Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2 and the azo-dye Acid Orange 7

Article abstract of DOI:10.1016/j.cattod.2015.10.011

The adsorption and catalytic properties are reported of aluminosilicate nanotubes (NTs) of the imogolite-type both bare (IMO, (OH)₃Al₂O₃SiOH) and Fe-doped. Iron-doped NTs were obtained by either direct synthesis (Fe-x-IMO) or post synthesis loading (Fe-L-IMO), with 0.70?wt.% Fe (Fe-0.70-IMO) and 1.4?wt.% Fe content (both Fe-1.4-IMO and Fe-L-IMO). The samples were characterized by X-ray diffraction (XRD); N₂ sorption isotherms at ?196?°C; diffuse reflectance (DR) UV–vis spectroscopy; IR spectroscopy, and electrophoretic mobility in water (ζ-potential). In water, the intrinsic acidity of Fe(OH)Al groups, formed by isomorphic substitution (IS) of Fe³⁺ for Al³⁺, is only marginally different from that of parent Al(OH)Al groups, as shown by ζ-potential measurements. Substitution of one octahedral Al³⁺ site, however, gives rise to the formation of three Fe–O–Al groups and three Fe(OH)Al groups: such process modifies the adsorption and catalytic properties of NTs outer surface, as shown by adsorption of the azo-dye Acid Orange 7 (NaAO7) from water solutions and by AO7^? catalytic degradation by H₂O₂. Interaction with AO7^? anions in water occurs in different ways: (i) with bare IMO, AO7^? anions preferentially adsorb by electrostatic interaction with the positively charged surface, and by H-bonding; (ii) with Fe-1.4-IMO, IS Fe³⁺ cations act as coordination centres for N atoms in the AO7^? moiety, through a ligand displacement mechanism; (iii) with Fe-L-IMO, AO7^? adsorption is hindered by the presence of Fe₂O₃ clusters. Concerning AO7^? degradation by H₂O₂, the outer surface of proper IMO is much more reactive than Fe-doped IMO towards hydrogen peroxide, by forming very reactive Al-OOH bridges. IS Fe³⁺ species are probably not accessible to H₂O₂, and (the expected) Fenton reaction between Fe³⁺ and H₂O₂ is not observed.

Full text of DOI:10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
Catalysis Today xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Reactivity of bare and Fe-doped alumino-silicate nanotubes
(imogolite) with H2O2 and the azo-dye Acid Orange 7
Ehsan Shafiaa, Serena Espositob, Marco Armandia, Elnaz Bahadoria, Edoardo Garronea,
Barbara Bonellia,∗
a Department of Applied Science and Technology and INSTM Unit of Torino Politecnico, Corso Duca degli Abruzzi 24, 10129 Turin, Italy
b Department Civil and Mechanical Engineering, Università degli Studi di Cassino e del Lazio Meridionale, Via G. Di Biasio 43, 03043 Cassino, FR, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 24 July 2015
Received in revised form 7 October 2015
Accepted 9 October 2015
Available online xxx
The adsorption and catalytic properties are reported of aluminosilicate nanotubes (NTs) of the imogolite-
type both bare (IMO, (OH)3Al2O3SiOH) and Fe-doped. Iron-doped NTs were obtained by either direct
synthesis (Fe-x-IMO) or post synthesis loading (Fe-L-IMO), with 0.70 wt.% Fe (Fe-0.70-IMO) and 1.4 wt.%
Fe content (both Fe-1.4-IMO and Fe-L-IMO). The samples were characterized by X-ray diffraction (XRD);
N2 sorption isotherms at −196 ◦C; diffuse reflectance (DR) UV–vis spectroscopy; IR spectroscopy, and
electrophoretic mobility in water (␨-potential).
Keywords:
Adsorption
In water, the intrinsic acidity of Fe(OH)Al groups, formed by isomorphic substitution (IS) of Fe3+ for Al3+
,
is only marginally different from that of parent Al(OH)Al groups, as shown by ␨-potential measurements.
Substitution of one octahedral Al3+ site, however, gives rise to the formation of three Fe–O–Al groups
and three Fe(OH)Al groups: such process modifies the adsorption and catalytic properties of NTs outer
surface, as shown by adsorption of the azo-dye Acid Orange 7 (NaAO7) from water solutions and by
AO7− catalytic degradation by H2O2. Interaction with AO7− anions in water occurs in different ways: (i)
with bare IMO, AO7− anions preferentially adsorb by electrostatic interaction with the positively charged
surface, and by H-bonding; (ii) with Fe-1.4-IMO, IS Fe3+ cations act as coordination centres for N atoms
in the AO7− moiety, through a ligand displacement mechanism; (iii) with Fe-L-IMO, AO7− adsorption is
hindered by the presence of Fe2O3 clusters.
Ligand displacement
Heterogeneous catalysis
Hydrogen peroxide
Azo-dyes
Alumina
Concerning AO7− degradation by H2O2, the outer surface of proper IMO is much more reactive than
Fe-doped IMO towards hydrogen peroxide, by forming very reactive Al-OOH bridges. IS Fe3+ species
are probably not accessible to H2O2, and (the expected) Fenton reaction between Fe3+ and H2O2 is not
observed.
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
SiO H ꢀ SiO− + H+
(2)
NTs behaviour in water is crucial: scientific interest for IMO ini-
tially indeed arose from its possible use for the removal of cations
and anions from polluted water, in that the former should inter-
act with the inner surface, and the latter with the outer surface
of NTs, respectively [7–11]. Nonetheless, for applications requir-
ing highly dispersed NTs (e.g. functionalization of the outer surface
to get imogolite/polymer composites) dispersion of IMO in acidic
solution is recommended [12], in order to get positively charged
NTs that repel each other, finally preventing aggregation.
Imogolite (IMO) is a hydrated alumino-silicate with formula
(OH)3Al2O3SiOH [1,2] either found in soils (mainly of volcanic ori-
gin) or obtained by sol–gel synthesis [3–5].
It occurs as single-walled nanotubes (NTs) with Al(OH)Al and
Al–O–Al groups at the outer surface, and silanols (SiOHs) at the
inner one. In water, at neutral pH, the outer surface is positively
charged, due to protonation of Al(OH)Al groups, whereas inner
SiOHs are partially dissociated [6]:
Considering NTs aspect ratio, NTs length varies between 400 nm
and several microns; inner diameter is constant at ca. 1.0 nm,
and outer diameter is ca. 2.0 in natural IMO, whereas it varies
around 2.5–2.7 nm in synthetic samples [1]. The presence of differ-
ent inter-tubes impurities may affect NTs packing [13], which also
depends on the synthesis temperature: the latter is usually 100 ◦C,
Al(OH)Al + H+ ꢀ Al(OH2)+Al
(1)
∗
Corresponding author.
E-mail address: barbara.bonelli@polito.it (B. Bonelli).
http://dx.doi.org/10.1016/j.cattod.2015.10.011
0920-5861/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
2
E. Shafia et al. / Catalysis Today xxx (2015) xxx–xxx
This work addresses the adsorption and catalytic properties of
both bare and Fe-doped IMO NTs in water. The surface properties
related to both Al(OH)Al and Fe(OH)Al groups are studied, by mea-
suring the ␨ potential in water and the interaction with the (bulky)
anion of the azo-dye Acid Orange 7 (NaAO7). The latter is a model
molecule for the many azo-dyes found as contaminants of both
wastewater and groundwater [37], and in virtue of its molecular
dimensions (Scheme 2a) is expected to mainly interact with the
outer surface of NTs, limiting parasitic interactions possibly deriv-
ing by diffusion within IMO inner pores [38].
The degradation of AO7− by H2O2 is studied as catalyzed by both
bare and Fe-doped samples. Several reports exist on the catalytic
activity of some transition aluminas, able to form reactive Al-OOH
groups in the presence of H2O2 [39–41], and so the outer surface
of IMO NTs is expected to have catalytic properties in the presence
of hydrogen peroxide. Moreover, in doped IMO Fe species could
trigger the Fenton reaction [42]:
Scheme 1. Ideal representation of three bundles formed by Imo NTs organized into
a nearly hexagonal packing, of which the cell parameter a is evidenced, correspond-
ing to the distance between the centre of two adjacent NTs. Pores A, B and C are
pointed out only in the right bundle, for the sake of clarity. The marker is 1.0 nm,
corresponding to the inner diameter of A pores.
Fe2+ + H2O2 → Fe3+ + OH− + HO·
Fe3+ + H2O2 → Fe2+ + H+ + HO2·
HO· + H2O2 → HO2· + H2O
(3)
(4)
(5)
(6)
(7)
but synthesis at 25 ◦C yields NTs with an outer diameter close to
that of natural IMO [14].
NTs arrange in bundles with nearly hexagonal packing. This
structure gives rise to three kinds of pores [15,16] shown in
Scheme 1:
HO2· + H2O2 → HO· + H2O + O2
AO7 + HO· → degradationproducts
Known to occur in solution, the Fenton reaction has been largely
documented also with heterogeneous catalysts [43–46], although,
when promoted by Fe3+ (Eq. (4)), as in the present case, is slower
as compared to Fe2+ (Eq. (3)) [42,45,46].
i) Pores A are intra-tubes pores, 1.0 nm in diameter, lined by
SiOHs, the related A surface being very hydrophilic and able
to interact with probes like H2O, NH3 and CO [17,18].
ii) Pores B, 0.3–0.4 nm wide, are those among three aligned NTs
within a bundle, hardly accessible even to small molecules, like
water, in the powder [15,16].
2. Experimental
2.1. Materials synthesis
iii) Pores C are disordered slit (meso)pores among bundles, the sur-
face of which may interact with larger molecules, like phenol
[17] and 1,3,5-triethylbenzene [18].
IMO was obtained according to Farmer et al. [3,4]: at 20 ◦C TEOS
(Tetra-ethoxysilane) and Al(s-butoxide)3 were added to a 75 mM
aqueous solution of HClO4 in the molar ratios Si:Al:HClO4 = 1.x:2:1.
A slight excess of TEOS was used, in order to prevent preferential
formation of aluminium hydroxide during hydrolysis. The solution
was stirred for 18 h, diluted to 20 mM in Al, autoclaved at 100 ◦C for
4 days, dialyzed for 4 days against de-ionized water and then dried
at 50 ◦C.
Fe-x-IMO samples (x = 0.70 or 1.4 wt.% Fe) were obtained by
direct synthesis: in a glove box, a proper amount of FeCl3*6H2O,
corresponding to either 0.70 or 1.4 wt.% Fe, was added to 20 mM
solution of HClO4, then TEOS (tetraetoxysilicate) and Al-sec-
butoxide were added, the pH of the resulting mixture being =4.0.
The aqueous mixture was stirred at room temperature for 18 h,
diluted to 20 mM in Al, autoclaved at 50 ◦C for 4 days, filtered,
washed and dried overnight in oven at 50 ◦C. A reddish-brown pow-
der was obtained, the solution recovered by filtration being instead
transparent.
The interest in IMO-related chemistry has spread recently, one
reason being the possibility of changing NTs chemical composi-
tion. Modifications of the inner surface usually concern either the
substitution of Si by Ge atoms [19,20] or the fabrication of hybrid
organic-inorganic NTs [21–23]. The outer surface has amphoteric
properties [17,18] related to the fact that may be considered as
a curled gibbsite (Al(OH)3) sheet, and may be modified by either
reaction with organic functionalities [12,24–30] or isomorphic sub-
stitution (IS) of Fe3+ for Al3+ [31–33].
The first experimental studies on Fe-doped IMO showed that
NTs formation is preserved up to an overall 1.4 wt.% Fe content.
More recent reports on Fe-doped IMO [33] and aluminium-
germanate analogues [34] showed that IS is actually limited to
1.0 wt.% Fe, since formation of Fe oxo-hydroxide clusters unavoid-
ably occurs at higher Fe content, due to the natural tendency of iron
to form Fe–O–Fe bridges [33,34].
Fe-L-IMO was prepared by contacting with preformed IMO
an amount of FeCl3*6H2O corresponding to 1.4 wt.% Fe in water
(resulting pH 4.0) for 24 h under stirring, and finally adding NH4OH
solution (33% by weight) to precipitate all Fe3+ as oxo/hydroxide
species. The recovered solid was filtered, washed and dried in oven
at 50 ◦C overnight.
The presence of Fe3+ species imparts the solid new chemical and
solid-state properties: Fe-doped IMO showed interesting catalytic
properties in partial oxidation reactions [31,32]. Both theoretical
calculations [35] and experimental findings [33] showed that the
presence of iron affects the Fermi level of the material, by shifting
the band gap of IMO well into the Vis range (from 4.7 eV to 2.8 eV
[33]).
Chemical analysis of the supernatant solution did not detect the
presence of residual iron with all the Fe-doped samples.
Nonetheless, the presence of Fe at the outer surface of NTs
modifies the adsorptive properties of IMO in water: Fe species in
an imogolite–Fe3O4 (magnetite) composite have been shown to
adsorb arsenate ions present in polluted water, through a ligand
displacement mechanism [36].
2.2. Characterization methods
Powder X-Ray diffraction (XRD) patterns were obtained on
a X’Pert Phillips diffractometer operating with Cu Ka radiation
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
E. Shafia et al. / Catalysis Today xxx (2015) xxx–xxx
3
Scheme 2. (a) Structure formula and molecular dimensions [38] of AO7−. (b) UV–vis spectrum of 0.67 mM AO7 aqueous solution: absorption bands due to both the azo- and
hydrazone forms of the dye are observed.
◦
(1.541874 A) in the 2.5–18 2ꢀ range (step width = 0.02◦, time per
˚
According to reaction (8), a 0.030 M H2O2 concentration roughly
corresponds to the stoichiometric amount of H2O2 necessary for
complete AO7− degradation, and 0.10 M H2O2 to an excess of
hydrogen peroxide.
During both adsorption and catalysis experiments, aliquots of
the suspension were collected at regular intervals of time, the
supernatant fraction was separated by centrifugation (ALC cen-
trifuge PK110, at 4000 rpm for 2 min), and the UV–vis spectrum
was measured in the 190–800 nm range on a Cary 5000 UV–Vis–NIR
spectrophotometer (Varian instruments), using a quartz cell with
1 mm path length. The concentration of (residual) AO7− in solution
was evaluated by the intensity of the 484 nm band (vide infra), after
a proper calibration procedure.
step: 2.00 s).
To determine BET SSA (Brunauer–Emmett–Teller specific sur-
face area) and porous volume, N2 isotherms were measured at
−196 ◦C on samples previously outgassed at 250 ◦C, in order to
remove water and other atmospheric contaminants, still preserving
NTs structure [13,47].
Diffuse reflectance (DR) UV–vis spectra of samples outgassed
at 250 ◦C were measured on a Cary 5000 UV–Vis–NIR spectropho-
tometer (Varian instruments).
For IR measurements, samples were shaped as self-supporting,
thin wafers (20 mg cm−2) and outgassed at 250 ◦C in a standard
vacuum frame (residual pressure below 10−3 mbar).
Electrophoretic mobility as a function of pH was measured at
25 ◦C by means of light scattering technique on a Zetasizer Nano-ZS
instrument (Malvern Instruments, Worcestershire, UK). The cor-
responding ␨-potential curves were calculated according to the
Henry’s equation UE = 2ε␨f(Ka)/3Á, where UE is the electrophoretic
mobility, ε is the dielectric constant, ꢁ is the zeta potential, f(Ka)
is the Henry’s function, and ␩ is the viscosity. The adopted value
of f(Ka) was 1.5, in agreement with the Smoluchowski approxima-
tion, usually applied to aqueous solutions of moderate electrolyte
concentration, as in the present case. Water suspensions were
obtained after 2 min sonication with an ultrasonic probe (100 W,
20 kHz, Sonoplus; Bandelin, Berlin, Germany); the pH of the sus-
pension was then adjusted by adding either 0.10 M HCl or 0.10 M
NaOH.
3. Results and discussion
3.1. Powders characterization
Fig. 1a reports XRD patterns of IMO, Fe-0.70-IMO, Fe-1.4-IMO,
and Fe-L-IMO, all corresponding to hexagonal packing of NTs. The
d100 diffraction is at 2ꢀ = 3.9◦ with both IMO and Fe-L-IMO, and at
slightly higher 2ꢀ values with both Fe-0.70-IMO (2ꢀ = 4.0◦) and Fe-
1.4-IMO (2ꢀ = 4.1◦). The cell parameter a (i.e. the distance between
two aligned NTs, Scheme 1) is 2.6 nm with both IMO and Fe-L-
IMO, and 2.5 nm with both Fe-x-IMO samples (Table 1). Previous
work showed that a slight increase of A pores diameter occurs
in the sequence IMO = Fe-L-IMO < Fe-0.70-IMO < Fe-1.4-IMO [33],
and therefore the smaller a value for Fe-doped IMO was assigned
to the presence, within inter-tubes space, of anions coming from
the synthesis bath, joining adjacent NTs though not entering the
IMO formula. During the synthesis of Fe-x-IMO, chloride ions likely
replace (larger) perchlorate anions, so decreasing the final a value
[33]. Nonetheless, Amara et al. showed that with aluminium-
germanates hexagonalization of NTs shape occurs, a process also
leading to smaller a values [48]. Though the latter phenomenon
was observed with imogolite analogues with different chemical
composition and structural features [48], including NTs pores size
and length, the hypothesis of NT hexagonalization may not be com-
pletely disregarded, in that the presence of Fe could be able to affect
NTs mechanical properties and deformation.
2.3. Adsorption and catalytic tests
Adsorption of NaAO7 (from Fluka) was studied by contacting the
powder (1 g L−1) with a 0.67 mM NaAO7 water solution (starting pH
value of 6.80).
NaAO7 degradation with H2O2 was studied by contacting
powder suspensions (1 g L−1) with 0.67 mM water solution and
either 0.030 M H2O2 or 0.10 M H2O2. For comparison, catalytic
tests were also run on a commercial ␥-Al2O3 (Sigma–Aldrich,
SSA = 145 m2 g−1). Both adsorption and degradation experiments
were run in the presence of an excess of IMO units (molar ratio
IMO/AO7− ≈ 10/1).
Concerning the amount of H2O2 used during catalytic tests, the
reaction leading to the complete degradation of the dye is:
N2 adsorption isotherms in Fig. 1b show the features expected
for microporous systems with some mesoporosity; analysis of total
and micropores volume values (Table 1) shows that with Fe-L-IMO
an increase of mesopores volume occurs, because the loading pro-
cedure mainly affects the C surface [33].
C
16H11N2SO4Na + 42H2O2 → 16CO2 + 46H2O
+ 2HNO3 + NaHSO4
(8)
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
4
E. Shafia et al. / Catalysis Today xxx (2015) xxx–xxx
Table 1
Samples textural properties as derived by N2 sorption isotherms at -196 ◦C and low angles XRD patterns.
Sample
Fe (wt%)
BET SSA (m2 g−1
)
Total pore volume (cm3 g−1
)
aMicro-pore volume, (cm3 g−1
)
d100 (nm)
ba (nm)
IMO
–
383
400
450
455
0.21
0.27
0.22
0.22
0.13
0.13
0.15
0.14
2.3
2.3
2.2
2.2
2.6
2.6
2.5
2.5
Fe-L-IMO
Fe-0.70-IMO
Fe-1.4-IMO
1.4
0.70
1.4
a
As obtained by applying the t-plot method.
Cell parameter, a, as calculated by applying Eq. a = 2d100
√
/ 3.
b
Fig. 2. DR UV–vis spectra of IMO (1), Fe-0.70-IMO (2), Fe-L-IMO (3) and Fe-1.4-IMO
(4).
groups are obtained, the latter more abundant with Fe-L-IMO. The
presence of isolated Oh Fe3+ sites in Fe-L-IMO indicates that ionic
exchange also occurred between structural Al3+ and Fe3+ ions in
water:
3+
[Al(OH)+Al](s) + Fe(H2O)63+ → [Fe(OH)+Al](s) + Al(H2O)6
(9)
This fact indicates that Al-bridges are very reactive and prone
to ionic exchange in water. On the other hand, the occurrence of
some Fe2O3 clusters in Fe-x-IMO indicates that the actual isomor-
phic substitution of Fe for Al in the IMO structure is lower than
that corresponding to 1.4% by weight. Actually, magnetic suscepti-
bility measurements reported elsewhere [33] allowed us to asses
that IS is limited to ca. 1.0 wt.% Fe, independently on the synthesis
procedure, and then Fe oxo-hydroxide clusters form, with slightly
larger dimensions in the loaded sample, according to HRTEM (high-
resolution transmission electron microscopy) observations [33].
The aforementioned magnetic measurements showed that IS was
the main process with Fe-0.70-IMO NTs, to which corresponds the
chemical composition (OH)3Al1.975Fe0.025SiOH.
Fig. 3 reports IR spectra of the bare samples after dehydra-
tion at 250 ◦C: in the 3800–2700 cm−1 range a broad absorption
is observed due to H-bonded hydroxyls [17,18]. With IMO (curve
1), bands are seen 1595 and 1465 cm−1 assigned to carbonate-
like species similar to those formed at the surface of aluminium
hydroxide [17]. The intensity of such bands decreases in the
series Fe-L-IMO > Fe-0.70-IMO > Fe-1.4-IMO (curves 3, 2 and 4,
respectively), in agreement with the fact that the outer surface is
progressively losing its aluminium-hydroxide character as far as
Al3+/Fe3+ IS takes place.
Fig. 1. (a) Low angles XRD patterns of IMO (1), Fe-L-IMO (2), Fe-0.70-IMO (3), and
Fe-1.4-IMO powder samples. (b) N2 sorption isotherms at −196 ◦C of IMO (squares),
Fe-L-IMO (circles), Fe-0.70-IMO (stars) and Fe-1.4-IMO (triangles) samples dehy-
drated at 250 ◦C. Full and hollow symbols refer to adsorption and desorption run,
respectively.
DR UV–vis spectra of dehydrated samples are reported in Fig. 2:
IMO has, as expected, a negligible absorption in the UV–vis region
(curve 1), whereas the Fe-doped samples show similar, though not
equal, UV–vis features. Both Fe-x-IMO samples (curves 2 and 4)
strongly absorb at 270 nm, and show a minor absorption band at
480 nm, almost negligible with Fe-0.70-IMO. According to the lit-
erature, the former signal is due to charge-transfer transitions (CT)
from O2− to isolated Oh Fe3+ ions, the latter to d–d transitions of
Fe3+ in Fe2O3 nanoclusters [49,50]. Fe-L-IMO spectrum (curve 3)
is similar to that of Fe-1.4-IMO, though slightly shifted to higher
wavelengths, with a component at 310 nm and more intense d–d
transitions, the component at 310 nm being assigned to CT tran-
sitions in iron oxide clusters [49]. UV–vis spectra point out the
preferential formation of Fe2O3 clusters by post-synthesis method,
although Fe-L-IMO also absorbs at 270 nm. At 1.4 wt.% Fe content,
both isolated octahedral (Oh) Fe3+ species and aggregated Fe–O–Fe
No other differences emerge from Fig. 3, because of the relatively
low Fe content and the plethora of hydroxyls occurring both within
and outside NTs, especially with Fe-0.70-IMO, its IR spectrum being
very similar to that of proper IMO.
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
E. Shafia et al. / Catalysis Today xxx (2015) xxx–xxx
5
Fig. 3. IR spectra of IMO (1), Fe-0.70-IMO (2), Fe-L-IMO (3), and Fe-1.4-IMO (4)
samples dehydrated at 250 ◦C.
Fig. 4. ꢁ potential curves of IMO (squares), Fe-1.4-IMO (triangles) and Fe-L-IMO
(circles).
3.2. Samples behaviour in water
Fig. 4 reports ꢁ potential curves for samples IMO, Fe-1.4-IMO and
Fe-L-IMO, as IR spectroscopy did not notice any relevant difference
in the surface properties of Fe-0.70-IMO with respect to bare IMO.
IMO (squares) is positively charged at low pH, with the point of zero
charge (PZC) at pH 9.8, a value close to those reported for alumina
[51,52], and in agreement with the chemical nature of IMO outer
surface, exposing only Al–O–Al groups and Al(OH)Al bridges, like a
hydrated aluminium oxide. The curve of Fe-1.4-IMO (triangles) lies
slightly below that of IMO over nearly the whole pH range, with
a marginally higher PZC, indicating a limited decrease in acidity
of Fe(OH)Al groups with respect to Al(OH)Al. The same holds for
Fe-L-IMO, i.e., in water, acidity of hydroxyls of iron oxo-hydroxide
nanoclusters does not differ much from that of Al(OH)Al species. As
a whole, the three samples have similar ꢁ potential curves: error
bars show that the net difference in surface charge is not relevant,
in that the samples are positively charged at low pH, and negatively
charged at higher pH, so they should be able to adsorb anions in a
similar way at low pH.
Fig. 5. (a) UV–vis spectra of the starting AO7 solution (1) and of the supernatant
solutions after 120 min in contact with IMO (2), Fe-L-IMO (3) and Fe-1.4-IMO (4).
(b) Percentage of AO7 removed by IMO (squares), Fe-L-IMO (circles) and Fe-1.4-IMO
(triangles); the inset shows the initial pH drop. (c) pH of the solutions during AO7
adsorption measurements.
undergoes an azo-hydrazone tautomerism, via an intra-molecular
proton transfer, so that both hydrazone and azo-form are simul-
taneously present. The two bands at 310 and 230 nm and the
shoulder at 256 nm are due to aromatic rings absorptions. The band
at 484 nm is due to the n–␲* transition involving the lone pair on N
atoms and the conjugated system extending over the two aromatic
Some differences, however, emerged concerning the properties
of the outer surface of NTs, which could affect adsorption proper-
ties of iron-doped IMO: Fig. 5 describes the adsorption experiments
of NaAO7, the UV spectrum of which is reported in Scheme 2b.
In water, the hydrazone form of AO7−, stable in the solid phase,
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
6
E. Shafia et al. / Catalysis Today xxx (2015) xxx–xxx
moieties and encompassing the N–N group of the hydrazone form.
The shoulder at 403 nm has a similar nature, involving this time the
N–N group of the azo form [37,53,54].
Fig. 5a reports the UV–vis spectra of the liquid phase after
120 min contact with the studied samples: in the reported exper-
iments, no new bands form with respect to the starting solution,
indicating that a mere adsorption process occurs. From the obtained
adsorption data, the trend in intensities concerning the samples
IMO, Fe-1.4-IMO and Fe-L-IMO reported in Fig. 5b is evaluated,
whereas section c of the Figure shows the parallel behaviour of
pH.
Fig. 5b shows with Fe-1.4-IMO a steep increase in the percentage
of AO7− removal: within 30 min, a plateau is reached correspond-
ing to removal of ca. 95% AO7−. With IMO, the curve has a different
shape, indicating the occurrence of two processes, a faster one
within the first 30 min, followed by a slower one. The shape of Fe-
L-IMO curve is a similar to that of IMO, but the overall amount of
removed AO7− is smaller. Different curves shape indicates that dif-
ferent adsorption processes occur on the three adsorbents. With
Fe-1.4-IMO, a strong interaction should occur between the dye and
outer Fe3+ ions. The interaction is strong, in that FeAO7− adducts
form, not by electrostatic interaction, but by coordination of Fe3+
ions by N atoms of the organic dye, through ligand displacement.
Simple stoichiometric calculation shows an AO7−/Fe3+ molar ratio
of 3/1, indicating that probably after one FeAO7− adduct forms,
other dye molecules are adsorbed by the latter species. With IMO,
such process does not occur, and the steep increase observed in
Fig. 5b within the first 30 min is ascribed to AO7− moieties electro-
statically interacting with the (positively charged) outer surface.
The process occurring at longer times is ascribed to another phe-
nomenon, like the slow diffusion of AO7− molecules within IMO A
pores (vide infra). The adsorbing behaviour of Fe-L-IMO is similar
to that of IMO, indicating that the adsorption process is again elec-
trostatic interaction. The lower adsorption capacity of Fe-L-IMO as
compared to IMO may be ascribed to the presence, at the outer
surface of NTs, of clusters of Fe2O3 [33], on which AO7− molecules
cannot adsorb, as reported by the literature [37].
The simultaneous pH change is reported in Fig. 5c: in the three
cases, two processes are seen to occur, i.e. one rapid, involving
a steep decrease in pH (inset to Fig. 5c), the other one time-
dependent, and implying a moderate pH increase.
The latter process is likely related to the difficult diffusion of
the dye within A pores, the size of which is of the order of the
transverse dimension of AO7−: the interaction is with inner SiOHs
through a process that decreases the concentration of hydronium
ions in the solution. The fast process, instead, takes place at the
outer C surface, involving either Al(OH)Al or Fe(OH)Al species, and
causes the release of hydronium ions in solution. With Fe-1.4-IMO,
where Fe(OH)Al species are present, the latter process is particu-
larly relevant.
Fig. 6. (a) UV–vis spectra of the starting 0.67 mM AO7 solution (1) and of the super-
natant solutions after 5 min in contact with 0.030 M H2O2 and IMO (2), Fe-0.70-IMO
(3), Fe-1.4-IMO (4) and Fe-L-IMO (5). (b) Percentage of AO7 conversion as obtained
with IMO (squares), Fe-0.70-IMO (stars), Fe-1.4-IMO (triangles) and Fe-L-IMO (cir-
cles).
With Fe-1.4-IMO a second interaction may take place with
Fe(OH2)+Al groups:
Fe(OH2)+Al + AO7− = FeAO7− + AlOH + H+
(12)
Reaction (12), however, contributes to the removal of a minor
fraction of AO7− molecules, indicating that actually only a small
fraction of Fe(OH)Al groups is protonated (ca. 2%, as calculated on
the basis of stoichiometric considerations).
As far as IMO is concerned, the following reactions seem to
account for the features observed:
On the contrary, with Fe-L-IMO, Fe(OH)Al groups mainly served
as crystallization centres for the growth of iron oxo-hydroxide
clusters [33], and therefore they are no more available to bind to
AO7− species, finally lowering the adsorption capacity of Fe-L-IMO
(Fig. 5b) and giving rise to moderate changes in pH (Fig. 5c).
Fig. 6 reports catalytic results obtained with 0.030 M H2O2 and
1 g L−1 catalyst: section a reports UV–vis spectra of the super-
natant solutions after 5 min contact with the catalysts, and section b
reports overall conversions (% of AO7− degradation). Similar results
were obtained with 0.10 M H2O2, the corresponding final conver-
sions being reported in Table 2.
SiO− + H+ + AO7− = SiOH· · ·AO7−
(10)
(11)
Al(OH2)+Al + AO7− = Al(OH)Al· · ·AO7− + H+
With IMO, the adsorption process implies H-bonding of the
AO7− anion with acidic hydroxyls, namely SiOHs and bridged
species. This implies in turn reversal of Eqs. (1) and (2), with oppo-
site variations in pH. Simple stoichiometric calculation on the initial
pH drop shows that reaction (11) contributes to the removal of
ca. 5% of AO7− molecules, indicating that the interaction between
AO7− and the outer surface of IMO NTs is mainly electrostatic.
Fig. 6a shows that after 5 min most of AO7− has been removed
in the presence of IMO (curve 2), whereas all the Fe-doped samples
(curves 3–5) gave markedly lower conversions. In particular, with
Fe-L-IMO a band is evident at 265 nm (asterisk) not due to AO7−, but
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
E. Shafia et al. / Catalysis Today xxx (2015) xxx–xxx
7
AO7−. This means that transition metal-doped IMO could be able
to remove organic contaminants from polluted water, and not only
inorganic ions, as already reported for magnetite–imogolite com-
posite materials able to remove arsenate ions with high efficiency
[36].
Table 2
Catalytic conversions as obtained with either 0.030 M or 0.10 M H2O2 and 0.67 mM
AO7 in water.
Catalyst
% Conversion with
0.030 M H2O2
% Conversion with
0.10 M H2O2
IMO
95
35
39
43
35
5.0
92
41
44
43
49
11
On the contrary, Fe-doping has a detrimental effect on catalytic
activity of imogolite in the presence of H2O2. The outer surface of
bare imogolite, instead, readily reacts with H2O2, likely by forming
very reactive Al-OOH groups, which release •OH radicals able to
attack AO7− molecules. The reaction in water occurs in the very first
minutes and about 100% AO7− conversion is reached. Although sev-
eral transition aluminas are known to form Al-OOH species in the
presence of H2O2, the commercial ␥-Al2O3 used for comparison in
this study, provided a lower activity than bare imogolite. This find-
ing leads to the idea that the curvature of the surface of imogolite
NTs provides some strain to Al(OH)Al and Al–O–Al groups, which
are more reactive towards hydrogen peroxide than those at the
surface of other aluminium oxo-hydroxides. The catalytic activity
towards H2O2 opens the possibility to exploit the (outer surface of)
imogolite NTs as heterogeneous catalysts in aqueous phase, and at
mild temperatures, so overcoming problems related to imogolite
thermal stability and diffusion within A pores, which limited so far
the use of imogolite as a catalyst.
Fe-L-IMO
Fe-0.70-IMO
Fe-1.4-IMO
␥-Al2O3
None
References
[1] P.D.G. Cradwick, V.C. Farmer, J.D. Russell, K. Wada, N. Yoshinaga, Nat. Phys.
Sci. 240 (1972) 187–189.
[2] N. Yoshinaga, A. Aomine, Soil Sci. Plant Nutr. 8 (1962) 22–29.
[3] V.C. Farmer, M.J. Adams, A.R. Fraser, F. Palmieri, Clay Miner. 18 (1983)
459–472.
[4] V.C. Farmer, A.R. Fraser, Proceedings of the International Clay Conference,
Elsevier, Amsterdam, 1978, pp. 547–554.
[5] S.I. Wada, A. Eto, K. Wada, J. Soil Sci. (1979) 347–355.
[6] J.P. Gustafsson, Clay Clay Miner. 49 (2001) 73–80.
[7] Y. Arai, M. McBeath, J.R. Bargar, J. Joye, J.A. Davis, Geochim. Cosmochim. Acta
70 (2006) 2492–2509.
[8] C.J. Clark, M.B. McBride, Clay Clay Miner. 32 (1984) 291–299.
[9] L. Denaix, I. Lamy, J.Y. Bottero, Colloids Surf. A 158 (1999) 315–325.
[10] J.B. Harsh, S.J. Traina, J. Boyle, Y. Yang, Clay Clay Miner. 40 (1992) 700–706.
[11] R.L. Parfitt, A.D. Thomas, R.J. Atkinson, R.St.C. Smart, Clay Clay Miner. 22
(1974) 455–456.
[12] K. Yamamoto, H. Otsuka, S.I. Wada, A. Takahara, Chem. Lett. 30 (2001)
1162–1163.
[13] C. Zanzottera, A. Vicente, M. Armandi, C. Fernandez, E. Garrone, B. Bonelli, J.
Phys. Chem. C 116 (2012) 23577–23584.
[14] S.I. Wada, Clay Miner. 35 (1987) 379–384.
[15] W.C. Ackerman, D.M. Smith, J.C. Huling, Y. Kim, J.K. Bailey, C.J. Brinker,
Langmuir 9 (1993) 1051–1057.
[16] M.A. Wilson, G.S.G. Lee, R.C. Taylor, Clay Clay Miner. 50 (2002) 348–351.
[17] B. Bonelli, I. Bottero, N. Ballarini, S. Passeri, F. Cavani, E. Garrone, J. Catal. 264
(2009) 15–30.
[18] B. Bonelli, M. Armandi, E. Garrone, Phys. Chem. Chem. Phys. 15 (2013)
13381–13390.
[19] S. Mukherjee, V.M. Bartlow, S. Nair, Chem. Mater. 17 (2005) 4900–4909.
[20] A. Thill, P. Maillet, B. Guiose, O. Spalla, L. Belloni, P. Chaurand, M. Auffan, L.
Olivi, J. Rose, J. Am. Chem. Soc. 134 (2012) 3780–3786.
[21] I. Bottero, B. Bonelli, S. Ashbrook, P. Wright, W. Zhou, M. Tagliabue, M.
Armandi, E. Garrone, Phys. Chem. Chem. Phys. 13 (2011) 744–750.
[22] D.-Y. Kang, N.A. Brunelli, G.I. Yucelen, A. Venkatasubramanian, J. Zang, J.
Leisen, P.J. Hesketh, C.W. Jones, S. Nair, Nat. Commun. 5 (2014) 3342.
[23] D.-Y. Kang, J. Zang, C.W. Jones, S. Nair, J. Phys. Chem. C 115 (2011) 7676–7685.
[24] L.M. Johnson, T.J. Pinnavaia, Langmuir 6 (1990) 307–311.
[25] L.M. Johnson, T.J. Pinnavaia, Langmuir 7 (1991) 2636–2641.
[26] X. Qi, H. Yoon, S.-H. Lee, J. Yoon, S.-J. Kim, J. Ind. Eng. Chem. 14 (2008) 136–141.
[27] C. Zanzottera, A. Vicente, E. Celasco, C. Fernandez, E. Garrone, B. Bonelli, J.
Phys. Chem. C 116 (2012) 7499–7506.
Scheme 3. Species present at the outer surface of an IMO NT, before and after Fe3+
IS of one Oh Al3+
.
to residual H2O2, not consumed by the reaction. Fig. 6b shows that
bare IMO has an overall higher activity, as compared to Fe-doped
samples, since ca. 90% AO7 is converted within the first 5 min: this
is likely related to the presence of very reactive aluminium sites at
the outer surface of IMO NTs. Such sites are probably more reac-
tive than similar ones present at the surface of transition aluminas,
which have been shown by the literature to be active in epoxida-
tion reaction [39–41]. Comparison with commercial ␥-Al2O3 shows
that only 35% degradation of AO7− is obtained, and so an effect of
the curvature of the outer surface of IMO NTs can be postulated in
order to explain the higher activity of IMO with respect to ␥-Al2O3.
Concerning the poorer activity of Fe-doped IMO, it has to be con-
sidered that, notwithstanding the low Fe content, when one Oh Al3+
is replaced by one Oh Fe3+, three Fe(OH)Al and three FeOAl groups
form (Scheme 3), and so the amount of Fe-related sites is higher
and able to actually modify the surface properties of IMO, in partic-
ular the hydroxyls population of the outer surface. Moreover, such
Fe3+ sites are not reactive towards H2O2 molecules, likely due to
their high coordination and, simultaneously, the presence of H2O2
hinders the interaction between Fe3+ sites and AO7− species, which
has been instead observed in the absence of hydrogen peroxide.
[28] W. Ma, M.O. Yah, H. Otsuka, A. Takahara, J. Mater. Chem. 22 (2012)
11887–11892.
[29] S. Park, Y. Lee, B. Kim, J. Lee, Y. Jeong, J. Noh, A. Takahara, D. Sohn, Chem.
Commun. (2007) 2917–2919.
4. Conclusion
[30] K. Yamamoto, H. Osaka, S.I. Wada, D. Sohn, A. Takahara, Polymer 46 (2005)
12386–12392.
[31] M. Ookawa, Y. Inoue, M. Watanabe, M. Suzuki, T. Yamaguchi, Clay Sci. 12
(2006) 280–284.
[32] M. Ookawa, Clay Minerals in Nature – Their Characterization, Modification
and Application, InTech, 2012, pp. 239–258.
Fe-doping has a positive effect on the adsorption properties of
imogolite, in that with Fe-1.4-IMO about 90% dye is removed from
solution within 5 min, since by their nature of transition element
Fe3+ ions can display coordinative capacity towards N atoms of
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011

G Model
CATTOD-9832; No. of Pages8
ARTICLE IN PRESS
8
E. Shafia et al. / Catalysis Today xxx (2015) xxx–xxx
[33] E. Shafia, S. Esposito, M. Manzoli, M. Chiesa, P. Tiberto, G. Barrera, G. Menard,
P. Allia, F.S. Freyria, E. Garrone, B. Bonelli, J. Nanopart. Res. 17 (2015) 336.
[34] A. Avellan, C. Levard, N. Kumar, J. Rose, L. Olivi, A. Thill, P. Chaurand, D.
Borschneck, A. Masion, RSC Adv. 4 (2014) 49827–49830.
[35] F. Alvarez-Ramírez, J. Chem. Theory Comput. 5 (2009) 3224–3231.
[36] N. Arancibia-Miranda, M. Escudey, C. Pizarro, J.C. Denardin, M.T.
García-González, J.D. Fabris, L. Charlet, Mater. Res. Bull. 51 (2014) 145–152.
[37] F.S. Freyria, B. Bonelli, R. Sethi, M. Armandi, E. Belluso, E. Garrone, J. Phys.
Chem. C 115 (2011) 24143–24152.
[38] X. Zhao, X. Bu, T. Wu, S.-T. Zheng, L. Wang, P. Feng, Nat. Commun. 4 (2013)
2344.
[39] B.S. Lane, K. Burgess, Chem. Rev. 103 (2003) 2457–2474.
[40] J. Sepulveda, S. Teixeira, U. Schuchardt, Appl. Catal. A 318 (2007) 213–217.
[41] D. Mandelli, M.C.A. van Vliet, U. Arnold, R.A. Sheldon, U. Schuchardt, J. Mol.
Catal. A 168 (2001) 165–171.
[45] M. Piumetti, F.S. Freyria, M. Armandi, F. Geobaldo, E. Garrone, B. Bonelli, Catal.
Today 227 (2014) 71–79.
[46] M. Piumetti, F.S. Freyria, M. Armandi, G. Saracco, E. Garrone, B. Bonelli, Chim.
Oggi 33 (2015) 40–45.
[47] K.J. MacKenzie, M.E. Bowden, J.W.M. Brown, R.H. Meinhold, Clay Clay Miner.
37 (1989) 317–324.
[48] M.S. Amara, S. Rouzière, E. Paineau, M. Bacia-Verloop, A. Thill, P. Launois, J.
Chem. Phys. C 118 (2014) 9299–9306.
[49] E. Borghi, M. Occhiuzzi, E. Foresti, I.G. Lesci, N. Roveri, Phys. Chem. Chem.
Phys. 12 (2010) 227–238.
[50] Y. Wang, Q.H. Zhang, T. Shishido, K. Takehira, J. Catal. 209 (2002) 186–196.
[51] B.M. Rotoli, P. Guidi, B. Bonelli, M. Bernardeschi, M.G. Bianchi, S. Esposito, G.
Frenzilli, P. Lucchesi, M. Nigro, V. Scarcelli, M. Tomatis, P.P. Zanello, B. Fubini,
O. Bussolati, E. Bergamaschi, Chem. Res. Toxicol. 27 (2014)
1142–1154.
[42] J.J. Pignatello, E. Oliveros, A. MacKay, Crit. Rev. Env. Sci. Technol. 36 (2006)
1–84.
[52] H.-Y. Shu, M.-C. Chang, H.-H. Hu, W.-H. Chen, J. Colloid Interface Sci. 314
(2007) 89–97.
[43] G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp, Handbook of Heterogeneous
Catalysis, Wiley John & Sons Incorporated, Weinheim, 2008.
[44] E. Neyens, J. Baeyens, J. Hazard. Mater. B 98 (2003) 33–50.
[53] J.A. Mielczarki, G.M. Atenas, E. Mielczarski, Appl. Catal. B 56 (2005)
289–303.
[54] C. Bauer, P. Jacques, A. Kalt, J. Photochem. Photobiol. A 140 (2001) 87–92.
Please cite this article in press as: E. Shafia, et al., Reactivity of bare and Fe-doped alumino-silicate nanotubes (imogolite) with H2O2
and the azo-dye Acid Orange 7, Catal. Today (2015), http://dx.doi.org/10.1016/j.cattod.2015.10.011