Hydrophobic effect of silica functionalized with silylated Ti-salicylaldimine complex on limonene oxidation by aqueous hydrogen peroxide

Article abstract of DOI:10.1007/s12039-015-0965-0

This research work describes the effect of hydrophobicity in inducing the diffusion of alkene substrates to the catalytic active sites. The aim of this research is to improve the catalytic activity by tailoring the degree of hydrophobicity of the catalyst. Silica functionalized with solid non-silylated Ti-salicylaldimine complex was prepared at room temperature by mixing imine ligand and Ti(IV) sulphate solution. The amorphous, solid complex formed was further silylated with octadecyltrimethoxysilane (OTMS) to induce hydrophobicity. The composition of the resulting silica functionalized with silylated Ti-salicylaldimine complex was varied with Ti:OTMS molar ratio in the range of 1:1/4 to 1:4 at room temperature. The successful attachment of alkylsilyl groups to silica functionalized with Ti-salicylaldimine complex was proven by the FTIR and ²⁹Si solid state NMR spectra. The FTIR spectra showed increasing peak area for sp ³ C-H stretching mode (ca. 2919 cm^-1 and 2850 cm^-1) and decreasing peak area for Si-OH band with increasing amount of OTMS. TGA showed less water content with higher amount of alkylsilyl groups in the catalyst. This is in agreement with the lower kinetic rate of water adsorption capacity for the hydrophobic catalysts prepared. It was observed that the hydrophobic, silica functionalized with silylated Ti-salicylaldimine complex exhibited higher substrate conversion and reusability compared to the non-silylated catalyst.

Full text of DOI:10.1007/s12039-015-0965-0

J. Chem. Sci. Vol. 127, No. 11, November 2015, pp. 1905–1917. ꢀc Indian Academy of Sciences.
DOI 10.1007/s12039-015-0965-0
Hydrophobic effect of silica functionalized with silylated
Ti-salicylaldimine complex on limonene oxidation by aqueous
hydrogen peroxide
a
a
b
c
LAI SIN YUAN , SHEELA CHANDREN , JON EFENDI , CHIN SIONG HO and
HADI NUR
a
a,d,∗
Centre for Sustainable Nanomaterials, Ibnu Sina Institute for Scientific and Industrial Research,
Universiti Teknologi Malaysia, 81310 UTM Skudai, Johor, Malaysia
b
Department of Chemistry, Faculty of Mathematics and Science, Universitas Negeri Padang,
Padang 25131, Indonesia
c
Low Carbon Asia Research Center, Faculty of Built Environment, Universiti Teknologi Malaysia,
8
1310 UTM Skudai, Johor, Malaysia
d
Department of Physics, Institut Sains dan Teknologi Nasional, Jl. Moh. Kahfi II, Jagakarsa,
Jakarta Selatan 12640, Indonesia
e-mail: hadi@kimia.fs.utm.my
MS received 27 May 2015; revised 14 July 2015; accepted 2 August 2015
Abstract. This research work describes the effect of hydrophobicity in inducing the diffusion of alkene sub-
strates to the catalytic active sites. The aim of this research is to improve the catalytic activity by tailoring the
degree of hydrophobicity of the catalyst. Silica functionalized with solid non-silylated Ti-salicylaldimine com-
plex was prepared at room temperature by mixing imine ligand and Ti(IV) sulphate solution. The amorphous,
solid complex formed was further silylated with octadecyltrimethoxysilane (OTMS) to induce hydrophobicity.
The composition of the resulting silica functionalized with silylated Ti-salicylaldimine complex was varied with
Ti:OTMS molar ratio in the range of 1:1/4 to 1:4 at room temperature. The successful attachment of alkylsilyl
29
groups to silica functionalized with Ti-salicylaldimine complex was proven by the FTIR and Si solid state
3
−1
NMR spectra. The FTIR spectra showed increasing peak area for sp C-H stretching mode (ca. 2919 cm and
2
−1
850 cm ) and decreasing peak area for Si-OH band with increasing amount of OTMS. TGA showed less
water content with higher amount of alkylsilyl groups in the catalyst. This is in agreement with the lower kinetic
rate of water adsorption capacity for the hydrophobic catalysts prepared. It was observed that the hydropho-
bic, silica functionalized with silylated Ti-salicylaldimine complex exhibited higher substrate conversion and
reusability compared to the non-silylated catalyst.
Keywords. Hydrophobicity; silylation; silica functionalized with silyated Ti-salicylaldimine complex;
limonene oxidation; octadecyltrimethoxysilane.
1
. Introduction
cyclic, branched and long chain olefins cannot diffuse
into the pores to react with the active sites. Thus,
mesoporous or polyoxometalate Ti-containing cata-
Numerous efforts have been made to improve the per-
formance of heterogeneous catalysts in various reac-
tions. One of the factors for such improvement is the
local coordination of the active sites, for instance,
tetrahedral local structure of titanium(IV), which is a
well-known catalytic active site for the epoxidation and
oxidation of organic compounds.1 Although Titanium
Silicalite-1 (TS-1) can achieve high catalytic selectivity
3
4
5
lysts, such as TiO -zeolite, Ti-SBA-15, Ti-POM, Ti-
2
6
MCM-41 and others, which can accommodate larger
size of substrates have been synthesized. Apart from
that, other approaches have been carried out as well i.e.,
modification of surface properties of heterogeneous cat-
,2
7
alysts, such as the sulphation of TS-1, fluorinated
8
9
TiO -ZrO , Nb O /TS-1, silylation of several cata-
lysts
2
8
2
2
5
1
,13–18
in epoxidations, it is microporous and hence, many
and many others, in order to improve the
catalytic activity.
In this research, we focused on the surface modifica-
tion of heterogeneous catalysts. Of late, surface mod-
ification has been widely utilized as it could solve a
∗
For correspondence
1905

1906
Lai Sin Yuan et al.
number of problems, such as leaching and metal poi- 2. Experimental
soning. Previous studies have proven that the catalytic
2
.1 Materials
activity of catalysts containing Ti(IV) can be greatly
enhanced for bulky and long chain organic com-
pounds. However, the leaching of Ti(IV) active sites
The chemicals used are salicylaldehyde (98%, Sigma-
Aldrich), absolute ethanol (Merck), 3-aminopropyltri-
methoxysilane (97%, Sigma-Aldrich), titanium(IV)
sulphate solution (24%, Kanto Chemical Co., Inc.), and
octadecyltrimethoxysilane (90%, Acros Organics) and
molecular sieves 3Å (25 g, Sigma-Aldrich).
10
still can occur and the efficiency of hydrogen per-
11
oxide (H O ) decreases in organic-aqueous oxida-
2
2
tion. Support materials, i.e. silica, contain high density
of silanol groups (SiOH) which is highly hydrophilic.
The water molecules produced from H O easily form
2
2
strong hydrogen bonds with Si-OH and compete with
12,13
H O .
Apart from that, water molecules are able
2
2
2
.2 Methods
to open the oxirane ring of epoxides to form diol,
which can strongly adsorb to the titanium(IV)’s active
2.2a Synthesis of silica functionalized with non-
12,13
sites, causing catalyst deactivation.
results were significantly affected and thereby, some
The catalytic
silylated Ti-salicylaldimine complex: Silica function-
alized with Ti-salicylaldimine complex was synthe-
sized as follows: first, salicylaldimine ligand, which
is generally known as Schiff base ligand, was pre-
pared under nitrogen flow. Salicylaldehyde (0.05 mol,
8
,13–18
researchers
have come out with various silylating
agents to be grafted or attached to the surface of catalyst
to induce hydrophobicity.
Previously, it was reported that the long hydropho-
bic chain of octadecyltrichlorosilane (OTS), which was
attached to fluorinated-TiO -ZrO (F-TiO -ZrO ), man-
5
.30 mL) in absolute ethanol (20.00 mL) and 3-
aminopropyltrimethoxysilane (0.05 mol, 9.30 mL) in
absolute ethanol (20.00 mL) were added drop-wised
into a three-neck round bottom flask containing molec-
ular sieves 3Å (25 g) and absolute ethanol (50.00 mL).
Salicylaldehyde 98% (0.05 mol, 5.3 mL) reacted with
2
2
2
2
aged to achieve higher TON for titanium(IV) active sites
as compared to F-TiO -ZrO in 1-octene oxidation by
2
2
8
aqueous H O . Another research showed that the short
2
2
hydrophobic chain of N-methyl-N-(trimethylsilyl)tri-
fluoroacetamide (MSTFA) has also increased the epox-
ide yields in the epoxidation of cyclohexene with
3
-aminopropyltrimethoxysilane (0.05 mol, 9.3 mL) to
form bright yellowish ligand under mild stirring at room
temperature for a few minutes. The mixture was cen-
trifuged at 4000 rpm for 5 min and then vacuumed to
remove the ethanol solvent. Dark yellowish ligand was
formed after the removal of the solvent. Titanium(IV)
sulphate solution (0.01 mol, 8.9 mL) in 10.0 mL of
absolute ethanol was added into the yellowish ligand
14
13,15–18
H O . Several researches
showed that cata-
2
2
lysts with surface modification of hydrophobic sily-
lated titanium(IV) catalysts performed remarkable
higher activity than the non-silylated counterpart of
titanium(IV) catalysts. These examples proved that
the surface properties1
3–18
of catalysts play signif-
(0.02 mol) in 40.0 mL absolute ethanol. The solution
icant roles in achieving high catalytic activity and
selectivity.
quickly turned into colloidal yellowish suspension. It
was stirred overnight at room temperature to complete
the reaction. The resulting powder was centrifuged at
Herein, this research highlights the types of silylating
agents and catalysts utilized in the oxidation of small
aromatic alkene. Most of the silylating agents used in
previous reports were of short alkylsilyl chain with
trimethylsilyl and hexamethylsilyl groups in the oxida-
tion of small alkenes by aqueous H O . To date, the
4000 rpm for 10 min and washed with ethanol, fol-
lowed by hexane, twice each. The amorphous com-
plex formed was dried at room temperature for a few
days. The weight of the silica functionalized with non-
silylated Ti-salicylaldimine complex was 0.58 g with
2
2
findings of silylation with long alkylsilyl chain are
lessj reported. Besides that, catalysts commonly used
to improve hydrophobicity are metal oxides, especially
2
4% of titanium(IV) in the catalyst.
12–14,17
silylated Ti-MCM-41 and silylated Ti-MCM-48.
2.2b Silica functionalized with silylated of Ti-
Thus, an attempt to produce a different type of cat-
alyst, i.e. silica functionalized with silylated Ti-salic-
yladimine complex is shown here. A series of varia-
salicylaldimine complex: The silylation of Ti-
salicylaldimine complex by octadecyltrimethoxysilane
(OTMS) with Ti:OTMS molar ratio of 1:4 was car-
tions in the amount of octadecylsilyl (C ) group is used ried out as follows: Ti-salicylaldimine complex (10.00
18
to silylate Ti-salicylaldimine complex for the oxida- mmol) was added with OTMS (2.50 mmol, 1.18 mL)
tion of limonene (small, aromatic alkene) by aqueous and dissolved in toluene (10.00 mL). The mixture
H O .
was then stirred overnight. To prevent hydrolysis of
2
2

Silica functionalized with silyated Ti-salicylaldimine complex
1907
OTMS, the preparation process was done in a glove determined by Thermogravimetric Analyzer Mettler
box. A similar procedure was also carried out to for the Toledo using platinum crucible.
silylation of Ti-salicylaldimine complex with different
Ti:OTMS molar ratio (1:1/3, 1:1/2, 1:1, 1:2, 1:3, and
2
.4 Catalytic activity of silica functionalized with
1
:4). The solid catalysts were centrifuged, followed by
non-silylated and silylated Ti-salicylaldimine
complexes
washing with ethanol, toluene and hexane, twice each.
The concentration of Ti(IV) sulphate solution was
maintained at 0.01 mol for the entire synthesis process,
regardless of the OTMS concentration.
◦
The catalytic testing was carried out at 70 C with (R)-
(
+)-Limonene (5.00 mmol, 0.83 mL), 30% H O (0.01
2
2
mol, 0.38 mL) and the catalyst (0.05 g) for 22 h under
magnetic stirring. The products of the reactions were
analyzed by Gas Chromatography, Shimadzu model
GC-2014 equipped with BPX-5 column (30 m length,
2.3 Characterizations
The solid catalysts obtained were characterized by
FTIR, diffuse reflectance UV-Vis spectroscopy (DR
0
.25 mm inner diameter, 0.25 μ m film thickness) and
a flame ionization detector (FID). The temperature pro-
gram of the column oven was performed with the initial
temperature of 60 C, heating rate of 10 C/min and the
temperature was maintained at 100 C for 1 min. After
that, the same heating rate (10 C/min) was applied till
29
UV-Vis), Si solid-state magic angle spinning NMR
29
(
Si MAS NMR) spectroscopy, scanning electron
◦
◦
microscopy (SEM) and thermogravimetric analysis
◦
◦
(
TGA). All samples were preheated at 60 C for a few
◦
hours to remove any adsorbed water. In order to con-
firm the successful removal of water content in the
samples, the weight of the samples were measured
before and after heating until a constant weight of sam-
ples was obtained. There were no fixed heating hours
because each of the catalysts gave their constant weight
at different times according to their degree of silyla-
tion. It was observed that shorter time was required
for catalysts with higher coverage degree of hydropho-
bic surface with alkylsilyl groups compared to the
more hydrophilic surface with lower coverage degree of
alkylsilyl groups.
◦
1
38 C and then maintained for 2 min. Following that,
◦
the heating rate was increased to 15 C/min by heat-
ing up to 150 C and maintained for 1 min to elute the
products left inside the column.
◦
Limonene (organic phase) and aqueous H O (aque-
2
2
ous phase) are an immiscible liquid-liquid system. In
the reaction process, stirring was required to remove the
phase-boundary between limonene and aqueous H O ,
2
2
in order to form one homogeneous phase. After the
reaction has been stopped, the liquids separated back
as two phases. The upper organic phase containing the
products (carveol, carvone, epoxides, etc.) was sepa-
rated and taken for gas chromatography analysis. Unde-
niably, some products might have dissolved in the aque-
ous phase, but the amount is insignificant because the
products are all nonpolar liquids. Thus, the products
possess higher solubility in the organic phase com-
pared to the aqueous phase. Leaching was not tested in
this experiment as this research serves as a preliminary
experiment to investigate the hydrophobicity effects
towards limonene oxidation reaction. Further investi-
gation will be carried out for the leaching in future
work.
FTIR spectra were collected on a Perkin Elmer Spec-
trum One spectrometer with 10 scans and resolution
−
1
−1
of 4 cm , in the range of 4000–400 cm . Potassium
bromide (KBr) pellet technique was used to examine
the functional groups of a series of silica functional-
ized silylated Ti-salicyladimine complex (solid), where
the samples were mixed with KBr in the weight ratio
of 1:100. The DR UV-Vis analysis was recorded by
Perkin Elmer Ultraviolet-visible Spectrometer Lambda
9
00 and plotted using Kubelka-Munk function, in the
range of 200–800 nm, using barium sulphate as the
standard. Meanwhile, the ligand (liquid) was recorded
in the range of 190–800 nm, using ethanol as the ref-
29
erence. The Si solid state MAS NMR experiments
were performed using Bruker Avance 400 MHz 9.4T
spectrometer. The spectra were recorded at 79.44 MHz
using 4 μs radio frequency pulses, a recycle delay of
3
. Results and Discussion
3
.1 Physical properties
6
0 s and spinning rate of 7.0 kHz using a 4 mm zir- Figure 1 shows the FTIR spectra of silica functionalized
29
conia sample rotor. Si MAS NMR chemical shifts with non-silylated Ti-salicylaldimine complex and sily-
were referred to external TMS at 0 ppm. The mor- lated Ti-salicylaldimine complex functionalized silica
phology of the solid catalyst was observed using SEM. by octadecyltrimethoxysilane (OTMS) with different
The amount of organic moiety in the catalysts was Ti:OTMS molar ratios.

1908
Lai Sin Yuan et al.
(
a)
b)
(
C=N (1633)
C=C (1504)
(
(
c)
d)
(
e)
OH bend
(
1661)
(
f)
C=N (1610)
(
g)
(
h)
CH
2
rock (720)
Ti-N (602)
Si-C (1226)
(
i)
Si-OH
995)
(
symmetric C-H
stretch (2850)
asymmetric C-H
stretch (2919)
CH
2
scissoring (1469)
Si-O-Si (1127, 1038)
1000
4000
3200
2400
1800
1400
600
Wavenumber / cm−
1
Figure 1. FTIR spectra of (a) salicylaldimine ligand, (b) silica function-
alized with Ti-salicylaldimine complex, and silica functionalized with Ti-
salicylaldimine complex silylated with Ti/OTMS molar ratio of (c) 1:1/4, (d)
1:1/3, (e) 1:1/2, (f) 1:1, (g) 1:2, (h) 1:3, (i) 1:4.
It is worth noting that all the spectra showed two suggested that the Si-OH bands located at ca. 995 cm 1
−
−1
−1
24
absorption peaks at 601 cm and 697 cm , corre- have higher frequency than those previously reported
19,20
sponding to the Ti-N and Ti-O bonds, respectively.
because Ti(IV) ion was not bonded to Si-O. The trend
−1
−1
The band at 1610 cm (figure 1 (b) – (i)) is assigned of Si-OH bands (995 cm ) decreased with increasing
to C=N stretching mode. The shifting of free azome- concentration of OTMS (figure 1 (c) – (i)).
thine group (figure 1 (a)) from higher frequency of
The peak areas of SiOH/Ti(IV), denoted as Peak
−1
−1
ca. 1633 cm to lower frequency of 1610 cm is Area /Peak Area , decreased in the order of low
caused by the partial attraction of electron density in to high concentration of silica functionalized with
[995]
[602]
4+
C=N towards Ti ion. This indicates that Ti(IV) has silylated Ti-salicylaldimine complex. This is due to
been successfully coordinated with azomethine nitro- the hydrogen atoms from OH groups that have been
21,22
14,25
gen atom,
which resulted in the weakening of replaced by long chain Si-C H .
On the contrary,
18
38
3
the azomethine group bonding. The strong absorp- the intensity of asymmetric and symmetric sp C-H
tion peaks that appeared at ca. 1127 cm and 1038 stretching displayed at ca. 2919 and 2850 cm , respec-
cm (figure 1 (b) – (i)) are typically attributed to the tively, increased with the amount of OTMS. This indi-
asymmetric and symmetric stretching bands of Si-O-Si, cates that OTMS has been successfully silylated on
−1
−1
−1
respectively.
Nevertheless, the phenolic oxygen bonded to Ti4+ results listed in the table 1 show the areas of C-H and Ti-
also exhibited frequency in the same range as that of the N bands, [Peak Area
/Peak Area ]. In addi-
the surface of Ti-salicylaldimine complex. The FTIR
[2919, 2850]
[602]
symmetric stretching of Si-O-Si, which was at 1050 – tion, there are two other prominent peaks that can be
−
1 23
−1
1
100 cm . This occurrence assisted in reinforcing the observed at ca. 1469 and 720 cm in the FTIR spec-
4+
argument that the complex containing Ti is bonded tra. These two important bands originated from CH
to the oxygen of the phenolic group instead of Si-O-Ti scissoring and CH rocking modes, respectively. All
bonds, as no peak at ca. 960 cm can be detected.
The absorption band of Ti coordinated with pheno- successfully anchored on the complexes and hence,
2
14
3
−1
24
the above phenomena indicate that OTMS has been
4+
−
lic oxygen anion (C H O ) was less evident because led to the enhancement in the hydrophobicity of the
6
5
it overlapped with Si-O-Si symmetric stretching. It is catalysts.

Silica functionalized with silyated Ti-salicylaldimine complex
1909
Table 1. FTIR peak area of silica functionalized with non- Ti-salicylaldimine complex. The bands of salicy-
silylated and silylated Ti-salicyladimine complex with differ-
laldimine ligand (figure 2(a)) at 227 and 248 nm
a
ent amounts of octadecyltrimethoxysilane (OTMS) .
can be assigned to the π-π* transition of ben-
zene ring and imine group (C=N), respectively. The
Ti-salicylaldimine complex
with the molar ratio
of Ti(IV):OTMS
b
Ratio of Peak area
/
bands at lower energy of ca. 314 and 402 nm were
[
2850;2919]
c
−
Peak area
attributed to the substituted phenol group (C
6
H
5
O )
[
602]
and imine group (C=N) with n-π* transitions. The
Ti:OMS 1:1/4
Ti:OMS 1:1/3
Ti:OMS 1:1/2
Ti:OMS 1:1
Ti:OMS 1:2
Ti:OMS 1:3
Ti:OMS 1:4
1.03
1.29
1.32
1.90
1.98
2.86
4.67
absorption bands of silylated salicylaldimine ligand
(
figure 2(b)) showed similar pattern as those of sal-
icylaldimine ligand (figure 2(a)). It is observed that
a small peak appeared at 270–280 nm, which can
be attributed to the ethanol solvent’s cut-off. This is
because salicylaldimine ligand is polar and it might
have induced the shifting of ethanol polar solvent
from shorter wavelength, i.e. 210 nm, to longer wave-
aOctadecyltrimethoxysilane (OTMS) was attached on silica
functionalized with Ti-salicylaldimine complex in order to
prove that the peak area of C-H stretching mode increased length, ca. 270 nm. In the spectra of the complexes
with increasing concentration of OTMS.
The peak area of symmetric and asymmetric sp C-H
(
figure 2(c) and (d)), a shoulder band in the range
b
3
of 200–250 nm can be assigned to the association of
the π-π* transitions of aromatic ring and azomethine
group.
stretching bands from OTMS can be seen at ca. 2919 and
−
1
2
850 cm , respectively.
c
−1
The absorption band at 602 cm is attributed to Ti-N from
silylated Ti-salicylaldimine complex.
Laporte-allowed LMCT of oxygen from phenyl
−
4+
group (C H O ) to metal (Ti ) can be proven by the
6
5
shift from 315 nm to 300 nm after the complexation
−
4+ 26
(
C H O Ti ). It is observed that the n-π* of imine
3
.2 Diffuse reflectance UV-Visible analysis
6 5
group, C=N, has shifted to lower frequency due to
Figure 2 shows the diffuse reflectance UV-Vis spectra of the coordination of imine nitrogen to metal (C=NTi )
a) salicylaldimine ligand, (b) silylated salicylaldimine to form complexes. The UV spectra are broader
4+
22,27
(
ligand, (c) silica functioanlized with Ti-salicylaldimine and stronger in the range of 350–500 nm for the silica
complex and (d) silica functioanlized with silylated functionalized with silylated Ti-salicyladimine complex
3
13 nm (LMCT of C
61 nm (n
6
H
6
O– to Ti4+
)
3
–
* of imine group)
3
04 nm (LMCT
–
6 6
of C H O
to Ti4+)
(
(
d) Silica functionalized with silylated Ti-
salicylaldimine complex
c) Silica functionalized with non-silylated
Ti-salicylaldimine complex
(
b) Silylated salicylaldimine ligand
2
27 nm (
–
* of benzene ring )
2
48 nm (
–
* of imine group )
–
3
14 nm (
–
* of C
6
H
6
O )
4
02 nm (n
–
* of imine group)
(
a) Salicylaldimine ligand
190
250
300
350
400
450
500
550
600
650
700
750
800
Wavelength / nm
Figure 2. Diffuse reflectance UV-Vis spectra of (a) salicylaldimine lig-
and, (b) silylated salicylaldimine ligand, (c) silica functionalized with non-
silylated Ti-salicylaldimine complex and (d) silica functionalized with silylated
Ti-salicylaldimine complex.

1910
Lai Sin Yuan et al.
compared to the silica functionalized non-silylated Ti-
salicyladimine complex. It is suggested that the silanol
groups (Si-OH) have transformed into cross-linking
Si-O-Si by OTMS,25 resulting in higher intensity of
28
absorption bands in that region.
3
.3 29Si solid state MAS NMR analysis
Figure 3, (a–h) depicts the 29Si solid state MAS
NMR spectra of the silica functionalized with non-
silylated Ti-salicylaldimine complex and silylated Ti-
salicylaldimine complexes with Ti:OTMS molar ratios
29
of 1:1/4 to 1:4. Si solid state MAS NMR spectroscopy
is employed to examine the surface coverage and degree
of cross-linking of the alkylsilyl groups. This analyti-
cal method can be a semi-quantitative tool in detecting
the environmental alteration of Si from non-silylated
to silylated Ti-salicylaldimine complex. Shoulder peaks
with low density near −57 ppm were assigned to par-
2
tial cross-linked R-Si(OSi) (OH) and designated as T
2
7
3
signals. The T signals at −67 ppm in the spectra
showed the complete cross-linking of R-Si(OSi) .
4
It should be noted that the signal intensity of partial
cross-linked R-Si(OSi) (OH) (−57 ppm) was higher
2
after the addition of OTMS to the complex (figure 3
(a) – (h)). This is caused by the increasing amount of
OH groups (–Si(OSi) -OH), which were not fully con-
2
densed to R-Si(OSi) . However, the amount of alkyl
3
groups (carbon atoms) cannot be detected in this spec-
troscopy method, hence the amount of alkyl groups
that were less than Si-OH groups cannot be deduced.
Thus, FTIR spectroscopy can be used as a supplemen-
tary tool to figure out whether the peak of alkyl groups
increased with decreasing Si-OH groups or vice-versa.
The FTIR spectra in figure 1 showed increasing peak
3
−1
area for sp C-H stretching mode (ca. 2919 cm and
2
−1
850 cm ) and decreasing peak area for Si-OH band
with increasing amount of OTMS.
2
When the peak area of T was compared to the peak
3
2
area of T , it can be seen that T has a smaller peak
area. This indicates that the hydrolysis and conden-
sation of OTMS and 3-aminopropyltrimethoxysilane
have been completed with the presence of R-Si(OSi)
cross-linking at signal T .
3
3
3.4 SEM analysis
Figure 3. 29Si solid state MAS NMR spectra of (a) silica
functionalized with Ti-salicylaldimine complex, (b) silica
Figure 4 displays that morphology of the two types functionalized with Ti-salicylaldimine complex with
of catalysts. Figures 4(a) and (b) are images of sil- Ti:OTMS molar ratios of (b) 1:1/4, (c) 1:1/3, (d) 1:1/2, (e)
1:1, (f) 1:2, (g) 1:3, and (h) 1:4.
ica functionalized with non-silylated Ti-salicyladimine
complex under magnification of x10000 and magnifi-
cation of x20000, respectively, while figures 4(c) and

Silica functionalized with silyated Ti-salicylaldimine complex
1911
(a)
(b)
(c
(d)
)
Figure 4. SEM micrographs of silica functionalized with Ti-salicylaldimine com-
plex with (a) magnification of x10000, (b) magnification of x20000, (c) silylation
of OTMS under magnification of x10000, and (d) silylation of OTMS under magnifi-
cation of x20000.
◦
(
d) are images of silica functionalized with silylated Ti- temperature of 450–700 C, owing to the mass loss
salicyladimine complex under magnification of x10000 of alkylsilyl groups. It is observed that the highest
and magnification of x20000, respectively. It can be amount of alkylsilyl groups (Ti:OTMS = 1:4) con-
seen that the catalysts are amorphous or in disordered tributed to the biggest mass loss, followed by the molar
structure, which might be due to the route of synthe- ratio of Ti:OTMS with 1:1 and the least mass loss is
sis. The main difference between the non-silylated and shown by the catalyst with the Ti:OTMS molar ratio of
silylated catalysts is that the latter one showed obvious 1:1/4.
agglomeration in the presence of OTMS.
1
00
3.5 TG analysis
8
0
The thermogravimetric curves in figure 5 show the dif-
ference in mass loss for the silica functionalized with
non-silylated Ti-salicyladimine complex and silylated
Ti-salicylaldimine complex. For silica functionalized
with non-silylated Ti-salicylaldimine complex, there
are two steps involved: first, the mass loss of adsorbed
60
40
◦
water happened between 0–110 C; while the second
(a) Ti-salicylaldimine complex functionalized silica
Silylated Ti-salicylaldimine complex functionalized silica with
step is the mass loss of ligand organic moiety at the
temperature between 300–700 C.
For silica functionalized with silylated Ti-
salicyladimine complex with Ti/OTMS molar ratio of
20
0
(b) Ti : OTMS = 1 : 1/4
◦
(c) Ti : OTMS = 1 : 1
(d) Ti : OTMS = 1 : 4
200
400
600
800
Temperature/ °C
1
:1/4, 1:1, and 1:4, three steps are involved. The mass
decomposition of the first two steps was same as the sil-
ica functionalized with non-silylated Ti-salicylaldimine
complex, i.e. water desorption and ligand organic
Figure 5. Thermogravimetric curves of (a) silica function-
alized with Ti-salicylaldimine complex, silica functional-
ized with silylated Ti-salicyladimine complex with Ti/OTMS
moiety decomposition. The third step happened at the molar ratio of (b) 1:1/4, (c) 1:1, and (d) 1:4.

1912
Lai Sin Yuan et al.
0
0
0
.16
.14
.12
Ti:OTMS = 1:4
Ti:OTMS = 1:3
Ti:OTMS = 1:2
Ti:OTMS = 1:1
Ti:OTMS = 1:1/2
Ti:OTMS = 1:1/3
Ti:OTMS = 1:1/4
Ti(IV) complex
0
.1
0.08
0.06
0.04
0.02
0
0
50
100
150
200
250
300
Time/ min
Figure 6. Water adsorption capacities (mg g 1) of silica functionalized with
−
silylated and non-silylated Ti-salicylaldimine complex functionalized silica.
For the first step in the TG analysis, i.e. water des- with the highest intensity, in comparison to those of the
orption, it is clearly seen that the least water con- silica functionalized with silylated Ti-salicyladimine
tent adsorbed as shown by silica functionalized with complex.
silylated Ti-salicyladimine complex with Ti/OTMS
As reported in table 2, the water adsorption capacities
−1
molar ratio 1:4. This is attributed to its higher degree (mg g ) of the silylated samples gradually decreased
of hydrophobicity compared to the others. On the with increasing amount of OTMS. Thus, it is evident
contrary, non-silica functionalized with silylated Ti- that the hydrophobicity of these catalysts increased. It
salicyladimine complex showed the highest water con- is observed that the water capacities of all the sam-
tent. The results of TG analysis are in correlation with ples gradually became constant and this was attributed
the water adsorption analysis, which will be explained to water saturation on the superficial silanol groups.
in section 3.6.
In order to investigate the effectiveness of the sily-
lated and non-silylated Ti-salicyladimine complexas
catalysts, the catalytic activity of these catalysts were
tested out in an aqueous-organic-solid biphasic sys-
tem. The comparison of the catalytic performances is
made based on the introduction of octadecylsilyl groups
3.6 Water adsorption analysis
Figure 6 represents the results from the water adsorp-
tion capacity experiment carried out for 5 h by mea-
suring the weight gained by the catalysts every 30 min.
It is interesting that Langmuir isotherm with first order
kinetic reaction is observed. This shows that the water
molecules adsorbed to a maximum of one monolayer
(
C H Si(OSi) –) to these complexes, which is pre-
18 37 n
dicted to show significant improvement in the catalytic
activity in the oxidation of cyclohexene and limonene
by aqueous H O .
2
2
−1
on the solid surface with the rate constant, k, of 2.0 h .
Table 2. Water adsorption capacities (mg g 1) of var-
ious Ti:OTMS molar ratios of OTMS silylation on Ti-
salicylaldimine complex.
−
The kinetic rate obtained for the first 30 min is 2.32 X
−2
−1
10
g h for silica functionalized with non-silylated
−3
−1
Ti-salicyladimine complex and 2.80 X 10 g h for
silica functionalized with silylated Ti-salicylaldimine Molar ratio of silylated
complex. This indicates that the non-silylated cata- Ti-salicylaldimine complex
lyst possesses higher water adsorption capacity and
Water adsorption
capacity (mg g )
−
1
functionalized silica
faster kinetics in reaction as compared to the silylated
catalysts.
Ti:OTMS = 1:4
31.00
48.50
62.00
64.50
82.00
97.00
111.50
142.00
Ti:OTMS = 1:3
As shown in figure 6 and table 2, silica functionalized
Ti:OTMS = 1:2
with non-silylated Ti-salicyladimine complex shows the Ti:OTMS = 1:1
−1
Ti:OTMS = 1:1/2
Ti:OTMS = 1:1/3
Ti:OTMS = 1:1/4
Ti:OTMS = 1:0
highest water adsorption capacity, i.e. 142.00 mg g ,
and this implies that the complex functionalized sil-
ica was highly hydrophilic. This is in agreement with
the FTIR spectra, where polar Si-OH groups appeared

Silica functionalized with silyated Ti-salicylaldimine complex
1913
3.7 Catalytic activity
hydroperoxide, which then undergoes rearrangement to
generate carveol and carvone. The formation of carvone
In this research, carveol and carvone are the main prod- might be due to the water molecules that extracted the
ucts in the oxidation of limonene by aqueous H O , hydroxide ion of the limonene hydroperoxide. Hence,
2
2
as these two products favour the free radical allylic the negative charge on the oxygen in C-O delocalizes to
auto-oxidation reaction pathway. The plausible reac- generate C=O. Meanwhile, the generation of carveol is
tion pathway is shown in figure 7. Hydroxyl radicals attributed to the hydrogen addition process by the water
•
(
OH ), either from H O or H O after their interac- molecules to the ketone group from carvone.
2 2 2
tion with Ti(IV) active sites, initiate limonene at the
It is suggested that the formation of [Ti(salicylal-
3
allylic carbon atom (sp –hybridized carbon in CH CH- dimine)]SO complex in silica functionalized with Ti-
2
4
CH branch chain), which is stabilized by the resonance salicylaldimine complex has taken place. Imine group
3
2
structure and a vinylic carbon (sp -hybridized carbon in (C=N) is neutral and contains lone pair electrons,
4
+
CH CH-CH branch chain). Then, it is oxidized by the which can coordinate with Ti by dative covalent bond.
2
3
−
2+
dissolved oxygen in solution to form limonene peroxo Meanwhile, two moles of C
intermediate. This limonene peroxo intermediate can gen ions can reduce the Ti to Ti . Thus, the cationic
react with the limonene substrate to produce limonene complex can be formed, before being neutralized by one
H
O bearing anionic oxy-
6
4+
6
29
Figure 7. The plausible mechanistic pathway for limonene conversion.

1914
Lai Sin Yuan et al.
2
−
160
mole of SO ion to form a neutral complex, as shown
4
1
40
20
in the equation below.
1
ꢀ
ꢁ
(salicylaldimine) +Ti +SO 24 − −→ Ti (salicylaldimine) SO4
−
4+
100
80
2
2
(1)
60
The catalytic activity of silica functionalized with
40
20
0
non-silylated and silylated Ti-salicyladimine complex
was carried out in the oxidation of limonene by aque-
ous H O . The highlight of this catalyst is that sil-
0
20
40
60
80
100
120
2
2
TON for Ti
ica functionalized with non-silylated Ti-salicyladimine
complex is in solid form without post-grafting onto Figure 8. The effect of hydrophobicity in silica functional-
ized with silylated Ti-salicyladimine complex on the turnover
any inorganic support. It is actually unexpected because
number for Ti (TON) in the oxidation of limonene by aque-
ous hydrogen peroxide.
there is usually a need to anchor metal complexes
onto inorganic supports to form heterogeneous cata-
lyst. This phenomenon might be due to the use of 24%
Ti(IV) sulphate solution which contains high amount of capacity) could result in decreasing number of Si-OH
water and thus the water assisted in complete hydrolysis and this suppresses the adsorption of water molecules
and condensation reactions among methoxysilyl groups onto the active sites of Ti. This is owing to the weak acid
themselves, to form siloxane linkages. Therefore, solid characteristic of Si-OH and H O, that is able to induce
2
silica functionalized with Ti-salicylaldimine complex the oxirane ring opening, before forming glycols, which
will then lead to the deactivation of the catalysts.25
It should be highlighted that the TON for Ti gradu-
was produced.
Silica functionalized with non-silylated and silylated
Ti-salicyladimine complex were investigated as cata- ally became constant as the water adsorption capacity
lysts in the oxidation of limonene by aqueous H O . decreased. This phenomenon could probably be caused
2
2
Figure 8 shows the hydrophobic effect towards the by the concentration of OTMS that reached the max-
turnover number (TON) for Ti in the series of cat- imum coverage of the sureface of the complex. Intro-
alysts prepared. It was observed that the products ducing excess OTMS onto the catalyst may hydrolyze
from the oxidation of limonene by aqueous hydro- and form octadecylsilyl groups, which comprised of
gen peroxide (H O ), such as 1,2-epoxylimonene, 8,9- C H – and –Si-OH. As a result, limonene and H O
2
2
18 37
2
2
epoxylimonene, carveol and carvone, were in accor- will be attracted to the hydrolyzed octadecylsilyl groups
dance with literature reports.3 The silica functional- and this may reduce the catalytic activity of the catalyst
ized with non-silylated Ti-salicyladimine complexhas in the reaction.
0
the lowest TON (20) compared to that of the analogous
Figure 9 shows the conversion of limonene and
silylated Ti-salicylaldimine complexes. This is due to selectivity of carveol, carvone, and other products
the presence of a large number of silanol groups, as after the catalytic reaction. The selectivity of other
proven in the FTIR spectra (figure 1). These silanol unidentified products (may include 1,2-epoxylimonene,
groups can easily attract water molecules, which can 8,9-epoxylimonene, diepoxy product, and glycols)
suppress the active sites of Ti(IV). Moreover, it is also decreased with increasing hydrophobicity.
probably due to the polar products, such as epoxides,
Table 3 shows the concentration of Ti in silica func-
which have better interaction with polar surface of non- tionalized with silylated Ti-salicyladimine complex in
silylated Ti-salicylaldimine complex functionalized sil- the oxidation of limonene. It is seen that Ti content in
ica. This can lead to the blocking of the transition metal 50 mg of catalysts has slightly decreased but it did not
active sites, thus hindering the oxidation.25
affect the results of TON for Ti and product yields. This
When the water adsorption capacity decreased, the indicates that the hydrophobicity from the alkylsilyl
TON for Ti increased, as shown in figure 8. The high- groups plays a significant role in the product yields.
est degree of hydrophobicity has contributed to high-
est TON of Ti (110) with conversion percentage of used cyclohexene as the reactant in the oxidation
1%. The conversion increased by around 26% as com- reactions for alkylsilylated catalysts. Despite this,
It is found that most of the previous studies
6
pared to those shown by silica functionalized with non- a few reports can be obtained for the elucida-
silylated Ti-salicylaldimine complex. This is due to tion of the silylation effect on the catalytic activity
31,32
the successful substitution of Si-OH by OTMS. Higher in the limonene oxidation.
degree of hydrophobicity (lower water adsorption ture design (Ti-salicylaldimine complex functionalized
The catalyst’s struc-

Silica functionalized with silyated Ti-salicylaldimine complex
1915
active sites on the surface of MCM-41 or diffuse
through the pores of MCM-41, to reach the Ti active
sites inside the porous structure. This structure could
provide a higher surface area for the reaction to occur
and hence, the higher catalytic activity. Silica function-
alized with Ti-salicylaldimine complex possesses lower
surface area as it is nonporous and the reaction can only
take place on the outer surface of the catalyst.
/
%
11.50
1
2.00
1
4.62
1
2.06
1
1.22
12.42
28.72
3
0.69
33.41
20.77
24.61
1
0.54
Second, the silylating agent used in the previ-
31,32
ous researches
was hexamethyldisilazane (HDMS)
16.43
1
2.54
which is a short alkylsilyl chain, while in this reaserch
octadecyltrimethoxysilane (OTMS), which is a long
alkylsilyl chain, was employed. The long alkylsilyl
groups might have caused steric hindrance to the diffu-
sion of limonene to the Ti active sites. On the contrary,
limonene, which is branched, can come into contact
with the Ti active sites more easily when shorter alkylsi-
lyl groups on the catalyst’s internal and outer surface
were used.
1
2.57
10.98
14.96
9.63
Molar ratio of Ti:OTMS/ mmol
Figure 9. The conversion of limonene and the yields of
carvone, carveol and other products after the catalytic reac-
tion. The reactions were carried out at 70 C for 22 h with
◦
(
(
R)-(+)-Limonene (5 mmol, 0.83 mL), 30% aqueous H2O2
Thus, the catalytic activity of the catalyst in this
10 mmol, 0.38 mL) and catalyst (50 mg) under stirring research is slightly lower than those previously reported
conditions.
due to the difference in structure design (surface area),
types of silylating agent and substrates used. How-
silica) for this research is essentially quite different ever, it was observed that in both the catalytic sys-
from those reported, i.e. Ti-MCM-41. Silylated Ti- tems, Ti-MCM-41 or silica functionalized with Ti-
MCM-41 achieved a better performance (79–92% con- salicylaldimine complex, the selectivity increased with
version) compared to silica functionalized with sily- increasing concentration of silylating agent added.
lated Ti-salicyladimine complex (61% conversion) in
Figure 10 displays the stability results of the sily-
limonene oxidation. It might be assigned to two possi- lated catalysts that were obtained by conducting the
ble reasons, which will be explained in the following reusability testing with the same reaction condition. It
paragraphs.
was observed that the limonene conversion of the sil-
First, Ti-MCM-41 contained Ti incorporated within ica functionalized with non-silylated Ti-salicyladimine
the framework of the hexagonal mesoporous solid pow- complex decreased tremendously for the second run
der, while silica functionalized with Ti-salicylaldimine due to the adsorption of the water molecules on the
complex is disordered and non-porous solid with Ti Ti active sites to form Ti-OH or opening of the epoxy
coordinated to ligand. Limonene is a small, aromatic ring to form glycols. On the contrary, those of silica
molecule which can adsorb, react and desorb on Ti functionalized with silylated Ti-salicyladimine complex
Table 3. The concentration of Ti in silica functionalized with silylated Ti-
salicyladimine complex catalysts.
Silylated Ti-salicylaldimine Weight of Ti in 50 mg
complex functionalized
silica
catalysts of catalysts Products TON for
a
(mg)
(mmol)
(mmol)
Ti
Ti:OTMS = 1:0
Ti:OTMS = 1:0.5
Ti:OTMS = 1:1
Ti:OTMS = 1:2
Ti:OTMS = 1:3
Ti:OTMS = 1:4
582.90
862.50
1069.30
1558.80
1505.20
1919.50
0.09
0.06
0.05
0.03
0.04
0.03
1.78
2.09
2.80
2.38
2.88
3.07
19.00
34.00
56.00
69.00
81.00
110.00
a
The concentration of Ti was calculated based on the amount of Ti(IV) sul-
phate added in the synthesis of the catalysts. A constant concentration of Ti(IV)
sulphate (1.07 mmol) was used throughout the synthesis of the catalysts.

1916
Lai Sin Yuan et al.
through Fundamental Research Grant Scheme (FRGS)
and Research University Grant (GUP), Postdoctoral
Fellowship Scheme by Research Management Centre
(RMC), Development of Low Carbon Scenarios for
Asia Region, Universiti Teknologi Malaysia.
References
1. Notari B 1993 Catal. Today 18 163
2
. Elías V R, Crivello M E, Herrero E R, Casuscelli S G
and Eimer G A 2009 Ind. Eng. Chem. Res. 48 9076
. Nur H, Ikeda S and Ohtani B 2001 J. Catal. 204 402
. Chu H, Wan Y and Zhao D 2009 Catal. Today 148 19
. Kholdeeva O A and Maksimovskaya R I 2007 J. Mol.
Catal. A Chem. 262 7
3
4
5
Figure 10. Comparison of the reusability of (R)-(+)-
st
nd
Limonene for 1 and 2 run in the oxidation of limonene by
6
7
8
. Lin K, Pescarmona P P, Vandepitte H, Liang D, Tende-
loo G V and Jacobs P A 2008 J. Catal. 254 64
. Nur H, Prasetyoko D, Ramli Z and Endud S 2004 Catal.
Commun. 5 725
. Nur H, Hau N Y, Misnon I I, Hamdan H and Muhid M
N M 2006 Mater. Lett. 60 2274
◦
aqueous H2O2 at 70 C for 22 h.
remained almost the same, with only slight decrease
in the range of 2–6%. Therefore, it is shown that the
hydrophobic surface of these catalysts has successfully
repelled the adsorption of water molecules on the Ti
active sites to prevent deactivation.
9. Prasetyoko D, Ramli Z, Endud S and Nur H 2005 Mater.
Chem. Phys. 93
1
1
1
0. Lin K, Pescarmona P P, Houthoofd K, Liang D,
Tendeloo G V and Jacobs P A 2009 J. Catal. 263 75
1. Capel-Sanchez M C, Campos-Martin J M and Fierro J L
G 2010 Catal. Today 158 103
2. Corma A, Domine M, Gaona A J, Jorda L J, Navarro T
M, Rey F, Perez-Pariente J, Tsuji J, McCulloch B and
Nemeth T L 1998 Chem. Commun. 20 2211
3. Corma A, Jordá J L, Navarro M T, Pérez-Pariente J, Rey
F and Tsuji J 2000 Stud. Surf. Sci. Catal. 129 169
4. Bu J and Rhee H-K 2000 Catal. Lett. 65 141
5. Igarashi N, Hashimoto K and Tatsumi T 2007 Microp-
orous Mesoporous Mater. 104 269
4
. Conclusions
The catalytic activity of silica functionalized with Ti-
salicylaldimine complex in the oxidation of limonene
was significantly improved by introducing hydropho-
bicity on the catalyst surface. These results strong
depended on the amount of octadecylsilyl organic moi-
ety immobilized on the catalyst. It was observed that
the TON for Ti increased when the rate of water adsorp-
tion decreased. It is significant to conclude that the
improved catalytic activity of the catalysts might due
1
1
1
16. Nur H, Hau N Y, Muhid M N M and Hamdan H 2004
Phys. J. IPS A7 0218
to the hydrophobic surface. The decrease in polarity of 17. Peña M L, Dellarocca V, Rey F, Corma A, Coluccia S
silanol groups and suppression of water molecules can
be an idea to further improve the catalyst’s design and
productivity in various applications for future works.
and Marchese L 2001 Microporous Mesoporous Mater.
4-45 345
8. Yuan L S, Razali R, Efendi J, Buang N A and Nur H
013 Appl. Catal. A 460-461 21
9. Nakamoto K 1986 In Infrared and Raman spectra of
organic and coordination compounds (4th ed.) (New
York: John Wiley and Sons) p. 14
4
1
1
2
Supplementary Information
2
0. Yuan L S, Efendi J, Razali N S H and Nur H 2012 Catal.
Commun. 20 85
All additional information pertaining to characteriza-
tion of the complexes using FTIR peak area data
21. Mostafa S I, Ikeda S and Ohtani B 2005 J. Mol. Catal.
A Chem. 225 181
(table 1), FTIR spectra (figure S1), electronic spec-
tra (figure S2), thermogravimetric graph (figure S3)
and GC chromatogram (figure S4) are given in the
supporting information. Supplementary Information is
available at www.ias.ac.in/chemsci.
2
2
2. Roy G B 2009 Inorg. Chim. Acta 362 1709
3. Hu C, Zhang W, Xu Y, Zhu H, Ren X, Lu C, Meng Q
and Wang H 2001 Transit. Metal Chem. 26 700
24. Luo Y, Lu G Z, Guo Y L and Wang Y S 2002 Catal.
Commun. 3 129
Acknowledgments
25. Yang Y, Guan J, Qiu P and Kan Q 2010 Appl. Surf. Sci.
2
56 3346
The authors gratefully acknowledge funding from
the Ministry of Higher Education (MOHE), Malaysia
˙
2
6. Ispir E 2009 Phosphorus Sulfur Silicon Relat. Elem.
184(12) 3160

Silica functionalized with silyated Ti-salicylaldimine complex
1917
2
7. Parida K M, Sahoo M and Singha S 2010 J. Catal. 276 30. Saikia L, Srinivas D and Ratnasamy P 2006 Appl. Catal.
61
A 309 144
8. García-González C A, Sampaio da Sousa A R, Argemí 31. Guidotti M, Batonneau-Gener I, Gianotti E, Marchese
1
2
A, López Periago A, Saurina J, Duarte C M M and
Domingo C 2009 Int. J. Pharm. 382 296
9. Cagnoli M V, Casuscelli S G, Alvarez A M, Bengoa J F,
L, Mignard S, Psaro R, Sgobba M and Ravasio
N 2008 Microporous and Mesoporous Mater. 111
39
2
Gallegos N G, Samaniego N M, Crivello M E, Ghione G 32. Cagnoli M V, Casuscelli S G, Alvarez A M, Bengoa J F,
E, Pérez C F, Herrero E R and Marchetti S G 2005 Appl.
Catal. A 287 227
Gallegos N G, Crivello M E, Herrero E R and Marchetti
S G 2005 Catal. Today 107–108 397