Epoxidation of limonene over low coordination Ti in Ti- SBA-16

Article abstract of DOI:10.1016/j.apcata.2017.01.001

Epoxidation of limonene was studied over Ti-SBA-16 catalysts prepared using two different post grafting techniques. The first one consisted of an in-situ synthesis of a titanium acetylacetone (ACAC) precursor and the other one used directly TiO(ACAC)₂. Reaction conditions such as temperature, concentration of TBHP and solvent polarity were examined. The reaction should be carried out at 75?°C in acetonitrile using a TBHP/limonene molar ratio of 11/6. The conversion of limonene reached 80% with a 1,2-limonene oxide selectivity of 79% for both impregnation methods using Ti-SBA-16 with a Ti/Si atomic ratio of 7.3%. Under the same conditions the 3D pore structure of SBA-16 support favors reactants and products mass transfer compared to the 2D pore structure of SBA-15. No differences in catalytic activity were observed as atomic ratio Ti/Si was varied from 5 to 13.4% in Ti-SBA-16. Repeated catalytic tests showed no change in activity over three process cycles.

Full text of DOI:10.1016/j.apcata.2017.01.001

Accepted Manuscript
Title: Epoxidation of limonene over low coordination Ti in Ti-
SBA-16
Author: <ce:author id="aut0005"
author-id="S0926860X17300017-
7b03994c357d6eca371044b31d9844e4"> Luc
Charbonneau<ce:author id="aut0010"
author-id="S0926860X17300017-
fdd1a4fc16790ad28b91eb33eccdf21b"> Serge
Kaliaguine
PII:
S0926-860X(17)30001-7
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2017.01.001
APCATA 16112
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
15-11-2016
31-12-2016
3-1-2017
Please cite this article as: Luc Charbonneau, Serge Kaliaguine, Epoxidation of
limonene over low coordination Ti in Ti- SBA-16, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2017.01.001
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Epoxidation of limonene over low coordination Ti in Ti- SBA-16
Luc Charbonneau and Serge Kaliaguine *
Department of Chemical Engineering, Laval University, Quebec, Canada, G1V 0A6
*serge.kaliaguine@gch.ulaval.ca
1

Graphical abstract
Highlights
 New robust active, selective and reusable Ti-SBA-16 with high content of low
coordination Ti
 Comparison 3D pore structure of Ti-SBA-16 with 2D Ti-SBA-15
 New mechanism for limonene epoxidation
Abstract
Epoxidation of limonene was studied over Ti-SBA-16 catalysts prepared using two
different post grafting techniques. The first one consisted of an in-situ synthesis of a
titanium acetylacetone (ACAC) precursor and the other one used directly TiO(ACAC)2.
Reaction conditions such as temperature, concentration of TBHP and solvent polarity
were examined. The reaction should be carried out at 75oC in acetonitrile using a
TBHP/limonene molar ratio of 11/6. The conversion of limonene reached 80% with a
1,2-limonene oxide selectivity of 79% for both impregnation methods using Ti-SBA-16
with a Ti/Si atomic ratio of 7.3%. Under the same conditions the 3D pore structure of
SBA-16 support favors reactants and products mass transfer compared to the 2D pore
structure of SBA-15. No differences in catalytic activity were observed as atomic ratio
Ti/Si was varied from 5 to 13.4% in Ti-SBA-16. Repeated catalytic tests showed no
change in activity over three process cycles.
Keywords: titanium catalysts, epoxidation, limonene, solvent effect
2

1. Introduction
The polymer industry is facing the new challenge of producing more green and
renewable polymers. Corma et al underlined that natural products could be alternative
sources to petroleum derived monomers [1]. Limonene is a low cost raw material easily
obtained by extraction from citrus oil. The annual production of limonene reached 70 000
tons per year in 2013 [2, 3]. Limonene oxide is the epoxidation product of limonene.
Limonene oxide based polycarbonates have been synthesised and have shown very
promising properties [4, 5]. Homogeneous catalysis using molybdenum or
trimethylrhenium complexes have shown high limonene conversion with high selectivity
to the 1,2-limonene epoxide [6-8]. Homogeneous catalysts are however difficult to
implement at industrial scale due to the need of separation from the reaction mixture and
problematic reusability.
In the early seventies, Sheldon reported that titanium supported on silica can epoxidize
olefins in the presence of a peroxide [9, 10]. Further studies of supported titanium have
shown high olefin conversion and selectivity, as high as homogeneous molybdenum
catalysts [11]. Sheldon thought that the activity of silica supported titanium is related to
its Lewis acid character. Based on his homogeneous catalysis observations, it was
believed that silica enhances the Lewis acid character of titanium which led to a higher
activity than pure TiO2. At this time, Sheldon was not aware of the importance of
titanium coordination for catalytic activity.
Berube et al. developped a soft-templating titanium post-grafting impregnation on SBA-
15 [12, 13]. This technique was shown to yield very active solid catalysts for the
epoxidation of cyclohexene. Recently some attempts were made to epoxide limonene
over Ti-SBA-15 but conversion and selectivity were still low [14-20]. One work
describes the use of MCM-41supported Ti for limonene epoxidation. No catalyst stability
was reported MCM-41 is known to be a thin walled not hydrothermally resistant
mesostructured silica material. The various solids produced were compared as catalysts in
the epoxidation reaction of limonene to limonene oxide [21]. In this work it was
attempted to prepare low coordination Ti-SBA-16 by the technique of Bérubé et al. and
3

to improve on this preparation method by directly impregnating the support with a
solution of titanium ACAC.
2.Experimental
2.1 Material Synthesis
SBA-16 synthesis is similar to the one reported by Kleitz et al. [22-24]. In a
typical synthesis, 5.35g of F127 (BASF) was dissolved in a stirred solution containing
257.2 g water and 10.6 g concentrated hydrochloric acid (35-37%) at 45oC. When the
F127 was dissolved, 16,1 g of n-butanol was added to the solution as a cosurfactant.
Finaly 25.45 g of tetraethylorthosilicate (TEOS, 98%, Aldrich) was added to the mixture.
The solution was agitated for 24 h. The resulting mixture was transferred to an autoclave
to allow hydrothermal aging for 24 h at 140oC. The solution was thereafter cooled down
to 5oC and the pH was adjusted to 10. For the preparation of Ti-SBA-16 by the method of
Berube et al (method 1) a mixture of Ti(OPr)4 (Alfa Aesar, 98%) and acetylacetonate
(Aldrich,) with a molar ratio of 1/3 in anhydrous ethanol was then added to the
suspension of SBA-16 in its mother solution [12, 13]. The solution was stirred for 2 h,
filtered and washed with a 300 mL anhydrous ethanol to remove any un-grafted
complexe. Finally, the material was dried for 24h at 100oC and calcined at 550oC (ramp
of temperature: 2oC/min) for 3 h. A new simpler method (method 2) has been developed
in this work. TiO(ACAC)2 (Strem 98%,) was dissolved in 200 mL toluene at 45oC. SBA-
16 was previously calcined at 550oC for 6 h, added to the solution and agitated for 2 h.
The solid was filtered, washed with 300 mL of toluene and dried in oven at 100oC for 24
h. Finally, the sample was calcined at 550oC (ramp of temperature: 2oC/min) for 3 h. In
order to avoid any confusion, the post grafting method is specified in parentheses; (1) for
method 1 and (2) for method 2.
2.2 Characterization
4

Nitrogen adsorption-desorption isotherms were measured at -196 °C using an
Autosorb-1C sorption analyzer. The pore size distributions and the cumulative surface
area curves were determined using the nonlocal density functional theory (NLDFT)
method 31 supplied with the Autosorb-11.55 software from Quantachrome Instruments.
The kernel selected was N2 sorption on silica at -196 °C assuming a cylindrical pore
geometry and implementing the model of (metastable) NLDFT adsorption isotherms
(adsorption branch).
Small angle X-ray scattering (SAXS) measurements were taken on a Nanostar U small
angle X-ray scattering system (Bruker, Germany) using Cu Kα radiation (40 kV, 35 mA).
The d-spacing values were calculated by the formula d=2π/q. Prior to elemental analysis,
the catalyst was digested in a mixture of 10% (V/V) hydrochloric acid and hydrofluoric
acid 1% (V/V). Elemental analysis was performed in a Perkin-Elmer Atomic Absorption
M1100B spectrometer. Titanium coordination was determined by diffuse reflectance UV-
Vis on a Varian Cary 300 Bio with a scan rate of 200 nm per second.
2.3 Catalytic tests
The catalytic tests were performed in a 50 mL round bottom flask with 300 mg of
catalyst, 10 mL of the appropriate solvent, 6.2 mmol of limonene and 11,2 mmol of
TBHP in decane solution at 75oC for 24 h with an agitation at 500 rpm using a Fisher
Scientific Thermoplate equipped with a thermocouple. The catalyst was separated from
the solution by centrifugation. The reaction mixture was analyzed using a CP-3800 gas
chromatograph (Varian Inc.) equipped with a flame ionization detector (FID), a
Stabilwax column (30 m × 0.53 mm × 1 μm) coupled with a 5 m guard column. The
limonene conversion was determined using Equation 1
퐶표푛푣 % = 푛 (푙푖푚표푛푒푛푒)−푛 (푙푖푚표푛푒푛푒) ∗ 100
(1)
표
푓
( )
푛 (푙푖푚표푛푒푛푒)
표
5

And the selectivity was calculated using Equation 2:
푛
푒푝표푥푖푑푒
( )
푆푒푙 % = 푛 (푙푖푚표푛푒푛푒)−푛 (푙푖푚표푛푒푛푒) ∗ 100
(2)
표
푓
The reaction by-products were identified using a GC-MS ionic trap Thermo Instrument
ITQ-900 Ultra Trace. The column used was a DB-5-MS (30 m × 0.28 mm × 1 μm).
3. Results and discussion
3.1 Catalysts characterisation
3.1.1 BET Analysis
Titanium acetylacetonate (ACAC) complexes were favored over other titanium
complexes in this study. The interests of using these compounds as Ti precursors to
functionalize SBA-16 is the requirement of obtaining high dispersion and low
coordination of the active Ti sites. Several catalysts were synthesized by method 1 with
different Ti/Si ratios ranging from 5 and 13.4%. In this method, the amount of ACAC,
temperature and pH of the solution are important parameters to control in order to
produce a monomeric Ti(ACAC)2(OPr)2 [12, 13]. In reference [13] this synthesis method
was used to prepare catalysts with Ti/Si ratios ranging from 0 to 13.4%. The epoxidation
activity was shown to increase as theas the ratio was raised up to 5% which is the upper
limit at which isolated tetracoordinated Ti species as produced. Above this value (5 to
13.4%) the activity was found to be constant but the materials showed increased stability.
Due to this Ti/Si range of density of tetracoordinated Ti remained essentially constant but
some high coordination (up to 5) Ti was also formed. The advantage of using the ACAC
ligand for catalyst preparation by method 2 is its ability to isolate the central titanium
atom on the silica support, allowing high dispersion. In both impregnation methods, the
two titanium ACAC complexes react with the hydroxyls of the surface and are converted
to titanium oxide surface species upon calcination. N2 sorption isotherms of pure SBA-16
6

and Ti-SBA-16 materials were determined. As can be seen on Fig.1, all samples exhibit a
type IV isotherm with an hysteresis loop, indicating the presence of well-defined
mesopores in the structure.
Calcined Ti-SBA-16 mesporous catalysts displayed a relatively high surface area, large
pore diameter and large pore volume (Table 1). The pore size was calculated by non
linear DFT using adsorption branch. Table 1 shows that the surface area slightly
decreases upon addition of higher titanium amount which still however preserves the
mesotructure of the material. The impregnation with toluene TiO(ACAC)2 (method 2)
yielded higher pore volume and pore diameter than the one of materials prepared by
method 1. It seems that this impregnation method results in a better dispersion compared
to the in-situ formation of the titanium ACAC complex (method 1).
3.1.2 Small angle X-ray diffraction
The XRD patterns, represented on Fig.2, for the calcined Ti catalysts, exhibit a
strong 110 reflection peak at 0.54o corresponding to the Im3m cubic geometry of SBA-16
[25]. As illustrated on Fig.2 the symmetry of mesoporous structure was essentially
maintained after calcination of the grafted titanium material.
In this study the hydrothermal aging of SBA-16 was set to be at the high temperature of
140oC which enhances the formation of larger pores in order to favor the diffusion of the
reactants. The catalytic activity of the epoxidation catalyst is mainly influenced by its
accessibility of the reactants to the active sites. Low temperature hydrothermal aging
could have resulted in high silanol density, mainly situated in the intrawall micropores
allowing high amounts of grafted titanium [12, 26]. On the other hand, during
calcination, the condensation of these silanols would have entrapped the active sites
located in the micropores, blocking their accessibility for the reactants [12]. Materials,
7

prepared at high temperature hydrothermal aging, present therefore higher density of
accessible active sites after calcination. This effect allows reaching higher catalytic
activities[12].
3.1.3 DR-UV
To be active in epoxidation reaction, Ti(IV) must be in a tetrahedral
coordination. Diffuse Reflectance UV–Visible (DR-UV-Vis) spectroscopy is a rapid
and effective technique for revealing the state of titanium species in porous silica
materials. In the range of 200-400 nm, the absorption of visible light is highly sensitive
to the coordination of the titanium states and to the presence of TiO2 clusters. Fig.3
shows the DR-UV-Vis spectra of Ti-SBA-16 obtained by impregnation with method 1,
the in-situ titanium ACAC synthesis and impregnation method 2.
All catalysts, prepared by both impregnation methods, show a peak centered at 220 nm
indicating that titanium is present in a tetrahedral coordination. For sample prepared by
method 1 with Ti/Si atomic ratios between 5 to 13.4%, the adsorption signal at 220 nm
does not increase which indicates that the amount of tetrahedral titanium reached a
constant value. These results are similar to those obtained for Ti catalysts based on SBA-
15 mesoporous silica [13]. Moreover, the adsorption spectrum is rather broad with
secondary peak at 320-340 nm. This suggests that small TiO2 clusters of penta-
coordinated titanium species are formed which could be ascribed to the formation of Ti-
O-Ti chemical bonds during the calcination process [27].
3.2 Catalytic performance
3.2.1 Effect of reaction temperature
The reaction of limonene with TBHP in the presence of Ti-SBA-16 catalysts may
result in different epoxide isomers represented on Fig.4.
8

In general, the rate of epoxidation increases with increasing substitution of the olefin with
alkyl groups, which is consistent with the electrophilic nature of the reagent [11]. Based
on this observation, the major products will be of course 1,2-limonene oxide. 1,2-8,9-
limonene dioxide would be produced by the double epoxidation of limonene.
Limonene oxide based polycarbonates are produced by reacting 1,2-limonene oxide with
CO2 [4, 5]. For industrial applications, it is crucial to obtain the highest possible yield of
1,2-limonene oxide. A series of catalytic tests were conducted using a catalyst prepared
by method 1 with an atomic ratio of Ti/Si of 7.3% in the interest of establishing the
optimal reaction parameters. First, the temperature effect on the conversion of limonene
and selectivity toward 1,2-limonene oxide was studied and the results are summarized on
Fig.5:
Fig.5 shows that a rise in temperature from 70oC to 75oC improves the conversion from
59% to 80% without affecting its selectivity. GC-MS analysis of the reaction medium
was performed for the determination of the reaction by-products. Only 8,9-limonene
oxide isomer as a secondary product was present in the reaction mixture. At 80oC, the
selectivity for 1,2-limonene oxide decreases drastically from 81% to 51% which is
explained by the thermal autodecomposition of tert-butyl hydroperoxide. At 45oC, TBHP
is considered to be indefinitely stable. On the other hand, heating the reaction medium
above this temperature favors the reaction of TBHP autodecomposition represented on
Scheme 1[28, 29]:
The first step of this reaction consists of the homolytic cleavage of the peroxo group
forming tert-butyl-O. and a radical OH.. Tert-butyl-O. then reacts with another tert-butyl
hydroperoxide molecule which produces tert-butanol and tert-butyl-OO. radical. The final
step of this radical mechanism consists of the conversion of tert-butyl-OO. radical into
9

molecular oxygen and tert-butyl-O.. These radicals present in the reaction medium react
with limonene and convert it into sub-products such as carvone, perillal and carveol.
Even though these molecules are often considered as undesired products, they could be
used in fragrance or perfume industries, and they also have real interest in polymer
industries.
3.2.2 Effect of TBHP
Tert-butyl hydroperoxide in decane solution was favored over aqueous TBHP and
hydrogen peroxide solutions in this work for several reasons. On the basis of their
electrophilic properties, there should not be difference between TBHP and H2O2 in
conjunction with metal catalysts epoxidation. TBHP is however far superior because of
its stability and ease of manipulation [11, 28]. The aqueous TBHP solution reduces the
conversion and selectivity since water molecules compete with the peroxide for
adsorption on the actives sites [30, 31]. Moreover, the presence of water in both H2O2 and
TBHP reduces the selectivity of epoxide, as when the latter is formed it is readily
converted to the corresponding diol [32]. When TBHP was used in aqueous solution, the
selectivity for the 1,2-limonene diol varied from 8 to 20% [15, 17, 33-35]. Even by
changing the hydrophilic properties of silica to make it hydrophobic, the 1,2- limonene
diol was still produced with a selectivity as low as 8% [33]. Only one reference reports
limonene epoxidation over SBA-15 supported Ti using both H2O2 and TBHP in aqueous
solution with apparent high selectivity to limonene oxide in contradiction with all other
published works [36]. Finally, TBHP is more beneficial to the selective production of
epoxide than H2O2. Transition state involving Ti–OOR and a cyclic alkene must be much
bulkier than that related to the Ti–OOH and alkene. This characteristic significantly
retards the reaction rate of cyclic alkenes with TBHP, further decreasing the possibility of
acid-catalyzed secondary ring-opening reaction of epoxide molecules [37].
Sheldon has suggested that epoxidation of alkenes under homogeneous conditions should
be performed at a high alkene/peroxide ratio to favor the selectivity for the epoxide and
10

avoid the competing radical reaction (scheme 1). On the other hand, under these
conditions, the yield for epoxides is very low and becomes less significant for industrial
needs [11]. It was later found that using 2 equivalents of anhydrous TBHP with 1
equivalent of olefin can yield as high as 90% of epoxide [11]. Different catalytic tests
were performed with different amounts of TBHP in acetonitrile at 75oC, the optimal
temperature determined in the previous section. The results are summarized on Fig.6.
The results in Fig.6 show that increasing the amount of TBHP from 2.75 to 8.25 mmol,
increases the conversion up to 45% with a selectivity remaining as high as 95%. When
the reaction was performed with 11 mmol of the TBHP, which corresponds to oxidizing
agent-olefin molar ratio of nearly 2:1, the conversion drastically increases to nearly 80%
as the selectivity for 1,2-limonene oxide has decreased to 81%. GC-MS analysis has
shown that 8,9-limonene oxide was present but no 1,2-8,9 limonene diepoxide was
detected in the reaction mixture. At 11 mmol of TBHP, it is then possible to conclude that
the selectivity for monoepoxides is 100%. The selectivity for the 1,2-epoxide decreases
from 81% to 54% when the reaction is carried out with 13.5 mmol of TBHP. At this
TBHP content, the autodecomposition reaction (scheme 1) becomes more favorable and
the OH˙ radical formation becomes more important which decreases the selectivity
toward the epoxide [28, 29].
3.2.3 Effect of solvent
The activity and selectivity of titanium catalysts for the epoxidation of alkenes is
related to the nature and polarity of solvents. For example, acetonitrile exhibits higher
conversion and selectivity when catalysts such as Ti-MWW and Ti-Beta are employed
compared to TS-1 which was shown to be more active in methanol medium [38, 39].
However, methanol and acetonitrile has similar polarity and they only differ by their
11

protic/aprotic properties. Polar protic solvent, especially alcohols, can form complexes
with the active center that help stabilizing the Ti-OOR with the formation of an hydrogen
bond [38, 39]. On the other hand, like water, alcohols such as methanol are less
interesting solvents due to their reaction with the epoxides which converts the later into
ether-glycol by-products [32]. In this, study three different aprotic solvents, ranging from
non-polar to highly polar, were used and compared. The results are displayed in Table 2.
Cyclohexane is the less interesting solvent. Its non-polar hydrophobic character does not
allow good diffusion of the substrate from the reaction medium to the active sites. In the
presence of titanosilicate materials, alkanes are known to be converted to alcohol by
radical reactions [[40]]. The high conversion of limonene in the presence of cyclohexane
is mostly related to the radical reaction from the tert-butyl hydroperoxide decomposition.
Ethyl acetate yields interesting conversion and selectivity. Carniato et al obtained similar
results by using TiO2 on amorphous SiO2 [16]. Acetonitrile, which has the highest
polarity, is by far the first choice solvent for epoxidation of alkenes such as limonene
using TiO2/SiO2 catalysts. This solvent was also found to be the first choice for
cyclohexene epoxidation and has yielded the highest conversion and selectivity [33, 35,
37-39]. For epoxidation reaction, solvent molecules are believed to be stabilising agents
of the Ti-OOR in order to maintain the complex and avoid decomposition of the
peroxide. Corma has proposed that a coordination chemical bond is formed between the
supported titanium peroxo complex with a lone electron pair of methanol. Then, a bond is
formed between one oxygen of Ti-OOR and the hydrogen from OH functional group.
This solvent-Ti-OOR structure helps stabilise the peroxo group and favors higher
conversion [39]. The relation between methanol and TiOOR is represented on Scheme
2a. Acetonitrile possesses the ability to complex with transition metals with its C-N π
electrons [41]. In this work, we propose that the coordination of acetonitrile with
12

supported titanium catalysts is similar to that of methanol. The complex between
acetonitrile and the supported titanium active site is represented on Scheme 2 b.
Similar to the coordination bond between the π electrons of C-N and titanium, a Van der
Waals interaction is formed between the electron poor nitrogen atom of acetonitrile and a
lone electron pair from the Ti-OOR. This bond enables to stabilise the peroxo group and
favors the conversion and selectivity for epoxides. Ethyl acetate does not possess the
ability to complex with titanium by strong coordination bonding which implies that the
conversion and selectivity is less with this solvent. Scheme 3, with the coordinating
solvent molecule not being represented for sake of simplicity, shows our proposed
mechanism for the epoxidation of limonene to limonene oxide.
The first step consists of the reaction between the supported titanium with TBHP which
forms a Ti-peroxo complex, the source of oxygen for the formation of the epoxide. The
Ti-O-Si chemical bond is broken and Ti-OOR and Si-OH are formed. Then limonene
reacts with the catalyst to form the epoxide by a concerted mechanism. The final products
of the reaction are limonene oxide isomers and tert-butanol. Finally, the Ti-OOR is
regenerated by reaction of Ti-OH with a molecule of TBHP present in the liquid medium.
3.2.4 Effect of pore structure and titanium loading
In this work, a comparison of SBA-15 and SBA-16 was also performed under the
optimal conditions determined in the previous section. A sample of Ti-SBA-15 was
synthesised by method 1 with physicochemical properties similar to the SBA-16 sample
with Ti/Si =7.3% shown in Table 1 (Ti/Si=8.2%, surface area of 637 m2/g, pore volume
of 0.94 cm3/g and pore diameter of 8.4 nm) [13].
13

The conversion and selectivity reported in Table 3 show that in absence of catalyst, the
reaction of epoxidation does not occur. It was found that Ti-SBA-16 gives higher
conversion and selectivity compared to the Ti-SBA-15 catalyst. This is likely due to
differences in the rate of deactivation by pore blockage by oligomerization of epoxide
products. In general, the tridimensional cage structure of SBA- 16 allows a much more
rapid diffusion of the products out of the catalyst particle than the straight cylindrical
pore shape of SBA-15 [23, 24, 42, 43]. Carniato et al. using ethyl acetate solvent also
found that the conversion of limonene is less important when SBA-15 is used compared
to amorphous SiO2 support [16]. The structure problems associated with the diffusion in
SBA-15 have been described in the literature [44].
As reported in Table 4, the Ti-SBA-16 with 7.9 % prepared using method 2
(TiO(ACAC)2) gives results in the epoxidation of limonene similar to those obtained with
Ti-SBA-16 prepared by the post-grafting technique developed by Berube et al (method 1)
[12]. This indicates that both materials with nearly the same amount of titanium possess
the same catalytic activities even though method 2 is simpler and more easily controlled.
Values of TBHP conversion and selectivity are obtained from GC traces are also shown
in Table 4. The selectivity values are conservative estimates as some TBHP thermal
decomposition may happened in the hot injector (260oC).
1,2-limonene oxide is obtained as a mixture of cis and trans isomers. Low molecular
weight polylimonene carbonate may be produced by the reaction of CO2 with a mixture
of cis and trans 1,2-limonene oxide, but high molecular weight polycarbonate is only
obtained from trans 1,2-limonene oxide [4, 5]. Table 5 shows the trans/cis ratio results
obtained by GC-MS with the different catalysts.
14

For Ti catalysts prepared by method 1 with Ti/Si of 5.0 and 7.3%, the isomer ratio
trans/cis does not seem to be affected. On the other hand, an increase of the trans isomers
is observed when the atomic ratio is 13.4%. The post grafting technique (method 2)
yielded higher trans/cis atomic ratio compared to the in-situ technique (method 1). It
seems that the proportion of the trans/cis isomers is affected by the atomic Ti/Si ratio and
that the source of ACAC also influences this proportion.
3.2.5 Catalysts stability
One important asset of heterogeneous catalysts is their reusability. It is shown on
Fig.7 that the conversion and selectivity remained almost unchanged for three repeated
cycles. Between cycles, the catalysts were removed from the liquid phase by
centrifugation, dried in air and finally calcined in air at 550oC.
(a)
(b)
The loss of titanium by leaching after three catalytic cycle was around 5% for all
catalysts impregnated by both methods except for Ti-SBA-16 with an atomic ratio of
13.4% which the leaching was less important and around 1.2%. One interesting
comparison to make is the reusability of the Ti-SBA-16 (Ti/Si = 5.0%) compared to the
Ti-SBA-15 (5.6%) by Berube et al for epoxidation of cyclohexene [13]. In our case
conversion is nearly the same for the three cycles while for Berube the conversion drops
by nearly 12%. The effect of the support structure is considered negligible. The only
difference is that in this work TBHP in anhydrous solvent was used in comparison to
Berube et al that used the aqueous THBP. The main advantage using TBHP in aqueous
solution is that it is commercially available in high concentration of the oxidizing agent.
On the other hand, besides the problems related to the epoxide yields mentioned in the
previous section, the presence of water is affecting catalyst reusability. It is then believed
15

that the hydrolysis of the Ti-O-Si when the epoxidation reaction in the presence of water
is more important than in anhydrous conditions. During the recalcination process, the
water molecules adsorbed on Ti sites favor the formation of titanium clusters with higher
Ti coordination structure, from tetrahedral toward octahedral, and then reduce the surface
density of actives sites which induces the decay of the olefin conversion [31, 45, 46].
4. Conclusion:
In conclusion, this work reports the catalytic epoxidation reaction of limonene over Ti-
SBA-16 catalysts. The 3D cage structure of SBA-16 was confirmed to favor the diffusion
of the reactants to the active sites compared to the 2D pore structure of SBA-15. The
epoxidation of limonene to limonene oxide was shown to be very sensitive to reaction
parameters such as temperature and amount of the oxidizing agent. These needs to be
appropriately controlled in order to avoid the autodecomposition of the TBHP which
favors the radical reaction. In this work, it has been established that the high polarity
solvents lead to higher limonene conversion and 1,2-epoxide selectivity. Acetonitrile is a
far better solvent than ethyl acetate. Finally, the catalysts were shown to be stable over
three catalytic tests in terms of both conversion or selectivity.
Acknowledgements:
Financial support from the Natural Sciences and Engineering Council of Canada
(NSERC). We thank Professor D.Zhao and W. Shuai from Fudan University for the
Small angles XRD measurements. We also thank Professor T.O. Do from Laval
University for DR-UV measurements.
16

References:
[1] A. Corma, S. Iborra, A. Velty, Chemical Reviews, 107 (2007) 2411-2502.
[2] P. Gallezot, Catalysis Today, 121 (2007) 76-91.
[3] R. Ciriminna, M. Lomeli-Rodriguez, P. Demma Cara, J.A. Lopez-Sanchez, M.
Pagliaro, Chemical Communications, 50 (2014) 15288-15296.
[4] C.M. Byrne, S.D. Allen, E.B. Lobkovsky, G.W. Coates, Journal of the American
Chemical Society, 126 (2004) 11404-11405.
[5] O. Hauenstein, M. Reiter, S. Agarwal, B. Rieger, A. Greiner, Green Chemistry, 18
(2016) 760-770.
[6] T. Michel, M. Cokoja, V. Sieber, F.E. Kühn, Journal of Molecular Catalysis A:
Chemical, 358 (2012) 159-165.
[7] T. Michel, D. Betz, M. Cokoja, V. Sieber, F.E. Kühn, Journal of Molecular Catalysis
A: Chemical, 340 (2011) 9-14.
[8] K. Ambroziak, B. Saha, R. Mbeleck, D.C. Sherrington, Journal of Ion Exchange, 18
(2007) 598-603.
[9] R.A. Sheldon, J.A. Van Doorn, C.W.A. Schram, A.J. De Jong, Journal of Catalysis,
31 (1973) 438-443.
[10] R.A. Sheldon, J.A. Van Doorn, Journal of Catalysis, 31 (1973) 427-437.
[11] R.A. Sheldon, Journal of Molecular Catalysis, 7 (1980) 107-126.
[12] F. Bérubé, A. Khadhraoui, M.T. Janicke, F. Kleitz, S. Kaliaguine, Industrial &
Engineering Chemistry Research, 49 (2010) 6977-6985.
[13] F. Bérubé, B. Nohair, F. Kleitz, S. Kaliaguine, Chemistry of Materials, 22 (2010)
1988-2000.
[14] M. Fukuda, N. Tsunoji, Y. Yagenji, Y. Ide, S. Hayakawa, M. Sadakane, T. Sano,
Journal of Materials Chemistry A, 3 (2015) 15280-15291.
[15] A. Wróblewska, Molecules, 19 (2014) 19907.
[16] F. Carniato, C. Bisio, L. Sordelli, E. Gavrilova, M. Guidotti, Inorganica Chimica
Acta, 380 (2012) 244-251.
[17] D. Marino, N.G. Gallegos, J.F. Bengoa, A.M. Alvarez, M.V. Cagnoli, S.G.
Casuscelli, E.R. Herrero, S.G. Marchetti, Catalysis Today, 133–135 (2008) 632-638.
17

[18] A. Wróblewska, E. Makuch, P. Miądlicki, Catalysis Today, 268 (2016) 121-129.
[19] J. Młodzik, A. Wróblewska, E. Makuch, R.J. Wróbel, B. Michalkiewicz, Catalysis
Today, 268 (2016) 111-120.
[20] A. Wróblewska, E. Makuch, P. Miądlicki, in: Polish Journal of Chemical
Technology, 2015, pp. 82.
[21] C. Berlini, M. Guidotti, G. Moretti, R. Psaro, N. Ravasio, Catalysis Today, 60
(2000) 219-225.
[22] F. Kleitz, L.A. Solovyov, G.M. Anilkumar, S.H. Choi, R. Ryoo, Chemical
Communications, (2004) 1536-1537.
[23] F. Kleitz, T.-W. Kim, R. Ryoo, Langmuir, 22 (2006) 440-445.
[24] O.C. Gobin, Y. Wan, D. Zhao, F. Kleitz, S. Kaliaguine, The Journal of Physical
Chemistry C, 111 (2007) 3053-3058.
[25] B. Nohair, S. MacQuarrie, C.M. Crudden, S. Kaliaguine, The Journal of Physical
Chemistry C, 112 (2008) 6065-6072.
[26] F. Zhang, Y. Yan, H. Yang, Y. Meng, C. Yu, B. Tu, D. Zhao, The Journal of
Physical Chemistry B, 109 (2005) 8723-8732.
[27] S. Shen, Y. Deng, G. Zhu, D. Mao, Y. Wang, G. Wu, J. Li, X. Liu, G. Lu, D. Zhao,
Journal of Materials Science, 42 (2007) 7057-7061.
[28] R. Hiatt, J. Clipsham, T. Visser, Canadian Journal of Chemistry, 42 (1964) 2754-
2757.
[29] W.F. Brill, N. Indictor, The Journal of Organic Chemistry, 29 (1964) 710-713.
[30] Y. Bu, Y. Wang, Y. Zhang, L. Wang, Z. Mi, Reaction Kinetics and Catalysis Letters,
90 (2007) 77-84.
[31] A. Hagen, K. Schueler, F. Roessner, Microporous and Mesoporous Materials, 51
(2002) 23-33.
[32] R.E. Parker, N.S. Isaacs, Chemical Reviews, 59 (1959) 737-799.
[33] M.V. Cagnoli, S.G. Casuscelli, A.M. Alvarez, J.F. Bengoa, N.G. Gallegos, M.E.
Crivello, E.R. Herrero, S.G. Marchetti, Catalysis Today, 107–108 (2005) 397-403.
[34] M. Sasidharan, A. Bhaumik, Journal of Molecular Catalysis A: Chemical, 328
(2010) 60-67.
18

[35] M.V. Cagnoli, S.G. Casuscelli, A.M. Alvarez, J.F. Bengoa, N.G. Gallegos, N.M.
Samaniego, M.E. Crivello, G.E. Ghione, C.F. Pérez, E.R. Herrero, S.G. Marchetti,
Applied Catalysis A: General, 287 (2005) 227-235.
[36] F. Chiker, F. Launay, J.P. Nogier, J.L. Bonardet, Green Chemistry, 5 (2003) 318-
322.
[37] Q.H. Xia, X. Chen, T. Tatsumi, Journal of Molecular Catalysis A: Chemical, 176
(2001) 179-193.
[38] W. Fan, P. Wu, T. Tatsumi, Journal of Catalysis, 256 (2008) 62-73.
[39] A. Corma, P. Esteve, A. Martınez, Journal of Catalysis, 161 (1996) 11-19.
́
[40] H. Fu, S. Kaliaguine, Journal of Catalysis, 148 (1994) 540-549.
[41] M.F. Farona, K.F. Kraus, Inorganic Chemistry, 9 (1970) 1700-1704.
[42] O.C. Gobin, Q. Huang, H. Vinh-Thang, F. Kleitz, M. Eić, S. Kaliaguine, The Journal
of Physical Chemistry C, 111 (2007) 3059-3065.
[43] V.-T. Hoang, Q. Huang, M. Eić, T.-O. Do, S. Kaliaguine, Langmuir, 21 (2005)
2051-2057.
[44] A.R. Menjoge, Q. Huang, B. Nohair, M. Eic, W. Shen, R. Che, S. Kaliaguine, S.
Vasenkov, The Journal of Physical Chemistry C, 114 (2010) 16298-16308.
[45] E. Gianotti, C. Bisio, L. Marchese, M. Guidotti, N. Ravasio, R. Psaro, S. Coluccia,
The Journal of Physical Chemistry C, 111 (2007) 5083-5089.
[46] W. Yu, Z. Zhang, H. Wang, Z. Ge, T.J. Pinnavaia, Microporous and Mesoporous
Materials, 104 (2007) 151-158.
19

Figures
Fig.1 N2-Sorption isotherms for Ti-SBA-16 catalysts
20

Fig.2 Small angle X-ray diffraction patterns of Ti-SBA-16 catalysts
Fig.3 DR-UV of selected Ti-SBA-16 Catalysts
21

Fig.4 Epoxides from the epoxidation of limonene
Fig.5 Temperature effect on conversion of limonene and selectivity for 1,2-limonene
oxide using Ti-SBA-16 Ti/Si = 7.3 %, 11 mmol of TBHP, 6 mmol of limonene in
acetonitrile
22

Fig.6 Conversion and selectivity to 1,2-limonene oxide as a function of TBHP initial
content using Ti-SBA-16-(1) Ti/Si = 7.3 %, at 75oC with 6 mmol of limonene in
acetonitrile
23

a)
b)
Fig.7 Conversion of limonene a) and selectivity to 1,2-limonene oxide b) upon repeated
catalytic tests
24

Scheme 1 Mechanism of TBHP decomposition
Scheme 2 Coordination of a) Methanol with supported titanium on silica according
Corma et al.[39] and b) Proposed coordination structure of acetonitrile with Ti-SBA-16
25

Scheme 3 Proposed mechanism for the epoxidation of limonene to 1,2-limonene oxide
26

Table 1 Physical properties of the catalysts
Pore
diameter
(nm)
Ti/Si
(%)
Surface
Pore Volume
(cm3/g)
Catalyst
(m2/g)
SBA-16
-
693
12.6
0.82
Ti-SBA-16-(1)
Ti-SBA-16-(1)
Ti-SBA-16-(2)
Ti-SBA-16-(1)
5.0
608
600
635
582
11.6
11.4
11.7
11.2
0.71
0.68
0.78
0.60
7.3
7.9
13.4
27

Table 2 Relation of solvent dielectric constant with conversion and selectivity
1,2-limonene oxide
Selectivity
(%)
Relative
polarity
Temperature
oC
Conversion
(%)
Solvent
Acetonitrile
Ethyl Acetate
Cyclohexane
0.460
0.228
0.006
75
75
75
80
65
58
82
68
5
Reaction conditions: 6.2 mmol of limonene, 11 mmol of THBP in decane, 0.3 g of Ti-SBA-16-(1) (Ti/Si
7.3%), 75oC, 10 mL of solvent, 24h
28

Table 3 - Marked Comparison of catalysts support in the epoxidation of limonene to
limonene oxide
Ti/Si
(%)
1,2-limonene
oxide Selectivity Conversion Selectivity
(%)
TBHP
TBHP
Conversion
(%)
Catalyst
(%)
(%)
No catalyst
Ti-SBA-16-(1)
Ti-SBA-15
0
47
80
38
3
7.3
82
75
80
45
20
8.2
72
Reaction conditions: 6.2 mmol of limonene, 11 mmol of THBP in decane, 0.3 g of catalyst, 75oC,10 mL
of acetonitrile, 24h
29