1
Electronic Supplementary Information
Extremely High Arsenic Removal Capacity for Mesoporous
Aluminium Magnesium Oxide Composites
Wei Li,‡a Dehong Chen,b Fang Xia,a,c Jeannie Z. Y. Tan,a,b Pei-Pei Huang,d Wei-Guo Song,d
Natalita M. Nursama,b and Rachel A. Caruso*a,b
a CSIRO Manufacturing, Clayton South, Victoria, 3169, Australia.
b Particulate Fluids Processing Centre, School of Chemistry, The University of Melbourne,
Melbourne, Victoria, 3010, Australia.
c School of Engineering and Information Technology, Murdoch University, Murdoch, West
Australia, 6150, Australia.
d Beijing National Laboratory for Molecular Sciences (BNLMS), Institute of Chemistry, Chinese
Academy of Sciences, Beijing, 100190, P. R. China.
‡ Present address: International Iberian Nanotechnology Laboratory (INL), Avenida Mestre José
Veiga, Braga, 4715-330, Portugal.
Corresponding author email: [email protected]
Electronic Supplementary Material (ESI) for Environmental Science: Nano.This journal is © The Royal Society of Chemistry 2015
2
Table S1 Physical properties of the mesoporous aluminium magnesium oxide composites calcined at 900 °C.
Sample SBET
[m2 g-1]a
PSD
[nm]b
Vsp
[cm3 g-1]c
Phase
meso-Al-900 172.8 10.80 0.45 γ-Al2O3
meso-90Al10Mg-900 174.9 10.84 0.47 γ-Al2O3 & MgAl2O4 spinel
meso-80Al20Mg-900 144.1 12.08 0.41 γ-Al2O3 & MgAl2O4 spinel
meso-70Al30Mg-900 91.6 13.52 0.31 γ-Al2O3 & MgAl2O4 spinel
meso-50Al50Mg-900 121.9 13.41 0.36 MgAl2O4 spinel & cubic MgO
meso-30Al70Mg-900 101.4 8.57 0.27 MgAl2O4 spinel & cubic MgO
a SBET = BET specific surface area obtained from nitrogen adsorption data in the P/P0 range from 0.05 to 0.20. b PSD = pore size distribution determined by using the BJH method from the adsorption branch. c Vsp = single point pore volume calculated from the adsorption isotherm at P/P0 = 0.98.
Table S2 As(V) adsorption kinetic parameters of the mesoporous aluminium magnesium oxide composites calcined at 400 °C.
Samples Pesudo-second-order kinetic parameters
k [g mg-1 min-1] qe [mg g-1] R2
meso-Al-400 1.02×10-4 147.71 0.999
meso-90Al10Mg-400 9.25×10-5 125.63 0.996
meso-80Al20Mg-400 7.14×10-5 178.57 0.998
meso-70Al30Mg-400 5.20×10-5 181.82 0.996
meso-50Al50Mg-400 5.76×10-5 165.84 0.998
meso-30Al70Mg-400 2.67×10-4 74.52 0.998
meso-Mg-400 1.62×10-4 207.90 0.999
Al-400 1.35×10-4 17.30 0.897
3
Table S3 As(V) adsorption isotherm parameters of the mesoporous aluminium magnesium oxide composites calcined at 400 °C.
Sample Highest adsorptioncapacity [mg g-1]a
Langmuir isotherm Freundlich isotherm
qm [mg g-1] b R2 KF 1/n R2
meso-Al-400 299.03 271.83 0.4057 0.9716 94.55 0.1806 0.8621
meso-90Al10Mg-400 430.92 333.29 0.2536 0.8485 100.43 0.2100 0.9437
meso-80Al20Mg-400 502.97 422.09 0.0590 0.8902 95.85 0.2456 0.9795
meso-70Al30Mg-400 487.56 438.80 0.0321 0.8916 89.69 0.2531 0.9838
meso-50Al50Mg-400 466.02 416.25 0.0292 0.8778 86.18 0.2486 0.9730
meso-30Al70Mg-400 396.06 587.45 0.0022 0.8630 18.33 0.4482 0.9212
meso-Mg-400 912.32 928.95 0.1296 0.9692 242.32 0.2346 0.8059
Al-400 72.20 72.48 0.0259 0.9120 15.42 0.2362 0.9635
a The highest adsorption capacity was achieved using an initial arsenic concentration of 1020 mg L-1.
Table S4 As(V) adsorption isotherm parameters of the mesoporous aluminium magnesium oxide composites calcined at 900 °C.
Sample Highest adsorption capacity [mg g-1]a
Langmuir isotherm Freundlich isotherm
qm [mg g-1] b R2 KF 1/n R2
meso-Al-900 79.21 76.36 0.5496 0.9748 38.29 0.1203 0.8421
meso-90Al10Mg-900 92.61 87.76 1.9477 0.9732 46.58 0.1134 0.8331
meso-80Al20Mg-900 70.94 58.45 1.4830 0.9254 32.65 0.1090 0.9206
meso-70Al30Mg-900 43.78 35.42 0.7144 0.8900 20.10 0.1036 0.8942
meso-50Al50Mg-900 97.56 91.32 0.5279 0.8189 46.71 0.1163 0.7739
meso-30Al70Mg-900 165.85 159.01 1.079 0.9087 71.23 0.1397 0.7750
a The highest adsorption capacity was achieved using an initial arsenic concentration of 1020 mg L-1.
Table S5 As(III) adsorption isotherm parameters of the mesoporous aluminium magnesium oxide composites calcined at 400 °C.
4
Sample Highest adsorptioncapacity [mg g-1]a
Langmuir isotherm Freundlich isotherm
qm [mg g-1] b R2 KF 1/n R2
meso-Al-400 114.85 120.49 0.1091 0.9815 15.37 0.3171 0.9497
meso-90Al10Mg-400 143.12 157.58 0.0192 0.9901 18.10 0.3336 0.8784
meso-80Al20Mg-400 390.59 424.26 0.0011 0.9320 2.19 0.8137 0.9861
meso-70Al30Mg-400 330.58 416.24 0.0018 0.9452 3.21 0.7288 0.9720
meso-50Al50Mg-400 240.78 394.71 0.0022 0.9919 4.18 0.6243 0.9918
meso-30Al70Mg-400 120.39 196.06 0.0019 0.9731 1.91 0.6264 0.9694
meso-Mg-400 812.84 848.72 0.1007 0.9214 162.8 0.2945 0.7509
a The highest adsorption capacity was achieved using an initial arsenic concentration of 820 mg L-1.
Table S6 Atomic ratios of the mesoporous aluminium magnesium oxide composites calcined at 400 °C before and after adsorption of 400 mg L-1 of As(V) at pH 3.0 or As(III) at pH 7.0 obtained from XPS analysis.
Sample Atomic ratio
Al Mg Al/Mg As
meso-Al-400 16.07 0 0
As(V)-meso-Al-400 25.85 0 4.61
As(III)-meso-Al-400 17.27 0 2.23
Meso-80Al20Mg-400 1.37 6.22 4.54 0
As(V)-meso-80Al20Mg-400 10.31 1.11 9.29 3.07
As(III)-meso-80Al20Mg-400 17.06 1.72 9.92 4.41
meso-Mg-400 0 10.39 0
As(V)-meso-Mg-400 0 20.6 5.9
As(III)-meso-Mg-400 0 13.86 3.38
Figure at 400 °C
S1 A photoC.
ograph of thhe as-preparred mesoporrous aluminnium magneesium oxide
5
es calcined
FPT7a
Figure S2 TParticle sizeTEM image70Al30Mg-9analyzed for
TEM imagee distributioes of (g) 900 and parr particle siz
es of (a) mon of (d) m
meso-Al-90rticle size dze distributio
meso-50Al50meso-50Al50
00, (h) medistribution on statistics
0Mg-400, (b0Mg-400, (eso-90Al10of (k) mesoby using so
b) meso-30e) meso-30
0Mg-900, (o-70Al30Moftware Nan
Al70Mg-400Al70Mg-40i) meso-80
Mg-900. Ovenomeasure 1
00 and (c) m00 and (f) m0Al20Mg-90er 150 nano1.25 and Ori
6
meso-Mg-4meso-Mg-400, (j) meoparticles wigin 9.0.
400. 400. eso-
were
Figure control s
Figure magnesi
0.0
Inte
ns
ity
(a.u
.)
a
S3 TEM imsample Al-4
S4 (a) Syncium oxide s
0.1
mage (a), X400 synthesi
chrotron SAamples with
0q (Å )-1
meso-Ameso-9meso-8meso-7meso-5meso-3meso-M
XRD patternized in the a
AXS and (bh varying M
0.2
Al-90090Al10Mg-90080Al20Mg-90070Al30Mg-90050Al50Mg-90030Al70Mg-900Mg-900
n (b) and thabsence of t
b) wide angMg/Al molar
10
Inte
ns
ity
(a.u
.)
b
0.3
00000
he nitrogen the P123 sof
le XRD patr ratios calci
20 302
gas sorptioft template.
tterns of meined at 900 °
0 40 502-Theta (deg
n isotherm
esoporous a°C.
0 60 70
meso-Al-900meso-90Al10meso-80Al20meso-70Al30meso-50Al50meso-30Al70meso-Mg-90
gree)
7
(c) of the
aluminium
0 80
00Mg-9000Mg-9000Mg-9000Mg-9000Mg-900
00
8
Figure S5 The full XPS survey of meso-Al-400, meso-80Al20Mg-400 and meso-Mg-400.
0 200 400 600 800 1000 1200 1400
meso-Al-400
meso-80Al20Mg-400
meso-Mg-400Mg 1s
O 1s
C 1s
Al 2p
Mg 2p
Al 2s
Mg 2s Mg 1s
Mg
Aug
er
O 2
s
Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
9
Figure S6 Nitrogen gas sorption isotherms of the mesoporous aluminium magnesium oxides with varying Mg/Al molar ratios calcined at 900 °C and (b) the corresponding pore size distribution derived from the adsorption branches based on the BJH model. Each subsequent curve is shifted up the y axis by 30 cm3 g-1 in (a) and 0.01 cm3 g-1 nm-1 in (b), for clarity.
0 20 40 60 80 100
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
meso-Mg-900meso-30Al70Mg-900meso-50Al50Mg-900meso-70Al30Mg-900
meso-80Al20Mg-900
meso-90Al10Mg-900meso-Al-900
Pore Diameter (nm)
dV
/dD
(cm
g n
m )
3-1
-1
b
Relative Pressure (P/P )0
Vo
lum
eA
dso
rbed
(cm
g
)3
-1
a
0.0 0.2 0.4 0.6 0.8 1.0
0
100
200
300
400
500meso-Mg-900meso-30Al70Mg-900meso-50Al50Mg-400meso-70Al30Mg-900meso-80Al20Mg-900meso-90Al10Mg-900meso-Al-900
Figure varying pH = 3.0
Figure SpermissiWiley-B
2A
s(V
) u
pta
ke (
mg
g
)-1
S7 AdsorptMg/Al mol
0, the initial
S8 Distribution [P. Rav
Blackwell, U
0
0
50
100
150
200
Eq
tion isotherlar ratios cal As(V) con
tion of As(Vvenscroft, HUSA 2009, p
200
uilibrium
rms of As(Valcined at 90centration r
V) and As(IH. Brammerpp. 25]. Cop
400
concentr
V) on meso00 °C. Experanged from
III) species r, K. Richarpyright 200
600 80
ration of A
oporous alumerimental co
m 10 to 1000
under differrds, Arsenic
09, John Wil
00 1000
As(V) (mg
minium maonditions: do mg L-1.
rent pH valuc Pollution:ley & Sons.
0 1200
L )-1
agnesium oxose = 0.5 g
ues. Reprod: A Global
meso-Al-900meso-90Al10meso-80Al20meso-70Al30meso-50Al50meso-30Al70
10
xides with L-1, initial
duced with Synthesis,
Mg-900Mg-900Mg-900Mg-900Mg-900
11
Figure S9 Effect of co-existing anions on the adsorption capacities of meso-Al-400, meso-80Al20Mg-400, meso-70Al30Mg-400 and meso-Mg-400 for As(V). Experimental conditions: dose = 0.5 g L-1, initial pH = 6.0, initial As(V) concentration = 50 mg L-1; initial concentration of co-existing anions: 10 mg L-1 of NO3
-, 200 mg L-1 of Cl-, 200 mg L-1 of SO42-, 50 mg L-1 of CO3
2-, 50 mg L-1 of SiO3
2- or 50 mg L-1 of PO43-.
0
20
40
60
80
100
120
meso-Al-400meso-80Al20Mg-400meso-70Al30Mg-400meso-Mg-400
0
20
40
60
80
100
As(V
) u
pta
ke (
mg
g )
As(V
) rem
oval p
erc
en
tag
e (%
)
As(V) only NO Cl SO CO SiO PO3- -
42-
32-
32-
43-
Co-existing anions
&&&&
-1
Figure 400 afte
Figure and As(400 and
20
Inte
nsi
ty (
a.u
.)
a
S10 TEM ier adsorption
S11 XRD pIII) at the in
d (c) meso-M
0 40
2-Theta (d
After ads
After ads
images of (n of 400 mg
patterns of tnitial pH of
Mg-400.
60 8
egree)
meso-Al-400
sorption of As(V)
sorption of As(III)
(a) meso-Alg L-1 of As(V
the adsorbenf 3.0 and 7.0
80 20
Inte
nsi
ty (
a.u
.)
b
l-400, (b) mV) solution
nts before a0, respectiv
40
2-Theta (degre
meso-8
After adsorp
After adsorp
meso-80Al20at an initial
and after advely: (a) me
60 80
ee)
0Al20Mg-400
ption of As(V)
ption of As(III)
Inte
nsi
ty (
a.u
.)
0Mg-400, (pH of 3.0.
sorption of so-Al-400,
20
2-T
c
(c) and (d) m
400 mg L-1
(b) meso-80
40 60
mes
After adsorption
After adsorption
Theta (degree)
MgOMg(OH)2
Unindexed p
12
meso-Mg-
1 of As(V) 0Al20Mg-
80
so-Mg-400
n of As(V)
of As(III)
phase
13
Figure S12 The full XPS survey of meso-Al-400, meso-80Al20Mg-400 and meso-Mg-400 after adsorption of 400 mg L-1 of (a) As(V) and (b) As(III). Experimental conditions: dose = 0.5 g L-1, initial pH for As(V) and As(III) adsorption is 3.0 and 7.0, respectively.
0 200 400 600 800 1000 1200 1400
meso-Al-400
meso-80Al20Mg-400
meso-Mg-400
O 1sC 1s
Mg 1s
Al 2p
As 3d
Al 2s
As Auger
Mg 2p N 1s
As 2p
Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
0 200 400 600 800 1000 1200 1400
meso-Al-400
meso-80Al20Mg-400
meso-Mg-400
O 1sC 1s
Mg 1s
Al 2p
As 3d
As Auger
Mg 2p
As 2p
Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
Mg 2s
Mg
Aug
er
a b
14
Figure S13 (a) Al 2p XPS peak of meso-Al-400 before and adsorption of As(V) and As(III), (b) As 3d peak of meso-Al-400 after adsorption of As(V), (c) As 3d peak of meso-Al-400 after adsorption of As(III), (d) Al 2p XPS peak of meso-80Al20Mg-400 before and after adsorption of As(V) and As(III), (e) As 3d and Mg 2p peak of meso-80Al20Mg-400 after adsorption of As(V), (f) As 3d and Mg 2p peak of meso-80Al20Mg-400 after adsorption of As(III), (g) and (h) As 3d and Mg 2p peak of meso-Mg-400 after adsorption of As(V) and As(III), respectively. Experimental conditions: dose = 0.5 g L-1, initial concentration of As(V) or As(III) = 400 mg L-1; initial pH for As(V) and As(III) adsorption is 3.0 and 7.0, respectively.
As shown in Figure S13a and d, for both meso-Al-400 and meso-80Al20Mg-400, the Al 2p peak shifts towards lower binding energy after adsorption of As(V), while the Al 2p peak moves slightly towards higher binding energy after As(III) adsorption. This difference in the Al 2p peak shift after adsorption of As(V) or As(III) over meso-Al-400 and meso-80Al20Mg-400 is likely to be due to the following reasons: 1. The initial pH is different for adsorption of As(V) or As(III) over the samples for the XPS measurement. In this work, the initial pH for As(V) and As(III) adsorption on both meso-Al-400
65 70 75 80 85
meso-Al-400
after adsorption of As(V)
after adsorption of As(III)
Al 2p
74.75 eV
74.81 eV
74.44 eV
Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
a
40 42 44 46 48 50 52Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
45.0 ev
As(V)-O45.9 eV
44.8 eVAs(III)-O
As 3db
40 42 44 46 48 50 52Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
As(V)-O
45.88 eV As 3dc
65 70 75 80 85
meso-80Al20Mg0-400
after adsorption of As(V)
after adsorption of AS(III)
74.83 eV
74.65 eV
74.51 eV
Al 2p
Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
d
40 42 44 46 48 50 52 54
Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
50.8 eVMg-O
44.7 eVAs(III)-O
45.6 eVAs(V)-O
44.96 eV
40 42 44 46 48 50 52 54Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
45.8 eVAs(V)-O
50.4 eVMg-O
e fAs 3d-Mg 2p
40 42 44 46 48 50 52 54Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
44.5 eVAs(III)-O
49.4 eV
Mg-O-H
50.3 eV
Mg-O
49.64 eV
40 42 44 46 48 50 52 54 56Binding Energy (eV)
Inte
nsi
ty (
a.u
.)
45.89 eVAs(V)-O
50.6 eVMg-O
g h
As 3d-Mg 2p
As 3d-Mg 2p As 3d-Mg 2p
15
and meso-80Al20Mg-400 is 3.0 and 7.0, respectively. The initial pH affects the different dominant arsenic speciation, As(V) (H2AsO4
-) and As(III) (H3AsO30), which in turn caused the
different shift of the Al 2p peak when forming Al-O-As bonds during the XPS measurement. 2. There are various models of surface complexes for the arsenic (including As(V) and As(III)) immobilization on aluminium oxy-hydroxides and other metal hydroxides/oxides. Most publications proved that arsenic can be adsorbed onto variable-charge adsorbent surfaces by inner-sphere complexation (ligand exchange to form chemical bonding) and/or outer-sphere complexation (electrostatic interaction or hydrogen bonding)(Geoderma 2001, 100, 303–319). The type of sorption mechanism for a particular ion is greatly affected by environmental factors such as pH and ionic strength. For the inner-sphere surface complex, there are four molecular configurations including bidentate binuclear, bidentate mononuclear, monodentate binuclear and monodentate mononuclear (Geochimica et Cosmochimica Acta 2012, 83, 205–216; Geochimica et Cosmochimica Acta 2001, 65, 1211-1217). For each configuration, acid-base or non-dissociative sorption could be included (Journal of Molecular Structure: THEOCHEM 2006, 762, 17–23; Geochimica et Cosmochimica Acta 2012, 83, 205–216). The specific mechanism would be elucidated by X-ray absorption spectroscopy (XAS) including EXAFS and XANES. Generally, the adsorption mechanism is considered to be different for As(V) or As(III) immobilized on the aluminium oxides, although there is no consensus on the exact mechanism applied for either As(V) or As(III). Most publications reported that As(V) predominantly forms inner-sphere bidentate binuclear complexes with the surface of aluminium oxide (Chemosphere 2004, 55, 1259-1270; Journal of Colloid and Interface Science 2001, 234, 204–216; Environmental Science & Technology 2005, 39, 5481-5487; Environmental Toxicology and Chemistry 2006, 25, 3118-3124; Journal of Colloid and Interface Science 2001, 235, 80–88; Applied Geochemistry 2013, 31, 79–83; Environmental Science & Technology 2011, 45, 9687–9692), while some papers claimed the inner-sphere monodentate mononuclear complexes (Environmental Science & Technology 2009, 43, 2537–2543) and co-existing inner-sphere, hydrogen bond and electrostatic interactions would dominate depending on the pH (Journal of Hazardous Materials 254– 255 (2013) 301– 309; Environmental Science & Technology 2006, 40, 7784-7789; Environmental Science & Technology 2005, 39, 3571-3579; Microporous and Mesoporous Materials 2014, 198, 101–114; Geochimica et Cosmochimica Acta 2008, 72, 1986–2004). The dominant complexes debated for As(III)-aluminium oxide surface complexation including outer-sphere (Chemosphere 2003, 51, 1001-1013; Journal of Colloid and Interface Science 2001, 234, 204–216; Soil Science Society of America Journal 70:2017–2027; Environmental Chemistry Letters 2013, 11, 289–294), non-dissociative inner-sphere complexes (Journal of Molecular
16
Structure: THEOCHEM 2006, 762, 17–23), inner-sphere bidentate binuclear (Geochimica et Cosmochimica Acta 2012, 83, 205–216; Microporous and Mesoporous Materials 2014, 198, 101–114), a combination of inner-sphere bidentate binuclear and bidentate mononuclear (Chemosphere 2014, 113, 151–157), as well as co-existing mixtures of several surface inner-sphere complexes and outer-sphere complexes (Journal of Colloid and Interface Science 2001, 235, 80–88). Because of the different kinds of surface complexes obtained through chemical bonding (inner-sphere configurations) and possible discernable intervention from electrostatic outer-sphere complexes, the XPS peak shift is likely different between As(V) at pH 3.0 and As(III) at pH 7.0, even if the As(V) and As(III) species are chemisorbed onto the surface of meso-Al-400 and meso-80Al20Mg-400 forming As-O-M bonds. The different kinds of surface complexes have been reported to cause either a positive or negative Al 2p peak shift in the Fe/Al hydroxide when adsorbing different speciations of As(V), indicating that the Al 2p peak shift is not necessarily consistent even though an As-O-Al bond is formed. (Journal of Hazardous Materials 2015, 293, 97–104)
Figure S14 FTIR spectra of the vacuum-dried samples before and after adsorption of arsenic: (a) meso-Al-400, (b) meso-80Al20Mg-400 and (c) meso-Mg-400. Experimental conditions: dose = 0.5 g L-1, initial concentration of As(V) or As(III) = 400 mg L-1; initial pH for As(V) and As(III) adsorption is 3.0 and 7.0, respectively.
4000 3500 3000 2500 2000 1500 1000 500
meso-80Al20Mg-400
After adsorption of 400 ppm of As(V)
After adsorption of 400 ppm of As(III)
Wavelength Number (cm )
Tra
nsm
itta
nce
-1
4000 3500 3000 2500 2000 1500 1000 500
meso-Al-400
After adsorption of 400 ppm of As(V)
After adsorption of 400 ppm of As(III)
Wavelength Number (cm )
Tra
nsm
itta
nce
-1
4000 3500 3000 2500 2000 1500 1000 500
meso-Mg-400
After adsorption of 400 ppm of As(V)
After adsorption of 400 ppm of As(III)
Wavelength Number (cm )-1
Tra
nsm
itta
nce
3700 cm-1
841 cm-1
819 cm-1
a b c
17
Figure S15 Effect of the adsorbent dose on the As(V) adsorption capacity of meso-Al-400. The mass of meso-Al-400 was 0.02 g, the initial pH was 3.0 ± 0.1 and the initial As(V) concentration is 420 mg L-1. The volume of As(V) solution was varied from 10 to 100 mL, resulting in the dose ranging from 0.2 g L-1 to 2.0 g L-1.
Please note that the adsorbent dose was 0.5 g L-1 throughout the work. Figure S15 demonstrates that 0.5 g L-1 is the optimal adsorbent dose with the highest adsorption capacity (260 mg g-1) for As(V). When the dose was increased to 1.0 and 2.0 g L-1, the adsorption capacity was decreased by 25.4% and 42.3%, respectively.
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.20
50
100
150
200
250
300
A
s(V
) u
pta
ke (
mg
g-1)
Dose (g L-1)