time of 720 minutes Two main observed trends are documented: the increase
has been reported by [40] to reach equilibrium adsorption. in fluoride removal efficiency with increasing adsorbent dose
.e effect of contact time on adsorption is not definitive and has been observed in other literature such as Kim et al.’s
is dependent on factors such as competing ions in aqueous work [41] and the decrease in removal efficiency from 6 g to
solution, particle size, pore structure, and nature of adsor- 10 g adsorbent dose..e first trend is in agreement with Kim
bent surface. et al. who also reported an increase in fluoride removal
efficiency from 25% to 98.5% at a fixed initial fluoride
concentration when the adsorbent (pyrophyllite) dosage was
3.3. Effect of AdsorbentDosage. .e effects of adsorbent dose increased [41]. Meenakshi et al. [42] and .akre et al. [40]
on the removal efficiency of the fluoride ions by the granules reported similar observations while experimenting on
were studied for the optimum conditions, and results are fluoride removal studies with kaolinite and bentonite and
presented in Figure 5. .e maximum percentage of fluoride the increment is due to enhancement of the number of active
removal for KB granule was found to be 86% at 25°C and for sites available for adsorption of fluoride ions [40, 42].
a contact time of 140 minutes, while that of KBNPs was 96% .e second observed trend which is the decrease in
at the same temperature but at a contact time of 120minutes. removal efficiency trend observed in the adsorbent dose (6 g
.e dosage of 6 g of KBNPs and KB granules recorded the to 10 g) in relation to fluoride removal efficiency can be
maximum removal efficiency. For example, from the 2 g to described by two main reasons. .ese are (1) better utili-
6 g dosage recorded an increase from 68% to 86% for KB zation of the available active sites at low adsorbent dose in
only. However, a reduction in the fluoride removal efficiency comparison to high adsorbent dose where too many sites are
6 Advances in Materials Science and Engineering
662 cm–1
–1
815 cm–1
794 cm
–1
3643 cm–1 1642 cm
3620 cm–1 720 cm–1
3684 cm–1
1008 cm–1910 cm–1
996 cm–1
1634 cm–1
3615 cm–1 3404 cm–1 3624 cm
–1 1625 cm–1
742 cm–1
662 cm–1
976 cm–1 1010 cm–1 793 
4000 3500 3000 2500 2000 1500 1000 500 4000 3500 3000 2500 2000 1500 1000 500
Wavenumber (cm–1) Wavenumber (cm–1)
Bentonite KB before adsorption
Kaolinite KB aer adsorption
(a) (b)
2981 cm–1
1646 cm–1 1015 cm–1 794 cm–1
(3645–3060 cm–1) 1621 cm–1
1015 cm–1
4000 3500 3000 2500 2000 1500 1000 500
Wavenumber (cm–1)
KBNps before adsorption
KBNps aer adsorption
(c)
Figure 3: FTIR spectra for (a) bentonite and kaolinite clays, (b) KB composites, and (c) KBNPs composites before and after the adsorption
experiments.
available for limited quantity of adsorbate and (2) reduced (equation (4)) kinetic models [45] and the plots are pre-
driving force for adsorption as high adsorbent dose causes sented in Figure 6. .e process kinetics described the solute
low equilibrium fluoride concentration. Similar trends were uptake rate, thereby allowing the estimated residence time
reported by Maiti et al. [43] for laterite and by Goswami and required to achieve a definite extent of fluoride removal
Purkait [44] for pyrophyllite [43]. Goswami and Purkait [44] [46, 47].
attributed the reduction to a decrease in adsorption capacity
due to possible aggregation, overlapping, and overcrowding klog( q − q 􏼁 � log q − 1 t, (3)
of adsorbent resulting in decreased available surface area e t e 2.303
[44].
t 1 t
� 2 + , (4)qt 􏼐k2qe􏼑 qe
3.4. Adsorption Kinetics and Isotherm Models. .e experi-
mental data for the fluoride adsorption was fitted to pseudo- where qe and qt are the amount of fluoride adsorbed (mg/g)
first-order (equation (3)) and pseudo-second-order at equilibrium and any time t, respectively, and k (min−11 )
Transmittance (a.u.)
Transmittance (a.u.)
Transmittance (a.u.)
Advances in Materials Science and Engineering 7
100
80
60
40
20
0
0 40 80 120 160 200 240
Time (mins)
K:B:Nps
K:B
Figure 4: Effect of contact time on removal of fluoride.
100
90
80
70
60
2 4 6 8 10
Adsorbent dossage (g)
KB
KBNps
Figure 5: Effect of adsorbent dose on removal efficiency.
and k (min−12 ) are rate constants of adsorption for pseudo- literature of modified clay minerals [40, 44, 48, 49]. .e
first-order (PFO) and pseudo-second-order (PSO) models, adsorptive capacity values computed for the PSO are also
respectively. From equation (3), a plot of log(qe − qt) versus comparable to the measured experimental adsorptive ca-
time, t, results in a negative slope with gradient being (k1)/ pacity values. .e calculated adsorption capacities at equi-
2.303 and log(qe) intercept on the vertical axis. Also, to librium for both samples for PSO are seen to be closer to the
deduce the parameters of adsorption in equation (4), t/qt is experimental adsorption capacity.
plotted against t with slope (1/qe) and intercept (1/k2q2e) on .e Freundlich and Langmuir models are the two most
the vertical (t) axis. widely used models to describe the mechanism for isotherm
.e computed PFO and PSO parameters for these ex- kinetics processes in adsorption processes. .e Freundlich
periments are shown in Table 2. .e correlation coefficient model is empirical, while the Langmuir equation assumes
(R2) value for the PSO model is higher than the PFO model that the maximum adsorption occurs when the surface is
in both samples. .is indicates that the fluoride adsorption covered by adsorbate in a monolayer [50]. Again, the
process on both granules can be described by chemisorption Langmuir model assumes that the point of valence exists on
or valence forces process kinetic as expressed in other the surface of the adsorbent and that each of these sites is
Removal efficiency (%) % fluoride removal
8 Advances in Materials Science and Engineering
8
Equation y = a + b∗x
–2 Weight No weighting7 Residual sum 31650.5553 121995.28
of squares 5 82
Pearson’s r 0.99943 0.99901
6 Adj. R-square 0.99874 0.99784–4 Value Standard error
Intercept 52.43525 30.54972
t/qt Slope 21.99517 0.235235 Intercept 201.65083 55.19815
–6 Slope 29.08724 0.39031
4
–8 3
Equation y = a + b∗x
–10 Weight No weighting 2Residual sum 16.91975 7.6564
of squares 4
Pearson’s r –0.8588 –0.9251 1
Adj. R-square 0.7113 0.8414
–12 Value Standard errorIntercept –4.5173 0.70634 0
ln (qe – qt) Slope –0.0288 0.00544
Intercept –3.7571 0.47515
Slope –0.0281 0.00366
–14 –1
0 50 100 150 200 0 50 100 150 200 250
Time (mins) Time (mins)
KBNps Linear fit of KBNps KBNps Linear fit of KBNps
KB Linear fit of KB KB Linear fit of KB
(a) (b)
Figure 6: .e kinetic models for KB and KBNPs adsorptions: first order (a) and second order (b).
Table 2: Kinetic models’ parameters for adsorption of fluoride onto granules.
Pseudo-first-order (PFO) Pseudo-second-order (PSO)
Adsorbent Experiment qe (mg/g) K1 (min−1) qe (mg/g) R
2 K2 (g/mg/min) qe (mg/g) R
2
KBNPs 0.045 0.066 0.011 0.77 9.22 0.046 0.99
KB 0.033 0.109 0.098 0.60 4.19 0.034 0.99
capable of adsorbing one molecule. Furthermore, it is as- related to the minimum adsorption capacity and adsorption
sumed that the adsorption sites have equal affinities for intensity, respectively, and Ce is the equilibrium concen-
molecules of adsorbate and that the presence of adsorbed tration (mg/L). In these isotherms, when the Langmuir
molecules at one site will not affect the adsorption of favorability factor (KL factor), 0<KL < 1, the adsorption is
molecules at an adjacent site [2]. Both models are applied to influenced by the Langmuir isotherm. However, when the
understand the mechanism and quantify the adsorption heterogeneity factor (n) is greater than 1 (n> 1), the ad-
capacity of the adsorbent for the removal of fluoride ions sorption is usually influenced by the Freundlich model.
from aqueous solutions. .e obtained experimental results in this study were
.e Langmuir and Freundlich equations are given in fitted with the Langmuir and Freundlich isotherm models
equations (5) and (6), respectively. .e computed param- and are presented in Figures 7(a) and 7(b). From Figure 7(a),
eters for both models are given in Table 3. it can be seen from the experimental results that the
equilibrium adsorption capacity increased with increasing
q K C
Original form: q m L e equilibrium concentration of the fluoride ions when KBe � ,1 + KLCe granules were used as the adsorbent. R
2 values of 0.87 and
(5) 0.86 were obtained when the experimental results were fitted
C 1 1
linearized form: e � + ∗C , with Freundlich and Langmuir adsorption isothermmodels,
q ee KLqm qm respectively. From the R
2 values, the adsorption of fluoride
ions on KB granules can be explained by a combined
Original form (1/n): q � K ∗C , mechanism of Langmuir and Freundlich adsorption modelse F e
1 (6) [51–54]. However, considering heterogeneity factor (n)
linearized form: log qe � logKF + logCe, value of 0.518 which is less than 1, the adsorption of fluoriden ions on KB by the Freundlich isotherm model may not be
where KL and qm are Langmuir constants related to ad- favorable. .e Langmuir favorability factor (RL factor) was
sorption intensity (L/mg) and maximum adsorption ca- estimated to be 0.444. Since this value is less than 1, ad-
pacity (mg/g), respectively; qe is the equilibrium adsorption sorption of fluoride ions by KB is favorable [55]. .erefore,
capacity (mg/g), KF and (1/n) are the Freundlich constants, the adsorption of fluoride ions by KB most likely followed
ln (qe – qt)
t/qt (min/mgL–1) × 103
Advances in Materials Science and Engineering 9
Table 3: Adsorption isotherm models’ parameters for adsorption of fluoride onto granules.
Langmuir isotherm model Freundlich isotherm model
Adsorbent
KL (L/mg) qmax (mg/g) R
2 KF (mg/g) (min) n (mg/g) (L/mg)
1/n R2
KBNPs 90.5 0.047 0.85 0.048 20.0 0.90
KB 0.625 0.035 0.86 0.056 0.518 0.87
0.06
0.0452
0.05
0.0450
0.04
0.03 0.0448
0.02
0.0446
0.01
0.0444
0.6 0.8 1.0 1.2 0.18 0.20 0.22 0.24 0.26
Ce (mg/L) Ce (mg/L)
Experimental Experimental
Freundlich Freundlich
Langmuir Langmuir
(a) (b)
Figure 7: Isotherm model description for adsorbents: (a) KB adsorption and (b) KBNPs adsorption.
the Langmuir model. .e Langmuir and Freundlich con- 3.5. Proposed Mechanism for Fluoride Adsorption.
stants obtained from the linearized Langmuir and Loganathan et al. [57] proposed fivemechanisms for fluoride
Freundlich equations (in equations (5) and (6)) are pre- adsorption. .ey are (1) van der Waals forces (outer-sphere
sented in Table 3. .e experimental results of the KBNps surface complexation), (2) ion exchange (outer-sphere
adsorption of fluoride ions were also fitted with the Lang- surface complexation), (3) hydrogen bonding (H-bonding)
muir and Freundlich isotherms and the results presented in (inner-sphere surface complexation), (4) ligand exchange
Figure 7(b). It can be inferred from Figure 7(b) that the (inner-sphere surface complexation), and (5) chemical
equilibrium adsorption capacity increased with increasing modification of the adsorbent surface [57]..e schematics of
equilibrium concentration of the fluoride ions. However, the the mechanisms are shown in Figure 8. .e first two
equilibrium adsorption capacity almost plateaued between mechanisms are governed by weak physical adsorption and
fluoride equilibrium concentrations of 0.22 and 0.26mg/L. are nonspecific to fluoride, where in the presence of other
By fitting the experimental data with the Langmuir and competing anions they cannot be used to remove fluoride.
Freundlich models, R2 values of 0.86 and 0.90 were, re- .e third and fourth mechanisms are governed by strong
spectively, obtained. .erefore, by incorporating iron oxide chemical adsorption specific to fluoride and capable of re-
nanoparticles into the KB granules, the adsorption mech- moving fluoride selectively in the presence of other anions
anism slightly shifted towards the Freundlich adsorption such as phosphates..e fifthmechanism is governed by both
isotherm model. With a recorded heterogeneity factor value specific and nonspecific adsorption.
greater than 1, it implies that the adsorption of fluoride ions In this study, we propose the mechanism in Figure 8 to
on KBNPs is favorable. .e adsorption of fluoride ions by explain the fluoride adsorption in the use of the granules.
KBNPs therefore followed the Freundlich isotherm model. Clays are known to have mainly silica and alumina, with
.is agrees with several scholarly papers that show that proportional quantities of various metal oxides. .e clay
fluoride adsorption unto clay minerals can best be described minerals were doped with magnetite nanoparticles, which
with Freundlich isotherm model [33, 55, 56]. .e Langmuir will modify both internal and exterior surface properties.
and Freundlich constants obtained from the linearized .e adsorption capacity of fluoride on adsorbents can be
Langmuir and Freundlich equations (in equations (5) and increased by chemical modification of adsorbent surfaces
(6)) are presented in Table 3. [58–60]. .is is particularly of advantage in the case of
qe (mg/g)
qe (mg/g)
10 Advances in Materials Science and Engineering
magnetite nanoparticles (KB and KBNPs). .e maximum
Me OH +2  + F– Me F + H2O percentage removal for KBNPs and KB adsorbents was 91%
after 120 minutes’ contact time and 68% after 140 minutes’
contact time, respectively. Both sets of granules were found
to have minimum percentage removal after 20 minutes’
Me OH + F– Me F + OH– contact time with the recorded fluoride removal efficiencies
of ∼64% and ∼48% for KBNPs and KB, respectively. .e
Figure 8: Adopted mechanism for fluoride adsorption ( -adsorbent; increase in removal efficiency is demonstrated via the ad-
Me, multivalent metallic cation) [57]. dition of magnetite nanoparticles. For optimum conditions,
the maximum percentage of fluoride removal for KB granule
was found to be 86% at 25°C (for adsorbent dosage of 6 g)
adsorbents possessing negative surface and when they are and for a contact time of 140 minutes, while that of KBNPs
impregnated onto the adsorbent to create positive charges was 96% at 25°C but at a contact time of 120 minutes. .e
on the adsorbent surface, they attract fluoride by Cou- adsorption of fluoride by both adsorbents can be best de-
lombic forces and produce adsorption sites capable of scribed with Freundlich isotherm model. .e pseudo-sec-
chemical interaction with fluoride. .ese metallic cations ond-order kinetic model had higher R-squared value and
act as a bridge in adsorbing fluoride onto the adsorbent. best describes the mechanism for both KB and KBNPs
Based on general mechanism proposed in Section 3.5, the adsorbents. .is work clearly shows a feasible sustainable
possible specific reaction mechanism for adsorption of approach for the design of filtration systems for the removal
fluoride onto granular ceramic can be hypothesized as of fluoride from groundwater using mainly locally sourced,
follows: accessible, and cheap components.
2Fe3O4 + H2O⇄ 3Fe
+
2O3 + 2H (R1) (7)
Data Availability
Alternatively, H+
.is article is part of an ongoing project and the data cannot
be shared at the moment. .e entire data will be available
when the project is completed.
Clay-Fe3+ -O-H + F– Clay-Fe3+-F + H2O Conflicts of Interest
.e authors declare that there are no conflicts of interest.
Fe2O3 + H2O⟶ Fe(OH)3 (R3) (9) Acknowledgments
Clay − Fe(OH)3−x + F
− ⟶ Clay − Fe2 − F + XOH (R4) .is project was funded by Cambridge–Africa Partnership
(10) for Research Excellence (CAPREx) Project and ALBORADA
research fund at the University of Ghana, Legon, Ghana.
.e parent material in the production of the granules is Benjamin Agyei-Tuffour acknowledges the support of the
mainly aluminosilicate materials, which is clay. Metal oxides,
3+ 3+ University of Ghana BANGA-Africa Programme.especially multivalent ions such as Al and Fe , are crucial
in the adsorption of fluoride. In reaction (R1), possible
breakdown is envisaged for the magnetite, which modifies References
the surface of the granule. Reaction (R2) explains iron (III) [1] World Health Organization, Guidelines for Drinking-Water
oxide in presence of water, protonation, and fluoride at- Quality, World Health Organization, Geneva, Switzerland,
tachment with multivalent ion. Alternatively, the formation 2004.
of iron hydroxide (in reaction R3) is also paramount to [2] M. Mahramanlioglu, I. Kizilcikli, and I. O. Bicer, “Adsorption
fluoride adsorption. During the formation of iron hydrox- of fluoride from aqueous solution by acid treated spent
ide, the fluoride ions present in the aqueous solution would bleaching earth,” Journal of Fluorine Chemistry, vol. 115, no. 1,
replace the hydroxyl ions in the iron crystal lattice without pp. 41–47, 2002.
disturbing the crystal structure of the compound as in re- [3] N. Chinoy, “Effects of fluoride on physiology of animals and
action (R4). A similar process would occur in the presence of human beings,” Indian Journal of Environmental Toxicology,
aluminum oxide. .is isoelectronic property is possible vol. 1, no. 1, pp. 17–32, 1991.
because F− and OH− ions are of similar size, with compa- [4] P. T. Harrison, “Fluoride in water: a UK perspective,” Journalof Fluorine Chemistry, vol. 126, no. 11-12, pp. 1448–1456,
rable ionic radii [61]. 2005.
[5] X. Fan, D. J. Parker, and M. D. Smith, “Adsorption kinetics of
4. Conclusion and Recommendations fluoride on low cost materials,” Water Research, vol. 37,
no. 20, pp. 4929–4937, 2003.
.is work demonstrates the feasibility of removing fluoride [6] D. Banks, C. Reimann, O. Røyset, H. Skarphagen, and
from drinking water using modified clay composites with O. M. Sæther, “Natural concentrations of major and trace
Advances in Materials Science and Engineering 11
elements in some Norwegian bedrock groundwaters,” Applied [26] V. Tomar and D. Kumar, “A critical study on efficiency of
Geochemistry, vol. 10, no. 1, pp. 1–16, 1995. different materials for fluoride removal from aqueous media,”
[7] W. B. Apambire, D. R. Boyle, and F. A. Michel, “Geo- Chemistry Central Journal, vol. 7, no. 1, p. 51, 2013.
chemistry, genesis, and health implications of fluoriferous [27] X. Dou, Y. Zhang, H. Wang, T. Wang, and Y. Wang, “Per-
groundwaters in the upper regions of Ghana,” Environmental formance of granular zirconium-iron oxide in the removal of
Geology, vol. 33, no. 1, pp. 13–24, 1997. fluoride from drinking water,”Water Research, vol. 45, no. 12,
[8] N. Reddy and K. Prasad, “Pyroclastic fluoride in ground pp. 3571–3578, 2011.
waters in some parts of Tadpatri Taluk, Anantapur district, [28] M. Mohapatra, S. Anand, B. K. Mishra, D. E. Giles, and
Andhra Pradesh,” Indian Journal of Environmental Health, P. Singh, “Review of fluoride removal from drinking water,”
vol. 45, no. 4, pp. 285–288, 2003. Journal of Environmental Management, vol. 91, no. 1,
[9] E. J. Reardon and Y. Wang, “A limestone reactor for fluoride pp. 67–77, 2009.
removal from wastewaters,” Environmental Science & Tech- [29] P. Kau, D. Smith, and P. Binning, “Experimental sorption of
nology, vol. 34, no. 15, pp. 3247–3253, 2000. fluoride by kaolinite and bentonite,” Geoderma, vol. 84,
[10] K. M. K. Kut, A. Sarswat, A. Srivastava, C. U. Pittman Jr., and no. 1–3, pp. 89–108, 1998.
D. Mohan, “A review of fluoride in African groundwater and [30] S. G. Wang, Y. Ma, Y. J. Shi, andW. X. Gong, “Defluoridation
local remediation methods,” Groundwater for Sustainable performance and mechanism of nano-scale aluminum oxide
Development, vol. 2-3, pp. 190–212, 2016. hydroxide in aqueous solution,” Journal of Chemical Tech-
[11] B. G. Survey, “Fluoride in groundwater,” March 2019. nology and Biotechnology: International Research in Process,
[12] M. Sutton, R. Kiersey, L. Farragher, and J. Long, “Health Environmental Clean Technology, vol. 84, no. 7, pp. 1043–
effects of water fluoridation,” Technical Report, Health Re- 1050, 2009.
search Board, Dublin, Ireland, 2015. [31] E. Tchomgui-Kamga, V. Alonzo, C. P. Nanseu-Njiki,
[13] L. R. Puka, Fluoride Adsorption Modelling and the Character- N. Audebrand, E. Ngameni, and A. Darchen, “Preparation
ization of Clays for Defluoridation of Natural Waters, University and characterization of charcoals that contain dispersed
of Johannesburg, Johannesburg, South Africa, 2008. aluminum oxide as adsorbents for removal of fluoride from
[14] N. Das, P. Pattanaik, and R. Das, “Defluoridation of drinking drinking water,” Carbon, vol. 48, no. 2, pp. 333–343, 2010.
water using activated titanium rich bauxite,” Journal of [32] E. Tchomgui-Kamga, E. Ngameni, and A. Darchen, “Evalu-
Colloid and Interface Science, vol. 292, no. 1, pp. 1–10, 2005. ation of removal efficiency of fluoride from aqueous solution
[15] H. Lounici, L. Addour, D. Belhocine et al., “Study of a new using new charcoals that contain calcium compounds,”
technique for fluoride removal from water,” Desalination, Journal of Colloid and Interface Science, vol. 346, no. 2,
vol. 114, no. 3, pp. 241–251, 1997. pp. 494–499, 2010.
[16] D. Mohapatra, D. Mishra, S. P. Mishra, G. R. Chaudhury, and [33] N. Chen, Z. Zhang, C. Feng, N. Sugiura, M. Li, and R. Chen,
R. P. Das, “Use of oxide minerals to abate fluoride from “Fluoride removal from water by granular ceramic adsorp-
water,” Journal of Colloid and Interface Science, vol. 275, no. 2, tion,” Journal of Colloid and Interface Science, vol. 348, no. 2,
pp. 355–359, 2004. pp. 579–584, 2010.
[17] A. Raichur andM. J. Basu, “Adsorption of fluoride ontomixed [34] A. Tembhurkar and S. Dongre, “Studies on fluoride removal
rare earth oxides,” Separation and Purification Technology, using adsorption process,” Journal of Environmental Science
vol. 24, no. 1-2, pp. 121–127, 2001. & Engineering, vol. 48, no. 3, pp. 151–156, 2006.
[18] M. Agarwal, K. Rai, R. Shrivastav, and S. Dass, “Deflour- [35] A. Vinati, B. Mahanty, and S. K. Behera, “Clay and clay
idation of water using amended clay,” Journal of Cleaner minerals for fluoride removal from water: a state-of-the-art
Production, vol. 11, no. 4, pp. 439–444, 2003. review,” Applied Clay Science, vol. 114, pp. 340–348, 2015.
[19] S. A. Wasay, M. J. Haran, and S. Tokunaga, “Adsorption of [36] B. Kebede, A. Beyene, F. Fufa, M. Megersa, and M. Behm,
fluoride, phosphate, and arsenate ions on lanthanum-im- “Experimental evaluation of sorptive removal of fluoride from
pregnated silica gel,” Water Environment Research, vol. 68, drinking water using iron ore,” Applied Water Science, vol. 6,
no. 3, pp. 295–300, 1996. no. 1, pp. 57–65, 2016.
[20] I. Abe, S. Iwasaki, T. Tokimoto, N. Kawasaki, T. Nakamura, [37] N..i Hue and N. Hoang Tung, “Study about doping ion La3+
and S. Tanada, “Adsorption of fluoride ions onto carbona- onto surface of pyrolusite ore for removing simultaneously
ceous materials,” Journal of Colloid and Interface Science, both fluoride and phosphate from wastewater,” Journal of
vol. 275, no. 1, pp. 35–39, 2004. Chemistry, vol. 2017, Article ID 4893835, 11 pages, 2017.
[21] Y. Çengeloglu, E. Kır, and M. Ersöz, “Removal of fluoride [38] D. Mehta, P. Mondal, and S. George, “Utilization of marble
from aqueous solution by using red mud,” Separation and waste powder as a novel adsorbent for removal of fluoride ions
Purification Technology, vol. 28, no. 1, pp. 81–86, 2002. from aqueous solution,” Journal of Environmental Chemical
[22] R. Piekos and S. Paslawska, “Fluoride uptake characteristics of Engineering, vol. 4, no. 1, pp. 932–942, 2016.
fly ash,” Fluoride, vol. 32, pp. 14–19, 1999. [39] W. Xiang, G. Zhang, Y. Zhang, D. Tang, and J. Wang,
[23] N. V. R. M. Rao and C. S. Bhaskaran, “Studies on defluor- “Synthesis and characterization of cotton-like Ca-Al-La
idation of water,” Journal of Fluorine Chemistry, vol. 41, no. 1, composite as an adsorbent for fluoride removal,” Chemical
pp. 17–24, 1988. Engineering Journal, vol. 250, pp. 423–430, 2014.
[24] S. V.Mohan, S. Ramanaiah, B. Rajkumar, and P. Sarma, “Removal [40] D. .akre, S. Rayalu, R. Kawade, S. Meshram, J. Subrt, and
of fluoride from aqueous phase by biosorption onto algal bio- N. Labhsetwar, “Magnesium incorporated bentonite clay for
sorbent Spirogyra sp.-IO2: sorption mechanism elucidation,” defluoridation of drinking water,” Journal of Hazardous
Journal of Hazardous Materials, vol. 141, no. 3, pp. 465–474, 2007. Materials, vol. 180, no. 1–3, pp. 122–130, 2010.
[25] J. Qiao, Z. Cui, Y. Sun, Q. Hu, and X. Guan, “Simultaneous [41] J.-H. Kim, C.-G. Lee, J.-A. Park, J.-K. Kang, N.-C. Choi, and
removal of arsenate and fluoride from water by Al-Fe (hydr) S.-B. Kim, “Use of pyrophyllite clay for fluoride removal from
oxides,” Frontiers of Environmental Science & Engineering, aqueous solution,”Desalination andWater Treatment, vol. 51,
vol. 8, no. 2, pp. 169–179, 2014. no. 16–18, pp. 3408–3416, 2013.
12 Advances in Materials Science and Engineering
[42] S. Meenakshi, C. S. Sundaram, and R. Sukumar, “Enhanced Loaded Natural Zeolite,” Separation Science and Technology,
fluoride sorption by mechanochemically activated kaolinites,” vol. 42, no. 9, pp. 2033–2047, 2007.
Journal of Hazardous Materials, vol. 153, no. 1-2, pp. 164–172, [59] S.-X. Teng, S.-G. Wang, W.-X. Gong, X.-W. Liu, and
2008. B.-Y. Gao, “Removal of fluoride by hydrous manganese oxide-
[43] A. Maiti, J. K. Basu, and S. De, “Chemical treated laterite as coated alumina: performance and mechanism,” Journal of
promising fluoride adsorbent for aqueous system and kinetic Hazardous Materials, vol. 168, no. 2-3, pp. 1004–1011, 2009.
modeling,” Desalination, vol. 265, no. 1–3, pp. 28–36, 2011. [60] Y. Zhou, C. Yu, and Y. Shan, “Adsorption of fluoride from
[44] A. Goswami and M. K. Purkait, “Kinetic and equilibrium aqueous solution on La3+-impregnated cross-linked gelatin,”
study for the fluoride adsorption using pyrophyllite,” Sepa- Separation and Purification Technology, vol. 36, no. 2,
ration Science and Technology, vol. 46, no. 11, pp. 1797–1807, pp. 89–94, 2004.
2011. [61] S. M. Maliyekkal, S. Shukla, L. Philip, and I. M. Nambi,
[45] Y. S. Ho and G. McKay, “Pseudo-second order model for “Enhanced fluoride removal from drinking water by mag-
sorption processes,” Process Biochemistry, vol. 34, no. 5, nesia-amended activated alumina granules,” Chemical Engi-
pp. 451–465, 1999. neering Journal, vol. 140, no. 1–3, pp. 183–192, 2008.
[46] J. Lin and L. Wang, “Comparison between linear and non-
linear forms of pseudo-first-order and pseudo-second-order
adsorption kinetic models for the removal of methylene blue
by activated carbon,” Frontiers of Environmental Science &
Engineering in China, vol. 3, no. 3, pp. 320–324, 2009.
[47] L. Pontoni and M. Fabbricino, “Use of chitosan and chitosan-
derivatives to remove arsenic from aqueous solutions-a mini
review,” Carbohydrate Research, vol. 356, pp. 86–92, 2012.
[48] N. Chen, Z. Zhang, C. Feng, M. Li, D. Zhu, and N. Sugiura,
“Studies on fluoride adsorption of iron-impregnated granular
ceramics from aqueous solution,” Materials Chemistry
Physics, vol. 125, no. 1-2, pp. 293–298, 2011.
[49] V. Masindi, “Application of cryptocrystalline magnesite-
bentonite clay hybrid for defluoridation of underground
water resources: implication for point of use treatment,”
Journal of Water Reuse and Desalination, vol. 7, no. 3,
pp. 338–352, 2017.
[50] H. Freundlich, “Over the adsorption in solution,”=e Journal
of Physical Chemistry, vol. 57, no. 1906, pp. 385–471, 1906.
[51] M. Dessalegne, F. Zewge, and I. Diaz, “Aluminum hydroxide
supported on zeolites for fluoride removal from drinking
water,” Journal of Chemical Technology & Biotechnology,
vol. 92, no. 3, pp. 605–613, 2017.
[52] W. M. Gitari, T. Ngulube, V. Masindi, and J. R. Gumbo,
“Defluoridation of groundwater using Fe3+-modified bentonite
clay: optimization of adsorption conditions,” Desalination and
Water Treatment, vol. 53, no. 6, pp. 1578–1590, 2015.
[53] R. Mudzielwana, M. W. Gitari, and T. A. Msagati, “Char-
acterisation of smectite-rich clay soil: implication for
groundwater defluoridation,” South African Journal of Sci-
ence, vol. 112, no. 11-12, pp. 1–8, 2016.
[54] S. Zhang, Y. Lyu, X. Su, Y. Bian, B. Yu, and Y. Zhang,
“Removal of fluoride ion from groundwater by adsorption on
lanthanum and aluminum loaded clay adsorbent,” Environ-
mental Earth Sciences, vol. 75, no. 5, p. 401, 2016.
[55] M. K. Uddin, S. S. Ahmed, and M. Naushad, “A mini update
on fluoride adsorption from aqueous medium using clay
materials,” Desalination and Water Treatment, vol. 145,
pp. 232–248, 2019.
[56] A. Tor, N. Danaoglu, G. Arslan, and Y. Cengeloglu, “Removal
of fluoride from water by using granular red mud: batch and
column studies,” Journal of Hazardous Materials, vol. 164,
no. 1, pp. 271–278, 2009.
[57] P. Loganathan, S. Vigneswaran, J. Kandasamy, and R. Naidu,
“Defluoridation of drinking water using adsorption pro-
cesses,” Journal of Hazardous Materials, vol. 248-249, pp. 1–
19, 2013.
[58] S. Samatya, Ü. Yüksel, M. Yüksel, and N. Kabay, “Removal of
fluoride from water by metal ions (Al3+, La3+ and ZrO2+)