Chemosphere 62 (2006) 80–88

www.elsevier.com/locate/chemosphere

Evaluation of boron removal from water

by hydrotalcite-like compounds
Odair Pastor Ferreira a,*, Sandra Gomes de Moraes b, Nelson Durán b,c
,
Lorena Cornejo d,e, Oswaldo Luiz Alves a
a
LQES—Laboratório de Quı́mica do Estado Sólido, Instituto de Quı́mica, Universidade Estadual de Campinas, UNICAMP,
Cidade Universitária Zeferino Vaz, s/n, C.P. 6154, CEP 13083-970, Campinas, SP, Brazil
b
Laboratório de Quı́mica Biológica, Instituto de Quı́mica, Universidade Estadual de Campinas, UNICAMP,
C.P. 6154, CEP 13083-970, Campinas, SP, Brazil
c
Núcleo de Ciências Ambientais, Universidade de Mogi das Cruzes, Mogi das Cruzes, SP, Brazil
d
Área de Quı́mica Analı́tica y Ambiental, Departamento de Quı́mica, Facultad de Ciencias, Universidad de Tarapacá,
Arica, Casilla 7D, Chile
e
Centro de Investigaciones del Hombre en el Desierto, Arica, Chile

Received 2 September 2004; received in revised form 2 February 2005; accepted 3 April 2005
Available online 3 June 2005

Abstract

Layered double hydroxides (LDHs) or hydrotalcite (HT)-like compounds with different kinds of metal ions (Mg–Al
and Mg–Fe) in the brucite-like sheets were prepared and their adsorption properties were studied in the boron removal
from aqueous solution under laboratory conditions. The hydrotalcites were synthesized by the coprecipitation method
and were characterized by chemical analyses, X-ray diffraction (XRD), Fourier Transform infrared spectroscopy
(FTIR) and specific surface area measurements (BET). The affinity of these materials with a mixture of B(OH)3 and
BðOHÞ
4 was studied as a function of contact time, initial pH of the solutions, HT quantity and B concentration
(adsorption isotherms). It was found that 120 min is enough time for the equilibrium state to be reached in boron
adsorption. Boron removal was independent of the initial pH of the solutions because of the high buffering capacity
of the LDHs. On the other hand, the adsorption capacity increases with increasing the adsorbent quantity. The adsorp-
tion isotherms, described by the Langmuir model, are of L-type, suggesting that BðOHÞ
4 is adsorbed preferentially on
HT-like materials. Besides, Mg–Al hydrotalcites showed higher adsorption capacity than Mg–Fe. We proposed that in
Mg–Al hydrotalcites, the boron removal occurs by both adsorption on external surface and ion exchange, whereas for
Mg–Fe it occurs only by surface adsorption. After treatment of a solution containing 5.2 mgB l
1 with Mg–Al hydro-
talcites the final boron concentration reached the recommended limit by WHO for drinking water (0.5 mg l
1).
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: Hydrotalcite (HT)-like compounds; Layered double hydroxides; Boron removal; Water treatment

1. Introduction
*
Corresponding author. Tel.: +55 19 3788 3394; fax: +55 19
3788 3023. Boron is naturally found in groundwater and its
E-mail address: odair@iqm.unicamp.br (O.P. Ferreira). concentration depends on the surrounding geological

0045-6535/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2005.04.009
 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88
characteristics. Its presence in surface water occurs
frequently as a consequence of the discharge of treated
sewage effluents. This element is generally found in nat-
ural water as boric acid, B(OH)3, and/or borate,
BðOHÞ
4 (Peak et al., 2003). Boron is an essential micro-
nutrient for plants, although it may be toxic at higher
levels. The range in which it is converted from a nutri-
ent to a contaminant is quite narrow. The boron
concentration recommended by the World Heath Orga-
nization (WHO) for drinking water is 0.5 mg l
1 (WHO,
1998), but some places in the world present a much
higher level of this element in water. Borate toxicity is
of extreme concern for agriculture in arid regions and
in poorly drained soils of high salinity. Therefore, boron
removal from water is of great importance in several
situations.
There is no universal method for the removal of bor-
on from water. One or more methods may be applied
according to the boron concentration in the medium.
The main removal processes studied so far involve some
of the following processes: precipitation–coagulation,
adsorption, ion exchange, solvent extraction after boron
complexation, ultrafiltration and the use of boron selec-
tive resins, particularly Amberlite XE243 (Okay et al.,
1985; Goldberg et al., 1993; Keren et al., 1994; Smith
et al., 1995; Rajakovic and Ristic, 1996; Lou et al.,
1999; Simonnot et al., 2000; Öztürk and Kavak, 2004;
Polat et al., 2004). These investigations have shown that
treatments using selective resins are suitable for drinking
water, despite their high application and regeneration
costs (Simonnot et al., 2000). Therefore, novel materials
and methods are needed for the treatment of waters con-
taminated with boron.
In recent years, some sorption (adsorption, bio-
sorption and ion exchange) processes have gained
importance as effective purification and separation tech-
niques for removal of toxic substances from water due to
the preparation of some innovative materials. One of
these promising materials belong to the family of syn-
thetic compounds called layered double hydroxides
(LDHs) or hydrotalcite (HT)-like compounds. The gen-
eral chemical formula of these lamellar solids is written
as [M2+1
xM3+x(OH)2](An
)x/n Æ mH2O, where M2+ is a
divalent cation (Mg2+, Co2+, Ni2+, Zn2+, Mn2+, Cd2+),
M3+ is a trivalent cation (Al3+, Fe3+, Cr3+, Ga3+), An

is an anion and x is defined as the M3+/M2+ + M3+ ratio,
whose value varies between 0.2 and 0.33. These solids
present a brucite-like structure, where metallic cations
are located in the center of an edge-sharing octahedral
network. Isomorphous substitution of divalent cations
by trivalent ones generates positive charges on the lay-
ers, which are compensated by anions located in the
interlayer region. Water molecules are also found in
the interlayer space. Both anions and water molecules
are arranged in a liquid-like configuration. Therefore,
HT-like compounds can be viewed as a stacking of
81
positive (brucite-like sheets) and negative (anions)
layers, consecutively (Cavani et al., 1991; Newman and
Jones, 1998; Rives and Ulibarri, 1999; Ferreira et al.,
2004).
LDHs can be readily synthesized by coprecipitation
methods under laboratory conditions (Reichle, 1985).
The physical–chemical properties of the LDHs depend
on the M2+ and M3+ nature, interlayer anion and x val-
ues (responsible for the charge in the layers). These
materials can interact with anions both by the interlayer
space and external surface. Due to the presence of a
huge number of exchangeable anions (2–5 mmol g
1)
and their relatively large surface areas (10–120 m2 g
1)
(Miyata, 1983), HT are good ion exchangers/adsorbents
for removal of toxic anions from contaminated waters.
LDHs have been studied as potential adsorbents for sev-
eral anionic species like dichromate, selenate, fluoride,
Phenols, dyes, pesticides, phosphate, among others (Uli-
barri et al., 1995; Goswamee et al., 1998; Inacio et al.,
2001; Das et al., 2002, 2003; Ferreira and Alves, 2002;
Seida and Nakano, 2002; You et al., 2002). The anion
exchange properties of LDHs are affected considerably
by the nature of the interlayer anions. LDHs generally
have greater affinities for anions with higher charge den-
sity, like carbonate, which is preferentially intercalated,
preventing further ion exchange (Miyata, 1983; Sato
et al., 1986).
HT-like compounds have already been used for boron
adsorption. Parker et al. (1995) utilized a commercial HT
(Mg6Al2(OH)16CO3 Æ 4H2O) and synthetic ones contain-
ing sulfate and chloride anions for the adsorption of dif-
ferent anions such as F
, Br
, BðOHÞ


4 , NO3 and
2

HPO4 . Zhang and Reardon (2003) showed possible bor-
on removal from aqueous solution by in situ precipita-
tion of hydrocalumite Ca4Al2(OH)12(OH)2 Æ 6H2O. In
the present work, we present a systematic study employ-
ing LDHs on boron removal [a mixture of B(OH)3 and
BðOHÞ
4 ], which was investigated by the batch adsorp-
tion methods. We prepared HT-like compounds with dif-
ferent kinds of metal ions (Mg–Al and Mg–Fe) in the
brucite-like sheets and evaluated the effect of some
parameters such as contact time, initial pH of the boron
solution, boron concentration (adsorption isotherms)
and HT quantity on the efficiency of boron removal. Fur-
thermore, through adsorption isotherms were evaluated
the maximum adsorption capacity for each adsorbent.
The results reported here provide evidence for the reme-
diation of boron present in waters using HT-like com-
pounds as adsorbent.
2. Materials and methods
All chemicals (reagent grade, Ecibra or Vetec) were
used as received, without further purification. All solu-
tions were prepared with deionized water.
82 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88
2.1. Preparation of hydrotalcites
Mg–Al and Mg–Fe hydrotalcites with Mg2+/M3+
(M = Al3+ or Fe3+) molar ratios equal to 2 were syn-
3+
thesized by the co-precipitation method at variable
pH. An aqueous solution (70 ml) containing 0.1 mol of
Mg(NO3)2 Æ 6H2O and 0.05 mol of Al(NO3)3 Æ 9H2O or
Fe(NO3)3 Æ 9H2O was added dropwise under vigorous
mechanical stirring into 150 ml of a 2.0 M NaOH solu-
tion. During this process, the temperature was kept at
45 ± 3 °C. Once the addition of the nitrates was com-
pleted, a thick slurry was obtained, which was heated
for 2 h at 85 ± 3 °C for ageing. The solid products were
separated by filtration and washed several times with
deionized water until pH 7–8. Finally, the solids were
dried at 60 °C for 24 h and crushed below 105 lm. The
materials will be named hereafter as HTFe for Mg–Fe
hydrotalcite and HTAl for Mg–Al hydrotalcite.
2.2. Adsorption experiments
All solutions were prepared and stocked in polyethyl-
ene or TeflonTM flasks. The adsorption experiments were
carried out in 50 ml polyethylene centrifuge tubes. Bor-
on stock solutions were prepared from Na2B4O7 Æ 10-
H2O. Most of the experiments were carried out at the
same pH of the stock solution (pH  9) and at room
temperature (around 25 °C).
The time-dependent adsorption of B on HTFe and
HTAl was carried out with 100 mg of the adsorbent
and 40 ml of 26 mg l
1 boron solution (2.5 grams of
adsorbent per liter of solution). The mixtures were stir-
red at low speed (100 rpm) for different time intervals
(10–120 min), centrifuged and the boron concentrations
in the supernatant were then determined.
The effect of the adsorbent quantity on boron re-
moval was investigated using 40 ml of 18 mg l
1 boron
solutions, with increasing amounts of HTAl and HTFe
(20–200 mg). The mixtures were stirred for 120 min, cen-
trifuged and the supernatant solutions were collected for
boron determination.
pH-dependent experiments were carried out by
adjusting the initial pH (in the 4–10 range) of the boron
solution (18 mg l
1) using 0.1 M HNO3 or 0.1 M NaOH
solutions. One hundred milligrams of HTAl and HTFe
were weighed into centrifuge tubes and 40 ml of the
stock solution with adjusted pH was added and stirred
for 120 min. Then, the pH was measured again, the mix-
tures were centrifuged and the supernatant boron con-
centrations were determined.
Adsorption isotherms were obtained by the batch
equilibration technique. One hundred milligrams of
HTAl and HTFe were weighed into centrifuge tubes
and then these were filled with 40 ml of aqueous solu-
tions of boron (pH  9) ranging in concentration from
0 to 50.0 mg l
1. The mixtures were stirred for 120 min
at 25 °C, centrifuged and the supernatant boron concen-
trations were determined.
2.3. Chemical analysis and physical measurements
Metal contents of the HDLs were determined by
Inductively Coupled Plasma Optical Emission Spec-
trometry (ICP OES), using a Perkin-Elmer instrument
model Optima 3000 DV, after dissolving the samples
in 0.1 M HCl. Boron contents in all supernatant solu-
tions were also determined by ICP OES, using a Jobin
Yvon instrument model JY 50P (Sah and Brown,
1997). Carbon and nitrogen were determined by elemen-
tal microanalyses using a Perkin-Elmer Model 2400
CHN Analyzer. Powder X-ray diffraction (XRD) pat-
terns were obtained with a Shimadzu XRD6000 diffrac-
tometer, using CuKa (k = 1.5406 Å) radiation at 30 mA
and 40 kV and a scanning rate of 1° min
1. Fourier
transform infrared spectroscopy (FTIR) spectra were re-
corded using KBr pellets on a Bomem MB spectrometer
over the 4000–400 cm
1 range. ATR-FTIR spectra were
recorded on a Bomem FTLA2000, equipped with a
MIRacleTM ATR accessory. The specific surface areas
(BET) were determined using a Flowsorb 2300 from
Micromeritics Instrument Corporation.
3. Results and discussion
3.1. Characterization of the hydrotalcites
The results of chemical analyses for Mg, Al and Fe
showed that the Mg/Al and Mg/Fe molar ratios were
2.1 and 1.9 for HTAl and HTFe, respectively. The molar
ratio values observed for the precipitated solids are very
close to the nominal ratios used for preparing the start-
ing solutions, thus showing that the precipitation pro-
cesses were efficient.
The X-ray diffraction patterns for the as-prepared
LDH are shown in Fig. 1a. The XRD patterns show
sharp and symmetric peaks at lower 2h values, which
are characteristic of HT-like compounds (lamellar com-
pounds), and also indicate a high degree of crystallinity
of the parent samples (Ferreira et al., 2004). All the dif-
fraction peaks can be indexed to the space group R3m,
thus revealing that the hydrotalcites crystallized in a
rhombohedral structure (JCPDS card 22-0700 and
JCPDS card 25-0521). The results agree well with
previous reports (Del Arco et al., 2000; Das et al.,
2002). The (0 0 3) and (0 0 6) peaks (Fig. 1a) are attrib-
uted to basal reflections, corresponding to successive
stacking of the brucite-like sheets. The remainder peaks
in Fig. 1a are attributed to non-basal reflections. Sec-
ondary phases such as Fe(OH)3, Al(OH)3 and Mg(OH)2
were not observed in the XRD patterns of the prepared
samples.
 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88 83

carbonate could be located in the interlayer spaces of

(a)
both samples.
(003)

(006)
The FTIR spectra of the as-prepared HTFe and
HTAl samples are shown in Fig. 1b. The broad band ob-
served in the 4000–3000 cm
1 range is attributed to the
O–H stretching mode (mO–H). The OH groups are pres-
ent in the layers and also in both adsorbed and interla-
mellar water. The mode at about 1640 cm
1 is assigned
(015)

(110)
Intensity

(113)
(012)

(018)
500 cps to HOH deformation (dH–O–H), thus confirming the
HTAl
presence of water molecules in the interlayer region of
the parent materials. The band observed at approxi-
(003)

mately 1400 cm
1 is actually formed by two peaks: a

sharp one centered at 1385 cm
1 and a broader one at
(006)

(012)

(110)
(113)
(015)

1360 cm
1 (marked with an upward arrow). These

(018)

250 cps

HTFe
modes are assigned to the m3 (asymmetric stretching
10 20 30 40 50 60 mode) of nitrate and carbonate anions, respectively
2θ (degrees) (Hernandez-Moreno et al., 1985; Das et al., 2002). The
two weak bands located at about 825 and 835 cm
1 have
(b) been identified as the m2 [out-of-plane deformation
mode] of nitrate anions. Therefore, FTIR results corrob-
HTAl orate with the CHN analysis, indicating the presence of
both carbonate and nitrate in the interlayer spaces of
HTFe and HTAl. The remaining bands below
T (%)

1000 cm
1 are due to metal oxygen vibrations (Hernan-

HTFe dez-Moreno et al., 1985).
Specific surface areas of the two compounds were
determined by the single-point BET method and were
found to be 5 and 20 m2 g
1 for HTAl and HTFe,
respectively.

4000 3600 3200 2800 1600 1200 800 400

3.2. Adsorption
Wavenumber (cm-1)

Fig. 1. (a) XRD patterns of HTFe and HTAl. (b) FTIR spectra It is well know that Na2B4O7 Æ 10H2O is an excellent
of HTFe and HTAl. buffer (pH  9), dissociating completely into equal con-
centrations of neutral boric acid, B(OH)3, and borate,
BðOHÞ
4 (Eq. (1)). In concentrated aqueous solutions
polyborates can be formed by combining B(OH)3 and
Using the (0 0 3) peaks and Bragg equation it is pos-
BðOHÞ
4 , but in dilute solutions the presence of poly-
sible to calculate the interlayer space (d-spacing, layer
borates is negligible (Momii and Nachtrieb, 1967; Smith
thickness + gallery height) of the HT-like compounds.
and Wiersema, 1972; Stevenson et al., 1997)
The interlayer space of HT-like compounds depends
on the geometry and the size of the anion, M2+/M3+ B4 O2


7 þ 7H2 O
2BðOHÞ3 þ 2BðOHÞ4 ð1Þ
molar ratio, and for some anions, on the degree of
hydration. The interlayer spacing calculated for HTAl Boric acid, B(OH)3, and its anion borate, BðOHÞ

4, be-
and HTFe are 8.89 and 7.85 Å, respectively, which are have quite differently from most other oxyanions. Bo-
typical values for nitrate and carbonate-containing rate can be formed by the addition of one hydroxyl
HT. Therefore, the synthesized HTAl has mainly nitrate group to boric acid (Eq. (2)). The anion concentration
in the interlayer spaces, whereas HTFe presents mainly in solution is pH-dependent. At pH values around 9.0,
carbonate (Cavani et al., 1991; Del Arco et al., 2000). the B(OH)3 and BðOHÞ
4 concentrations are practically
Nitrate is present in the metal salts used for the synthesis the same, while at higher pH values (11.0), BðOHÞ
4
and carbonate must have resulted from CO2 present in is the predominant species (pKa = 9.2) (Lou et al.,
aqueous solution, since these syntheses were not pro- 1999; Peak et al., 2003)
cessed with decarbonated water and under inert atmo-
BðOHÞ3 þ OH

BðOHÞ

4 ð2Þ
sphere (Cavani et al., 1991; Newman and Jones, 1998).
However, CHN analyses (not shown here) showed N The pH of the borate solutions studied were approxi-
and C in all samples, suggesting that both nitrate and mately 9.0, suggesting a buffering behavior. In all
84 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88
experiments, only the presence of B(OH)3 and BðOHÞ

4 Comparing the kinetic curves shown in Fig. 2, it is
species in solution was considered.
3.3. Effect of contact time
The variation of boron adsorption as a function of
time is shown in Fig. 2. The kinetic curves show that
the equilibrium time and adsorption capacity of HTAl
and HTFe are quite different. We observed that adsorp-
tion of boron is a fast process on HT-like compounds.
This behavior has already been reported in the adsorp-
tion of anionic species on HT-like compounds (Her-
mosı́n et al., 1996; Inacio et al., 2001; Das et al.,
2002). Using HTAl as adsorbent, the removed amount
of boron presents an initial step (10 min) and after that
it increases slowly until approximately 80 min. The
amount of adsorbed boron was 9.20 mg g
1. For HTFe,
only 10 min were necessary for reaching equilibrium and
the maximum amount of adsorbed boron was
3.13 mg g
1.
It is well known that HT-like compounds interact
strongly with anionic species due to the existence of po-
sitive charges on the lamellae and also over their exter-
nal surface (Inacio et al., 2001). This property makes
possible the occurrence of two different mechanisms of
anion removal: (i) adsorption on external surface (faster)
and (ii) ion exchange (slower) (Cavani et al., 1991; New-
man and Jones, 1998). The adsorption process depends
on the variable charge density on the surface and of
the surface area of the adsorbent. On the other hand,
ion exchange has its occurrence conditioned to the na-
ture of the interlayer anions present in the sample (Mi-
yata, 1983; Sato et al., 1986). Generally, the adsorption
process is faster than ion exchange because of the strong
interaction (Coulomb forces) between negative ions and
the positive external surface and due to the high exposi-
tion of the external surface of the hydrotalcites, while
ion exchange is a diffusion process.
10
8
Removal (mg B/g)
HTAl
6 HTFe
4 pH buffering capacities (Fig. 3a). Thus, when the ini-
2 of the HT-like compounds. When the initial pH was
0 directly from the solution by LDHs (Hermosı́n et al.,
0 20 40 60 80 100
Time (min)
Fig. 2. Effect of time on boron adsorption (C0 = 26 mgB l
1
and pH = 9.2).
probable that boron is only adsorbed on the external
surface of HTFe. Miyata (1983) showed that divalent
anions interact more strongly with the brucite-like sheets
than monovalent ones. The strong interactions between
carbonate anions and hydrotalcite lamellae inhibit anio-
nic exchange for HTFe, since this adsorbent have mainly
carbonate in the interlayer space (Sato et al., 1986). Due
to the smaller surface area showed by HTAl, one could
expect lower adsorption capacity for this sample. How-
ever, Fig. 2 shows the opposite, and we believe that the
most probable reason for that is the occurrence of both
removal mechanisms commented above, adsorption fol-
lowed by anion exchange. In this adsorbent, anion ex-
change can occur due to the presence of nitrate as the
main interlayer species (Miyata, 1983). The ion ex-
change of nitrate by others anionic species in HT-like
compounds has been abundantly described in the litera-
ture (Del Arco et al., 2000; Komarneni et al., 2003;
Tamura et al., 2004).
The experiments of contact time suggest that 120 min
is enough to reach the equilibrium state in the boron
adsorption on HT-like compounds studied.
3.4. Effect of solution pH
The effect of the initial pH on boron removal for
HTFe and HTAl samples are shown in Fig. 3a. 0.1 M
HNO3 and NaOH solutions were used for pH adjust-
ments. Buffer solutions were not used in order to avoid
contamination of foreign anions, since it is well recog-
nized that secondary anions extensively affect adsorp-
tion (You et al., 2002; Das et al., 2003). Miyata (1983)
showed that the ion selectivity in HT-like compounds
is in the following order: OH
> F
> Cl
> Br
>
NO3
> I
and divalent anions have higher ion selectiv-
ity than monovalent ones. Based on these previous stud-


ies, we believe that NO
3 and OH have a minor effect
on boron adsorption compared to conventional buffer
solutions.
The amount of boron removed was higher for HTAl
than for HTFe (Fig. 3a). However, the maximum
amount of boron removed for each adsorbent remains
practically constant in the pH range studied. In other
words, the adsorption process was pH independent.
The relationship between the initial and equilibrium
pH for boron adsorption suggests that LDHs have high
tial pH was low, the pH increased due to dissolution
high, the decrease in pH was related to OH
adsorption
120
1996). Therefore, boron removal was carried out at only
one pH value (equilibrium pH) and, as a consequence,
the amount of boron removed remained constant irre-
spective to the pH of the starting boron solution due
 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88 85

100 12 3.5. Adsorbent dose

HTAl (a)

80
HTFe
10 The amount of boron removed from 18 mg l
1 boron
solutions increases by increasing the adsorbent quantity
% of adsorpition

HTAl for both adsorbents (Fig. 4a). For HTAl, the amount of

Final pH
8
60
boron adsorbed reaches a plateau at 2.5 g l
1 with maxi-
6 mum adsorption around 92%. On the other hand, for
40 HTFe a plateau was not observed in the studied dose
HTFe 4 range and the maximum adsorption was around 33%.
20 For HTAl (from 2.5 g l
1 on), the increase in the
2 number of adsorption sites (increase of adsorbent mass)
0 did not result in the increase of boron removed. This
4 5 6 7 8 9 10 observation indicates that the chemical equilibrium
Initial pH
existing between the species of boron in solution and
50 the equilibrium in the solid-solution interface may be
(b) responsible for such behavior. We performed ATR-
FTIR measurements in the HTAl compound submitted
Conc. of Mg2+ (mg/l)

40
HTAl
HTFe
to the treatment of the boron-containing water and we
30 verified the presence of BðOHÞ
4 on its surface (Fig.
4b). BðOHÞ
4 and B(OH)3 present distinct bands in
20 the infrared spectra (Su and Suarez, 1995; Peak et al.,
2003). BðOHÞ
4 presents a characteristic band in 955
10 cm
1 attributed to the asymmetric stretching of tetrahe-
dral boron, while B(OH)3 presents a band in 1410 cm
1
0 assigned to the B–O asymmetric stretching of trigonal
4 5 6 7 8 9 10
Initial pH

Fig. 3. (a) Effect of pH on boron removal. (b) Concentration of 100

(a)
Mg2+ after boron removal.
80
% of adsorption

to the buffering effect of the adsorbent. Other authors 60 HTAl

have also observed the buffering effect of hydrotalcite- HTFe
like compounds. You et al. (2001) observed this studying 40
SeO2
3 adsorption in Mg–Al and Zn–Al hydrotalcites,
where the amount of anion removed was almost con- 20
stant in the pH range where this effect was observed. Sei-
da and Nakano (2002) observed the same effect studying 0
phosphate adsorption in Mg–Fe and Ca–Fe 0 1 2 3 4 5
hydrotalcite. Adsorbent quantity (g/l)
An important point to be evaluated is the stability of
the adsorbent in the adsorption process. If metals are re- HTAl-containing adsorved boron (b)
leased in the solution, it would be important that this
amount of metals keeps below the levels that are harm- 955
ful or to choose low toxicity cations. The amount of
T (%)

magnesium ion released from the compounds decreased HTAl

with increasing initial pH (Fig. 3b). This result shows
higher stability of the brucite-like sheets at higher pH
values which is expected, since these materials are
hydroxides. The amount of Mg2+ released from HTFe
was smaller than from HTAl. Fig. 3b provides some
interesting points: (i) adsorption performed in pH values
higher than 7 induces a decrease in the amount of re- 3900 3600 3300 1800 1500 1200 900
Wavenumber (cm-1)
leased metals (natural waters usually present pH in the
range of 7–9); (ii) the nature of the metals in the bru- Fig. 4. (a) Adsorption of boron as a function of adsorbent
cite-like sheets would be also an important issue in the quantity. (b) ATR-FTIR spectra of HTAl and HTAl contain-
control of the releasing of metals. ing adsorbed boron.
86 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88
boron. In Fig. 4b we clearly observe a band at 955 cm
1
(marked with up arrow) of BðOHÞ
4.
When BðOHÞ
4 is adsorbed in the HTAl surface, the
equilibrium described in Eq. (2) would be shifted to
the right. However, in a limit value of concentration
there would not have the conversion of B(OH)3 to
BðOHÞ
4 and, because of that we would still have boron
in the solution. This way, even if we increase the number
of adsorption sites, the amount of boron removed would
be the same, once hydrotalcite-like compounds can not
remove B(OH)3.
For the HTFe sample, as the amount of adsorbed
BðOHÞ
4 is smaller in each step, the equilibrium would
not reach its extreme conditions, and consequently by
increasing the number of adsorption sites we observed
the increase in the amount of boron removed.
3.6. Adsorption isotherms
The boron adsorption isotherms on HTFe and HTAl
at 25 °C are shown in Fig. 5. Isotherm profiles can give
information about the nature of the adsorbed species
(Hermosı́n et al., 1996). According to the classification
proposed by Giles et al. (1960), both are L-type (sub-
group 2) isotherms, indicating that the interaction sor-
bate–sorbent is much stronger than solvent–sorbent at
the adsorption sites. Isotherms with this profile are typ-
ical of systems where the monofunctional adsorbate is
strongly attracted by the adsorbent, mostly by ion–ion
interaction, reaching a saturation value given by the iso-
therm plateau. These results again suggest that BðOHÞ
4 a maximum adsorption was found at 2.5 g l
1 (92%).
anions are preferentially removed. L-type isotherms
have been reported by others authors for the adsorption
of inorganic anions on HT-like compounds (You et al.,
2001).
The data obtained from the adsorption process was
fitted into the linearly transformed Langmuir adsorption
isotherm. Our data nicely fit the Langmuir model, with
r2 = 0.999 and r2 = 0.994, for HTFe and HTAl, respec-
15
HTAl
12
qe (mg/g)
9
6 boron removal performance was attributed to the fact
HTFe
3 be described as a pure adsorption, but as a combination
0 tion was the only process responsible for B removal.
0 10 20 30 40
Ce (mg B/l)
Fig. 5. Adsorption isotherms on HTAl and HTFe.
tively. Using this model, it was possible to calculate
the maximum capacity of boron removal as 14.0 mg g
1
for HTAl and 3.6 mg g
1 for HTFe. Therefore, the
adsorption capacity of HTAl is much higher than that
of HTFe. The process may not be described as a pure
adsorption in the case of HTAl, but as a combination
of adsorption and ion exchange (Hermosı́n et al., 1996).
It is important to comment that when HTAl was used
for the treatment of water containing 5.2 mg l
1 of bor-
on (in the adsorption isotherms experiments) it was pos-
sible to reach the limit recommend by WHO to boron
concentration in drinking water: 0.5 mg l
1. This result
suggests that HT-like compounds can either be used as
adsorbent for boron removal in natural water or partic-
ipate in the purification process.
4. Conclusions
In this study we showed that Mg–Fe (HTFe) and
Mg–Al (HTAl) hydrotalcite-like compounds can be used
as adsorbents for boron removal [BðOHÞ
4 ] from aque-
ous media. It was found that 120 min is enough for
the equilibrium state to be reached in the adsorption
experiments.
The HT-like compounds synthesized showed high pH
buffering capacities and, consequently, the amount of
boron removed was independent of the initial pH. On
the other hand, the boron removal was dependent of
the adsorbent dose. When HTAl was used as adsorbent,
After this point, the increase in adsorbent dose (increase
in the number of adsorption sites) did not result in the
increase of the boron removal capacity. However, when
HTFe was used as adsorbent, the boron removal in-
creased linearly with increasing adsorbent dose, but
the maximum removal capacity of removal was only
33%.
Adsorption isotherms were performed for initial B
concentrations ranging from 0 to 50 mg l
1. The iso-
therms are L-type for both adsorbent indicating an
ion–ion interaction and that BðOHÞ
4 is the chemical
species adsorbed. The adsorption can be described by
the Langmuir model and the maximum adsorption
capacity was 14.0 and 3.6 mgB g
1 for HTAl and HTFe,
respectively. Therefore, the adsorption capacity of HTAl
is much higher than that of HTFe. This difference in the
that in the case of HTAl the removal process may not
of adsorption and ion exchange, while for HTFe adsorp-
HTAl have mainly nitrate in the interlayer space, which
can be exchanged by others anionic species, whereas
HTFe present mainly carbonate, which interacts
strongly with the brucite-like sheets and therefore can
 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88 87

hardly be exchanged by other anions. In the treatment of References

boron solutions (5.2 mg l
1) using HTAl, it was possible
to reduce the boron content to the limit recommended Cavani, F., Trifirò, F., Vaccari, A., 1991. Hydrotalcite-type
by WHO for drinking water (0.5 mg l
1). In summary, anionic clays: preparation, properties and applications.
this work shows that HT could be used either as a novel Catal. Today 11, 173–301.
Das, J., Das, D., Dash, G.P., Parida, K.M., 2002. Studies on
material for the treatment of natural waters containing
Mg/Fe hydrotalcite-like compounds (HTlc)—I. Removal of
boron or to participate in one of the stages of the treat-
inorganic selenite (SeO2

3 ) from aqueous medium. J. Colloid
ment process. Interf. Sci. 251, 26–32.
Das, D.P., Das, J., Parida, K., 2003. Physicochemical charac-
terization and adsorption behavior of calcined Zn/Al
Acknowledgements hydrotalcite-like compounds (HTLc) towards removal of
fluoride from aqueous solution. J. Colloid Interf. Sci. 261,
The authors are grateful to CNPq, FAPESP and 213–220.
PRONEX (Proc. No. 11P22810-97) for financial sup- Del Arco, M., Guitiérrez, S., Martı́n, C., Rives, V., Rocha, J.,
port. This is a contribution of the Millennium Institute 2000. Effect of the Mg:Al ratio on borate (or silicate)/nitrate
exchange in hydrotalcites. J. Solid State Chem. 151,
for Complex Materials.
272–280.
Ferreira, O.P., Alves, O.L., 2002. Material poroso bidimen-
sional para a descoloração de efluentes têxteis contendo
Appendix A corantes aniônicos e sua reciclagem. PI 0200354-6, Brazilian
Patent.
Ferreira, O.P., Alves, O.L., Gouveia, D.X., Filho, A.G.S., de
Al(OH)3 aluminum hydroxide
Paiva, J.A.C., Filho, J.M., 2004. Thermal decomposition
Al(NO3)3 Æ 9H2O aluminum nitrate nonahydrate and structural reconstruction effect on Mg–Fe based hydro-
Å Angström talcite compounds. J. Solid State Chem. 177, 3058–3069.
ATR-FTIR attenuated total reflectance Fourier trans- Giles, C.H., MacEwan, T.H., Nakhwa, S.N., Smith, D., 1960.
form infrared spectroscopy Studies in adsorption. Part XI. A system of classification of
BðOHÞ
4 borate solution adsorption isotherms, and its use in diagnosis of
B(OH)3 boric acid adsorption mechanisms and in measurement of specific
CO2 carbon dioxide surface areas of solids. J. Chem. Soc., 3973–3993.
CHN carbon–hydrogen–nitrogen Goldberg, S., Forster, H.S., Heick, E.L., 1993. Boron adsorp-
tion mechanism on oxides, Clay Minerals, and soils inferred
°C degree Celsius
from ionic strength effects. Soil Sci. Soc. Am. J. 57, 704–708.
g l
1 gram per liter
Goswamee, R.L., Sengupta, P., Bhattacharyya, K.G., Dutta,
h hour D.K., 1998. Adsorption of Cr(VI) in layered double
HT hydrotalcite hydroxides. Appl. Clay Sci. 13, 21–34.
ICP OES inductively coupled plasma optical emission Hermosı́n, M.C., Pavlovic, I., Ulbibarri, M.A., Cornejo, J.,
spectrometry 1996. Hydrotalcite as sorbent for trinitrophenol: sorption
Fe(OH)3 iron hydroxide capacity and mechanism. Water Res. 30, 171–177.
Fe(NO3)3 Æ 9H2O iron nitrate nonahydrate Hernandez-Moreno, M.J., Ulibarri, M.A., Rendon, J.L., Serna,
LDHs layered double hydroxides C.J., 1985. IR characteristics of hydrotalcite-like com-
Mg(OH)2 magnesium hydroxide pounds. Phys. Chem. Miner. 12, 34–38.
Inacio, J., Taviot-Guého, C., Forano, C., Besse, J.P., 2001.
Mg(NO3)2 Æ 6H2O magnesium nitrate hexahydrate
Asdorption of MCPA pesticide by Mg–Al-layered double
HTFe Mg–Fe hydrotalcite
hydroxides. Appl. Clay Sci. 18, 255–264.
HTAl Mg–Al hydrotalcite Keren, R., Grossl, P.R., Sparks, 1994. Equilibrium and kinetics
mg milligrams of borate adsorption–desorption on pyrophyllite in aqueous
mg l
1 milligrams per liter suspensions. Soil Sci. Soc. Am. J. 58, 1116–1122.
mg g
1 milligrams per gram Komarneni, S., Newalkar, B.L., Li, D.S., Gheyi, T., Lopano,
ml milliliter C.L., Heaney, P.J., Post, J.E., 2003. Anionic clays as
mmol g
1 millimol per gram potential slow-release fertilizers: nitrate ion exchange.
min minutes J. Porous Mater. 10, 243–248.
HNO3 nitric acid Lou, J., Foutch, G.L., Jung, W.N., 1999. The sorption capacity
of boron on anionic-exchange resin. Separ. Sci. Technol. 34,
M mol per liter
2923–2941.
rpm rotation per minute
Miyata, S., 1983. Anion-exchange properties of hydrotalcite-
NaOH sodium hydroxide like compounds. Clays Clay Miner. 31, 305–311.
Na2B4O7 Æ 10H2O sodium tetraborate decahydrate Momii, R.K., Nachtrieb, N.H., 1967. Nuclear magnetic reso-
m2 g
1 square meter per gram nance study of borate–polyborate equilibria in aqueous
WHO World Heath Organization solution. Inorg. Chem. 6, 1189–1191.
88 O.P. Ferreira et al. / Chemosphere 62 (2006) 80–88
Newman, S.P., Jones, W., 1998. Synthesis, characterization and
applications of layered double hydroxides containing
organic guests. New J. Chem. 22, 105–115.
Okay, O., Güçlü, H., Soner, E., Balkas, T., 1985. Boron
pollution in the Simav river, Turkey and various methods of
boron removal. Water Res. 19, 857–862.
Öztürk, N., Kavak, D., 2004. Boron removal from aqueous
solutions by adsorption on waste sepiolite and activated
sepiolite using full factorial design. Adsorption 10, 245–257.
Parker, L.M., Milestone, N.B., Newman, R.H., 1995. The use
of hydrotalcite as an anion absorbent. Ind. Eng. Chem. Res.
34, 1196–1202.
Peak, D., Luther, G.W., Sparks, D.L., 2003. ATR-FTIR
spectroscopic studies of boric acid adsorption on hydrous
ferric oxide. Geochim. Cosmochim. Acta 67, 2551–2560.
Polat, H., Vengosh, A., Pankratov, I., Polat, M., 2004. A new
methodology for removal of boron from water by coal and
fly ash. Deslination 164, 173–188.
Rajakovic, Lj.V., Ristic, M.Dj., 1996. Sorption of boric acid
and borax by activated carbon impregnated with various
compounds. Carbon 34, 769–774.
Reichle, W.T., 1985. Catalytic reactions by thermally activated,
synthetic, anionic clay-minerals. J. Catal. 94, 547–557.
Rives, V., Ulibarri, M.A., 1999. Layered double hydroxides
(LDH) intercalated with metal coordination compounds
and oxometalates. Coord. Chem. Rev. 181, 61–120.
Sah, R.N., Brown, P.H., 1997. Boron determination—a review
of analytical methods. Microchem. J. 56, 285–304.
Sato, T., Wakabayashi, T., Shimada, M., 1986. Adsorption of
various anions by magnesium aluminum-oxide (Mg0.7Al0.3-
O1.15). Ind. Eng. Chem. Prod. Res. Dev. 25, 89–92.
Seida, Y., Nakano, Y., 2002. Removal of phosphate by layered
double hydroxides containing iron. Water Res. 36, 1306–
1312.
Simonnot, M.-O., Castel, C., Nicolai, M., Rosin, C., Sardin,
M., Jauffret, H., 2000. Boron removal from drinking water
with a boron selective resin: is the treatment really selective?
Water Res. 34, 109–116.
Smith Jr., H.D., Wiersema, R.J., 1972. Boron-11 nuclear
magnetic resonance study of polyborate ions in solution.
Inorg. Chem. 11, 1152–1154.
Smith, B.M., Todd, P., Bowman, C.N., 1995. Boron removal by
polymer-assisted ultrafiltration. Separ. Sci. Technol. 30,
38–59.
Stevenson, K.J., Hatchett, D.W., White, H.S., 1997. Electro-
chemical deposition of polyborate monolayers at Ag(1 1 1)
electrodes. Langmuir 13, 6824–6828.
Su, C., Suarez, D.L., 1995. Coordination of adsorbed boron: a
FTIR spectroscopic study. Environ. Sci. Technol. 29, 302–
311.
Tamura, H., Chiba, J., Ito, M., Takeda, T., Kikkawa, S., 2004.
Synthesis and characterization of hydrotalcite-ATP interca-
lates. Solid State Ion. 172, 607–609.
Ulibarri, M.A., Pavlovic, I., Hermosı́n, M.C., Cornejo, J., 1995.
Hydrotalcite-like compounds as potential sorbents of phe-
nols from water. Appl. Clay Sci. 10, 131–145.
WHO, 1998. Guidelines for Drink-Water Quality, second ed.,
Addendum to Vol. 2. World Heath Organization, Geneva.
You, Y., Vance, G.F., Zhao, H., 2001. Selenium adsorption on
Mg–Al and Zn–Al layered double hydroxides. Appl. Clay
Sci. 20, 13–25.
You, Y., Zhao, H., Vance, G.F., 2002. Adsorption of dicamba
(3,6-dichloro-2-methoxy benzoic acid) in aqueous solution
by calcined-layered double hydroxides. Appl. Clay. Sci. 21,
217–226.
Zhang, M., Reardon, E.J., 2003. Removal of B, Cr, Mo, and Se
from wastewater by incorporation into hydrocalumite and
ettringite. Environ. Sci. Technol. 37, 2947–2952.