Immobilization of gold on short-channel mesoporous SBA-15 functionalized with thiol and hydrophobic groups for oxidation reactions

Article abstract of DOI:10.1016/j.cattod.2019.09.014

Nanosized gold entities immobilized on short-channel SBA-15 mesoporous materials functionalized solely with mercaptopropyl groups and together with propyl and octyl moieties have been prepared following a two-steps procedure. These materials have been used as catalysts for the oxidation of cyclohexene with molecular oxygen in liquid phase at atmospheric pressure. First, SBA-15 materials functionalized only with mercaptopropyl groups and with a combination of these groups with two hydrophobic moieties, namely propyl and octyl, have been synthesized in the presence of the non-ionic surfactant P104. Small particles having short channel length have been identified by TEM and SEM. In order to study the influence of the nature of the sulphur functional group, these S-bearing materials were also oxidized with hydrogen peroxide or dimethyldioxirane (DMD) to sulfonic groups prior to gold immobilization. The effectiveness of these oxidation methods was assessed by ²⁹Si MAS NMR, ¹³C CP MAS NMR, XPS, chemical and thermal analysis, and it has been found that DMD oxidized efficiently thiol groups to sulfonic groups, but H₂O₂ leaves a fraction of unreacted thiol. In a second step, nanosized gold entities have been prepared by a two-liquid phases route involving rosemary oil as organic phase and an aqueous phase formed by dissolving gold in a solution of ammonium chloride in concentrated nitric acid (aqua regia). Following this method, a fraction of the Au dissolved in the aqua regia solution is spontaneously reduced and transferred to the essential oil phase. Upon contacting the gold-bearing organic layer with the mesoporous materials, gold is actually immobilized on them, rendering metal contents in the range 1.1-0.2 wt.percent, being those functionalized with alkyl chains the least efficient in capturing gold from the organic phase. No surface plasmon resonance band at 520 nm characteristic of gold nanoparticles has been detected by UV–vis spectroscopy in these Au-containing materials. All of them are active and selective for the allylic oxidation of cyclohexene, but their specific activity varies as a function of the nature of the functional groups. It has been found that the most active catalysts are those pre-oxidized with DMD. The presence of hydrophobic octyl groups increases substantially the turnover number of the reaction (TON), while the short-chain propyl moieties hardly affect the activity. It has been found that the nanosized gold entities initially present in the catalysts evolve in the reaction medium towards the formation of nanoparticles.

Full text of DOI:10.1016/j.cattod.2019.09.014

CatalysisTodayxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Immobilization of gold on short-channel mesoporous SBA-15 functionalized
with thiol and hydrophobic groups for oxidation reactions
R. de la Serna Valdés, J. Agúndez, C. Márquez-Álvarez, J. Pérez-Pariente⁎
Instituto de Catálisis y Petroleoquímica (ICP-CSIC), Marie Curie 2, 28049, Cantoblanco, Madrid, Spain
A R T I C L E I N F O
A B S T R A C T
Keywords:
Nanosized gold entities immobilized on short-channel SBA-15 mesoporous materials functionalized solely with
mercaptopropyl groups and together with propyl and octyl moieties have been prepared following a two-steps
procedure. These materials have been used as catalysts for the oxidation of cyclohexene with molecular oxygen
in liquid phase at atmospheric pressure. First, SBA-15 materials functionalized only with mercaptopropyl groups
and with a combination of these groups with two hydrophobic moieties, namely propyl and octyl, have been
synthesized in the presence of the non-ionic surfactant P104. Small particles having short channel length have
been identified by TEM and SEM. In order to study the influence of the nature of the sulphur functional group,
these S-bearing materials were also oxidized with hydrogen peroxide or dimethyldioxirane (DMD) to sulfonic
groups prior to gold immobilization. The effectiveness of these oxidation methods was assessed by 29Si MAS
NMR, 13C CP MAS NMR, XPS, chemical and thermal analysis, and it has been found that DMD oxidized effi-
ciently thiol groups to sulfonic groups, but H2O2 leaves a fraction of unreacted thiol. In a second step, nanosized
gold entities have been prepared by a two-liquid phases route involving rosemary oil as organic phase and an
aqueous phase formed by dissolving gold in a solution of ammonium chloride in concentrated nitric acid (aqua
regia). Following this method, a fraction of the Au dissolved in the aqua regia solution is spontaneously reduced
and transferred to the essential oil phase. Upon contacting the gold-bearing organic layer with the mesoporous
materials, gold is actually immobilized on them, rendering metal contents in the range 1.1-0.2 wt.%, being those
functionalized with alkyl chains the least efficient in capturing gold from the organic phase. No surface plasmon
resonance band at 520 nm characteristic of gold nanoparticles has been detected by UV–vis spectroscopy in these
Au-containing materials. All of them are active and selective for the allylic oxidation of cyclohexene, but their
specific activity varies as a function of the nature of the functional groups. It has been found that the most active
catalysts are those pre-oxidized with DMD. The presence of hydrophobic octyl groups increases substantially the
turnover number of the reaction (TON), while the short-chain propyl moieties hardly affect the activity. It has
been found that the nanosized gold entities initially present in the catalysts evolve in the reaction medium
towards the formation of nanoparticles.
Short-channel SBA-15
Dimethyldioxirane
Gold clusters
Cyclohexene
Oxidation
Hydrophobicity
1. Introduction
phases system reported in mid eighteenth century scientific literature
[5], involving the use of essential oils as organic phase and a solution of
Chemical and catalytic reactivity of gold nanoparticles is affected by
a variety of different factors, as their size and shape, and, if they are
supported on a solid material, also by the specific characteristics of the
support, as they can control not only the metal-support interaction, but
also the diffusion and adsorption/desorption of reagents and products
under actual reaction conditions [1–3]. It is possible in a first approach
to gain some control on these properties by using specific synthesis
procedures of both the gold entities and the support on which they
would eventually be immobilized. On this regard, we have shown [4]
that it is possible to prepare nanosized gold by using a two-liquid
gold in a mixture of concentrated nitric acid and ammonium chloride
(this a variety of aqua regia). In this system, a fraction of the gold is
spontaneously reduced and transferred to the essential oil phase, which
contains small disordered Au clusters of about 1–1.5 nm in size, to-
gether with Au nanoparticles (AuNPs) in the range of 2–20 nm, but 90%
of them in the range of 2–8 nm [4]. Hence, this method predates by one
century the most widely known two-liquid phases methodology re-
ported by Michael Faraday in the nineteenth century [6,7]. Moreover, it
has been found that the formation of nanoparticles can be suppressed
by adjusting the gold concentration in the aqueous phase, and therefore
⁎ Corresponding author.
E-mail address: jperez@icp.csic.es (J. Pérez-Pariente).
https://doi.org/10.1016/j.cattod.2019.09.014
Received 21 December 2018; Received in revised form 19 June 2019; Accepted 11 September 2019
0920-5861/©2019ElsevierB.V.Allrightsreserved.
Pleasecitethisarticleas:R.delaSernaValdés,etal.,CatalysisToday,https://doi.org/10.1016/j.cattod.2019.09.014

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
this is a suitable and simple method to obtain gold entities with size
below ∼2 nm [8].
where TMOS stands for tetramethyl orthosilicate (Sigma-Aldrich, >
99%); MPTMS for 3-mercaptopropyltrimethoxysilane (Sigma-Aldrich,
95%); R for either propyltrimethoxysilane (TMPS, Sigma-Aldrich, >
99%), or octyltriethoxysilane (OTES, Alfa Aesar, 95%), and x = 0 or
0.05; P104 for Pluronic P104, the triblock co-polymer
(EO)27(PO)6(EO)27, (Sigma-Aldrich); and HCl for hydrochloric acid
(Panreac, 37 wt.%). The HCl and H2O content of the gel containing
TMPS were optimized to 5.41 and 178 mol per mole of TMOS, respec-
tively, in order to improve the ordering of the recovered SBA-15 ma-
terial. The procedure to prepare the material functionalized with
MPTMS was as follows. 170 mL of 1.1 M HCl were placed in a pyrex
bottle provided with a stirring bar, and 3 g of P104 were added. Once
the surfactant was dissolved, 4.4 mL of TMOS and 0.260 mL of MPTMS
were added. For the preparation of propyl- and octyl-functionalized
materials, the corresponding amount of TMPS and OTES were added in
this step. Then, the mixture was stirred for 24 h at room temperature.
After this, the closed bottle was heated at 80 °C (or 60 °C for the TMPS
gel) for 5 days. After that, the bottle was cooled down to room tem-
perature and its content filtered, washed with water and ethanol and
dried at room temperature overnight. The dried sample was treated
with ethanol (100 mL of ethanol per g of sample) under stirring in a 1 L
round-bottom flask at 90 °C for 4 h, in order to remove the surfactant.
The resulting samples were denoted as SH for the as-made material
functionalized with SH groups only; SPr for that functionalized with
both SH and propyl groups, and SOc for the material resulting from gels
containing both SH and octyl groups. In all cases, the suffix “ext” is
added after the name for the samples treated with ethanol to remove
the surfactant, i.e, SH-ext and so on.
Ordered mesoporous silica materials are very convenient supports
for the immobilization of nanosized gold, owing to their tunable (within
certain limits) porous structure, large pore size and surface area, and
the possibility to decorate the surface with functional groups with the
potential capability of interacting with deposited gold nanosized enti-
ties, rendering high dispersion and preventing them from agglomera-
tion and leaching [9]. Moreover, it is well-known that gold has strong
affinity for thiols [10,11]. For these reasons, the nanosized Au entities
prepared by the essential oil-mediated route described above have been
immobilized on mesoporous SBA-15 materials functionalized with
propylthiol moieties as gold binding groups [8]. However, it has been
found that the Au-thiol catalysts are inactive for the liquid phase oxi-
dation of cyclohexene with molecular oxygen, and the reaction is
triggered only by those catalysts for which the thiol groups are spon-
taneously oxidized to sulfonic groups during the gold immobilization
process, i.e., upon contacting the gold-containing essential oil phase
with the SH-SBA-15 material. Moreover, it has been observed in these
cases that the increase in catalytic activity runs parallel with the ag-
glomeration of gold clusters to render nanoparticles [8].
These studies evidence the complexity of the overall catalytic re-
action, and brought us to investigate further the influence of some
properties of the catalyst that might be determinant for the reaction
output. First, we wondered whether an SBA-15 morphology other than
the conventional rod-like particles we have reported previously [8]
might have an influence on the overall catalyst performance. In these
particles, the channels extend along their main axis. In this way, and in
order to facilitate the diffusion in the interior of the mesopores, we have
prepared functionalized SBA-15 materials possessing short channels by
using the non-ionic block copolymer surfactant Pluronic P104
((EO)27(PO)6(EO)27, where EO and PO stands for ethylene and propy-
lene oxide, respectively), instead of the higher molecular weight
Pluronic P123 used previously, in conditions similar to those previously
reported for all-silica materials [12,13]. In this case, as it will be shown
below, SBA-15 particles with plate-like morphology have been
achieved, where the channels run perpendicular to the basal plane of
the platelets. Second, as far as oxidized sulphur species seem to be re-
quired for Au activity [8], we have intentionally pre-oxidized the SBA-
15 materials bearing -SH groups before contacting them with the gold
solution by using two different oxidation reagents and procedures,
namely hydrogen peroxide, which has been used previously to oxidize
thiol to sulfonic acid in MCM-41 [14,15] and SBA-15 [16], and di-
methyldioxirane, which is able to oxidize thiol groups to sulfones and
eSO3− groups [17,18]. Finally, it has to be considered that the catalytic
cyclohexene oxidation reaction involves molecules with very different
polarity, i.e. cyclohexene and its two main and more polar oxidation
products, cyclohexenol and cyclohexenone [8]. Therefore, it could be
hypothesized that the modification of the hydrophobicity of the cata-
lysts would have an impact on their performance. In order to explore
this hypothesis, SBA-15 precursor materials of desired morphology
were synthesized by a co-condensation method by using mercapto-
propyl and alkyl (propyl and octyl) hydrophobic groups. The Au-con-
taining materials derived from these precursors have been tested as
catalysts for the oxidation of cyclohexene with oxygen in liquid phase at
atmospheric pressure.
2.2. Oxidation of thiol-SBA-15 supports with hydrogen peroxide
0.5 g of the extracted SBA-15 materials were oxidized with 15 mL of
a 15 wt.% H2O2 solution in water under stirring for 24 h, and the re-
sulting solid was filtered and washed with water, treated with diluted
sulfuric acid, washed again with water and dried at room temperature.
The resulting samples were denoted as SH-hp, SPr-hp and SOc-hp,
where “hp” stands for hydrogen peroxide.
2.3. Oxidation of thiol-SBA-15 supports with dimethyldioxirane
A solution of dimethyldioxirane (DMD) in acetone was first pre-
pared according to the following procedure [18]: a 2 L three-necked
round bottom flask was provided with a magnetic bar for magnetic
stirring, introduced in an ice bath and connected to a two-necked re-
ceiving flask. Then, 254 g of water, 192 mL of acetone and 58 g of
NaHCO3 were added. Under vigorous stirring, 120 g of oxone
(2KHSO5.KHSO4.K2SO4, Alfa Aesar, 99.9%) were added slowly in por-
tions at 3 min intervals. After this, a moderate vacuum was applied, the
ice bath removed and the collecting two-necked flask introduced in a
dry ice/ethanol bath, and the dimethyldioxirane/acetone solution
gently distilled and collected in the cooled flask (−78 °C). The product
is stored at 23 °C. Then, 0.7 g of the extracted SBA-15 supports were
treated with 15 mL of the DMD acetone solution under stirring at room
temperature for 24 h, in a 25 mL Erlenmeyer flask covered with Par-
afilm. The reaction mixture was then filtered, washed with acetone and
dried at room temperature. The samples were denoted SH-DMD, SPr-
DMD and SOc-DMD.
2. Experimental
2.4. Synthesis of the gold entities
2.1. Synthesis of thiol-functionalized short-channel SBA-15 silica
The gold nanosized entities were prepared from a two-liquid phases
system according to a method previously described [4,8]. In a typical
procedure, a gold lump (0.094 g, Johnson-Matthey, 99.99%) was dis-
solved under gentle stirring in 34 g of aqua regia, prepared by mixing
(4:1, w/w) nitric acid (Panreac, 65 wt.%) and ammonium chloride
(Sigma-Aldrich, > 98 wt.%), while heating at 40 °C in a sand bath. After
Synthesis of short-channel SBA-15 silica functionalized solely with
thiol and with both thiol and alkyl groups was carried out by a method
adapted from that reported for pure silica materials [12,13]. These
hybrid SBA-15 materials were prepared from gels with molar compo-
sition: 1.0 TMOS: 0.05 MPTMS: x R: 0.077 P104: 6.27 HCl: 303 H2O,
2

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
cooling, the resulting golden yellow solution was placed in a 100 mL
decanting funnel, and then 17 g of rosemary essential oil were gently
added, which remained as a top layer over the gold solution. The
system was then left at room temperature undisturbed, i.e., the two
liquid phases were not mixed, and aliquots of the organic layer were
taken at the desired time to prepare the Au-SBA-15 materials, as it is
described below. The rosemary essential oil was supplied by the
Spanish company El Granero Integral (The Integral Barn) and its che-
mical composition [4,8] was determined by GC–MS employing a gas
chromatograph (Agilent 6890) coupled with a mass spectrometer
(Agilent 5973N) using a capillary column made of methylpolysiloxane
(30 m × 0.25 mm × 0.25 μm), heating from 70 to 290 °C at 6 °C /min.
The oil composition (in wt%) was: 24.9% 1,8-cineole, 21.9% alpha-
pinene, 20.91% camphor, 9.06% camphene, 3.81% borneol, 3.34%
verbenone, 2.59% myrcene, 2.41% beta-pinene, 2.01% caryophyllene,
2.00% p-cymene, 1.16% alpha-humulene, 0.98% bornyl acetate, 0.94%
gamma-terpinene, 0.63% 4-terpineol, 0.63% alpha-terpineol, 0.29%
alpha-terpinolene, 0.28% fenchone.
solution (TBHP, 5 wt.% referred to cyclohexene; ∼5.5 M in decane,
Sigma-Aldrich), 4.2578 g of toluene (75 wt.% of the cyclohexene;
Panreac, > 99.5%), and 0.070 g of catalyst. O2 (1.8 mL/min) was
bubbled through the stirred reaction mixture. The catalysts were pre-
viously heated at 100 °C for 1 h in the reaction flask equipped with a
tube containing molecular sieve 5A in order to remove the traces of
water coming from the catalysts, subsequently cooled down to the re-
action temperature of 65 °C, and then the reagents were added. The trap
was removed and replaced by the condenser before the reaction started.
We have decided to follow ref. [19] on the use of TBHP as initiator,
although the same group reported later that oxidation reactions can be
carried out free of initiator [20] provided the cyclohexene does not
contain a chemical stabilizer. However, the subject is still controversial
(see [21] and references therein). As we cannot guarantee our cyclo-
hexene source is stabilizer-free, we have preferred to keep using TBHP
as in our previous investigation [8]. Aliquots of 0.20 mL of the reaction
mixture were taken at given time intervals, filtered through a 4 mm
PTFE filter of 0.45 μm and analysed by GC in a Varian CP-300 instru-
ment, by using a FactorFour™ (Varian VF-1 ms) dimethylpolysiloxane
capillary column of 15 m of length and 0.25 mm of i. d. Octane was
used as internal standard. Five reaction products were identified, two of
them resulting from the addition of oxygen to the cyclohexene double
bond, namely cyclohexene epoxide and cyclohexanediol, and three
coming from the allylic oxidation of the cyclohexene ring: 2-cyclo-
hexen-1-ol, 2-cyclohexen-1-one and 2-cyclohexenyl hydroperoxide.
Cyclohexene conversion was calculated from the yields of these five
products.
2.5. Immobilization of the gold entities on SBA-15
Aliquots of 3.75 mL of the organic layer were taken after 24 h of
addition of the rosemary oil to the gold solution and each of them was
mixed with 18.75 mL of ethanol. To this solution, 0.5 g of the SBA-15
materials were added and the mixture was stirred at room temperature
for 4 h. After that, the solid was separated by centrifugation and washed
with four portions of 40 mL of ethanol each. The corresponding Au-
containing samples were denoted as SH-Au, SH-hp-Au, SPr-Au, and the
like. The routes followed to prepare the catalysts are displayed in
Scheme 1.
2.7. Characterization techniques
Powder X-ray diffraction was carried out using a PANalytical X’pert
Pro instrument (Cu Kα radiation). Gold content of the solid was de-
termined by inductively coupled plasma (ICP-OES) spectrometry with
an ICP PlasmaQuant PQ 9000 Analytik Jena spectrometer.
Thermogravimetric analyses were performed in a Perkin-Elmer TGA7
instrument, in air (40 mL/min) with a 20 °C/min heating ramp from 25
to 900 °C. CHNS elemental analyses were done in a LECO CHNS-932
analyser provided with an AD-4 Perkin-Elmer scale. Nitrogen adsorp-
tion-desorption isotherms were measured in a Micromeritics ASAP
2420 apparatus at the temperature of liquid nitrogen (−196 °C). The
samples were degassed in situ at 70 °C in vacuum for 16 h prior analysis.
Surface area was determined using the BET method. The pore volume
and the average pore diameter were calculated by applying the BJH
protocol to the adsorption branch of the isotherm. Diffuse reflectance
2.6. Catalytic tests
The oxidation of cyclohexene with molecular oxygen was carried
out by the procedure described in [8], which was developed from that
reported in [19]. The catalytic reaction was carried out in a 50 mL glass
four-neck round-bottom flask, provided with a condenser through
which water at 5 °C was circulated to minimize the evaporation of re-
agents and products. The flask was immersed in a silicone oil bath to
keep the reaction temperature at 65 °C, measured inside the reaction
mixture by a thermometer inserted in one of the necks. The experi-
mental set-up was not protected against environmental light. The re-
action mixture was made of 5.6770 g of cyclohexene (0.0676 mol,
Sigma-Aldrich, 99%), 0.5677 g of octane (10 wt.% referred to cyclo-
hexene; Sigma-Aldrich, > 99%), 0.2839 g of a tert-butyl hydroperoxide
UV–vis (UV–vis) spectra were recorded on
a Cary 5000 Varian
Scheme 1. Routes followed to prepare SBA-15 samples functionalized with mercaptopropyl (and alkyl) groups, their oxidized counterparts and corresponding gold-
containing catalysts.
3

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
spectrophotometer equipped with an integrating sphere with the syn-
thetic polymer Spectralon as reference. The data were expressed in
absorbance units. MAS NMR spectra were recorded with a Bruker AV
400 WB spectrometer. 1H to 13C cross-polarization (CP) spectra were
recorded using a π/2 rad pulses of 4.5 μs for 1H, a contact time of 5 ms
and a recycle delay of 3 s. For the acquisition of the 13C spectra, the
samples were span at a rate of 5–5.5 kHz. 29Si MAS NMR spectra were
recorded at 79.5 MHz using a 4 mm probe. Transmission electron mi-
croscopy studies were carried out using a JEOL 2100F electron micro-
scope operating at 200 kV. The catalysts were dispersed in ethanol and
dropped onto a holey carbon copper grid for the observations. Scanning
electron microscopy was carried out using a Field Emission Phillips
XL30 S-FEG instrument. The samples were coated with Cr before ob-
servation.
X-ray photoelectron spectra (XPS) have been collected using a
SPECS instrument with UHV system (pressure in the range of 10−10
mbar), equipped with a PHOIBOS 150 9MCD energy analyzer. A non-
monochromatic Mg Kα (1253.6 eV) X-ray source was used, working at a
power of 200 W, with an acceleration voltage of 12 kV. High-resolution
regions were recorded with a pass energy of 20 eV. For analysis, the
powder samples were pressed into self-supporting wafers and stuck on
the sample holder with double-sided adhesive conductive carbon tape.
Spectra were referenced against the O 1s emission line (binding energy
set to 532.9 eV) to correct for charging effect. Decomposition of ex-
perimental peaks into components (70% Gaussian, 30% Lorentzian)
was done using a non-linear, least squares fitting algorithm and a
Shirley baseline. Relative atom ratios were calculated from the sum
area of the core-level components of Si, S and Au, using the relative
sensitivity factors provided by Casa XPS software.
Table 1
Structural properties of the as-made materials and structural and textural
characteristic of the extracted samples.
Sample d-Spacing
d100(nm)
Unit cell parameter
a0(nm)
SBET
Vp (cm3/ Dp (nm)
g)
(m2/g)
SH
SH-ext
SPr
SPr-ext
SOc
SOc-ext 9.6
8.72
8.69
7.79
7.60
10.2
10.1
10.0
9.0
8.8
8.9
661
497
426
0.55
0.34
0.40
4.5
2.6
3.2
8.3
hydrocarbon chain length increases. The pore size of SH is 4.6 nm, and
it is reduced to ∼3 nm for the samples prepared from propyl (SPr) and
octyl (SOc) containing gels, which again suggests their effective in-
corporation into the pores, decreasing in this way their average free
diameter.
The inspection of the SH-ext sample by TEM reveals the presence of
some particles with thin plate shape and developed (001) faces
(Fig. 2a). Moreover, it shows a very well ordered hexagonal arrange-
ment of pores, as evidenced also by the electron diffraction pattern
obtained along the channel direction (inset in Fig. 2a). Long rod-like
particles have never been observed. The platelets exhibit a small
thickness, with a maximum of 300–400 nm, consistent with the thick-
ness of the all-silica SBA-15 platelet particles and the nanorods forming
those particles described previously [12,13]. The images taken per-
pendicular to the channel direction (Fig. 2b) show that the channels run
along the short dimension of the platelets. TEM images of the two alkyl-
containing samples, SPr-ext and SOc-ext, show the presence of very tiny
particles of irregular morphology whose size is < 200 nm, with a rela-
tively disordered arrangement of pores, in agreement with their XRD
pattern discussed above. SEM images (Fig. 3) confirm the presence of
particles of irregular shape in the SPr and SOc samples, with sizes in
general below 200 nm, being somewhat larger in the former. Moreover,
no rod-like particles have been found. Therefore, regardless the per-
fection of the hexagonal pore arrangement and the fact that for the two
samples functionalized with propyl and octyl groups no plate-like
particles have been formed, in all these cases SBA-15 particles having
short channels have been achieved, which was one of the main goals of
the research. However, further investigation would be needed in order
to improve the pore arrangement and particle morphology for these
functionalized materials.
The presence of sulphur in the three samples suggests the effective
incorporation into the solid of the mercaptopropyl groups present in the
synthesis gel, but it can be observed that the amount of this element
slightly decreases from 1.78 wt.% for SH, to 1.51 wt.% and 1.39 wt.%
for the samples prepared from gels containing octyl (SOc) and propyl
(SPr) groups, respectively (Table 2). This evidences a competition be-
tween the two functional groups for their incorporation into the SBA-15
structure. It can be seen also in Table 2 that the C/S ratio is higher than
that of the propylthiol moiety, which suggests that other carbon com-
pounds are present in the materials. Indeed, the 13C CP MAS NMR
spectra that will be discussed below show the presence of ethoxy
groups, as well as some minor amount of residual surfactant. TG ana-
lysis of sample SH (Fig. 4) shows a weight loss at T < 100 °C attributed
to water desorption, and a sharp one at higher temperature, centered at
320 °C, which can be assigned to the desorption/combustion of pro-
pylthiol groups in SBA-15 materials [8,16]. The samples functionalized
with alkyl groups in addition to mercaptopropyl present TG patterns
qualitatively similar to that of SH (Fig. S1 of Supplementary informa-
tion for sample SPr and Fig. S2 for sample SOc), but the weight loss
centered at ∼300 °C increases with the chain length of the hydrocarbon
moiety (Table 2). This increase of weight loss is more relevant if it is
taken into account that the amount of sulphur decreases in the same
way. This behavior suggests that the mercaptopropyl and alkyl moieties
3. Results and discussion
3.1. Extracted samples
The XRD pattern of the extracted thiol-containing SH material
(Fig. 1) shows the profile characteristic of SBA-15 mesostructure, with
an intense d100 reflection and unit cell parameter (a0) of 10.1 nm
(Table 1). The introduction of propyl and octyl groups together with
mercaptopropyl produces a visible broadening of the d100 reflection,
which suggests that the alkyl groups have been effectively incorporated
in the co-condensation process of the corresponding alkoxysilanes and
are present in the pore surface in addition to mercaptopropyl moieties,
perturbing somehow the micellar arrangement of the P104 molecules
that built up the structure. The surface area decreases with the in-
corporation of alkyl groups and the effect is more pronounced as the
Fig. 1. XRD patterns of samples SH-ext, SPr-ext and SOc-ext.
4

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 2. (A) TEM Images of sample SH-ext along the channel direction showing the hexagonal packing of the pores. Low magnification particle morphology and
electron diffraction pattern are included in the insets corroborating that the channels are running along the short dimension of the crystals, which are close to
hexagonal plates. (B) TEM image of sample SH-ext perpendicular to the channels, (C) sample SOc-ext and (D) sample SPr-ext.
are desorbed/decomposed at roughly the same temperature, and that at
least part of the extra-weight loss of these two samples as compared to
SH could be attributed to alkyl chains present in the structure, although
the co-existence of several organic species in the material preclude a
quantitative determination of their content. 13C CP MAS NMR will show
the effective incorporation of hydrocarbon moieties in the samples.
formed by reaction of the hot ethanol used to remove the surfactant
with the silanol groups present in the as-made material. Finally, the
band at 70 ppm corresponds to residual surfactant [23]. Oxidation of
this sample with hydrogen peroxide largely removed the 26 ppm band,
while new ones appeared at 54 ppm and 18 ppm. The former has been
assigned to the α carbon atom attached to the sulfur atom in the sul-
fonic acid group, and the latter to the β carbon atom of the propyl
sulfonic group [23], while the C atom bonded to Si is responsible for the
11 ppm resonance. The bands at 16 ppm and 70 ppm are associated to
residual surfactant. It is interesting to notice that the oxidation treat-
ment largely decreases the intensity of the bands associated to ethoxy
groups, in agreement with the chemical analysis. If the sample is oxi-
dized with DMD, the band at 26 ppm corresponding to the mercapto-
propyl group is completely removed, while a new and intense one ap-
pears at 54 ppm, as well as another one at 18 ppm, while that at 11 ppm
remains unchanged. It has been reported that dimethyldioxirane is a
very efficient agent for the oxidation of organic sulfides and thiol
groups to sulfone and sulfonic groups [18], and hence these three bands
at 54, 18 and 11 ppm should correspond, respectively, to the C atoms in
α and β positions of the propylsulfonic moiety and to the C atom linked
to the Si atom. Moreover, the DMD treatment removes all the ethoxy
groups, as the resonances of the corresponding C atoms have vanished.
The oxidation of the thiol groups has also been monitored by XPS.
Fig. 6 shows the S 2p core level spectra of sample SH-ext before and
after DMD oxidation. The broad peak of the SH-ext sample is asym-
metric in the low binding energy (BE) side, which evidences that it
3.2. Oxidized samples
Oxidation of the extracted samples with either hydrogen peroxide or
dimethyldioxirane (DMD) produced significant changes in the nature of
the samples that can be monitored by a combination of chemical and
thermal analysis, XPS and NMR spectroscopy. Significant differences
have been found in the way the two oxidation procedures affect the
starting materials. The first observation common to all oxidized samples
is that none of the oxidation procedures removed sulphur in a sig-
nificant manner, as the content of this element is quite similar to that of
the extracted samples (Table 2). The consequence of the oxidation
process on the nature of the functional groups present in the solid can
be conveniently monitored by 13C CP MAS NMR spectroscopy (Fig. 5).
In the spectrum of sample SH, resonance signals at 11 ppm and
∼26 ppm correspond, respectively, to the carbon atom attached to Si in
the mercaptopropyl moiety and the two remaining carbon atoms of the
propyl chain [22,23]. The signals at 59 ppm and 16 ppm correspond,
respectively, to the C atom attached to oxygen and the terminal methyl
groups in the eOeCH2eCH3 ethoxy group [23]. These groups are
5

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
thiol to sulphur species in high oxidation state. Moreover, the BE value
−
is closer to that of eSO3 than to sulfone. In order to distinguish be-
tween both groups, the acidity of the solid has been determined by base
titration, and it has been found to be quite similar to that of the sample
treated with hydrogen peroxide, which contains sulfonic groups.
−
Therefore, eSO3 species would be present in the sample treated with
DMD. It can be seen in Table 3 that the S/Si atom ratio at the surface is
nearly half that of the bulk, which indicates a concentration gradient of
sulphur across the SBA-15 particles that would have a S-rich inner core.
The occurrence of silane units in the extracted and oxidized samples
can also be monitored by 29Si MAS NMR (Fig. 7). The signals appearing
in the spectra at −92, −101 and −110 ppm correspond to Q2, Q3 and
Q4 species, respectively, where Qn refers to [Si(OSi)n(OH)4-n] environ-
ments, with n = 0,1, 2, 3 and 4. It can be noticed the high concentration
of silanol SieOH groups, as evidenced by the intensity of the corre-
sponding −101 ppm signal. The resonance at -66 ppm is attributed to
T3 sites, silane species (those having CeSi bonds) in environments de-
scribed as RSi(OSi)3 [29,30].
The oxidation of the samples also leaves its footprint in the TGA
profile (Fig. 4), which contributes to clarify further the outcome of the
oxidation process. For SH sample, oxidation with hydrogen peroxide
gives rise to a new weight loss centered at ∼250 °C, while that corre-
sponding to thiol groups decreases in intensity but it is still clearly
visible as a sharp thermal event centered at 330 °C. Thermal events at
230 °C–260 °C are associated to the presence of sulfonic moieties in
mesoporous materials and sulfonic resin Amberlyst-15 [8,16]. The total
weight loss at T > 140 °C is smaller than that of the extracted material,
owing to the removal of the ethoxy groups, in agreement with the NMR
results. It can be also noticed that the amount of adsorbed water
(weight loss at T < 140 °C) increases with the treatment. From these
results, it could be concluded that the treatment of this sample with
hydrogen peroxide only oxidized partially the thiol to sulfonic acid.
However, oxidation with DMD produces deeper changes in the TGA
pattern (Fig. 4, SH-DMD sample). In this case, two high-temperature
thermal events are identified, centered at ∼270 °C and 500 °C, being
the former more intense than the latter, while no event associated to the
presence of thiol groups is observed. Moreover, the most intense weight
loss in mesoporous SBA-15 materials containing sulfonic groups occurs
at T ∼ 350 °C [8,13]. Then, this result confirms the effective oxidation
of thiol to sulfonic groups, in agreement with the XPS results discussed
above.
Fig. 3. SEM images of samples SPr-ext (top) and SOc-ext (bottom).
Table 2
Chemical composition (wt.%) and TG analysis results (wt.%) of supports.
Sample
C
H
S
C/Sa
Weight loss
T < 140 °C
Weight loss
T > 140 °C.
As we have seen that the chemical changes produced in the thiol
groups by the oxidation treatments can be conveniently monitored by
TG, we have used this methodology to check their effect on the two
samples functionalized with propyl (SPr) and octyl (SOc) groups in
addition to mercaptopropyl. The TGA pattern of the SPr sample oxi-
dized with hydrogen peroxide (SPr-hp in Fig. S1) strongly resembles
that of sulfonic acid-containing materials, with thermal events asso-
ciated to the decomposition/oxidation of organic material detected at
250 °C, 370 °C (being this one the most intense) and 510 °C. This evi-
dences an efficient oxidation of thiol to sulfonic acid groups. The TG
profile of the sample treated with DMD is quite similar to that of sample
SH-DMD, suggesting again the presence of sulfonic groups, although the
main weight loss shifts to slightly higher temperature and is centered at
320 °C. Finally, the treatment of sample SOc with H2O2 does not reveal
significant changes in the TGA pattern (Fig. S2), other than reducing
the weight loss at T > 140 °C. This evidences that the treatment is not
suitable for oxidizing the thiol groups in this case. The presence of long
C8 hydrophobic alkyl chains in the material might reduce the ability of
hydrophilic hydrogen peroxide to reach the thiol groups located at the
surface, under the conditions described in this work. The higher hy-
drophobicity of this sample as compared with that of SH and SPr ma-
terials is evidenced by its much lower amount of adsorbed water, as
revealed by the weight loss at T < 140 °C, (Table 2). However, DMD is
as efficient as it was for the two other samples to oxidize thiol to sul-
fonic groups, being the TGA pattern quite similar to that of sample SH-
SH-ext
SH-hp
SH-DMD
SPr-ext
SPr-hp
SPr-DMD 7.67
SOc-ext
SOc-hp
6.38
8.18
7.95
2.48 1.78
2.7 1.67
3.47 1.47
9.56
5.75
17.07
15.89
20.90
21.30
14.31
18.68
28.05
24.47
27.65
13.05 8.47
14.43 14.72
20.30 6.59
12.77 7.15
14.59 9.37
26.44 4.32
23.48 5.06
26.79 3.57
10.57 3.20 1.39
5.99
2.35 1.25
2.97 1.40
14.98 3.54 1.51
13.30 3.28 1.51
SOc-DMD 14.37 3.50 1.43
a
Mol ratio.
envelopes the S 2p3/2 and S 2p1/2 components with 2:1 intensity ratio
due to the characteristic spin-orbit splitting [24]. The BE at the max-
imum intensity of the peak, which is mostly contributed by the S 2p3/2
component, is 163.5 eV, which corresponds to SH groups [25]. After
oxidation of this sample with DMD, this peak disappears completely
and it is replaced by a new one at higher BE, also asymmetric and tailed
at the low BE side, centered at 168.8 eV. Unresolved S 2p peaks at
168.0 eV have been assigned to sulfone -S(O)2- group in poly(ether
sulfone) [26]. S 2p3/2 peaks at BE in the range 168.1–168.5 have been
−
assigned to sulfonic [25] or sulfonate group eSO3 [27], and it has
been also reported that the BE of the sulfonate group is sensitive to the
countercation [28]. Therefore, the BE of 168.8 eV can be assigned to S
(VI) chemical species and evidences that DMD oxidizes quantitatively
6

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 4. TGA/DTG curves for SH samples (left) and the corresponding Au-containing materials (right).
DMD.
trend is found for oxidized SOc material containing octyl groups, as in
this case the amount of gold in SOc-DMD sample is further reduced to
only 0.23 wt.%, and pre-oxidation with H2O2 already decreases it to
0.7 wt.%. The sulphur content remains nearly unchanged after Au im-
mobilization, and the S/Au ratio of the samples is in the range 8–40. For
some of them, a small increase in the total carbon content is found
(Table 4).
The XPS spectrum of sample SH-Au (Fig. 6) shows the presence of
two S 2p peaks at BE energies 163.6 eV and 168.1 eV, being the in-
tensity of the last one 57% of the total S intensity. The low BE peak
corresponds to thiol groups still present in the sample, while the high
BE peak is assigned to S in eSO3−, as it was discussed above. This
evidences that a major fraction of the thiol groups present on the
3.3. Gold-containing materials
SH-ext and all the oxidized samples put in contact with aliquots of
the rosemary oil solution as described in the experimental section are
able to immobilize gold, whose amount varies for different materials
(Table 4). It is nearly 1.1 wt.% for sample SH, and it slightly decreases
for its corresponding oxidized materials, but it is similar for both of
them. Pre-oxidation of sample SPr with H2O2 does not change its ability
to capture gold from the rosemary solution, as the metal content is quite
similar to that of sample SH. However, pre-oxidation with DMD clearly
reduces the gold content (0.46 wt.%). A similar though more marked
7

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 5. 13C MAS NMR spectra of selected samples before (left) and after (right) gold immobilization.
sample surface has been oxidized upon contacting the material with the
Au-containing essential oil layer. The oxidant activity of the essential
oil that has been in contact with the solution of gold in aqua regia
should be linked to the changes that we have observed in its colour as a
function of time. The oil starts to acquire a deep reddish coloration after
one hour, becoming brown-red after one day (Fig. S3). These changes
strongly suggest that the terpenic compounds initially present in the oil
would react with the strongly acidic and oxidant solution of gold in
aqua regia, and the compounds resulting from this reaction (which have
not been identified yet) should be able to oxidize thiol to sulfonic
groups. The S 2p spectrum of the SH-DMD-Au (Fig. 6) is quite similar to
the corresponding sample before Au immobilization, evidencing the
presence of eSO3−. We do not observe in the spectrum any S 2p signal
at BE < 163.0 eV that would eventually be associated to S-linked to Au
[24], within the resolution power and sensitivity of our XPS instrument.
It will be shown below that the S/Au surface ratio is > > 1, and then
most of the S atoms would not be involved in Au binding, and therefore
the overwhelming majority of S-free groups would overshadow S atoms
eventually involved in Au binding.
The gold immobilization process produced in general noticeable
changes in the TGA patterns of the samples, which can be interpreted
on the basis of the changes in the chemical nature of the S-bearing
groups revealed by the XPS spectra. The TG profiles of the SH samples,
whether pre-oxidized or not, are quite similar (Fig. 4). The weight loss
at T < 160 °C is composed of two thermal events, one at lower T due to
water desorption, and the other centered at T ∼ 100 °C, probably
arising from the desorption of more strongly held water or eventually
from volatile organic products deposited in the materials during the
immobilization of gold. Several thermal events can be distinguished at
higher temperature, centered at ∼240 °C, ∼330 °C and ∼520 °C, being
Fig. 6. S 2p core-level spectra of samples (a) SH-ext, (b) SH-Au, (c) SH-DMD
and (d) SH-DMD-Au. Scatter points correspond to raw data and lines to curve-
fitting results.
Table 3
XPS results.
Sample
BE (eV)
S 2p
Surface atomic ratio
Bulk atomic ratio
S/Si
Au 4f7/2
Au 4f5/2
S/Si
Au/Si
Sox/Sa
Au/Si
SH-ext
SH-Au
163.5
163.6
168.1
168.8
168.5
–
84.5
–
88.1
0.020
0.018
–
0
0.57
0.043
0.043
–
0.0012
0.0046
SH-DMD
SH-DMD-Au
–
84.6
–
88.2
0.024
0.015
–
1
1
0.043
0.041
–
0.0028
0.0038
a
Fraction of oxidized sulphur species.
8

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 7. 29Si MAS NMR spectra of selected samples before (left) and after (right) gold immobilization.
groups are still present in this sample, in agreement with what was
concluded from TGA. A signal at 59 ppm previously assigned to the C
atom attached to O in the ethoxy group in the extracted samples is
observed in all the spectra, which suggests that these groups have been
restored due to the treatment of the samples with the ethanolic solution
of the Au-containing rosemary oil in the presence of the sulfonic groups.
In the spectrum of the SOc-hp-Au sample, four additional resonance
signals are detected at 12.4 ppm, 22.8 ppm, 29.6 ppm and 32.5 ppm,
assigned [31] to the C atoms of the octyl chain indicated in Fig. 7. The
29Si MAS NMR spectrum of this sample (Fig. 7) shows the presence of
two signals at −65 and −54.7 ppm, being the former more intense
than the latter. They are attributed respectively to T3 and T2 sites,
where Tm corresponds to silane species (those having CeSi bonds) in
environments described as [RSi(OSi)m(OH)3-m], m = 1, 2 and 3 [29].
These signals are more intense than those of the SH samples (Fig. 7),
while its sulphur content is only slightly lower, which clearly evidences
the presence of Si-C8 units at the channel surface.
To summarize, the immobilization of gold produces the oxidation of
the S-bearing functional groups initially present in the materials to
sulfonic acid, although minor amounts of thiol groups might still be
present in some of the samples that were not pre-oxidized or they were
so with hydrogen peroxide prior gold immobilization. This information
is relevant for a proper understanding of the catalytic activity of these
materials, as it will be discussed below.
Fig. 8 collects the UV–vis spectra of the Au-containing materials. In
general, two bands of low intensity are observed at ∼250 nm and
∼300 nm, although they are hardly observed in some of them. The
latter band is in the 300–450 nm range reported for 0.8 nm gold clusters
stabilized with thiol groups [32], while UV bands at ∼250 nm have
been assigned to gold clusters having a partial positive charge [33].
However, it has to be considered that SPr-hp support also presents a
band of weak intensity at ∼250 nm, which could eventually be due to
minor amounts of organic species adsorbed during gold deposition and
still retained after washing. Nevertheless, the surface plasmon re-
sonance band at 520 nm characteristic of Au nanoparticles is not de-
tected, save for sample SPr-hp-Au, where a band of very weak intensity
is found at this wavenumber. This sample is the one with the highest
gold content, although it is still low, 1.17 wt.%, and very similar to the
1.09 wt.% of sample SH–Au, where no plasmon band is detected.
Moreover, it has not been detected either in conventional SBA-15
samples with 3.5 wt.% of sulphur and 2.57 wt.% of Au prepared by the
same procedure described here [8]. It is then possible that the alkyl
groups present in the SPr material facilitate some agglomeration of gold
species during the immobilization process. As no AuNPs are detected in
Table 4
Chemical composition (wt.%) and TG analysis results (wt.%) of Au-containing
materials.
Samplea
C
H
S
Au
C/Sb
Weight loss
T < 140 °C
Weight loss
T > 140 °C
SH-Au
10.13 2.91 1.67 1.09 16.17 10.65
10.27 2.73 1.66 0.82 16.49 6.64
8.35
8.25
15.46
17.28
17.46
15.22
18.80
22.05
19.95
SH-hp-Au
SH-DMD-Au
SPr-hp-Au
SPr-DMD-Au 8.94
SOc-hp-Au
2.98 1.55 0.87 14.37 11.42
2.74 1.32 1.17 16.64 10.93
3.02 1.41 0.46 16.89 10.61
13.40 3.26 1.51 0.70 23.66 7.41
SOc-DMD-Au 12.67 3.36 1.48 0.23 22.80 7.16
a
Traces of N ranging from 0.1 wt.% to 0.5 wt.% are found in the samples.
Mol ratio.
b
most pronounced the second one. These events are characteristic of
sulfonic groups, as it has been discussed above. However, only in
sample SH-DMD-Au all S-bearing groups have been shown by XPS to
−
belong to eSO3 species, and therefore it should be associated to the
smooth weight loss observed at ∼330 °C. In contrast to this, a more
intense 330 °C step is observed in sample SH-Au, which still contains
thiol groups as evidenced by XPS, whereas for sample SH-hp-Au the
second weight loss is even steeper and closer to 300 °C. These TGA
features suggest that part of the thiol groups might still be present in the
two last samples. We have reported previously [8] that the rosemary oil
phase put in contact with the gold aqua regia solution for one day is not
able to oxidize the mercaptopropyl groups present in conventional SBA-
15 materilas having rod-shape morphology, containing therefore very
long channels. Then, it can be concluded that the oxidation of thiol
groups is facilitated by the much shorter channels existing in the SH
sample, although the oxidation is not complete, as thiol groups remain
in the Au-materials. TGA profiles of the propyl-(SPr) and octyl-(SOc)
samples are qualitatively similar to those of SH materials (Figs. S1 and
S2), although sample SOc-hp-Au still shows a narrower weight loss
centered at 300 °C, which would suggest the presence of some un-
reacted thiol groups. 13C CP MAS NMR spectra of SH samples reveal
further details of the functional groups present in the Au-containing
materials (Fig. 5). Sample SH-DMD-Au presents an intense resonance
signal at 54 ppm corresponding to the C atoms attached to the sulfonic
group in the propyl chain, as it has been discussed above and in
agreement with the XPS spectrum. However, the intensity of this signal
in the SH-hp-Au sample is lower and the signal at 26 ppm characteristic
of the mercaptopropyl group is still detected, which indicates that thiol
9

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 9. Au 4f core-level spectra of samples (a) SH-Au and (b) SH-DMD-Au.
Scatter points correspond to raw data and lines to curve-fitting results.
nanoparticles in the size range of 1–3 nm linked to dodecanethiol [7]. A
shift of the core-level energy of Au clusters has been reported as a
function of the Au nanocluster size, in such a way that the XPS BE as
well as peak width have been found to increase with decreasing cluster
size [24,35,36]. Therefore, the BE found here for these two Au-bearing
materials would be consistent with the presence of small gold na-
noclusters, but no AuNPs, in agreement with the UV–vis spectra.
However, the presence of some Au(I) species at the cluster surface that
−
would interact with the eSO3 groups cannot be excluded.
3.4. Catalytic activity
The Au-samples were used as catalysts for the oxidation of cyclo-
hexene in liquid phase by using molecular oxygen as oxidant. Fig. 10
depicts the conversion of cyclohexene versus reaction time for the SH
catalysts (left) and alkyl-functionalized SPr and SOc materials (right).
The activity of sample SH prior gold deposition has also been included
for comparison purposes. It can be noticed that the activity of sample
SH-Au approaches 40% at 24 h reaction, and keeps increasing up to
60% at 48 h. This behavior is in agreement with the presence of sulfonic
acid groups, as it has been reported that Au-SBA-15 materials functio-
nalized with thiol groups are inactive for this reaction under the con-
ditions here described [8]. As the Au content of the several samples is
different, it would be convenient to use the turnover number (TON,
defined as the number of moles of cyclohexene converted per mole of
Au) for comparison purposes, which are collected in Table 5 for 24 h of
reaction. TON of SH-Au catalyst is nearly three times that of Au im-
mobilized over rod-shape SBA-15 containing sulfonic groups prepared
by the same two-liquid phases procedure describe here [8], which
evidences the benefits of using the short-channel SBA-15 material as
support. It is also interesting to notice that the catalyst is still active at
long reaction times, a behavior that has not been seen before [8], nor it
is shared by the rest of catalysts reported here. The activity (TON) of the
sample pre-oxidized with DMD is slightly higher than that of SH at 24 h,
but it does not increase when keeping the reaction running for an ad-
ditional 24 h period. In contrast to this beneficial DMD oxidation, pre-
oxidation with H2O2 led to less active catalysts, although the difference
in activity was small (Table 5).
Fig. 8. UV–vis spectra of Au-containing materials before and after (samples
labeled as -React) reaction.
the SOc samples, containing octyl groups and very low amount of gold,
this suggests that the metal aggregation could be regulated by the
density of sulfonic acid groups and their interaction with the gold
species, the content of metal and the eventual presence of hydrophobic
groups that would interact only weakly with the immobilized gold
entities.
XPS spectra of the Au 4f spectra of SH-Au and SH-DMD-Au provide
additional clues on the nature of the gold species present in these ma-
terials (Fig. 9). Two broad and symmetrical signals of Au 4f7/2 and Au
4f5/2 are detected, respectively, at BE energies of 84.5 eV and 88.1 eV
for sample SH-Au, and at 0.1 eV higher BE for SH-DMD-Au. There is no
evidence of more than one oxidation state of Au. The BE of these signals
is about +0.5 eV higher than that commonly found for Au(0), which is
The effect of pre-oxidation with DMD or H2O2 of the catalysts
containing C3 and C8 alkyl groups varies according to the chain length
(Table 5). For the former (SPr catalysts), DMD is also a preferred oxi-
dant, but the corresponding TON values are roughly the same as those
of alkyl-free SH samples. However, the TON values of C8-functionalized
catalysts (SOc) are clearly higher, being again DMD the most favorable
pre-oxidation agent. It was discussed above that DMD can convert
83.9
0.1 eV for Au 4f7/2 [24], but it is also far from the band at
84.9 eV reported for Au(I) in a gold octanethiol complex [34]. More-
over, a BE of 83.8 eV has been reported for the Au 4f7/2 core-level of Au
10

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 10. Cyclohexene conversion as a function of reaction time. Left: SH catalysts. Right: SPr and SOc catalysts.
point. The selectivity to enone increases continuously and it becomes
the dominant product for conversion > ∼20%. The complex evolution
pattern of the several reaction products as a function of conversion
suggests a complex reaction network, where the enol seems to be a
primary product, while the hydroperoxide is a reaction intermediate
that leads to the enone and enol.
Table 5
TON (x103) values at 24 h. Catalyst L1-8 reported in [8] has been included for
comparison.
SH-Au SH-
SH-
SPr-
SPr-
DMD-Au
SOc-
hp-Au
SOc-
DMD-Au
L1-8 (Ref.
[8])
hp-Au DMD-Au hp-Au
6.5
5.4
7.3
5.4
6.8
8.9
18.9
2.3
4. Conclusions
quantitatively the thiol groups to eSO3−, and this would explain the
higher TON of the supports pre-oxidized with this reagent for each
series of supports, SH, SPr and SOc. In this last case, the TON value
increases by 2.5 times as compared with that of SH-DMD-Au catalyst.
This effect could be most probably attributed to the higher hydro-
phobicity of this material, in agreement with the starting hypothesis
that an increase in hydrophobicity would increase the overall activity
by enhancing the adsorption of the hydrophobic cyclohexene. On the
other hand, these results also evidence the relevance of the adsorption/
diffusion characteristics of the porous support on the performance of
the Au-catalysts.
It has been observed that the color of the catalysts evolves in the
reaction medium, being purple colored at the end. Indeed, UV–vis
spectra of the used catalysts show for all of them the presence of the
characteristic surface plasmon resonance band at ∼525 nm (Fig. 8),
and the presence of AuNPs has been further confirmed by TEM in the
case of sample SH-DMD-Au (Fig. 11). This gradual aggregation process
of the gold clusters initially present in the catalysts that takes place in
the course of the aerobic oxidation of cyclohexene has already been
observed previously by us [8] and reported also for triphenylphosphine-
stabilized Au clusters deposited on SiO2. In the last case, the catalytic
activity has been linked to the formation of metallic particles larger
than 2 nm [37]. It is also remarkable on this regard that, save for SOc
samples, the TON of the different catalysts are relatively close to each
other, which would mean that, in spite of the transient nature of the
metal entities present in the reaction medium, they should be quite
similar for all catalysts.
SBA-15 materials functionalized with mercaptopropyl and alkyl
hydrocarbon groups possessing short channels in the length range of
200–400 nm can be conveniently prepared by using P104 tri-block
copolymer surfactant and appropriate synthesis conditions. The thiol
groups can be efficiently pre-oxidized with dimethyldioxirane (DMD) to
produce sulfonic groups, but the oxidation with hydrogen peroxide is
less efficient. The simultaneous presence of octyl groups prevents to a
large extent the effect of H2O2 on the thiol groups, although it does not
affect the activity of the stronger oxidant DMD. Both pre-oxidized as
well as pristine SH-bearing samples are able to immobilize gold na-
noclusters synthesized by a two liquid-phases method involving the use
of rosemary essential oil as organic phase and a solution of gold in aqua
regia as aqueous phase. In addition, the immobilization process pro-
duces the partial oxidation of S-functional groups to sulfonic groups.
The presence of alkyl groups decreases the amount of gold immobilized
on the catalysts. No AuNPs are in general detected, the metal being
present as small clusters, save in one sample where they coexist with a
small amount of AuNPs. The activity of these materials in the oxidation
of cyclohexene is regulated mainly by the gold content, but their
average activity is much larger than that found previously for Au-cat-
alysts prepared by the same method reported here but supported on
conventional rod-shape SBA-15 possessing long channels, which evi-
dences the benefits of reducing the diffusion pathway through the pore
system. Pre-oxidation of the samples with DMD improves the turnover
number (TON), an effect that has been attributed to the presence of
sulfonic groups in the catalysts. Moreover, it has been found that the
presence of hydrophobic octyl groups also increases notably the TON
values. It has been observed for all catalysts that the gold clusters
evolve in the reaction medium to form gold nanoparticles. This work
evidences the influence of factors affecting the catalytic activity not
directly related to the metal active phase in the course of the cyclo-
hexene oxidation in liquid phase, in particular the chemical composi-
tion, morphology and pore length of the host support.
Selectivity to the five different products identified in this reaction is
depicted in Fig. 12 as a function of total conversion for all catalysts. It
can be seen that allylic oxidation to produce 2-cyclohexen-1-one, 2-
cyclohexen-1-ol and 2-cyclohexenyl hydroperoxide is the preferred re-
action pathway, which accounts for nearly 90% of the total products. At
low conversion (∼5%) the enol is the major product, but its selectivity
decreases rapidly with conversion and it stabilizes around 20% for
conversion > ∼15%. At this conversion, the selectivity to hydroper-
oxide reaches a maximum of ∼40% and decreases rapidly beyond this
11

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 11. TEM images of catalysts SH-DMD-Au before (top) and after (below) reaction.
Fig. 12. Selectivity vs conversion for the five reaction products.
12

R. de la Serna Valdés, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Acknowledgements
[16] D. Margolese, J.A. Melero, S.C. Christiansen, B.F. Chmelka, G.D. Stucky, Chem.
Mater. 12 (2000) 2448–2459.
[17] R.W. Murray, R. Jeyaraman, J. Org. Chem. 50 (1985) 2847–2853.
[18] W. Adam, L. Hadjiarapoglou, Top. Curr. Chem. 64 (1993) 45–61.
[19] M.D. Hughes, Y.-J. Xu, P. Jenkins, P. McMorn, P. Landon, D.I. Enache, A.F. Carley,
G.A. Attard, G.J. Hutchings, F. King, E.H. Stitt, P. Johnston, K. Griffin, C.J. Kiely,
Nature 437 (2005) 1132–1135.
The authors acknowledge the current Spanish Ministry of Science,
Innovation and Universities (formerly the Ministry of Economy and
Competitiveness) for the funding of this study through project
MAT2016-77496-R. The help of Dr. Isabel Díaz with the TEM study is
greatly appreciated.
[20] H. Alshammari, P.J. Miedziak, T.E. Davies, D.J. Willock, D.W. Knight,
G.J. Hutchings, Catal. Sci. Technol. 4 (2014) 908–911.
[21] E. Beletskiy, M.C. Kung, H.H. Kung, Catalytic properties of gold nanoparticles, in:
C. Louis, O. Pluchery (Eds.), Gold Nanoparticles for Physics, Chemistry and Biology,
2nd ed., World Scientific, New Jersey, 2017, pp. 285–318.
[22] I. Diaz, C. Márquez-Álvarez, F. Mohíno, J. Pérez-Pariente, E. Sastre, J. Catal. 193
(2000) 283–294.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at https://doi.org/10.1016/j.cattod.2019.09.014.
[23] J.D. Webb, T. Seki, J.F. Goldston, M. Pruski, C.M. Crudden, Microporous
Mesoporous Mater. 203 (2015) 123–131.
[24] M.-C. Bourg, A. Badia, R.B. Lennox, J. Phys. Chem. B 104 (2000) 6562–6567.
[25] E. Cano-Serrano, G. Blanco-Brieva, J.M. Campos-Martin, J.L.G. Fierro, Langmuir 19
(2003) 7621–7627.
References
[26] J. Hopkins, J.Ps. Badyal, Macromolecules 27 (1994) 5498–5503.
[27] M.M. Nasef, H. Saidi, Appl. Surf. Sci. 252 (2006) 3073–3084.
[28] G. Zotti, S. Zecchin, G. Schiavon, F. Louwet, L. Groenendaal, X. Crispin,
W. Osikowicz, W. Salaneck, M. Fahlman, Macromolecules 36 (2003) 3337–3344.
[29] A.C. Cattaneo, C. Ferar, D.C. Villa, S. Angioni, C. Milanese, D. Capsoni, S. Grandi,
P. Mustarelli, V. Allodi, G. Mariotto, S. Brutti, E. Quatarone, Microporous
Mesoporous Mater. 219 (2016) 210–220.
[30] P. Wu, P. Bai, Z. Lei, K.P. Loh, X.S. Zhao, Microporous Mesoporous Mater. 141
(2011) 222–230.
[31] S. Dirè, V. Tagliazucca, E. Callone, A. Quarante, Mat. Chem. Phys. 126 (2011)
909–917.
[32] G.H. Woehrle, M.G. Warner, J.E. Hutchinson, J. Phys. Chem. B 106 (2002)
9979–9981.
[33] N. Bogdanchikova, A. Pestryakov, I. Tuzovskaya, T.A. Zepeda, M.H. Farias,
H. Tiznado, O.A. Martyniuk, Fuel 110 (2013) 40–47.
[34] A. McNeille, D.H. Brown, W.E. Smith, M. Gibson, L. Watson, J. Chem. Soc. Dalton
Trans. (1980) 767–770.
[1] A.S.K. Hashmi, G.J. Hutchings, Angew. Chem. Int. Ed. 45 (2006) 7896–7963.
[2] A. Corma, A. Leyva-Perez, M.J. Sabater, Chem. Rev. 111 (2011) 1657–1712.
[3] A. Corma, H. García, Chem. Soc. Rev. 37 (2008) 2096–2126.
[4] A. Mayoral, J. Agúndez, I.M. Pascual-Valderrama, J. Pérez-Pariente, Gold Bull. 47
(2014) 161–165.
[5] N. Lemery, Cours de Chymie, L.-C. d’Houry, Paris, 1757.
[6] M. Faraday, Philos. Trans. R. Soc. Lond. 147 (1857) 145–181.
[7] M. Brust, M. Walker, D. Bethell, D.J. Schiffrin, R. Whyman, J. Chem. Soc. Chem.
Commun. (1994) 801–802.
[8] J. Agúndez, L. Martin, A. Mayoral, J. Pérez-Pariente, Cat. Today 304 (2018)
172–180.
[9] L.-F.- Gutiérrez, S. Hamoudi, K. Belkacemi, Catalysts 1 (2011) 97–154.
[10] R.G. Nuzzo, D.L. Allaro, J. Am. Chem. Soc. 105 (1983) 4481–4483.
[11] J.L. Love, L.A. Estroff, J.K. Kriebel, R.G. Nuzzo, G.M. Whitesides, Chem. Rev. 105
(2005) 1103–1169.
[12] P. Linton, J.-C. Hernandez-Garrido, P.A. Midgley, H. Wennerström, V. Alfredsson,
Phys. Chem. Chem. Phys. 11 (2009) 10973–10982.
[13] P. Linton, H. Wennerström, V. Alfredsson, Phys. Chem. Chem. Phys. 12 (2010)
[35] S. Peters, S. Peredkov, M. Neeb, W. Eberhardt, M. Al-Hada, Surf. Sci. 608 (2013)
129–134.
[36] A.A. Tal, W. Olovsson, I.A. Abrikosov, Phys. Rev. B 95 (2017) article number
245402.
3952–3958.
[14] W.W. Van Rhijn, D.E. De Voss, B.F. Sels, W.D. Bossaert, P.A. Jacobs, Chem.
Commun. (1998) 317–318.
[15] I. Diaz, F. Mohino, J. Pérez-Pariente, E. Sastre, Thermo. Acta 413 (2004) 201–207.
[37] B.G. Donoeva, D.S. Ovoshchnikov, V.B. Golovko, ACS Catal. 3 (2013) 2986–2991.
13