Adsorption mechanism of selenate and selenite on the binary oxide systems.
ABSTRACT Removal of selenium oxyanions by the binary oxide systems, Al- or Fe-oxides mixed with X-ray noncrystalline SiO(2), was previously not well understood. This study evaluates the adsorption capacity and kinetics of selenium oxyanions by different metal hydroxides onto SiO(2), and uses X-ray absorption spectroscopy (XAS) to assess the interaction between selenium oxyanions and the sorbents at pH 5.0. The binary oxide systems of Al(III)- or Fe(III)-oxides mixed with SiO(2) were prepared, and were characterized for their surface area, point of zero charge (PZC), pH envelopes, X-ray diffraction analysis (XRD), and then macro-scale adsorption isotherm and kinetics of selenite and selenate, micro-scale adsorption XAS. The adsorption capacity of selenite and selenate on Al(III)/SiO(2) is greater than on Fe(III)/SiO(2). Adsorption isothermal and kinetic data of selenium can be well fitted to the Langmuir isotherm and pseudo-second-order kinetic models. Based on simple geometrical constraints, selenite on both the binary oxide systems forms bidentate inner-sphere surface complexes, and selenate on Fe(III)/SiO(2) forms stronger complexes than on Al(III)/SiO(2).
- SourceAvailable from: Wen-Hui Kuan[show abstract] [hide abstract]
ABSTRACT: Reactions of Al(III) at the interface between SiO2(s) and aqueous solution were characteristically and quantitatively studied using electrophoretic methods and applying a surface complexation/precipitation model (SCM/SPM). The surface and bulk properties of Al(III)/SiO2 suspensions were determined as functions of pH and initial Al(III) concentration. Simulated modeling results indicate that the SCM, accounting for the adsorption mechanism, predicts sorption data for low surface coverage only reasonably well. Al(III) hydrolysis and surface hydroxide precipitation must be invoked as the Al(III) concentration and/or pH progressively increase. Accordingly, the three processes in the Al(III) sorption continuum, from adsorption through hydrolysis to surface precipitation, could be identified by the divergence between the SCM/SPM predictions and the experimental data. SiO2(s) suspensions with low Al(III) concentrations (1 x 10(-4) and 1 x 10(-5) M) exhibit electrophoretic behavior similar to that of a pure SiO2(s) system. In Al(III)/SiO2 systems with high Al concentrations of 1 x 10(-3), 5 x 10(-3) and 1 x 10(-2) M, three charge reversals (CR) are observed, separately representing, in order of increasing pH, the point of zero charge (PZC) on the SiO2 substrate (CR1), the onset of the surface precipitation of Al hydroxide (CR2), and at a high pH, the PZC of the Al(OH)3 coating (CR3). Furthermore, in the 1 x 10(-3) M Al(III)/SiO2(s) system, CR2 is consistent with the modeling results of SCM/SPM and provides evidence that Al(III) forms a surface precipitate on SiO2(s) at pH above 4. SiO2(s) dissolution was slightly inhibited when Al(III) was adsorbed onto the surface of SiO2(s), as compared to the dissolution that occurs in a pure SiO2(s) suspension system. Al hydroxide surface precipitation dramatically reduced the dissolution of SiO2(s) because the Al hydroxide passive film inhibited the corrosion of the SiO2(s) surface by OH- ions.Journal of Colloid and Interface Science 05/2004; 272(2):489-97. · 3.17 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Selenate (SeO4(2-)) is an oxyanion of environmental importance because of its toxicity to animals and its mobility in the soil environment. It is known that iron(III) oxides and hydroxides are important sorbents for SeO4(2-) in soils and sediments, but the mechanism of selenate adsorption on iron oxides has been the subject of intense debate. Our research employed Extended X-ray absorption fine structure and attenuated total reflectance-Fourier transform infrared spectroscopies to determine SeO4(2-) bonding mechanisms on hematite, goethite, and hydrous ferric oxide (HFO). It was learned that selenate forms only inner-sphere surface complexes on hematite but forms a mixture of outer- and inner-sphere surface complexes on goethite and HFO. This continuum of adsorption mechanisms is strongly affected by both pH and ionic strength. These results suggest that adsorption experiments should be conducted on several different iron oxides and over a wide range of reaction conditions to accurately assess the reactivity of oxyanions on iron oxides.Environmental Science and Technology 05/2002; 36(7):1460-6. · 5.26 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: This paper evaluates the use of upflow anaerobic sludge bed (UASB) bioreactors (30 degrees C, pH=7.0) to remove selenium oxyanions from contaminated waters (790 microg Se L(-1)) under methanogenic and sulfate-reducing conditions using lactate as electron donor. One UASB reactor received sulfate at different sulfate to selenate ratios, while another UASB was operated under methanogenic conditions for 132 days without sulfate in the influent. The selenate effluent concentrations in the sulfate-reducing and methanogenic reactor were 24 and 8 microg Se L(-1), corresponding to removal efficiencies of 97% and 99%, respectively. X-ray diffraction (XRD) analysis and sequential extractions showed that selenium was mainly retained as elemental selenium in the biomass. However, the total dissolved selenium effluent concentrations amounted to 73 and 80 microg Se L(-1), respectively, suggesting that selenate was partly converted to another selenium compound, most likely colloidally dispersed Se(0) nanoparticles. Possible intermediates of selenium reduction (selenite, dimethylselenide, dimethyldiselenide, H(2)Se) could not be detected. Sulfate reducers removed selenate at molar excess of sulfate to selenate (up to a factor of 2600) and elevated dissolved sulfide concentrations (up to 168 mg L(-1)), but selenium removal efficiencies were limited by the applied sulfate-loading rate. In the methanogenic bioreactor, selenate and dissolved selenium removal were independent of the sulfate load, but inhibited by sulfide (101 mg L(-1)). The selenium removal efficiency of the methanogenic UASB abruptly improved after 58 days of operation, suggesting that a specialized selenium-converting population developed in the reactor. This paper demonstrates that both sulfate-reducing and methanogenic UASB reactors can be applied to remove selenate from contaminated natural waters and anthropogenic waste streams, e.g. agricultural drainage waters, acid mine drainage and flue gas desulfurization bleeds.Water Research 05/2008; 42(8-9):2184-94. · 4.66 Impact Factor
Adsorption mechanism of selenate and selenite on the binary
Ya Ting Chana, Wen Hui Kuanb, Tsan Yao Chenc, Ming Kuang Wanga,*
aDepartment of Agricultural Chemistry, National Taiwan University, Taipei, 10617, Taiwan
bDepartment of Environmental and Safety Engineering, Ming-Chi University of Technology, Taipei County, 243, Taiwan
cDepartment of Engineering and System Sciences, National Tsing Hua University, Hsin-Chu, 30043, Taiwan
a r t i c l e i n f o
Received 6 January 2009
Received in revised form
29 June 2009
Accepted 30 June 2009
Published online 4 July 2009
Binary oxide systems
Pseudo-second-order kinetic model
X-ray absorption spectroscopy (XAS)
a b s t r a c t
Removal of selenium oxyanions by the binary oxide systems, Al- or Fe-oxides mixed with
X-ray noncrystalline SiO2, was previously not well understood. This study evaluates the
adsorption capacity and kinetics of selenium oxyanions by different metal hydroxides onto
SiO2, and uses X-ray absorption spectroscopy (XAS) to assess the interaction between
selenium oxyanions and the sorbents at pH 5.0. The binary oxide systems of Al(III)- or
Fe(III)-oxides mixed with SiO2were prepared, and were characterized for their surface area,
point of zero charge (PZC), pH envelopes, X-ray diffraction analysis (XRD), and then macro-
scale adsorption isotherm and kinetics of selenite and selenate, micro-scale adsorption
XAS. The adsorption capacity of selenite and selenate on Al(III)/SiO2is greater than on
Fe(III)/SiO2. Adsorption isothermal and kinetic data of selenium can be well fitted to the
Langmuir isotherm and pseudo-second-order kinetic models. Based on simple geometrical
constraints, selenite on both the binary oxide systems forms bidentate inner-sphere
surface complexes, and selenate on Fe(III)/SiO2forms stronger complexes than on Al(III)/
ª 2009 Elsevier Ltd. All rights reserved.
Selenium (Se) is an essential micronutrient for humans and
animals, but both Se toxicity and Se deficiency occur in
different parts of the world (Frankenberger and Benson, 1994).
Selenium can increase activity of the free hydroxyl radicals
(OH-) that cause high oxidation stress harmful to living beings
(Zhao et al., 2008). The essential or toxic character of Se in
living beings depends not only on its concentration in the
circumstances, but also on the chemical compound, which
directly affects absorption and bioavailability (Mikkelsen
et al., 1989; Lenz et al., 2008). Selenate and selenite are the
dominant species in aqueous systems (Jacobs, 1989), however,
selenite is easily adsorbed by soil minerals and a higher
concentration of selenate in industrial waste water is the
major selenium compound polluting waters.
Aluminum, iron, and silica oxides are ubiquitous minerals
in the earth’s crust and they are usually employed in
removing the pollutants in waters, because Al- and Fe-oxides
have high surface areas and point of zero charge (i.e., 8–9.2)
(Sparks, 2003). However, single oxide minerals are scarcer
than binary oxide minerals in natural environments. The
interaction of particles in binary oxide systems could influ-
ence the adsorption mechanism of adsorbents (Honeyman,
1984; Goh et al., 2008; Zhang et al., 2008). Lo and Chen (1997)
and Kuan et al. (1998) showed that the adsorption capacity of
selenium oxyanions on Al- and Fe-oxides coating quartz
sands was about 1.1–1.5 mg Se per gram quartz. Kuan et al.
* Corresponding author. Tel.: þ8862 3366 4808; fax: þ8862 2366 0751.
E-mail address: email@example.com (M.K. Wang).
0043-1354/$ – see front matter ª 2009 Elsevier Ltd. All rights reserved.
Available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/watres
water research 43 (2009) 4412–4420
(2004) reported Al–Si complexation and precipitate forms on
the amorphous SiO2surfaces, and discrete Fe(OH)3particles
form in Fe(OH)3/SiO2 (Meng and Letterman, 1993, 1996).
Therefore, it is necessary to study the adsorption capacities
and mechanisms of selenate and selenite by the binary oxide
systems, Al(III)- or Fe(III)-oxides mixed with X-ray non-
X-ray absorption spectroscopic studies provide informa-
tion to help understand the interaction between oxyanions
and mineral surfaces, and this information can also improve
both theoretical and practical understanding of surface
complexation reactions (Sparks, 2003). Hayes et al. (1987)
previously compared selenate and selenite adsorption on
various mineral surfaces and showed that selenate generally
formed relatively weakly-bound complexes, while selenite
Fe-oxides can strongly affect their surface chemistry and
reactivity with oxyanions. For example, sulfate formed
inner-sphere monodentate complexes on a hematite surface
(Hug, 1997), but it can be formed as inner- and outer-sphere
since pH values can strongly affect surface properties of Fe-
oxides and oxyanions, it is necessary to discuss the change of
mineral surfaces and oxyanions at different pH values. Man-
ceau and Charlet (1994) reported that the changes in surface
structure of HFO, goethite, and akaganeite led to different
ratios of edge and corner sharing surface complexes for arse-
nate, selenate, and selenite at acidic pH. Peak (2006) reported
that adsorption mechanisms of selenium oxyanions on
mineral surfaces are strongly influenced by sorbent surface
properties, and observed that selenite forms outer- and inner-
sphere complexes on hydrous aluminum oxides, while
hydrous aluminum oxides form outer-sphere complexes with
sulfate and selenate, but inner-sphere monodentate surface
complexes are formed between sulfate and selenate and
a-Al2O3. The overall conclusion is that selenite forms inner-
sphere complexes on aluminum oxide minerals and selenate
there have been few studies using XAS to examine the inter-
actions of amorphous binary oxide systems and oxyanions.
The objectives of this study were to evaluate the adsorp-
tioncapacityand kinetics ofseleniumoxyanionsby the binary
oxide systems, and to understand the interaction between
selenium oxyanions and the adsorbents using XAS.
2. Materials and methods
binary oxide systems
Preparation and characterization of the
The binary oxide systems of Al(III) or Fe(III) mixed with X-ray
noncrystalline SiO2 were carried out in 1-L polypropylene
bottles (with caps). The silica (SiO2) purchased from Cab-O-Sil
M5 (Cabot Corp., Tuscolca, IL) was an X-ray noncrystalline and
fine silica with BET surface area of 200 ? 25 m2g?1(Kuan et al.,
2004). Before synthesis of the binary oxide systems, the
suspension of SiO2was aged at 25?C under an N2atmosphere
for 24 h. The concentrations of binary oxide suspensions were
5 ? 10?3M for Al(NO3)3or Fe(NO3)3mixed with 1 g L?1SiO2at
pH 5.0, monitoring with pH-stat (TIM865 Titration Manager,
Radiometer Analytical). The suspensions of the binary oxide
systems were synthesized at 25?C for sorbents in the study,
adjusted to the desired pH by NaOH and HNO3solutions, and
shaken at 300 rpm for 24 h. After the synthesis process,
a subsample collected from the suspension was centrifuged
(Hitachi, 18PR-52) at 21,400 ? g for 15 min. The precipitate was
washed several times with double deionized water (DDW) and
freeze-dried (Millitorr Elemech FD-101) prior to X-ray and
specific surface area analyses.
The precipitates preparedat variouspH werecharacterized
by X-ray diffraction (XRD). The samples were examined with
an X-ray diffractometer (Rigaku Miniflex) using CuKa radia-
tiongeneratedat 30 KV and15 mA.The specificsurfacearea of
the binary oxide systems was determined by a specific area
(Micrometry, Tristar, FL), using the N2-BET method. The
Zetasizer (Malvern Zetasizer 3000 HS) was used to determine
the PZC by the zeta potential function of pH in 0.01 M NaNO3.
All adsorption experiments were carried out at 25?C, and the
background electrolyte concentrations were 0.1, 0.01 and
0.001 M NaNO3solution, as monitored by pH-stat. Sample pH
was adjusted by HNO3or NaOH for the entire reaction process.
After all reactions, the suspensions were passed through
a membrane filter (Millipore filter, 0.22 mm), and then the
filtrates were analyzed by inductively coupled plasma atomic
2000DV). The selenite solutions had already been analyzed by
ion chromatography (IC), and we observed no selenite
oxidizing to selenate during that operation. Besides that, our
sample solutions were analyzed as soon as possible and all
samples were kept in a 4?C refrigerator. Our XANES results
showed that no oxidizing occurs. Thus, analyzing the solu-
tions by ICP-AES is more efficient.
Perkin Elmer, Optima
2.3. Equilibrium adsorption
For the pH-edge experiment, the initial concentration of
selenium oxyanions was 1 mM, and the pH of the binary oxide
suspensions was maintained within the range of 2–10. The
adsorption isotherm was studied by 0–3 mM selenium oxy-
anions reacted with the binary oxide systems at 25?C and at
pH 5 for 24 h. The isothermal results were fitted by Langmuir
isotherm, described by Eq. (1):
1 þ KLCe
where Ceis the equilibrium concentration (mg L?1) in solution,
Qeis the amount of selenite and selenate adsorbed at equi-
librium (mg g?1), and KLis the Langmuir constants related to
adsorption capacity and energy of adsorption, respectively.
For adsorption kinetic experiments, the initial concentration
of selenite was 1 mM, and selenate was 0.15 mM to react with
the binary oxide systems, individually. The suspensions were
set at pH 5.0 and shaken at 3000 rpm for 240 min. Subsamples
water research 43 (2009) 4412–4420
were taken at different times. The pseudo-second-order
kinetic model can be solved with Eq. (2). The kinetic rate
equation is expressed as follows:
dqt=dt ¼ k2
where qeqis the sorption capacity at equilibrium and qtis the
solid-phase loading of selenium at time t. k2 (g mg?1h?1)
represents the pseudo-second-order rate constant for the
kinetic model (Jang et al., 2003; Saha et al., 2004). By inte-
grating Eq. (2) with the boundary conditions of qt¼ 0 at t ¼ 0
and qt¼ qt at t ¼ t, the following linear equation can be
where V0(mg g?1h?1) is the initial sorption rate. Therefore,
the V0 and qeq values of kinetic tests can be determined
experimentally by plotting t/qtversus t.
2.5.X-ray absorption spectroscopy analysis
Samples for XAS analysis were obtained by mixing the binary
oxide systems suspensions with 10?3M selenium oxyanions
solution at pH 5.0 (adjusted by NaOH or HNO3solution) in
0.01 M NaNO3 solution. After being shaken for 1 day, the
sample suspensions were centrifuged at 21,400 ? g for 20 min,
and the paste washed several times by DDW to remove excess
salts. XAS data included X-ray absorption near-edge struc-
tures (XANES) and extended X-ray absorption fine structure
(EXAFS) spectroscopy at the Se K-edge (12.658 keV) were
collected on a Wiggler 20 beamline BL-17C at the National
Synchrotron Radiation Research Center (NSRRC), Hsin-Chu,
Taiwan. The electron storage ring operated at 1.5 GeV with
a fixed current of 250 mA. The K-edge spectra of Se reacted
with the binary oxide systems were recorded in transmission
and fluorescence mode, respectively, and absorbance of the
incident X-rays was collected by the ionization chambers It
and If(Lytle Detector) (Lytle et al., 1984), separately. In order to
attenuate scattered principle energy X-rays from entering the
fluorescence detector, soller slits, and an absorbing filter (As
for Se atom) were placed between sample and Ifdetector. The
experiments were carried out from 12.63 to 12.70 keV at
ambient temperature. In addition, the Au L3-edge spectrum
was monitored by the Irchamber simultaneously with Itand If
chambers, serving as the reference to calibrate energy shift
due to monochromator drifts.
The XAS data was normalized with careful energy cali-
bration by the standard strategies (Koningsberger and Prins,
1998). In addition, the IFEFFIT package, including Athena and
Artemis was used for various data processing operations and
analysis of the EXAFS spectra (Newville, 2001; Ravel and
Newville, 2005). All of the XAS were normalized by Athena in
the ifeffit package, transformed from electron energy to
photoelectron wave number (k, A˚?1) and weighted by k3to
generate k3ø(k) spectra to better demonstrate the contribution
of different shells of the EXAFS oscillation function. Fourier
transformation (FT) was further processed to produce radial
structure functions (RSF). The FEFF8.2 program (Cernohorsky,
1960; Rehr and Albers, 2000; Ankudinov et al., 1998) was
employed to create theoretical phases and amplitude func-
tions representing photoelectron scattering paths of Se-O by
inputting standard structure parameters of reference Na2SeO4
and NaHSeO3from the Inorganic Crystallography Standard
Databases (Inorganic Crystal Structure Database, 1981). The
model fitting was conducted by the Artemis program,
producing structure parameters such as coordination number
(CN), interatomic distance (R), and Debye–Waller factor (s2).
The phase and amplitude shift of EXAFS spectra between local
structure of the experimental sample and the theoretical
atomic model were adjusted by Debye–Waller factors and
amplitude reduction factor. A fixed value of 0.85 was used for
the global amplitude reduction factor (S0
3.Results and discussion
3.1.Characterization of the binary oxide systems
X-ray diffraction patterns show that both Al(III)/SiO2 and
Fe(III)/SiO2are X-ray noncrystalline mixed oxides at pH 5.0.
The BET surface area of Al(III)/SiO2is decreased to less than
that of X-ray noncrystalline silica oxide (SiO2), however, that
of Fe(III)/SiO2is increased (Table 1). Stumm and Wollast (1990)
reported that a surface-coordinated metal ion, such as Cu(II)
or Al(III), can block an oxide surface group, and the passive Al
hydroxide inhibited the corrosion of the SiO2(s)surface by OH?
ions, Al-oxide particles may block the surface pores of silica
oxide. Meng and Letterman (1993, 1996) reported discrete
Fe(OH)3particles formed in Fe(OH)3/SiO2, and discrete Fe(OH)3
particles increase the total surface area of the binary oxide
systems. The surface charge of both binary oxide systems
decreases with increasing pH value because of increasing OH?
ions (Fig. 1). The zeta potential of Fe(III)/SiO2is less than Al(III)/
SiO2, and Meng and Letterman (1993) reported the zeta
potential of Fe(III)/SiO2as the overall results contributed by
negatively charged SiO2and positively charged Fe(OH)3. The
silicate adsorption on Fe(OH)3 may decrease the surface
potential of Fe(OH)3. Stumm (1992) reported that the pHzpc
range of aluminum and iron oxides is 7.8–9.1, however, the
interaction of Al with amorphous silica suggests Al has
covered the SiO2particles, and the negative silica sites could
not cause the pHzpcof Al(III)/SiO2to be less than 9.1 (Kuan
et al., 2004), thus the pHzpcof the Fe(III)/SiO2system is lower
Table 1 – The surface area and pore size of the binary
oxide systems synthesized at pH 5.0 and X-ray
noncrystalline silica oxide.
System BET surface
Al(III)/SiO2, pH 5.0
Al(III)/SiO2, pH 8.0
Fe(III)/SiO2, pH 5.0
Fe(III)/SiO2, pH 8.0
The BET surface area of SiO2is 200 m2g?1(Kuan et al., 2004).
water research 43 (2009) 4412–4420
For Al(III)/SiO2and Fe(III)/SiO2prepared at 25?C, pH 5.0, the
amount of selenium oxyanions adsorbed by the binary oxide
systems surface is dependent on the anionic characteristics
and the type of adsorption sites. To determine the optimum
pH for adsorption of selenium over the binary oxide systems,
adsorption of selenium oxyanions as a function of pH was
studied. Removal of selenium in the range of pH 2–12 is shown
in Fig. 2. The removal selenium efficiencies of both binary
oxide systems decreases with increasing pH value. However,
the amphoteric effect of aluminum and negative surface
charge of SiO2reduce the amounts of removal selenium oxy-
anionsamount intheAl(III)/SiO2systemat pH < 4.0.Whenthe
pH value is increased, the OH?ions can compete with sele-
nium oxyanions forfixed surface adsorption sites. The highest
removal efficiency of selenite and selenate for both binary
oxide systems is around pH 4.0, while Al3þis dissolved in the
range of pH < 4.0. Ghosh et al. (1994) observed the removal
was 100% at a pH of 5.5 or less for all initial concentrations of
selenite on hydrous alumina, and virtually all of the selenite
was adsorbed, even at pHs higher than the PZC (8.1) of
alumina. However, the adsorption of selenate was more
strongly dependent on pH than that of selenite. Jordan et al.
(2009a) reported sorption of selenite onto magnetite decreases
when the pH increases. Jordan et al. (2009b) reported that
selenite adsorption onto hematite depended on the pH of the
suspension, and the maximum sorption of selenite is in the
acidic pH range. Sorption of selenite onto hematite decreases
with increasing pH.
At any given pH value, both the binary oxide systems have
higher affinities for selenite than for selenate, and the sele-
nium oxyanion removal capacity of the binary oxide systems
is strongly influenced by the surface charge and the environ-
mental pH. This is consistent with the results derived from
Fig. 1 – Zeta potential of the different binary oxide systems
changes at different pH in 0.01 M NaNO3.
Fig. 2 – The pH adsorption edge of (a) selenite, (b) selenate on Al(III)/SiO2and (c) selenite, (d) selenate on Fe(III)/SiO2at pH 5.0
in different concentration of the background electrolytes.
water research 43 (2009) 4412–4420
distinguishing various anion affinities based on the first
acidity constant (Hayes, 1987; Hayes et al., 1988). Furthermore,
removal of selenite remained effective when the system pH is
high, suggesting that strong complexation occurs between
selenite and both binary oxide surfaces.
The pH adsorption envelope of selenate on both binary
oxide systems is affected by the concentration of the back-
ground electrolyte, as shown by experimental evidence, but
selenite is not. The selenite and selenate adsorption by Fe-
oxides both formed inner-sphere complexes, which was
provided by the EXAFS (Hayes et al., 1988). There is no doubt
that selenium adsorption on Fe(III)/SiO2is not influenced by
theelectrolytesolutions. Peak and Sparks(2002)indicated that
selenate forms only inner-sphere surface complexes on
hematite but forms a mixture of outer- and inner-sphere
surface complexes on goethite and HFO. The selenate
adsorption mechanism is strongly affected by both pH and
ionic strength. Although the adsorption envelope of both
selenite and selenate are dependent on electrolyte concen-
tration for the Al system in previous studies, Ghosh et al.
(1994) observed the removal was 100% at a pH of 5.5 or less for
all initial concentrations of selenite on hydrous alumina. At
very low ratios, virtually all of the selenite was adsorbed, even
at pHs higher than the PZC (8.1) of alumina. And the adsorp-
tion of selenate was more strongly dependent on pH than that
of selenite. Peak (2006) and Peak et al. (2006) reported that
yaluminosilicate formed both inner-sphere and outer-sphere
complexes, selenate formed outer-sphere complexes with
complexes with a-Al2O3. It may be argued that selenium
adsorption by Al(III)/SiO2 formed outer-sphere or inner-
sphere complexes. However, selenium adsorption may be
influenced by the electrolyte solutions in Al(III)/SiO2. From our
EXAFS results, selenite on Al(III)/SiO2 formed inner-sphere
complexes. Although Fig. 2 showed that selenite adsorption
on Al(III)/SiO2was affected by electrolyte concentrations in
the range of pH 4–6, the effect was not significant with
increasing pH. Hayes (1987) reported that selenite adsorbed
onto the adsorbent, goethite (a-FeOOH), forms an inner-
sphere complex, and selenate forms a weak outer-sphere, and
thenselenateonadsorbents isaffectedby lytropiceffectinthe
electric double layer theory. Fig. 2 shows that the removal
efficiencies of selenite and selenate on Al(III)/SiO2are greater
than on Fe(III)/SiO2in the Al(III)/SiO2system. When the basic
solution was added into the Al(III)/SiO2system, Al formed
complexes with SiO2to increase the adsorption sites on the
negatively charged surface of SiO2. Thus a pH 5.0 and 0.01 M
electrolyte solution was chosen for isothermal, kinetics, and
To investigate the adsorption capacity, a series of selenite and
selenate solutions were shaken with both the binary oxide
systems that were prepared at pH 5.0 for 24 h. Adsorption data
is fitted by Langmuir isotherm well (r2> 0.858, P < 0.01 (n ¼ 7))
(Fig. 3, Table 2). The correlation coefficients suggest that the
Langmuir isotherm model is suitable for describing the
adsorption equilibrium of selenite and selenate by the binary
oxide systems. The fitting data shows that the maximum
adsorption capacity (Qmax) for selenite and selenate by Al(III)/
SiO2was 32.7 and 11.3 mg/g, and the Qmaxfor selenite and
selenate by Fe(III)/SiO2are 20.4 and 2.4 mg g?1, respectively. In
consideration of the adsorption capacities for selenium
species, Al(III)/SiO2and Fe(III)/SiO2appear to be much higher
than coarse Al- or Fe-oxide coated sand (Lo and Chen, 1997;
Fig. 3 – Adsorption isotherms of (a) selenite and (b) selenate
on the binary oxide systems at pH 5.0, 25 8C in 0.01 M
NaNO3as background electrolytes.
Table 2 – The isotherm adsorption parameters of selenite
and selenate on the binary oxide systemsat pH 5.0 for the
Langmuir isothermal fitting model.
Adsorbate KL(L mg g?1)aQmax(mg g?1)b
5.50 ? 10?2
7.06 ? 10?2
1.68 ? 10?1
3.03 ? 10?1
a Langmuir constants (L mg g?1).
b The maximum adsorption capacity (mg g?1).
** P < 0.01 (n ¼ 7).
water research 43 (2009) 4412–4420
Kuan et al., 1998). Both of the binary oxide systems can be
attributed to fine particles that increase the contact between
selenium species and the adsorbents.
The adsorption capacity of selenite on both binary oxide
systems is more than selenate, which is consistent with the
results derived from distinguishing various anion affinities
based on the first acidity constant (Hayes, 1987; Hayes et al.,
1988). The chemical characteristics and the geometry struc-
ture of selenite are similar to phosphate, whereas selenate is
similar to sulfate (Hayes, 1987; wijnja and Schulthess, 2000),
and adsorbed selenite on the both binary oxide systems is
greater than selenate.
The Al(III)/SiO2system shows a much greater capacity for
selenite and selenate compared with the Fe(III)/SiO2system,
resulting from high affinity of the more positive charge of the
Al(III)/SiO2 system for selenium oxyanions (Fig. 1). The
precipitation of iron particles on SiO2cannot increase any
adsorption site, and the negatively charged surface of SiO2
increases the repulsion between selenium oxyanions and
SiO2. Thus the Qmaxof selenium on Al(III)/SiO2is much higher
than that of Fe(III)/SiO2(Table 2). As for the mechanisms by
which selenite and selenate are adsorbed on the binary oxide
systems, previous spectroscopic studies show that selenium
is adsorbed to Al- and Fe-oxides by forming inner-sphere
surface complexes by ligand exchange with hydroxyl groups
at the mineralsurface(Peak, 2006). Tostudy themechanismof
the adsorption process, we subjected the samples to XANES
and EXAFS spectra of selenium adsorbed to the binary oxide
The kinetics of adsorption that describe the solute adsorption
rate governing the residence time of the sorption reaction is
one of the important characteristics that define the efficiency
of adsorption. Fig. 4 shows the adsorption data of selenium by
the both binary oxide systems at different time intervals and
the simulation of the pseudo-second-order kinetic model.
Adsorption of both selenite and selenate on the binary oxide
systems approach pseudo-equilibrium more rapidly than
those on Al or Fe coatings on quartz sands (Lo and Chen, 1997;
Kuan et al., 1998). For selenite on Al(III)/SiO2and Fe(III)/SiO2,
95% removal of selenite was achieved within 2 h of contact,
and the adsorption equilibrium was approached (Fig. 4a). The
removal percentage for selenate individually reached 99 and
96% on Al(III)/SiO2and Fe(III)/SiO2in 30 min, and the adsorp-
tion approached equilibrium too (Fig. 4b). The greater surface
area and smaller particle size of both binary oxide systems
raise the probability of the adsorption reactions. The kinetics
data has been fitted by the zero-, first-, pseudo-second-,
second-order, parabolic, and Elovich models, evaluating the r2
and P-value, and the total fitted data was not shown. The
pseudo-second-order kinetic model is the best model to
describe the kinetic data in the binary oxide systems. Table 3
shows the determination coefficients (r2) and the other
parameters obtained from the pseudo-second-order kinetic
model by the plot of t versus t/qtto determine the V0and qeq
values for all the media. The pseudo-second-order model fits
the kinetic data of the both binary oxide systems very well
(r2> 0.999, P < 0.001, n ¼ 18). The initial adsorption rate of
selenite over Al(III)/SiO2is 6.25 mg g?1h?1, which is less than
that of selenate (57.5 mg g?1h?1), and the rate of selenite and
selenateoverFe(III)/SiO2is8.36and4.89 mgg?1h?1. Thefitting
data shows that Al(III)/SiO2has a higher affinity for selenate,
and the adsorption capacity of selenate is greater than sele-
nate on Fe(III)/SiO2. Unlike selenite, in which the supernatant
pH is stable, the pH increases from 5.0 to 8.0 for the selenate
Fig. 4 – Adsorption kinetics of (a) selenite and (b) selenate
on the binary oxide systems at pH 5.0, at 25 8C in 0.01 M
NaNO3as background electrolytes.
Table 3 – Kinetic parameters of a pseudo-second-order
kinetic model fitting selenium oxyanions adsorption by
the binary oxide systems at pH 5.0.
1.24 ? 10?2
3.04 ? 10?2
* V0is the initial sorption rate.
*** P < 0.001 (n ¼ 18).
water research 43 (2009) 4412–4420
equilibrium solution, indicating that hydroxide ions are
released in the adsorption process. Since selenite strongly
bonds to the metal oxide surface via the formation of stronger
inner-sphere complexes, selenite takes more time to reach
equilibrium in the slow process (Balistrieri and Chao, 1987;
Hayes, 1987; Neal et al., 1987; Zhang and Sparks, 1990; Scott
and Morgan, 1996).
3.4.XAS data analysis
From the adsorption reaction of the XANES, there is no change
in oxidation state following interaction between the selenium
species and the binary oxide systems. Fig. 5 and Table 4 show
the results of Se K-edge EXAFS conducted on samples of
selenite adsorbed on the binary oxide systems at pH 5.0. The
radial structure functions (RSFs) of the Fourier-transformed
data are shown in Fig. 5a,b. In all the selenite samples, there is
clear evidenceof twoshells in the RSFs:a firstshellSe-O (fitted
with 3O at 1.72 A˚) and a second shell Se-Al or Se-Fe in both
binary oxide systems (Fig. 5a). The Se-O distance is in good
agreement with previous EXAFS experiments conducted on
selenite, which have shown Se-O bond distances of 1.68–
1.72 A˚for aqueous and adsorbed selenite (Manceau and
Charlet, 1994; Peak et al., 2006; Peak, 2006). Based on simple
geometrical constraints, the best assignment of the bonding
environmentforseleniteonAl(III)/SiO2(Se-Alwas fittedwith 2
Al at 3.22 A˚) at pH 5.0 is a bidentate–binuclear surface complex
(the range of the bond length is w3.2 A˚). This is in good
agreementwith previousstudies (Hayes et al.,1987) of selenite
adsorption on other aluminum-bearing mineral surfaces,
which also reported bidentate–binuclear surface complexa-
tion. The presence of inner-sphere selenate on goethite is
consistent with the work of Manceau and Charlet (1994), and
they reported a similar Se-Fe bond distance of 3.29 A˚, as
compared to 3.31 A˚in this work (Table 4). On hydrous ferric
oxide (HFO), however, they found that their data were best
described with 0.4 Fe at 2.80 A˚ and 1.8 Fe at 3.29 A˚. They
attributed the distance of 2.80 A˚ to a bidentate mononuclear
surface complex and the distance of 3.29 A˚to that of a biden-
tate–binuclear surface complex. The result of selenite on
Fe(III)/SiO2shows the distance of 3.05 A˚ to form a bidentate
mononuclear surface complex.
The local structure of selenate on the binary oxide
systems was conducted by EXAFS fitting (Table 4 and
Fig. 5b). In all the selenate samples, the first shell, Se-O, in
the RSFs (fitted with 4O at 1.66 A˚) is consistent with
Fig. 5 – RSF profiles of EXAFS signal for selenium species
standard: (a) selenite and (b) selenate sorbed on the binary
oxide systems at pH 5.0. Solid lines and open circles
represent fitted and experimental data, respectively.
Table 4 – Structure parameters of selenium chemicals and selenium oxyanions adsorbed on the binary oxide systems
derived from the EXAFS spectrum fitting.
1st shell: Se-O
2nd shell: Se-metal
R (A˚) CN
a Amplitude reduction factor.
b Interatomic distance (?0.02 A˚).
c Coordination number (?20%).
d Debye–Waller factor (disorder parameter).
water research 43 (2009) 4412–4420
previously published Se-O distances of aqueous and adsor-
bed selenate (Hayes et al., 1987; Peak and Sparks, 2002). The
CN ¼ 1) of the Se-Al distance in the selenate sorption
samples suggest that the coordination environment of
selenate on the Al(III)/SiO2 surface is an inner-sphere
surface complex in the study. However, the Se-Fe bond
length (3.04 A˚) shows that selenate on Fe(III)/SiO2 forms
bidentate mononuclear complexes (Hug, 1997; Peak and
Sparks, 2002). The EXAFS fitting data shows that selenate
binding to Fe(III)/SiO2 is stronger than to Al(III)/SiO2.
According to the EXAFS principles, the surrounding atoms at
a distance from the X-ray absorbing atom is the bond
distance. The longer the bond distance is, the weaker the
interaction between the center and surrounding atoms. The
Se-metal bond distances for selenite and selenate for the
Fe(III)/SiO2 system are virtually identical. The fit quality
estimated accuracy of bond distances is ?0.02 A˚. Peak and
Sparks (2002) indicated selenate forms only inner-sphere
surface complexes on hematite but forms a mixture of
outer- and inner-sphere surface complexes on goethite and
HFO. The EXAFS results showed that selenate on Fe(III)/SiO2
formed inner-sphere complexes. Although both selenate
and selenite form inner-sphere complexes on the binary
oxide systems at pH 5.0, the data suggests that selenite on
the binary oxide systems forms stronger surface complexes
number(3.32 A˚ and
According to the results, we suggest that Al- or Fe-oxide can
modify the negative surface charge of SiO2in the binary oxide
systems, and apply the binary oxide systems to anion
that Al(III)/SiO2 and Fe(III)/SiO2 are efficient at removing
selenium oxyanions from aqueous solutions. However, the
stronger association between Al(III) and the SiO2surface cau-
ses the overall surface charge to be more positive than Fe(III)/
SiO2, and the selenium adsorption capacity of Al(III)/SiO2to be
more than Fe(III)/SiO2. At pH 5.0, selenite on Al(III)/SiO2and
Fe(III)/SiO2forms inner-sphere complexes, including biden-
tate and monodentate ones, respectively. However, selenate
on Al(III)/SiO2 forms weaker inner-sphere monodentate
complexes, the same as selenate on Fe(III)/SiO2. According to
the bond distance of selenate, the higher affinity between
selenateand Fecauses theshorterbonddistance.The findings
obtained in the present study are of fundamental significance
in advancing the frontiers of knowledge on geochemistry of
the binary oxide systems.
We thank the National Science Council, Taiwan for the
financial support (NSC 91-2313-B-002-361, 92-2313-B-002-090,
93-2313-B-002-008, 94-2313-B-002-091, 95-2313-B-002-041 and
96-2313-B-002-021) of nano-particle projects.
r e f e r e n c e s
Ankudinov, A.L., Ravel, B., Rehr, J.J., Conradson, S.D., 1998. FEFF8.
20: real-space multiple-scattering calculation and
interpretation of x-ray-absorption near-edge structure. Phys.
Rev. B 58, 7565–7576.
Balistrieri, L., Chao, T.T., 1987. Adsorption of selenium by
amorphous iron oxyhydroxide and manganese dioxide.
Geochim. Cosmochim. Acta 54, 739–751.
Cernohorsky, M., 1960. The ratio method for absolute
measurements of lattice parameters with cylindrical cameras.
Acta Crystallogr 13, 823–826.
Frankenberger Jr., W.T., Benson, S. (Eds.), 1994. Selenium in the
Environment. Marcel Dekker, New York, p. 472.
Ghosh, M.M., Cox, C.D., Yuan-Pan, J.R., 1994. Adsorption of
selenium on hydrous alumina. Environ. Prog 13, 79–88.
Goh, K.H., Lim, T.T., Dong, Z., 2008. Application of layered double
hydroxides for removal of oxyanions: a review. Water Res. 42,
Hayes, K.F., 1987. Equilibrium, spectroscopic and kinetic studies
of ion adsorption at the oxide/aqueous interface. PhD
dissertation. Stanford University, CA.
Hayes, K.F., Roe, A.L., Brown Jr., G.E., Hodgson, K.O., Leckie, J.O.,
Parks, G.A., 1987. In situ X-ray adsorption study of surface
complexes: selenium oxyanions on a-FeOOH. Science 238,
Honeyman, B.D., 1984. Cation and anion adsorption at the
oxide/solution interface in systems containing binary
mixtures of adsorbents. PhD dissertation. Stanford
University, Stanford, CA.
Hug, S.J., 1997. In Situ Fourier transform infrared measurements
of sulfate adsorption on hematite in aqueous solutions. J.
Colloid Interface Sci. 188, 415–422.
Inorganic Crystal Structure Database (ICSD), 1981. In:
Kurchhof, A., Pebler, A., Warkenten, E. (Eds.), Gmelin-Institut
fu ¨r Anorganische Chemie. Springer-Verlag, Berlin.
Jacobs, L.W., 1989. Selenium in Agriculture and the Environment.
American Society of Agronomy, Inc., Madison, WI.
Jang, M., Shin, E.W., Park, J.K., Choi, S.I., 2003. Mechanisms of
arsenate adsorption by highly-ordered nano-structured
silicate media impregnated with metal oxides. Environ. Sci.
Technol 37, 5062–5070.
Jordan, N., Lomenech, C., Marmier, N., Giffaut, E., Ehrhardt, J.J.,
2009a. Sorption of selenium(IV) onto magnetite in the
presence of silicic acid. J. Colloid Interface Sci. 329, 17–23.
Jordan, N., Lomenech, C., Marmier, N., Giffaut, E., Ehrhardt, J.J.,
2009b. Competition between selenium (IV) and silicic acid on
the hematite surface. Chemosphere 75, 129–134.
Koningsberger, D.C., Prins, P., 1998. X-ray Absorption: Principles,
Applications, Techniques of EXAFS, SEXAFS, and XANES. John
Wiley & Sons, New York.
Kuan, W.H., Lo, S.L., Wang, M.K., Lin, C.F., 1998. Removal of Se(IV)
and Se(VI) from water by aluminum-oxide-coated sand. Water
Res. 32, 915–923.
Kuan, W.H., Lo, S.L., Wang, M.K., 2004. Modeling and
electrokinetic evidences on the processes of the Al(III)
sorption continuum in SiO2(s)suspension. J. Colloid Interface
Sci. 272, 489–497.
Lenz, M., Van Hullebusch, E.C., Hommes, G., Corvini, F.F.X.,
Lens, P.N.L., 2008. Selenate removal in methanogenic and
sulfate-reducing upflow anaerobic sludge bed reactors. Water
Res. 42, 2184–2194.
Lo, S.L., Chen, T.Y., 1997. Adsorption of Se(IV) and Se(VI) on an
iron-coated sand from water. Chemosphere 35, 919–930.
Lytle, F.W., Greefgor, R.B., Sandstrom, D.R., Marques, E.C.,
Wong, F., Spiro, C.L., Huffman, G.P., Huggins, F.E., 1984.
Measurement of soft X-ray absorption spectra with
water research 43 (2009) 4412–4420
a fluorescent ion chamber detector. Nucl. Instrum. Methods
Phys. Res. A 226, 542–548.
Manceau, A., Charlet, L., 1994. The mechanism of selenate
adsorption on goethite and hydrous ferric oxide. J. Colloid
Interface Sci. 168, 87–93.
Meng, X., Letterman, R.D., 1993. Effect of component oxide
interaction on the adsorption properties of mixes oxides.
Environ. Sci. Technol 27, 970–975.
Meng, X., Letterman, R.D., 1996. Modeling cadmium and sulfate
adsorption by Fe(OH)3/SiO2mixed oxides. Water Res. 30,
Mikkelsen, R.L., Page, A.L., Bingham, F.T., 1989. Factors affecting
selenium accumulation by agricultural crops. In: Jacobs, L.W.
(Ed.), Selenium in agriculture and the environment. Soil
Science Society of America. SSSA Special Publication 23,
Madison, WI, pp. 65–94.
Neal, R.H., Sposito, G., Holtzclaw, K.M., Traina, S.J., 1987.
Selenite adsorption on alluvial soil: I. Soil composition and
pH effects; II. Solution composition effects. Soil Sci. Soc. Am.
J 51, 1161–1169.
Newville, M., 2001. IFEFFIT: interactive XAFS analysis and FEFF
fitting. J. Synchrotron Radiat 8, 322–324.
Peak, D., 2006. Adsorption mechanisms of selenium oxyanions at
the aluminum oxide/water interface. J. Colloid Interface Sci.
Peak, D., Sparks, D.L., 2002. Mechanisms of selenate adsorption
on iron oxides and hydroxides. Environ. Sci. Technol 36,
Peak, D., Saha, U.K., Huang, P.M., 2006. Selenite adsorption
mechanisms on pure and coated montmorillonite: an
EXAFS and XANES spectroscopic study. Soil Sci. Soc. Am. J
Ravel, B., Newville, M., 2005. ATHENA, ARTEMIS, HEPHAESTUS:
data analysis for X-ray absorption spectroscopy using IFEFFIT.
J. Synchrotron. Radiat. 12, 537–541.
Rehr, J.J., Albers, R.C., 2000. Theoretical approaches to x-ray
absorption fine structure. Rev. Mod. Phys. 72, 621–654.
Saha, U.K., Liu, C., Kozak, L.M., Huang, P.M., 2004. Kinetics of
selenite adsorption on hydroxyaluminum– and
hydroxyaluminosilicate–montmorillonite complexes. Soil Sci.
Soc. Am. J 68, 1197–1209.
Scott, M.J., Morgan, J.J., 1996. Reactions at oxide surfaces. II.
Oxidation of Se(IV) by synthetic birnessite. Environ. Sci.
Technol 30, 1990–1996.
Sparks, D.L., 2003. Environmental Soil Chemistry, second ed.
Academic Press, Amsterdam.
Stumm, W., 1992. Chemistry of the Solid–Water Interface. John
Wiley & Sons, New York.
Stumm, W., Wollast, R., 1990. Kinetics of the surface-controlled
dissolution of oxide minerals. Rev. Geophys. 28, 53–69.
Wijnja, H., Schulthess, C.P., 2000. Vibration spectroscopy study of
selenate and sulfate adsorption mechanisms on Fe and Al
(hydr)oxide surfaces. J. Colloid Interface Sci. 229, 286–297.
Zhang, P., Sparks, D.L., 1990. Kinetics of selenate and selenite
adsorption/desorption at the goethite/water interface.
Environ. Sci. Technol 24, 1848–1856.
Zhang, N., Lin, L.S., Gang, D., 2008. Adsorptive selenite removal
from water using iron-coated GAC adsorbents. Water Res. 42,
Zhao, L., Zhao, G.H., Du, M., Zhao, Z.D., Xiao, L.X., Hu, X.S., 2008.
Effect of selenium on increasing free radical scavenging
activities of polysaccharide extracts from a Se-enriched
mushroom species of the genus. Eur. Food Res. Technol 226,
water research 43 (2009) 4412–4420