Efficient fluoride adsorption by mesoporous hierarchical alumina microspheres

Article abstract of DOI:10.1039/c5ra27371d

Mesoporous Hierarchical Alumina Microspheres (HAM) with high efficiency for fluoride removal have been synthesized and characterized. Two types of HAM, differing mostly in crystallinity, surface area and pore size have been obtained. Fluoride adsorption s

Full text of DOI:10.1039/c5ra27371d

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Efficient fluoride adsorption by mesoporous
hierarchical alumina microspheres†
Cite this: RSC Adv., 2016, 6, 42288
Sara Gracia Lanas,a Manuel Valiente,b Eleonora Aneggi,a Alessandro Trovarelli,a
`
a
a
*
Marilena Tolazzi and Andrea Melchior
Mesoporous Hierarchical Alumina Microspheres (HAM) with high efficiency for fluoride removal have been
synthesized and characterized. Two types of HAM, differing mostly in crystallinity, surface area and pore size
have been obtained. Fluoride adsorption studies have been carried out by means of potentiometry and
Isothermal Titration Calorimetry (ITC). The latter method has been applied for the first time to obtain
direct determination of the adsorption enthalpy (DHads) of Fꢀ ion on HAM. The kinetics of the reaction
revealed a two-step process for fluoride adsorption on the adsorbent material. The DHads values
obtained are clearly negative for the different samples investigated. Experimental adsorption data are
well fitted by a Langmuir isotherm. The adsorption constant obtained for type A is 1 order of magnitude
higher than for type B, showing that the synthetic protocol has a remarkable effect on this parameter.
The highest defluoridation capacity reaches 26 mmol gꢀ1 after 1 hour of equilibration for the amorphous
HAM, which is higher than for other adsorbents reported in the literature.
Received 21st December 2015
Accepted 18th April 2016
DOI: 10.1039/c5ra27371d
www.rsc.org/advances
water, since it offers simplicity of plant design and lower energy
and operative costs.16
Introduction
Several materials have been tested for water deuoridation,
for example alumina (Al2O3),7,17 biosorbents,18 clays19 and
composite materials.20 However, most of uoride sorption
methods are unable to reach the concentration limit requested
by international regulations (1.5 mg Lꢀ1)1 thus, the optimization
of materials for water deuoridation is still an actual challenge.
Alumina is recognized as one of the most effective materials
for uoride removal because of the high surface area and
porosity, thermal stability and low solubility in a wide pH
range.21 For this application, several mesoporous alumina-
based materials have been developed by modifying the orig-
inal substrate and showed interesting deuoridation
capacity.6,9,22–24
Hierarchically structured metal oxides have been recently
employed in a wide range of applications from adsorption,12,25,26
drug delivery,27 catalysis,28 to sensors29 and electronic conver-
sion and storage.30 Among them, alumina-based hierarchical
structures received a special attention due to its low cost and
eco-friendly properties.12
Groundwater contamination by uoride is recognized as
a serious problem worldwide, as uoride can be toxic to
humans in the case of chronic exposure to elevated concentra-
tions.1 Minerals containing uoride are used by industries for
several purposes such as glass and ceramic production, semi-
conductor manufacturing, brick and iron works, aluminum
smelters and phosphate fertilizers.1,2 Aside from natural
geological uoride enrichment of groundwater, anthropogenic
contamination can elevate uoride concentrations from ten to
thousands of mg Lꢀ1, much higher than the standard level
recommended by the World Health Organization (1.5 mg Lꢀ1).1
Fluoride absorbed by consumption of contaminated water is
rapidly distributed through the body and integrated into the
teeth and bones. The prolonged assumption of uoride can
cause several diseases, such as skeletal uorosis.1 Many health
issues caused by uoride consumption are irreversible,3 thus it is
necessary to control the uoride concentration in drinking water.
The available water deuoridation methods include:
adsorption,3–12 precipitation,13 electro-coagulation14 or ion-
exchange.15 Among these technologies, adsorption seems to be
most powerful method for uoride removal from contaminated
The scope of this work is to study uoride adsorption by high
surface hierarchical alumina microspheres (HAM) and charac-
terize the adsorption process in terms of loading capacity and
thermodynamics.
HAM have been prepared and characterized by modication
of standard methods, while adsorption studies have been
carried out using a unique experimental approach, which
combines potentiometry and Isothermal Titration Calorimetry
(ITC).
a
`
Dipartimento Politecnico, Laboratori di Tecnologie Chimiche, Universita di Udine, via
del Cotonicio 108, 33100 Udine, Italy. E-mail: andrea.melchior@uniud.it
b
`
Departamento de Quimica, Centre GTS, Universitat Autonoma de Barcelona, Campus
Bellaterra Edicio CN, Barcelona, Spain
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c5ra27371d
42288 | RSC Adv., 2016, 6, 42288–42296
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
In particular, ITC is an extremely powerful and high sensitive
RSC Advances
Textural characteristics of all fresh samples were measured
technique able to directly provide the heat evolved in a chemical according to the Brunauer–Emmett–Teller (BET) equation by
ꢁ
process and then obtain useful information, such as stability nitrogen adsorption isotherms at ꢀ196 C. Pore size distribu-
constants and reaction enthalpy values. This technique has tions were calculated by applying the Barrett–Joyner–Halenda
been hitherto widely used for the study of chemical equilibria in (BJH) method to the isotherms desorption branch. These
solution31–35 and biomolecular interactions,36 while only measurements were carried out in a Tristar 3000 gas adsorption
a limited number of examples of its application in adsorption analyzer (Micromeritics). Prior to the adsorption measure-
studies can be found in the literature.37–44
ments, the samples were degassed at 150 C during 1 h.
ꢁ
Accurate data on the enthalpy associated to the adsorption
The particle size distribution was determined by dynamic
process allow the design of the best conditions both for the light scattering (DLS) on a Horiba LB-550 Particle Size Analyzer.
uptake and for the eventual successive release of a given Around 5 mg of HAM A and B respectively were suspended in 10
chemical species. In previous works, uoride adsorption has mL of TISAB solution. The suspensions were sonicated during
been oen modeled with Langmuir isotherm5,7 and the associ- 15 min previously to the DLS measurement.
ated enthalpy (DHads) has been calculated by the van't Hoff
X-ray powder diffraction (XRD) patterns were determined
equation (using the temperature dependence of the Langmuir using a Philips PW3040/60 PRO instrument (equipped with an
adsorption constant). However, many studies considered the x'celerator detector) operating at 40 kV and 40 mA with Ni-
discrepancies between enthalpy values obtained directly (ITC) ltered Cu Ka radiation. Diffraction proles were collected in
and from van't Hoff equation and evidenced the large uncer- the 2q range of 5–100 ꢁC with a step of 0.02 and counting time of
tainties associated to the latter method.45–47 In this work, ITC is 15 s per step.
applied for the rst time to obtain direct determination of DHads
The pH value of zero point charge (pHzpc) was determined by
for uoride ion adsorption by HAM to provide independent and solid addition method2 using 0.1 M potassium chloride (KCl)
more robust thermodynamic parameters.
solution with 20 mg Lꢀ1 Fꢀ. 20 mL of the prepared solution
(0.1 M KCl with uoride) were added into a series of vessels.
Then, the initial pH (pHi) of the solutions was adjusted in the
range of 3.0–10.0. Aer that, 15 mL of any solution were added
into another vessel with 0.3 mg of the adsorbent material. The
suspensions were then equilibrated for 48 h.
Materials and methods
Synthesis of adsorbent material
Two types of HAM were synthesized using the methodology re-
ported previously48–50 with some modications. All the reagents
were of the analytical grade and used without any further puri-
cation. In a typical synthesis, a solution 0.05 M in aluminum
sulfate and 0.1 M in urea was prepared in 100 mL of distilled water
and then stirred thoroughly during 15 minutes. Then 35 mL of the
prepared solution were added in a Teon-lined digestion vessel of
100 mL capacity and then placed on a turntable for uniform
heating using a microwave digestion system. The microwave
treatment was conducted in a temperature-controlled mode at
180 ꢁC by non-pulse heating time for 3 min (type A) or 20 min (type
B), using 2.45 GHz microwave radiation under power range of
0–1000 W (using maximum power). Aerwards the sample was le
for cooling until ambient temperature was reached. The pH of the
resultant solution was adjusted at ꢂ9 with NaOH. The synthesized
material was then collected by centrifugation and washed with hot
distilled water and ethanol several times. The precipitate was
dried in an oven at 80 ꢁC for 12 h. In order to convert the resulting
g-AlOOH obtained in this synthesis into Al2O3, the powder was
calcinated in a muffle furnace at 600 ꢁC for 2 hours.
Once the equilibration time was reached, the nal pH (pHf)
of the solutions was measured again. The difference between
the nal and the initial pH (DpH ¼ pHi ꢀ pHf) was plotted
versus the pHi. The point of zero charge is given for the point
where the curve intersect with abscissa, at this point DpH ¼ 0.
In order to conrm our results, the experiment was performed
at different ionic strength, 0.1 M and 0.01 M KCl.
Potentiometry
Potentiometric studies for uoride removal were performed
adding 20 mL of 0.5 g Lꢀ1 HAM suspension into a vessel and
then titrating with sodium uoride (NaF) solution 80 mM. 0.5
mL of uoride solution was added to the suspension and once
the equilibrium time was expired (1 h), the uoride concen-
tration in solution was measured using a uoride selective
electrode (Really-Flow™ Solid-State combination uoride-
selective Electrode, Weiss Research) before the consecutive
addition. The suspension was kept under stirring during the
experiment and in total around 9–13 equal-volume additions
were made for each experiment.
Material characterization
The experiments were performed in total ionic strength
The particles were analyzed by transmission electron microscopy adjustment buffer (TISAB), which was prepared following the
(TEM) using a Transmission Electronic Microscope JEM-2011, procedure described.51 At the pH xed by TISAB solution, the
resolution: 0.18 nm at 200 kV, accelerating voltage: 80–200 kV, formation of hydrouoric acid (HF) is negligible and the
equipped with camera: CCD GATAN 895 USC 4000, detector EDS concentration of hydroxyls (OHꢀ), the only other anion that the
Oxford LINCA with energy resolution: 136 eV. Scanning electron electrode responds to, is insignicant.
microscopy (SEM) analyses were performed by MERLIN FE-SEM,
Potentiometric data have been tted both with Langmuir
ZEISS with the detector EDS Oxford LINCA X-Max and EBSD and Freundlich isotherms at different temperatures (298.15 and
analysis Oxford Nordlys II.
318.15 K).5–8,52 The Langmuir isotherm describes an
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 42288–42296 | 42289

View Article Online
RSC Advances
Paper
homogeneous sorption, where all active sites have equal affinity was set to 60 min, which was long enough to let the system to
for the adsorbate, and makes possible to obtain the value of the reach thermal equilibrium.
adsorption constant (b) and loading capacity (Qmax) of the
The values of b and Qmax calculated with Langmuir isotherm
from potentiometric data have been used to calculate [Fꢀ]e for
each titrant addition by numerically solving eqn (1) in the
experimental conditions of the calorimetric titrations. The
DHads (kJ molꢀ1) value has been calculated to best t the
experimental heat for the uoride adsorption according to the
modied isotherm42 shown below:
HAM53,54 according to the equation:
Qmaxb½Fꢀꢃ
½Fꢀꢃ
¼
(1)
1 þ b½Fꢀꢃe
ads
e
where: [Fꢀ]e ¼ equilibrium uoride concentration (M), [Fꢀ]ads
¼
uoride adsorbed at equilibrium (mol gꢀ1), Qmax ¼ maximum
quantity of adsorbed Fꢀ per gram of HAM (mol gꢀ1) and b ¼
Langmuir constant (Mꢀ1).
Qmaxb½Fꢀꢃ
qcum
¼
e DHads
(3)
1 þ b½Fꢀꢃ
e
Freundlich isotherm species the adsorption for heteroge-
neous surfaces characterized by the heterogeneity factor 1/n where: qcum ¼ total heat involved in the reaction per gram of
represented by equation:
adsorbent (kJ gꢀ1). Dilution heat (qdil) was also determined to
correct the total heat measured (qmeas) by the instrument.
Thereby qcum ¼ qmeas ꢀ qdil, represents only the heat involved on
the adsorption reaction. Data analysis has been done within the
MS-Excel using Solverstat and EST utilities.55,56
1
n
½Fꢀꢃ ¼ Kf ½Fꢀꢃ
Kf is the Freundlich constant ((L mmolꢀ1 gꢀ1 1/n
).
Additional batch titrations were performed for sample A
only. In a typical experiment, 20 mL of uoride solutions with
different initial concentrations (0.5–16 mM) were mixed with 10
mg of HAM for a specied contact time (1 h) at 298.15 K. Aer
the equilibrium, uoride concentration was then measured
using the uoride selective electrode.
(2)
ads
e
)
Results and discussion
Characterization of the adsorbent material
The SEM and TEM micrographs obtained for the two samples
(type A and B) do not reveal signicative difference. The SEM
images in Fig. 1A-i and B-i reveal ower-like hollow micro-
spheres with a highly textured surface and smooth inner wall
surface. The microspheres have a diameter of ꢂ0.9 mm, which is
Isothermal titration calorimetry
Fluoride samples and the standard solutions for calorimetric signicantly smaller than those obtained by the published
studies were also prepared in TISAB buffer. ITC experiments methodology (5 mm).48 In Fig. 1A-ii and B-ii TEM images show
have been carried out at 298.15 K using a TAMIII thermostat (TA a magnication of the particles, where the detailed morphology
Instruments) equipped with a nanocalorimeter (1 mL cell of the ower-like microspheres (consisting in a randomly
volume) and an automatic titration syringe. The sample was assembly of nanoplatelets) can be seen. The detailed TEM
stirred continuously at 90 rpm and the reference cell was lled images given in Fig. 1A-iii and B-iii, evidence the hierarchical
with 0.8 mL of distilled water throughout all experiments.
structure of the particles formed by the packing of nanoplatelets
In a typical experiment, about 10 additions of 17.5 ml of 80 about 200 nm length and thin thickness. In Fig. 1A-iv and B-iv
mM uoride solution were added to 0.7 mL of ꢂ0.5 g Lꢀ1 HAM an enlargement of the nanoplatelets evidencing the smooth
suspension. The delay time between two consecutive injections surface is also shown.
Fig. 1 SEM (A-i and B-i) and TEM (A-ii, A-iii, A-iv, B-ii, B-iii and B-iv) images of synthesized HAM of type A and B.
42290 | RSC Adv., 2016, 6, 42288–42296
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
No clear peaks were observed in the diffractogram of type A
(Fig. 4), indicating that the HAM are amorphous while the
microspheres of type B showed some characteristic peaks for
aluminum oxide in accordance with Al2O3 reference standard
diffractogram (JCDPS card no. 10-0425). However, the low
intensity of the signals evidenced that the material is a poorly
crystalline solid.
It has been previously suggested that the crystallinity of
alumina based adsorbents to have an inuence on the uoride
adsorption capacity60 and that low crystallinity of alumina-
based materials seems to be an advantage for their adsorption
capacity.12,60–62 According to these observations, type A, in
principle, would be the most efficient material for a uoride
adsorption.
The point of zero charge for both type A and B is pHzpc ¼ 9.0
at different ionic strength (0.1 M and 0.01 M KCl). At the
working pH ¼ 5.5, the surface of the alumina is positively
charged, therefore the interaction of alumina surface with the
solution containing Fꢀ could be represented by the following
reactions:3,8,51
Fig. 2 Particle size distribution of HAM for type A (blue) and B (green)
determined by DLS (suspension in TISAB).
In Fig. S1 (ESI†) representative SEM and TEM images of the
uncalcined type B HAM are reported. It can be seen that the
microspheres have the same characteristic structure of the
calcined HAM, indicating that the hierarchical morphology is
stable, even aer high temperature treatment.
The particle size distribution of the two samples analyzed by
DLS is represented in Fig. 2. In Fig. 2A it can be seen that the
HAM suspension of type A was obtained with an average
diameter between 0.6 and 1 mm. This value corresponds
approximately with the diameter of the particles, which means
that the particles aggregation is almost absent in our working
conditions. The distribution of type B is shown in Fig. 2B, which
indicates an average size larger than type A.
^AlOH + H+ 4 ^AlOH2
(4)
(5)
+
+
^AlOH2 + Fꢀ 4 ^AlF + H2O
The physical features of samples A and B are summarized in
Table S1.†
The adsorption–desorption isotherms (Fig. 3) can be classi-
ed as Type IV,57 characteristic of mesoporous materials.57 The
H3 type-hysteresis loop58 at high relative pressures (over pres-
sure P/Po of 0.7 in both types A and B) indicates the presence of
mesoporous formed between the assembly of the nano-
platelets.59 The BET surface area of the synthesized material was
found to be 254.1 m2 gꢀ1 and 241.6 m2 gꢀ1 for type A and B
respectively. A mesoporous material normally contains pores
with diameters between 2 and 50 nm. The BJH pore size was
determined to be 12.96 nm with a pore volume of 1.07 cm3 gꢀ1
for sample A and 25.55 nm and 2.32 cm3 gꢀ1 for sample B,
which conrms the mesoporosity of our material. Because of
the characteristic downy surface of the particles, the synthe-
sized alumina has excellent porous properties and seems to
exhibit a great potential for uoride adsorption.
Fig. 4 X-ray diffraction (XRD) patterns of HAM for type A (blue) and B
(green) samples.
Fig. 3 N2 adsorption (green triangles) and desorption (red circles) isotherms of HAM for type A (Fig. 2A) and sample B (Fig. 2B).
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 42288–42296 | 42291

View Article Online
RSC Advances
Paper
Fig. 5 Fluoride adsorption kinetics on HAM (type A, 0.5 g Lꢀ1, pH ¼ 5.5,
T ¼ 298.15 K, [Fꢀ]0 ¼ 16 mM) fitted with mono- (green line) and bi-
(dashed red line) exponential functions.
Fluoride adsorption
In Fig. 5 the [Fꢀ]e/[Fꢀ]0 ratio vs. time (h) plot for an adsorbent
dose of 0.5 g Lꢀ1 and initial concentration of 16 mM of uoride
is shown (type A). Experimental data were tted with both mono
and bi-exponential eqn (6) and (7):
ꢀ
½F ꢃ ¼ a1eꢀk t
(6)
(7)
1
e
ꢀ
½F ꢃ ¼ a1eꢀk t þ a2eꢀk t
1
2
e
Fig. 6 ITC for fluoride adsorption into HAM of type A (blue) and type B
(green). Each peak corresponds to an injection of 17.5 ml of 80 mM
fluoride solution to 0.7 mL of 0.5 g Lꢀ1 HAM suspension.
The tting results obtained are summarized in Table S3† in
the ESI.† The bi-exponential model better ts the experimental
data with respect to the mono-exponential and suggests the
presence of a fast rst reaction followed by a much slower
process (k2 is 3 orders of magnitude lower than k1, Table S3†).
This has also been observed for uoride absorption by MgAl–
CO3 layered double hydroxides.63 The fast adsorption step
occurs during the rst hour, then the rate decreases and, aer 3
h, 75% of the initial uoride is removed.‡
In Fig. 6, the heat evolved for each addition of uoride to
HAM suspension (for type A and B) is shown. The heat signal
reaches the baseline in 60 minutes that is, the heat exchanged
related to the Fꢀ adsorption on the HAM expires in 1 hour.
Although the equilibrium is not yet reached according to the
overall reaction model (Fig. 5), in ITC experiments it seems that
only the rst process produces a detectable thermal effect.
Therefore, calorimetric experiments were analyzed by using
[Fꢀ]e values calculated from potentiometric data with the same
delay time between additions (60 min) as in calorimetry
(Table 1).
Experimental points are better represented by the Langmuir
than the Freundlich model suggesting the presence of a homo-
geneous anion adsorption, forming a monolayer on the posi-
tively charged surface of the microspheres. The Langmuir
constant obtained for type A is signicantly higher than for type
B (ꢂ1 order of magnitude, Table 1) showing that the synthetic
protocol has a remarkable effect on this parameter. While the
difference in the BET surface areas of both samples is not
remarkable, DLS experiments evidenced higher particle aggre-
gation in type B (Fig. 2 and Table S1†). However, the higher
efficiency in uoride removal for type A than type B seems also
to be related to the different crystallinity of the two samples, as
previously found.12,60–62
The maximum loading capacity of the material (Qmax) results
higher for sample A than sample B, being 0.026 and 0.020 moles
of uoride per gram of material, respectively. These values are
higher than those found for other alumina-based adsorbents
which are reported in Table 22,3,5–7,9–12,64–67 and for other adsor-
bent materials reported in Table S4 in the ESI.†20,68–72
Potentiometric data for uoride adsorption by HAM (type A
and B) were tted both with the Langmuir and Freundlich
isotherms (Fig. 7) and the resulting parameters are reported in
Table 1. Potentiometric batch titrations carried out for type A
sample conrmed the values obtained (Fig. S2 and Table S2 in
the ESI†).
The high efficiency of the HAM could be associated to the
structure of the hierarchical material. As suggested,70 the mes-
oporosity and the elevated surface area gives to the material
a high number of active sites exposed to the surface. In addi-
tion, the particularity of the hierarchical materials avoids the
aggregationꢀof the nanoplatelets, exposing their entire surface
to adsorb F .70
‡ ai ¼ initial concentration for each exponential (M), ki ¼ constant rates (hꢀ1), t ¼
time (h).
42292 | RSC Adv., 2016, 6, 42288–42296
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
Table 1 Langmuir and Freundlich isotherm parameters for fluoride adsorption by HAM of type A and B with different experimental conditions
Langmuir equation
Qmax (mol gꢀ1 b (Mꢀ1
Freundlich equation
Kf (L mmolꢀ1 gꢀ1
Delay
Type Temperature (K) time (h)
)
)
log b
R2
n
)
log Kf
R2
A
A
A
B
318.15
298.15
298.15
298.15
1
1
12
1
0.039 ꢄ 0.002 4165 ꢄ 794 3.62 ꢄ 0.19 0.961
0.026 ꢄ 0.001 4563 ꢄ 486 3.66 ꢄ 0.11 0.979
0.055 ꢄ 0.001 4820 ꢄ 394 3.68 ꢄ 0.08 0.999
3.1 ꢄ 0.5 31 ꢄ 1
5.9 ꢄ 0.5 20.2 ꢄ 0.3
4.4 ꢄ 0.6 42 ꢄ 1
2.3 ꢄ 0.1 7.6 ꢄ 0.3
1.49 ꢄ 0.03 0.721
1.31 ꢄ 0.01 0.871
1.62 ꢄ 0.02 0.847
0.88 ꢄ 0.04 0.895
0.020 ꢄ 0.001
598 ꢄ 100 2.78 ꢄ 0.17 0.967
Similar potentiometric experiments with 12 h delay between less negative than the value obtained for type A (DHads¼ꢀ17.7 ꢄ
additions have been also performed (type A) to study the 0.6 kJ molꢀ1). The DHads values obtained previously by van't
adsorption of uoride at long times (Fig. S3†). The resulting Hoff equation are in most cases positive2,5,7 and spread over
adsorption constant is quite similar to that obtained with a 1 h a wide range (6.8–36.6 kJ molꢀ1). Nevertheless, also negative
delay, while the loading capacity is ꢂ50% larger.
DHads have been reported for other alumina-based adsor-
The experimental and calculated qcum values for type A and B bents,3,64 natroalunite microtubes,11 nanosized iron oxides68 or
are shown in Fig. 8. In type B, aer the third injection the heat UiO-66-NH2 (values ranging from ꢀ0.8 to ꢀ94.7 kJ molꢀ1
,
72
ow measured for the remaining injections corresponds to the Table 2 and S4†).
dilution heat indicating that the adsorption process is nished,
To compare our method for the determination of DHads with
while for type A 6 injections are needed. It is also evidenced that the standard van't Hoff interpolation, normally used in previous
the total heat evolved in uoride adsorption in type A is works,2,3,5,7,11,20,64,68,72 the adsorption parameters were deter-
signicantly higher than in type B, which conrms that type A is mined for HAM (type A) also at 318 K (Table 1).
a more efficient adsorbent for uoride removal than type B.
The DHads value calculated by the van't Hoff equation (DHads
The process is clearly exothermic for both type A and B ꢂ ꢀ4 kJ molꢀ1), is less negative than that obtained by ITC
samples (Table 2) and, as can be observed, the adsorption measurements. However, when DHads is calculated with the
enthalpy obtained for type B (DHads ¼ ꢀ13.3 ꢄ 0.9 kJ molꢀ1) is van't Hoff equation, the propagated uncertainty on the DHads
values can be large. Thus, an ideal DHads determined by van't
Hoff equation should be calculated using very precise values of
equilibrium constants and in a large range of temperature
where, on the other hand, the heat capacity is not rigorously
constant. Therefore, the direct determination of the heat
exchanged by ITC seems to be the sole way to obtain reliable
heats of adsorption and calculate DHads values.
It can be noticed that the TDSads terms (obtained from the
DGads calculated with the b values at 298.15 K in Table 1 and the
DHads values in Table 2) are nearly the same for the two mate-
rials (+3.2 and +2.6 kJ molꢀ1 for HAM A and B respectively). This
positive entropy change is in agreement with previous results
adsorption of metal cations39,43 and reects the increase
disorder due to charge neutralization and desolvation of the
uoride anion.
Conclusions
In this work, two new HAM adsorbents have been synthesized
and used for uoride removal. SEM, TEM, BET and XRD
diffraction studies reveal a high porous structure of amorphous
alumina. The two types show similar physical characteristics
with small differences in pore size and specic surface area. The
main difference between the two samples is the presence of
crystalline phase, smaller pore size and higher surface area in
the one prepared with a shorter microwave treatment (type A).
The kinetic study reveals a bi-exponential model indicating
a two-step process for uoride adsorption.
Fig. 7 Experimental fluoride adsorption isotherms presenting the
Fluoride adsorption experiments evidenced that the amor-
experimental data fitted with Langmuir (green line) and Freundlich
(dashed red line) models with 1 h between injections for type A and B. phous HAM (type A) have a superior affinity for Fꢀ ion than type
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 42288–42296 | 42293

View Article Online
RSC Advances
Paper
42294 | RSC Adv., 2016, 6, 42288–42296
This journal is © The Royal Society of Chemistry 2016

View Article Online
Paper
RSC Advances
¨¨
2 E. Kumar, A. Bhatnagar, U. Kumar and M. Sillanpaa, J.
Hazard. Mater., 2011, 186, 1042–1049.
3 M. G. Sujana, G. Soma, N. Vasumathi and S. Anand, J.
Fluorine Chem., 2009, 130, 749–754.
4 B. D. Turner, P. Binning and S. L. S. Stipp, Environ. Sci.
Technol., 2005, 39, 9561–9568.
5 X. Zhao, J. Wang, F. Wu, T. Wang, Y. Cai, Y. Shi and G. Jiang,
J. Hazard. Mater., 2010, 173, 102–109.
6 W. Li, C. Y. Cao, L. Y. Wu, M. F. Ge and W. G. Song, J. Hazard.
Mater., 2011, 198, 143–150.
7 W. X. Gong, J. H. Qu, R. P. Liu and H. C. Lan, Chem. Eng. J.,
2012, 189–190, 126–133.
8 L. Chai, Y. Wang, N. Zhao, W. Yang and X. You, Water Res.,
2013, 47, 4040–4049.
9 C. Yang, L. Gao, Y. Wang, X. Tian and S. Komarneni,
Microporous Mesoporous Mater., 2014, 197, 156–163.
10 M. Vithanage, A. U. Rajapaksha, M. S. Bootharaju and
T. Pradeep, Colloids Surf., A, 2014, 462, 124–130.
11 B.-S. Zhu, Y. Jia, Z. Jin, B. Sun, T. Luo, X.-Y. Yu, L.-T. Kong,
X.-J. Huang and J.-H. Liu, Chem. Eng. J., 2015, 271, 240–251.
12 D. Kang, S. Tong, X. Yu and M. Ge, RSC Adv., 2015, 5, 19159–
19165.
13 D. Thakre, S. Jagtap, A. Bansiwal, N. Labhsetwar and
S. Rayalu, J. Fluorine Chem., 2010, 131, 373–377.
14 M. Bennajah, B. Gourich, A. H. Essadki, C. Vial and
H. Delmas, Chem. Eng. J., 2009, 148, 122–131.
15 N. Chubar, J. Colloid Interface Sci., 2011, 357, 198–209.
16 L. H. Velazquez-Jimenez, E. Vences-Alvarez, J. L. Flores-
Fig. 8 Experimental calorimetric data fitted with calorimetric Lang-
muir isotherm (line) for type A (blue) and B (green) HAM.
˜
Arciniega, H. Flores-Zuniga and J. R. Rangel-Mendez, Sep.
Purif. Technol., 2015, 150, 292–307.
B, being the Langmuir constant about 1 order of magnitude
higher. Adsorption experiments have shown that synthesized
materials have a high efficiency towards deuoridation, with
a maximum loading capacity of 0.026 moles of uoride per
gram of material aer 1 h contact time, which is signicantly
better than most other materials reported in the literature.
Therefore our results show that the synthesized HAM have high
deuoridation capacity, which makes them a potential candi-
date for water treatment.
The combined potentiometric-ITC method has been used to
calculate the adsorption parameters and, for the rst time, to
obtain the DHads value by direct measurement of the heat
associated to the process. Contrarily to the data presented in
most previous works, clearly negative enthalpy values for uo-
ride adsorption are obtained here. Therefore, ITC can be
a useful tool for an accurate and relatively fast screening of
sorbent materials.
17 J. Chilton, E. Dahi, M. Lennon and P. Jackson, Fluoride in
Drinking-water, World Health Organization (WHO), Geneva,
2006.
18 V. Tomar, S. Prasad and D. Kumar, Microchem. J., 2014, 112,
97–103.
19 N. Hamdi and E. Srasra, Desalination, 2007, 206, 238–244.
20 S. Wen, Y. Wang and S. Dong, RSC Adv., 2015, 5, 89594–
89602.
21 Y. Kim, C. Kim, I. Choi, S. Rengaraj and J. Yi, Environ. Sci.
Technol., 2004, 38, 924–931.
22 Y. Q. He, N. N. Zhang and X. D. Wang, Chin. Chem. Lett.,
2011, 22, 859–862.
23 J. Han, Z. Du, W. Zou, H. Li and C. Zhang, Chem. Eng. J.,
2015, 262, 571–578.
24 M. Nazari and R. Halladj, J. Taiwan Inst. Chem. Eng., 2014, 45,
2518–2525.
25 J. Zhang, S. Liu, J. Lin, H. Song, J. Luo, E. M. Elssfah,
E. Ammar, Y. Huang, X. Ding, J. Gao, S. Qi and C. Tang, J.
Phys. Chem. B, 2006, 110, 14249–14252.
Acknowledgements
26 W. Cai, J. Yu and M. Jaroniec, J. Mater. Chem., 2010, 20, 4587.
27 W. Wei, G. Hu, D. Yu, T. Mcleish, Z. Su and Z. Shen, J. Am.
Chem. Soc., 2008, 130, 15808–15810.
The present work has been partially supported by the Spanish
MINECO grant CTM2012-30970.
28 H. Guo, Y. He, Y. Wang, L. Liu, X. Yang, S. Wang, Z. Huang
and Q. Wei, J. Mater. Chem. A, 2013, 1, 7494–7499.
29 C. Peng, J. Guo, W. Yang, C. Shi, M. Liu, Y. Zheng, J. Xu,
P. Chen, T. Huang and Y. Yang, J. Alloys Compd., 2016,
654, 371–378.
References
1 Environmental Health Criteria, Fluorides, World Health
Organization (WHO), Geneva, 2002.
This journal is © The Royal Society of Chemistry 2016
RSC Adv., 2016, 6, 42288–42296 | 42295

View Article Online
RSC Advances
Paper
30 P. Yu, X. Zhang, D. Wang, L. Wang and Y. Ma, Cryst. Growth 52 S. G. Wang, Y. Ma, Y. J. Shi and W. X. Gong, J. Chem. Technol.
Des., 2009, 2008, 528–533. Biotechnol., 2009, 84, 1043–1050.
31 L. Cavallo, S. Del Piero, J.-M. Ducere, R. Fedele, A. Melchior, 53 G. Limousin, J. P. Gaudet, L. Charlet, S. Szenknect,
´ ´
`
G. Morini, F. Piemontesi and M. Tolazzi, J. Phys. Chem. C,
2007, 111, 4412–4419.
V. Barthes and M. Krimissa, Appl. Geochem., 2007, 22, 249–
275.
32 F. Endrizzi, A. Melchior, M. Tolazzi and L. Rao, Dalton 54 K. Y. Foo and B. H. Hameed, Chem. Eng. J., 2010, 156, 2–10.
Trans., 2015, 44, 13835–13844.
55 C. Comuzzi, P. Polese, A. Melchior, R. Portanova and
33 A. Melchior, E. Peralta, M. Valiente, C. Tavagnacco,
M. Tolazzi, Talanta, 2003, 59, 67–80.
F. Endrizzi and M. Tolazzi, Dalton Trans., 2013, 42, 6074– 56 S. del Piero, A. Melchior, P. Polese, R. Portanova and
6082. M. Tolazzi, Ann. Chim., 2006, 96, 29–49.
34 P. Di Bernardo, P. L. Zanonato, A. Melchior, R. Portanova, 57 S. Lowell, J. E. Shields, M. A. Thomas and M. Thommes,
M. Tolazzi, G. R. Choppin and Z. Wang, Inorg. Chem.,
2008, 47, 1155–1164.
35 P. Di Bernardo, P. L. Zanonato, F. Benetollo, A. Melchior,
M. Tolazzi and L. Rao, Inorg. Chem., 2012, 51, 9045–9055.
36 J. E. Ladbury and M. L. Doyle, Biocalorimetry 2: Applications
Characterization of Porous Solids and Powders: Surface Area,
Pore Size and Density, Kluwer Academic Publishers,
Dordrecht, 2004.
58 K. S. W. Sing and R. T. Williams, Adsorpt. Sci. Technol., 2004,
22, 773–782.
of Calorimetry in the Biological Sciences, Wiley, Chichester, 59 W. Cai, J. Yu, S. Gu and M. Jaroniec, Cryst. Growth Des., 2010,
2004. 10, 3977–3982.
37 M. E. Strayer, J. M. Binz, M. Tanase, S. M. Kamali Shahri, 60 J. Du, D. A. Sabatini and E. C. Butler, Chemosphere, 2014, 101,
R. Sharma, R. M. Rioux and T. E. Mallouk, J. Am. Chem.
Soc., 2014, 136, 5687–5696.
38 L. N. H. Arakaki, M. G. da Fonseca, E. C. da Silva Filho,
21–27.
61 M. B. McBride and L. G. Wessellnk, Environ. Sci. Technol.,
1988, 22, 703–708.
A. P. D. M. Alves, K. S. de Sousa and A. L. P. Silva, 62 Y. H. Li, S. Wang, A. Cao, D. Zhao, X. Zhang, C. Xu, Z. Luan,
Thermochim. Acta, 2006, 450, 12–15.
39 L. N. H. Arakaki, V. L. S. A. Filha, A. F. S. Germano,
D. Ruan, J. Liang, D. Wu and B. Wei, Chem. Phys. Lett., 2001,
350, 412–416.
S. S. G. Santos, M. G. Fonseca, K. S. Sousa, 63 L. Lv, J. He, M. Wei, D. Evans and Z. Zhou, Water Res., 2007,
´
J. G. P. Espınola and T. Arakaki, Thermochim. Acta, 2013,
41, 1534–1542.
556, 34–40.
64 S. Jagtap, M. K. N. Yenkie, N. Labhsetwar and S. Rayalu,
Microporous Mesoporous Mater., 2011, 142, 454–463.
65 H. Cai, G. Chen, C. Peng, L. Xu, X. Zhu, Z. Zhang, Y. Dong,
G. Shang, F. Ke, H. Gao and X. Wan, RSC Adv., 2015, 5,
29266–29275.
66 H. M. Cai, G. J. Chen, C. Y. Peng, Z. Z. Zhang, Y. Y. Dong,
G. Z. Shang, X. H. Zhu, H. J. Gao and X. C. Wan, Appl. Surf.
Sci., 2015, 328, 34–44.
40 N. Welsch, Y. Lu, J. Dzubiella and M. Ballauff, Polymer, 2013,
54, 2835–2849.
41 P. R. S. Braga, A. A. Costa, J. L. de Macedo, G. F. Ghesti,
M. P. de Souza, J. A. Dias and S. C. L. Dias, Microporous
Mesoporous Mater., 2011, 139, 74–80.
42 C. W. Chronister and R. S. Drago, J. Am. Chem. Soc., 1993,
115, 4793–4798.
43 R. S. Drago, S. C. Dias, M. Torrealba and L. de Lima, J. Am. 67 J. Zhu, X. Lin, P. Wu, Q. Zhou and X. Luo, Appl. Surf. Sci.,
Chem. Soc., 1997, 119, 4444–4452. 2015, 357, 91–100.
44 W. Lin, J. Walter, A. Burger, H. Maid, A. Hirsch, W. Peukert 68 M. Mohapatra, K. Rout, P. Singh, S. Anand, S. Layek,
and D. Segets, Chem. Mater., 2015, 27, 358–369.
45 Y. Liu and J. M. Sturtevant, Protein Sci., 1995, 4, 2559–2561.
H. C. Verma and B. K. Mishra, J. Hazard. Mater., 2011, 186,
1751–1757.
46 Y. Liu and J. M. Sturtevant, Biophys. Chem., 1997, 64, 121– 69 N. A. Medellin-Castillo, R. Leyva-Ramos, E. Padilla-Ortega,
126.
R. O. Perez, J. V. Flores-Cano and M. S. Berber-Mendoza, J.
Ind. Eng. Chem., 2014, 20, 4014–4021.
47 L. S. Mizoue and J. Tellinghuisen, Biophys. Chem., 2004, 110,
15–24.
70 Z. Jin, Y. Jia, T. Luo, L. T. Kong, B. Sun, W. Shen, F. L. Meng
and J. H. Liu, Appl. Surf. Sci., 2015, 357, 1080–1088.
71 S. Dong and Y. Wang, Water Res., 2016, 88, 852–860.
48 W. Cai, J. Yu and S. Mann, Microporous Mesoporous Mater.,
2009, 122, 42–47.
49 X. Wu, B. Zhang and Z. Hu, Mater. Lett., 2012, 73, 169–171. 72 K. Y. A. Lin, Y. T. Liu and S. Y. Chen, J. Colloid Interface Sci.,
50 X. Wu, B. Zhang and Z. Hu, Mater. Lett., 2013, 91, 249–251.
51 S. S. Tripathy, J. L. Bersillon and K. Gopal, Sep. Purif.
Technol., 2006, 50, 310–317.
2016, 461, 79–87.
42296 | RSC Adv., 2016, 6, 42288–42296
This journal is © The Royal Society of Chemistry 2016