Molecular Catalysis 528 (2022) 112511

Contents lists available at ScienceDirect

Molecular Catalysis
journal homepage: www.journals.elsevier.com/molecular-catalysis

Synthesis of heterocycles compounds from condensation of limonene with

aldehydes using heteropolyacids supported on metal oxides
Julián E. Sánchez-Velandia a, *, Herme G. Baldoví b, A. Yu Sidorenko c, Jaime A. Becerra d,
Fernando Martínez O a, *
a
Centro de Investigaciones en Catálisis-CICAT, Universidad Industrial de Santander Escuela de Química, Km 2 vía El Refugio, Piedecuesta, Santander, Colombia
b
Departamento de Química, Universidad Politécnica de Valencia, Camino de Vera s/n, 46022, Spain
c
Institute of Chemistry of New Materials of National Academy of Sciences of Belarus, Skaryna str, 36, Minsk 220141, Belarus
d
Environmental Catalysis Research Group, Chemical Engineering Department, Engineering Faculty, Universidad de Antioquia UdeA, Calle 70 No. 52-21, Medellín,
Colombia

A B S T R A C T

In this contribution, heterogeneous materials based on heteropolyacids (HPA, H3PW12040 hydrate) supported over several metal oxides (TiO2, SiO2, SBA-15, Al2O3)
were prepared using wetness impregnation, characterized by XRD, FTIR, SEM, acidity by TPD, nitrogen physisorption, RAMAN spectroscopy, 31P NMR and then
evaluated in the catalytic condensation of limonene with benzaldehyde. After incorporation of HPA over each support, no structural changes were observed as it was
suggest by FTIR, RAMAN and XRD. Acidity analysis by TPD-NH3 showed that increasing the HPA amount increases the acidity of the catalyst until it is constant
(20–30%). HPA over alumina, silica and SBA-15 also showed high values of acidity but less than typical HPA/P25 catalyst (HPA over commercial Degussa P25). This
last was one of the most selective materials (surface area of 39 m2/g and acidity of 187 µmol NH3. g−cat1) for the synthesis of 3-oxabyciclo[3.3.1]nonane (up to 80%)
with a low selectivity to the limonene isomers by-products. The use of water as a solvent decreased considerably the catalytic activity. The same observation was
found when the volume of the solvent increased to 30 mL. Besides, catalytic condensation can be performed at free solvent conditions and at room temperature;
however, when decrease temperature, a higher number of isomers was observed rather than the desired product. In addition, catalytic condensation was also
evaluated in biomass of essential oils and mixtures with sesquiterpenes showing good to excellent results. Typical LHHW mechanism was evident after kinetic
analysis. HPA/P25 was shown to be a robust material since can be reused up to two times with a possible decrease in its catalytic activity.

1. Introduction heterocyclic products with various types of structures like chromene

[13,14], isobenzofuran [4,15], as well as complex polycyclic molecules
The molecules containing agro-industrial wastes can be used as an [16]. It is known that some of these products exhibit high biological
alternative sources for the production of chemicals like fragrances, food activity (analgesic, antiviral, antitumor), and can also be used in fine
additives, biofuel, vitamins, antioxidants, etc. [1]. Chemical molecules chemistry and perfumery [17,18]. Thus, 3-oxabicyclo[3.3.1]nonane
from biomass are structurally different from the carbon forms in fossil compounds or their derivatives are one type of molecules that can be
resources and then, their transformation into valuable products is a obtained through condensation of monoterpenes with aldehydes. The
current challenge. Monoterpenes are molecules considered as an excellent properties as a biological precursor, as well as metabolites (or
important and valuable start platforms of chemicals that can be con­ as fragrance), reinforce their importance to be investigated [19]. Even,
verted into a lot of substances with potential and attractive attention in these compounds are important synthons that offer an important plat­
the fine chemistry industry (pharmaceutical, medical, agrochemical, form for the construction of biologically/medicinally substances with
etc.) [2,3]. In addition, these compounds and their derivatives play an numerous stereocenters [20]. It should be noted that some of the mol­
important role in the creation of new biologically active substances [4]. ecules with a 3-oxabicyclo[3.3.1]nonane structure are potent inhibitors
The transformation of monoterpenes and monoterpenoids has been of the TDP1 enzyme, a target for anticancer therapy [15]. Heteroge­
addressed using different chemical routes, including isomerization [5, neous catalysts can be used as greener, sustainable, and more efficient
6], biotransformation [7], condensation [8] oxo-transfer [9], alkylation routes for the synthesis of oxabicyclic or isobenzofuranes-like com­
[10], hydrolysis [11], pyrolysis [12], among other. Condensation of pounds. There are a lot of advantages for using heterogeneous materials
terpene compounds with aldehydes makes it possible to obtain for the transformation of chemicals into high-added value substances

* Corresponding authors.
E-mail addresses: [email protected] (J.E. Sánchez-Velandia), [email protected] (F. Martínez O).

https://doi.org/10.1016/j.mcat.2022.112511
Received 1 June 2022; Received in revised form 9 July 2022; Accepted 12 July 2022
2468-8231/© 2022 Elsevier B.V. All rights reserved.
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

including economic impact and robustness (among others) which could example, titanium dioxide is a material of choice in several applications
improve many of the current processes. Then, the use of heterogeneous which include cell attachment and proliferation [26]. Together with
catalysts is an efficient alternative for the synthesis of HPA could be applied in both pharmaceutical and medical industries.
isobenzofurans-like compounds. Considering the advantages of titanium dioxide together with the use of
It has been reported that isobenzofurans derivatives, with a phar­ heteropolyacids for increasing the acidity and for using in condensation
maceutical potential, can be achieved from renewable 2-carene over reactions, these kinds of materials could be an alternative strategy to
halloysite nanotubes (Fig. 1a) [21]. The highest selectivity to the desired propose rather than previous heterogeneous HPA/SiO2 [8]. In addition,
target (ca. 70%, at soft reaction conditions: 50 ◦ C, 1.0 g of catalyst, 0.1 g and considering the lack of metal oxides used for the condensation of
of monoterpene, 0.1 of anisaldehyde) was related to the weak Brønsted limonene with benzaldehyde, other supports based on metal oxides
acid sites, which promote the condensation rather than parallel isom­ could be also tested.
erization (side reaction). Similarly, heteropoly acid catalysts (and its Prompted by the properties and acidity of heteropolyacids catalysts,
acidic Cs salt: Cs2.5H0.5PW12O40) were tested for the synthesis of 3-oxa­ in this contribution, we have evaluated the catalytic activity of the HPA
bicyclo[3.3.1]nonane (Fig. 1b) with a wide class of monoterpenes supported on inert oxides such as titanium oxide, silica nanoparticles
(including α-pinene, β-pinene, limonene, nerol, α-terpineol) and benz­ and SBA-15 in the selectivity synthesis of 3-oxabyciclo[3.3.1]nonane
aldehyde, cuminaldehyde and trans-cinnamaldehyde [8]. High yields like compounds from the condensation of limonene and benzaldehyde.
were observed at relatively benign conditions (dimethyl carbonate or The parameters evaluated were the amount of HPA impregnated on the
2-methyltetrahydrofuran and 50 ◦ C). However, it is seeming dubious best support, amount of aldehyde, temperature, and nature of acid
that the synthesis of these kinds of compounds from bicycle α- and support. Limonene was chosen because of its availability and besides
β-pinene is because of their strainer internal structure which must favor that is a large-tonnage waste juice production [27,28]. The relationship
the isomerization instead of condensation. In this case, the monovalent between surface textural properties and acidity was elucidated. Finally,
cation of Keggin’s structure, Cs2.5H0.5PW12O40, was applied because of kinetics was also included in this research incorporating the parameters
its strong Brønsted acidity, larger surface and high thermal stability. of both pseudo-homogeneous and heterogeneous models fitted to
Taking this into account, condensation of bicycle monoterpenes (such as experimental tests by using a genetic algorithm. For the best of our
3-carene and pinenes) with aromatic aldehydes should be favored with knowledge, no detailed study on this topic has been regarding until now.
weak Brønsted acid sites while condensation of cyclic monoterpenes
(such as limonene) is stand up for strong Brønsted acid sites. Then, 2. Methodology
Brønsted acid sites should be considered for these types of C-C
condensation reactions as an important parameter to be studied. 2.1. Synthesis of HPA impregnated on inert oxides
Heteropolyacids catalysts (HPA) with a typical Keggin’s structure
have been applied for numerous reactions: esterification and biomass Heteropolyacid (H3PW12040 hydrate from Aldrich) supported on
conversion [22], selective sulfoxidation [23], methacrolein oxidation inert oxides (TiO2-P25 from Sigma Aldrich (SBET = 34 m2 g− 1) SBA-15,
[24], among others, and their excellent features as heterogeneous cat­ titanium nanopowder from Sigma-Aldrich (SBET = 45 m2 g− 1), Alumina
alysts is because their acidity which allows obtaining selectively (SBET = 40 m2g− 1), silica (SBET = 28 m2 g− 1) were prepared by wetness
different targets. Its Brønsted acidity is due to the ionic structure which impregnation. Briefly, 0.12 g of HPA (20 wt%) was dispersed on 0.6 of
permits high proton mobility. However, their main disadvantage is each support (previously dried at 110 ◦ C for 2 h) and then, 30 mL of
related to their easy lixiviation in polar solvents. It is for this reason that water was added. The mixture was constantly stirred for 2 h and after­
dispersion/anchoring/impregnation of this type of HPA on inert sup­ ward dried at 110 ◦ C for 17 h. Finally, the solid was calcined at 300 ◦ C
ported materials should enhance their robustness and then, their recy­ for 5 h (2 ◦ C min− 1 under static air). At this temperature, Keggin’s
clability in catalytic reactions. Also, increasing their thermal stability structure is maintained to assure the impregnation of HPA on the sup­
and improving acid strength in inert solids [25]. Among of the supports, port. The solids were named HPA/S, where HPA stands for hetero­
metal oxides have shown to be interesting because of their high surface polyacids while S means support (used during each impregnation).
area, good mechanical stability, and besides they are biocompatible. For

Fig. 1. (a) Isobenzofurane derivative (b) 3-oxabicyclo [3.3.1] nonane, products of the condensation of monoterpenes with aldehydes. R means aromatic
derived aldehyde.

2
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

2.2. Characterization techniques limonene, 16% others) was acquired from MultiQuímicos (Bucar­
amanga, Colombia) while Rosmarinus oil (19% α-pinene, 11% β-pinene,
X-ray powder diffraction (XRD) data were measured with a Bruker 7% camphene, 16% limonene, 2% p-cymene, 45% others) was pur­
PANalytical Empyrean diffractometer (Malvern Instruments Limited, chased from Perfumes&Aromas (Colombia). The corresponding 3-oxa­
Malvern, UK) (Cu Kα radiation) in transmission geometry. The analysis byciclo[3.3.1]nonane was isolated and purified from the reaction
by FTIR was performed in a Perkin Elmer Frontier spectrum 65 model mixture using preparative thin layer chromatography with petroleum
with a resolution of 4 cm− 1 and 64 scans. The morphology and the ether and ethyl acetate as eluent solvents. The product was well iden­
composition of each material were characterized using a high-resolution tified using NMR (1H and 13C) (reported previously in [29] and which
field emission scanning electron microscope (HRFESEM from Zeiss in­ correspond to 3-oxabyciclo[3.3.1]nonane; supporting information)
struments, Model GeminiSEM 500), equipped with Energy dispersive X-
ray detector, EDS (OXFORD INSTRUMENTS) for composition analysis.
2.4. Kinetic assessments
Nitrogen adsorption/desorption isotherms at 77 K were measured on a
Micromeritics Tristar 3000 apparatus at -196 ◦ C. Before analysis, the
For the kinetic analysis of the condensation of limonene with benz­
samples were degassed under vacuum at 150 ◦ C for 12 h. The specific
aldehyde, reaction rates were calculated by the initial reaction rate
surface area was determined from the linear part (0 - 0.23 P/P0) of the
method from experimental data of concentration in a liquid-phase batch
BET plot. The total pore volume was measured from the isotherms at P/
reactor at several conditions shown in Table 1. The details of the cal­
P0 = 0.95 and the mean pore diameter was determined by the BJH
culations of the reaction rate by initial reaction rate method is presented
method applied to the adsorption branch. Raman spectra were collected
in Supporting Information. The kinetic parameters were obtained
with a Renishaw In Via "Reflex" Raman Spectrometer equipped with an
through an optimization of the sum of squared errors in an objective
Olympus microscope and a CCD Detector using a 325 nm laser with a
function fitting experimental reaction rate calculations with some ki­
power of 15 mW as the excitation source. The spectra were collected by
netics expressions incorporating the Genetic Algorithm (GA) in Matlab
averaging 10 scans at a resolution of 3 cm− 1. The acidity of the catalysts
software. The GA is a method for solving constrained and unconstrained
was determined by temperature-programmed pulsed adsorption and
optimization problems (with discontinuous, nondifferentiable, stochas­
desorption of ammonia (NH3-TPD) experiments utilizing a Micro­
tics or highly nonlinear objective functions) based on natural selection,
meritrics Autochem II paired with an OmniStar Balzers Instrument mass
randomly modifying a population of individual solutions. The proposed
chromatograph. The sample amount of 0.95 g was prepared to remove
kinetics expressions stablished from heterogeneous models based on the
any adsorbed solvent, water or gas adsorbed, by pre-treating it (in
Langmuir-Hinshelwood-Hougen-Watson (LHHW) and Eley-Rideal (ER)
vacuum) with Ar flow of 10 mL/min at 200 ᵒC for 1 h and a He flow of
mechanisms, along with a power law based pseudo-homogeneous (pH)
10 mL/min at 200 ᵒC for 20 min. Then, analysis of the samples was
model.
performed after cooling down to room temperature starting with a
pulsed adsorption of ammonia with a He flow of 100 mL min− 1 at 100
3. Results and discussion
ᵒC. Each pulse of ammonia lasted 10 min and NH3 was introduced to the
system via a 0.5 mL loop at 100 ᵒC. When the sample was saturated with
3.1. Catalyst characterization
NH3, the desorption process started with a He flow of 100 mL/min and
heating up to 500 ᵒC for 5 min (heating rate of 10 ᵒC min− 1). gasses
3.1.1. FTIR measurements
desorbed were monitored by online QMS (Quadrupole Mass Spectros­
Fig. S1 (in Supporting Information) shows the IR spectrum for all
copy) analysis. The total acidity of the catalyst was calculated by inte­
HPA synthetized and impregnated catalysts in comparison with the HPA
gration of the NH3 desorption profile referred to as the QM signal, at
(H3PW12040 hydrate) pattern. The spectrum of the HPA pattern clearly
mass 15. Such mass value was preferred to the mass signal at 17, to
shows the characteristic bands of the [PW12O40]3− Keggin anion at 1080
exclude any contribution due to water fragmentation. Thermogravi­
(υas(P-Oa)), 982 (υas (W-Od)), 890 (υas (W-Ob-W)), and 800 cm− 1
metric analyses were performed on a TGA/SDTA851e METTLER
(υas(W-Oc-W)) [30]. It seems that in all supported catalysts, no bands
TOLEDO station. Analysis was carried out under O2 atmosphere and
related to the typical vibration of Keggin anion were observed which
heated up to 900 ᵒC (rate of 10 ᵒC min− 1)
could be associated with a homogeneous dispersion of this poly­
oxometalate acid in all supports or because of the H-bonding in­
2.3. Catalytic reactions
teractions with the supports which could be disturbed by dipole-dipole
interactions [31]. On the contrary, typical vibrations of each support
Initial catalytic tests were carried out with limonene and benzalde­
were noticed. For example, for SBA-15 material, bands at around 960
hyde (purchased from Sigma-Aldrich with a purity of at least 98%) and
and 1090 cm− 1 corresponding to Si-OH vibration and Si-O-Si asym­
they were used as received at least otherwise. Catalytic reactions were
metric stretching vibration of SiO2 were observed [32]. In the case of
performed in 15 mL batch reactors under magnetic stirring. In a typical
HPA/SBA-15, the same bands that parent SBA-15 were found. In addi­
experiment, the selected amount of catalyst (10, 20, 30 mg) was added
tion, for HPA supported on silica nanoparticles HPA/np-SiO2 the same
to different amounts of limonene and benzaldehyde (Typically 1 mmol
vibrations were achieved. For titanium dioxide patterns and their
and 3 mmol, respectively) and constantly stirred for 5 h and 50 ◦ C.
modification with HPA (that includes titanium dioxide HPA/P25 and
Catalysts were ground for obtaining a particle size of <100 µm and the
final mixture was stirred at 750 rpm, with the objective to ensure no
Table 1
diffusional problems related to internal and external mass transfer. The
Set of experiments and reaction conditions used for the kinetic analysis.
products were identified by gas chromatography with a mass selective
(GC–MS) and FID, Agilent 6890, and HP-1 column (100 m x 250 μm x Test set nA/nB mA (g) mcatalyst (g) V (mL)
1 7.3873 0.0658 0.0310 0.4515
0.5 µm). The carrier gas was N2 (23.80 mL min− 1) and the split ratio was
2 2.4298 0.2000 0.0306 0.6110
15:1. The oven temperature was kept at 50 ◦ C for 3 min and then it was 3 3.6230 0.1331 0.0305 0.5286
raised to 180 ◦ C at 15 ◦ C min− 1 and maintained at that temperature for 2 4 2.9883 0.2659 0.0301 0.9264
min. Finally, the oven temperature was again raised to 300 ◦ C at 15 ◦ C 5 8.6758 0.0134 0.0305 0.1052
and maintained for 2 min. n-Heptane was used as an internal standard. ni: mol of specie i (A: limonene, B: benzaldehyde), all tests with HPA/P-25, V:
Balance of mass was verified checking that it was at around 100% volume of the reaction. Each set test comprises 7 individual experiments at
(95–100%). Reactions with essential oils were carried out in the same different times of reaction: 5, 15, 30, 45-, 60-, 120-, and 300-min. Conditions:
way. Turpentine oil (72% α-pinene, 9% β-pinene, 1% camphene, 2% 750 rpm, 50 ◦ C.

3
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

nano powder HPA/NP-TiO2) show bands at 500 and 600 cm− 1due to Table 2
typical Ti-O and Ti-O-Ti vibrations [33]. For both Al2O3 and HPA/Al2O3, Textural properties for the synthetized HPA supported catalysts.
it seems the typical Al-O stretching band in octahedral structure at 459, Catalyst SBET (m2 Smeso (m2 Pore volume (cm3 Dp
595 and 656 cm− 1. On the other hand, Al-O and Al-O-H bands at 715 and g− 1) g− 1) g− 1) (nm)
1072 cm− 1are related to symmetrical bending mode. In addition, the HPA 3 - - -
HPA/SiO2 19 - 0.10 9
broad peak between 900 and 1100 cm− 1 are assigned to O-H deforma­
P25 49 - - -
tion vibrations [34]. 5-HPA/P25 37 - - -
10-HPA/P25 36 - - -
3.1.2. X-ray diffraction 20-HPA/P25 39 - - -
To obtain information about the crystallographic structure of each NP-TiO2 73 20 0.25 13
HPA/NP- 62 9 0.13 10
material, XRD patterns are shown in Fig. 2. In the case of the homoge­
TiO2
neous Keggin’s acid, it shows the typical shifts which are assigned to SBA-15 555 481 0.40 25
H3PW12O40⋅14H2O according to JCPDS data [35]. It is well known that HPA/SBA-15 444 386 0.23 7.2
Keggin anion in the heteropolyacid structure is composed of a central Al2O3 129 100 0.60 12
HPA/Al2O3 81 77 0.41 9.1
PO4 tetrahedron unit surrounded by 12 edge- and corner- sharing
metal-oxygen octahedral WO6. Secondary structures are formed by
protonated species of water (e.g., H5O+ 2 ) which connect four heteropoly amounts of HPA are incorporated. Although these differences are not
anions using a hydrogen bond with the terminal M=O atoms (M is metal significant, it is observed a decrease in the surface area could be
and O is oxygen). When heteropolyacid was impregnated on different attributed to the block of the pores or the formation of small crystal (or
metal oxides, the typical signals attributed to the HPA disappeared while aggregates) that appears to change structurally the catalyst structure
the shifts of each support appear slightly. For example, the well-defined [41]. On the other hand, different charges of HPA over P25 do not affect
peaks for HPA/SiO2 show the signals of the SiO2 together with the the surface area of the material. This fact could be associated with the
typical heteropolyacid shifts. Similarly, for both HPA supported on ti­ same crystal size which is directly related to the surface area of the
tanium dioxide nanopowder and P25 showed the anatase and rutile material [42]. Besides, when preparation of the catalyst was scaled up to
phases in all the ranges at the shifts: 25.4◦ , 38.0◦ , 48.0◦ and 54.7◦ 10 g, it was found that the surface area is almost the same (34 m2 g− 1)
(JCDPS No. 21-1272) [36,37]. For the material modified over alumina, allowing to verify the high reproducibility of the synthesis of the
the typical γ-phase of alumina appears with the (3 1 1), (2 2 2), (4 0 0) material.
and (4 4 0) reflections [38,39]. In general, there were no intense peaks Comparatively, when titanium dioxide nano powder, NP-TiO2, was
corresponding to the crystalline structure of the HPA phase detected for used as support, the surface also decreased after incorporation of the
all the catalysts. This clearly indicates that HPA could be highly HPA unit, but its surface area is higher, again, with respect to typical
dispersed on the support of amorphous WOx species (Because of the Degussa P25. On the other hand, other materials with different geom­
calcination time) which were decomposed during the synthesis. etries and structures were also used: SBA-15 and γ-alumina. Similarly,
when HPA is impregnated over both supports, the surface area decreases
3.1.3. Textural properties significantly up to 444 and 81 m2 g− 1, respectively. In the case of SBA-15
Table 2 shows the textural properties measured for all heterogeneous materials, the apparent mesoporous surface area is drastically affected
HPA catalysts. It is seeming that the surface area for HPA homogeneous by the incorporation of kegging’s unit. In general, changes in surface
(5 m2 g− 1) is lower in comparison to other materials. It has been re­ area in all supports are affected with a minimum of 15% and a maximum
ported the low surface area of non-supported Kegging’s structure [40, of 37% with respect to the support without modification.
41]; however, changes are observed when different supports are
impregnated within heteropolyacids. For example, when P25 (titanium 3.1.4. Scanning electronic microscopy (SEM)
dioxide from Degussa P25, composed mainly for Anatase) is used as Some selected HPA supported on P25, SBA-15 and SiO2 are showed
support a decrease in the surface area is observed when different in Fig. 3. For HPA/P25 (Fig. 3 (a, b)), it seems a large distribution of
boundary rugosity together with small particles on the surface. Some
cylindrical geometries with irregular shape together with jagged edges
were also observed. In the case of HPA/SBA-15 (Fig. 3 (c, d)) several
ropes like shape particles were seen and it appears that the size of the
ropes is relatively uniform. On the other hand, these thin ropes can be
aggregates making macroscopic structure with a well wheat like struc­
ture defined. Similar observations have been reported previously [43].
For HPA/SiO2, among of the particles observed were of similar shape
and even of size that such obtained for HPA/SBA-15 (Fig. 3 (e, f)). White
spots over the surface of silicon oxide are observed with a good
dispersion. These particles are mainly associated with the deposition of
heteropolyacid over the typical structure of the material.

3.1.5. Raman spectroscopy

Typical Raman spectra were acquired for almost all the samples, and
they are showed in Fig. S2 (Supporting information). In the case of HPA/
P25 it seems typical bands of the rutile phase at 448 cm− 1 and at 610
cm− 1 which are characteristics of the Eg and A1g vibrational modes [44].
The same bands and behavior were observed for HPA supported on ti­
tanium dioxide nanotubes. Interestingly, for both samples no band at
1010 cm− 1 was observed (characteristic for the Keggin structure) [45].
Contrary, this band was detected for both HPA/Al2O3 and HPA/SiO2
materials showing a slight dispersion of the heteropolyacid on each
Fig. 2. XRD pattern of HPA supported on several oxides.

4
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

Fig. 3. SEM images of: (A, B) HPA/P25 (C, D) HPA/SBA-15 (E, F) HPA/NP-SiO2.

support. A broad peak at around 930–950 cm− 1 could be attributed to Bu

Table 3
mode of γ-Al2O3 [46]. For HPA/SBA-15 and HPA/SiO2 a band at 1600
Acidity for different heteropolyacids supported materials.
cm− 1 is associated to typical vibration of Si-O-Si which is in line with the
observations achieved by using FTIR (Section 3.1.1). Material Total acidity (µmol g−cat1)

HPA/SBA-15 78
3.1.6. NMR measurements 5-HPA/P25 83
Solid-state nuclear magnetic resonance spectroscopy is an atomic- HPA/SiO2 109
level method to determine the chemical structure among contain 10-HPA/P25 112
atoms such as phosphorus, aluminum, etc. [47]. Analysis of solid-state 15-HPA/P25 121
HPA/Al2O3 145
of 31P NMR spectrum can indicates a chemical interaction of PW unit HPA/P25 187
with the support. In the literature it is well reported two kinds of peaks 30-HPA/P25 183
when HPA is supported on different oxides: a peak at around -15 ppm HPA/NP-TiO2 214
related with intact Keggin structure and another in -14 ppm associated
with species such as H6P2W18O62 or H6P2W21O71 [48]. Also, because of
constant (183–187 µmol/g-cat). In fact, the increases of heteropolyacid
the standard (H3PO4 in many of the cases), this shift can be improved at
increases the acidity of typical titanium dioxide.
least 0.01 ppm, making better the process of identification of the species
In the same way, at the same loading (20%) titanium dioxide
[49]. Fig. S3 (in supporting information) shows the 31P NMR spectrum
nanoparticles generates more acid sites in comparison with Degussa
for the homogeneous heteropolyacids (H3PWO40) and for HPA sup­
P25. Although both materials present the same composition, the crys­
ported on SBA-15, SiO2 and titanium dioxide Degussa P25. In all sam­
tallographic phases and even the surface area is different (62 m2 g− 1 for
ples, a wide signal at around -15.4 ppm is observed evidencing the
HPA/NP-TiO2 and 39 m2 g− 1 for HPA/P25). This fact could be associ­
presence of the Keggin structure after its incorporation over the support
ated with the surface area can even change the distribution, coverage,
or also the presence of several coexisting species such as α- [PW12O4O]3−
and density of the acid sites in the catalyst. Then, materials with low
[49]. Apparently after incorporation of the monolayer of hetero­
surface area will exhibit a high density of the acid sites which opposite
polyacids over each support, the characteristic peak which suggest the
trend will have less density of the acid sites. On the other hand, HPA/
presence of the fundamental unit of the heteropolyacids is not shifted,
SiO2 and HPA/SBA-15 exhibited a difference of 31 µmol/g-cat which
however, in the case of HPA/P25 it seems a broad band close to the
could be owning with the typical geometry and the surface area of both
region of -10 to -5 ppm. This band (peak) is mainly associated with
supports (19 vs 444 m2/g, respectively). In general, the order of total
structural changes of Keggin unit and possible defects in the molecule
acidity for the most relevant materials was the following: HPA/NP-TiO2
which are not possible to detect because of the conditions of the
> HPA/P25 > 30-HPA/P25 > HPA/Al2O3 > 15-HPA/P25 > 10-HPA/
experiment. Considering these results and the previous one based on
P25 > HPA/SiO2 > HPA/SBA-15.
FTIR and XRD, it could be possible to suggest that possibly hetero­
polyacids is dispersed in each support as its fundamental cations rather
3.2. Catalytic activity
than the typical crystal structure.

3.2.1. Screening of heterogeneous catalysts

3.1.7. Temperature programmed desorption of ammonia (TPDA)
All synthetized materials were evaluated in the catalytic condensa­
Catalyst acidity is an important feature to be considered, being a
tion of limonene with benzaldehyde for obtaining 3-oxabyciclo[3.3.1]
variable that can be measured by different ways; one of them is the
nonane. Table 4 shows the results obtained for each HPA supported
desorption of ammonia at different temperatures. This is a procedure
material. After 5 h, total conversion was achieved for the homogeneous
that indicates the strength and the amount of acid sites in a material. The
HPA and HPA supported on silica, nevertheless for these cases, selec­
values of total acidity are showed in Table 3. As can be seen, the catalyst
tivity to the desired target was 80 and 70%, respectively. For HPA
with the highest content of acidity (up to 214 µmol g−cat1) was the HPA/
supported on different morphologies of the TiO2, a different behavior
NP-TiO2, while the HPA/P25 with both 20% and 30% wt had similar
was observed. At 5 h, conversion of limonene decreased in the following
values of acidity. Interestingly, the variation of the amount of hetero­
order: HPA/P25 > HPA/NP-TiO2. Typical distribution of the acid sites
polyacid on titanium dioxide increases its acidity (5–20%), however, at
could be responsible of this behavior (specially, distribution of Brønsted
high loading of HPA (25%, 30%) the total acidity appears to keep
acid sites). When density of the acid sites (defined as number of acid sites

5
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

Table 4
Catalytic activity of the HPA supported catalysts.
Catalyst -r0 (mmol Conversion of Selectivity to TON**
g−cat1 min− 1) limonene after 5 h oxabicyclo [3.3.0] (mol/
(%) nonane* mol)
HPA 0.4 > 99 80 165
HPA/ 3.6 > 99 70 165
SiO2
HPA/P25 2.6 92 72 153
HPA/NP- 0.3 40 50 67
TiO2
HPA/ 2.0 91 60 152
SBA-15
HPA/ 0.0 3 - 5
Al2O3

Reaction conditions: 50 ◦ C, CH2Cl2 as solvent, 750 rpm. *Selectivity at 20% of

conversion. Average rate and calculated as -r0 = (n0-nt)/t.mcat where n0 are the
initial mole, nt the moles at the time t, and mcat is the mass of catalyst. **TON
was calculated as mmol of limonene converted over mmol of HPA used in the
synthesis.

divided by surface area) of these catalysts was compared, it was possible

to stablish that the same order of density of acid sites (e.g., 3.45 µmol
m− 2 for HPA/NP-TiO2 vs 4.80 µmol m− 2 for HPA/P25) was achieved.
Then, density of the acid sites could be crucial for conversion of limo­
nene and further synthesis of 3-oxabyciclo[3.3.1] nonane as it has been
reported previously for condensation of a well-known reaction of
monoterpenes with aldehydes [50,51].
In addition, HPA dispersed on alumina did not achieve activity for
this reaction. Although this material exhibited high acidity, the typical
geometry (architecture) of each alumina could not favor the trans­
formation of limonene with benzaldehyde. On the other hand, HPA
appears to decompose on WO3 phases at temperatures higher than
600 ◦ C [52], however, our hypothesis points out to describe that either
decomposition of HPA on these supports or the strength of the acid sites
could be responsible for the non-catalytic activity in this reaction.
Selectivity to 3-oxabyciclo[3.3.1]nonane at 20% of conversion was ob­
tained in the next order: HPA > HPA/P25 > HPA/SiO2 > HPA-SBA-15.
As selectivity depends on the conversion of limonene in each hetero­ Fig. 4. Conversion of limonene (a) and selectivity to 3-oxabicyclo [3.3.1]
geneous catalyst, acidity and dispersion should be a critical factor for the nonane (b) using several HPA supported materials. Reaction conditions: 30 mg
synthesis of the desired target. Even though over silica, the excellent of catalyst, CH2Cl2 as the solvent, 750 rpm, 1 mmol of Limonene, 3 mmol of
yield was achieved (TON = 165), this type of material has been reported benzaldehyde. The dashed lines were included to guide the eyes.
previously [8] and it was tested for comparison. HPA over alumina only
achieved 3% of conversion with a poor value of TON. Apparently, their HPA/P25 as heterogeneous catalyst. Fig. 5 (a) shows the effect of the
type and strength of the acid sites are not appropriate to drive the excess of benzaldehyde within the conversion and selectivity. Interest­
condensation of limonene with benzaldehyde. In general, it appears that ingly when increases the amount of benzaldehyde, increases slightly
HPA/P-25 catalyst is an active and selective catalyst for this reaction. conversion with time, however, no significant changes are observed. At
Fig. 4 shows that the selectivity to 3-oxabyciclo[3.3.1]nonane and 5 h of reaction, no differences in the conversion were seen when
the conversion of limonene change over reaction time for the HPA decreased the amount of benzaldehyde (from 2.9 to 2.0 mmol), never­
supported catalysts. HPA supported on alumina, silica and titanium di­ theless, a decrease to 92% was achieved with 3.6 mmol of the aldehyde.
oxide nanotubes were the less active materials. On the other hand, it is The same behavior was noted when the amount of aldehyde increased.
interesting to appreciate that HPA/SBA-15 displayed a 70% of conver­ With respect to selectivity, significant changes were observed when
sion and a selectivity of 70% at 1 h of reaction. Nevertheless, when time decreased the mmol of benzaldehyde at 1 h reaction while at 3 and 5 h
increases, a decreasing of 3-oxabyciclo[3.3.1]nonane is observed. This no major differences were seen. When increased the time, selectivity to
fact is attributed to an increasing of isomerization side reaction, which 3-oxabyciclo[3.3.1]nonane decreased slightly when 3.0 mmol of benz­
occur typically in the strong acid sites of the material. Then, considering aldehyde was used. On the contrary, for 2.9 mmol of benzaldehyde
these results and for further experiments, HPA/P25 was chosen as het­ selectivity to the desired target increased while that for 2.0 mmol
erogeneous catalyst. Although HPA/SiO2 was also an active material, remained constant. It seems that the optimum amount of benzaldehyde
this catalyst was not chosen because have been reported previously [29] to obtain a maximum yield to 3-oxabicyclo[3.3.1]nonane is 2.9 mmol
and in fact, would not give new insight into the perspective of our which was the same amount used for the screening catalytic experi­
research. For this reason, this catalyst was only tested for comparison. ments. An excess of benzaldehyde (3.6 mmol) decreases the interaction
of this molecule within the acid sites of the catalyst, and then limonene
3.2.2. Effect of reaction conditions tends to isomerize decreasing the selectivity to the desired oxabicyclic
To further evaluate different reaction conditions for the highly se­ target. Maybe activation of the aldehyde and coverage are not appro­
lective synthesis of 3-oxabyciclo[3.3.1] nonane, different amounts of priate to the further nucleophilic attack of limonene in the C=O bond of
benzaldehyde, the volume of solvent and temperature were evaluated benzaldehyde.
directly in the conversion and selectivity to the desired target using On the other hand, Fig. 5 (b) shows the effect of the solvent volume

6
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

larger volume of solvents (e.g., 10 and 30 mL) which is due to the

production of another product (side isomerization reaction) and to a
competition of limonene within the active sites of the catalyst. As
dichloromethane is a Lewis acid [53], with a larger amount of solvent it
could react with limonene, generating a block of the active sites of the
catalyst. An aspect to highlight is that when no solvent is used the
maximum conversion and selectivity were reached in comparison to
when a different volume of solvent was used. In addition, with a
decrease in the initial concentration (when increases the solvent vol­
ume), the condensation rate will decrease and then the contribution of
isomerization will increase. This hypothesis is in line with the experi­
mental results obtained in this research.
Considering the insight obtained previously, the effect of tempera­
ture was also explored in the catalytic conversion of limonene with
benzaldehyde over HPA/P25. For this, two temperatures were used for
further comparison (298 and 323 K, Fig. 6); it seems that at 323 K, the
maximum conversion is achieved at only 1 h while that at room tem­
perature and at 5 h, conversion of around 80% is achieved. However,
selectivity to the desired 3-oxabicyclico[3.3.1] nonane decreased as
temperature decreased. At room temperature, the major products are
isomers of limonene while at 323 K the major product is 3-oxabicyclico
[3.3.1] nonane. A possible reason for this fact could be associated with
the activation of the benzaldehyde during its interaction with the acid
sites of the catalyst. In general, temperature appears to affect some of the
kinetics for different condensation reactions becoming more active
when temperature increases [54,55]. Further investigations about how
temperature affects the kinetics performance of this reaction are
currently in progress.

3.2.3. Effect of the HPA loading over P25

It has been reported that HPA loading affects drastically the catalytic
efficiency of HPA/P25 and the acid sites (Section 3.1.7). In this research,
this effect was evaluated in the condensation of limonene with benzal­
dehyde for further synthesis of 3-oxabicyclico[3.3.1] nonane. Consid­
ering this aspect, in this research the effect of HPA loading over P25 was
evaluated in a range of 5 up to 30% and the results are indicated in
Fig. 7. With loading up to 10% no conversion and selectivity were ob­
Fig. 5. Conversion of limonene and selectivity to 3-oxabicyclo[3.3.1]nonane
tained while when increases the amount of HPA on P25 (from 15% to
with the variation of benzaldehyde (a) and with different volume of solvent
(b). X = Conversion, S = Selectivity. Reaction conditions: 30 mg of catalyst
30%) both conversion and selectivity are increased. It is seeming that
(HPA/P25), CH2Cl2 as the solvent, 750 rpm, 1 mmol of Limonene, 3 mmol of the optimum amount of HPA which further obtains the desired selec­
benzaldehyde. tivity to 3-oxabyciclo[3.3.1] nonane and conversion of limonene is 20%
wt. No significant differences in conversion nor selectivity were ach­
on the catalytic activity of HPA/P25. For solvent volume up to 5 mL ieved when HPA loading was 25% and 30%, respectively. Differences in
conversion is not affected significantly, while selectivity decreased to the catalytic activity for loadings below of 15% could be associated to
70%. A decrease in both conversion and selectivity is observed for a two factors: (1) a deficiency in the coverage of HPA over P25 and (2) the

Fig. 6. Effect of temperature in the conversion of limonene (a) and selectivity to 3-oxabicyclo [3.3.1]nonane (b) at different times. Reaction conditions: 30 mg of
catalyst, CH2Cl2 as the solvent, 750 rpm, 1 mmol of limonene, 3 mmol of benzaldehyde.

7
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

products on the acid sites of the material as well as to the isomerization

of α-pinene, β-pinene and limonene (which decreased significantly after
catalytic condensation).
In the case of orange essential oil, which contains as major product
limonene (92%), their composition varies in a range of monoterpenes
such as α-pinene (0.5%), sabinene (2%), myrcene (0.5%) and other (5%,
with similar molecular weight). In this research, after condensation with
benzaldehyde, the large amount of limonene is transformed into 3-oxa­
bicyclico[3.3.1] nonane changing significantly the composition of the
essential oil. These results are like that obtained with limonene reactant
grade because of their similar purity. Differences in both cases are
attributed to impurities in the samples. On the other hand, these results
also show that, even, in crude biomass such as essential oils, the reaction
between limonene and benzaldehyde appears to carry out successfully
with the HPA/P25 catalyst.

3.2.5. Laboratory scaling up

Because of the excellent yield and behavior of the HPA/P25 catalyst,
Fig. 7. Effect of HPA loading over P25 in the conversion of limonene and the synthesis of the desired target was evaluated at different laboratory
selectivity to oxabicycle [3.3.0]nonane at 5 h. Reaction conditions: 30 mg of scales and the results are shown in Fig. 9. It is possible to see that
catalyst, CH2Cl2 as the solvent, 750 rpm, 1 mmol of limonene, 3 mmol of regardless of the amount of catalyst, the same conversion and selectivity
benzaldehyde. are achieved showing that this process is highly robust and sustainable
at the conditions evaluated. Considering this fact, further configuration
amount and density of acid sites quantified in the surface of P25. of the process to carry out continuously is shown in Fig. S4 (in the
Although even with 25 and 30% wt a slight increase of conversion was
achieved, the value of TON decreased and then the sustainability of the
process seems to be affected.

3.2.4. Catalytic condensation with essential oils and a mixture of

monoterpenes/sesquiterpenes
To further evaluate the catalytic potential of HPA/P25 in the selec­
tive synthesis of 3-oxabyciclo[3.3.1]nonane, even, in complex mixtures
(see Table S1 in the supporting information), different additional re­
actions were evaluated at the same conditions that for catalytic
condensation of benzaldehyde with limonene. Fig. 8 shows the weight
composition of two essential oils: Rosmarinus and orange essential oils at
the beginning and after the reaction with benzaldehyde and HPA/P25.
In the case of Rosmarinus essential oil, it seems that the initial compo­
sition has different monoterpenes which include α-pinene (19%),
β-pinene (12%), camphene (8%), p-cymene (4%), limonene (16%)
among other (such as terpinolene, terpinene, etc which their sum is
41%). After condensation with benzaldehyde and the HPA/P25 catalyst,
the composition varies significantly, giving 3-oxabicyclico[3.3.1] non­
ane as the major product (up to 25%), with a different composition of Fig. 9. Laboratory scaling up of the synthesis of 3-oxabyciclo[3.3.1]nonane
monoterpene (α-pinene: 13%, p-cymene:7%, limonene 4%, other 47%). from limonene and benzaldehyde with different amounts of HPA/P25. Reac­
An increased in the amount of p-cymene is attributed to dehydration tion conditions: 50 ◦ C, solvent-free, 750 rpm,.

Fig. 8. Initial and final (24 h) compositions of Rosmarinus (a) and orange essential oils (b) with the catalytic condensation with benzaldehyde and HPA/P25 catalyst.
Reaction conditions: 30 mg of catalyst, CH2Cl2 as the solvent, 750 rpm, 50 ◦ C.

8
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

supporting information) using cinnamaldehyde as aldehyde for the left in constant stirring for another 55 min (up to 1 h). Although a slight
condensation. This compound was chosen because of its fragrance increase in the conversion was observed (from 15% to 19%), selectivity
properties. Besides, cinnamaldehyde can be extracted from natural to the desired was constant. In comparison when the catalyst was not
sources, increasing the sustainability of the process. Further analysis for removed from the medium, conversion increased obtaining the same
application in additives is currently performed in our laboratories. catalytic activity that those reported in previous sections. In general, at
1 h of reaction and at the reaction conditions tested in this manuscript,
3.2.6. Stability and reuse of the catalyst no evident lixiviation was achieved. To further verify the reasons for the
The reuse of heterogeneous catalysts is an important parameter to decreasing of both conversion and selectivity in the reaction of limonene
evaluate in order to show the robustness of the material. In this way, and benzaldehyde, characterization by XRD, SEM, physisorption and
HPA/P25 was removed from the suspension and reused again in another TGA was carried out and analyzed elsewhere for the catalyst fresh and
catalytic reaction. For this purpose, the catalyst was separated by using reused after its 3rd activity.
centrifugation and then washed twice with dichloromethane (x 2) and Fig. S5 (in the supporting information) shows the XRD pattern of
with ethyl acetate (x 1). Afterward, the catalyst was dried at room HPA/P25 catalyst after different reuses. As can be seen typical structure
temperature, grinding at the particle size (for avoiding possible internal of titanium dioxide (anatase phase) do not change after reuse of the
mass transfer limitation, Section 3.2.9.1) and reused under the same material, however intensity decreased in some of the planes (2θ =
reaction conditions. Results are presented in Fig. 10 (a). As can be 25◦ ,48◦ ,63◦ and 75◦ ). But, in general, it appears that typical crystalline
observed, both conversion of limonene and selectivity to 3-oxabicyclico structure of the material does not change significantly. Even, when SEM
[3.3.0] nonane remains constant until the second use of the catalyst. micrographs were taken (Fig. S6 in the supporting information) in both
Then, both conversion and selectivity decreased to 70% and 30%, fresh and reused catalyst, it seems that no changes in the typical
respectively. In the fourth use, the conversion of the limonene slightly morphology are observed.
decreased while selectivity decreased considerably (19%). In the third Considering the feature that neither morphology and crystallo­
and fourth reuse, a high distribution of limonene isomers was obtained graphic structure did not change after reuse of the material, thermog­
rather than 3-oxabyciclo[3.3.1]nonane. ravimetric analysis was performed, and the results are showed in Fig. 11.
When the lixiviation test was performed, Fig. 10 (b), the catalyst was As can be seen two different events are well stablished, the first one at
removed (using hot filtration) from the solution for the first 5 min and temperatures lower than 100 ◦ C and another between 200 and 400 ◦ C.
The first one is well stablished to be associated with physiosorbed water
while the second should correspond to organic species on the catalyst.
This difference is well significative in the reused and fresh materials. It
can be observed that for fresh catalyst no thermic events after 200 ◦ C
were seen, while that for the reused materials, clear events for temper­
atures between 200 and 400 ◦ C related with adsorption of organic
compounds is evident. After calculations, TG analysis showed a loss of
mass because of the organic compounds at around 2–3% which could
explain why catalyst loss activity after its second reuse. Another reason
of the loss of activity could be also associated with the loss of acid sites
during the performance of reaction. To further studies we are trying to
study this phenomenon using in-situ and in operando equipment.

3.3. Kinetic assessments

3.3.1. Internal mass transfer limitation

Internal mass transfer limitations can be evaluated by the analysis of
the reaction at different particle sizes and through the calculation of the
Weisz-Prater criterion (CWP), Eq. (1). This parameter is used to check
whether the reaction is limited or not by internal diffusion [56].
Experimentally, different aliquots can be taken from the reaction me­
dium (at the same reaction conditions, particularly at different particle

Fig. 10. Reuse of HPA/P25 on the selective synthesis of 3-oxabyciclo[3.3.1]

nonane (a). Lixiviation test (b). Reaction conditions: 200 mg of catalyst, 1
mmol of limonene, 3 mmol of benzaldehyde, 50 ◦ C, 750 rpm, 24 h. Experiments
were performed by triplicated. Fig. 11. Thermographs of fresh and reused HPA/P25.

9
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

sizes or at the same particle size and varying the initial concentrations of Matlab software. Additionally, some stats parameters such as the coef­
the reactants) and then reaction rate can be obtained. If CWP is <<1 the ficient of determinations (R2 and R2adj), variance s2, root mean square
internal diffusion is negligible. However, a more precise criterion is deviation Rmds and the 95% confidence interval CI95 included in Table 6
CWP≤ 3b, where b is the maximum decrease in concentration gradient discriminate between the different models proposed.
in pores, which is a function of the reaction order. The b values of 0.1, ∑( ′ )2
0.2 and 2 are estimated for the second, first and zero-order reaction, (4)
′
OF = rA0,exp − rA0
respectively [57]. In this research, the reaction order is closer to 1.0 (see
a comparison in Fig. S7 in the supporting information), then the values The suggested kinetics expressions analyzed in this study were
of CWP must be lower than 0.6. stablished from heterogeneous reaction theory based on the Langmuir-
Hinshelwood-Hougen-Watson (LHHW) and Eley-Rideal (ER) mecha­
CWP =
− rA0 ρcat Rcat 2
(1) nisms, along with a power law based pseudo-homogeneous (pH) model.
DAB CA,cat Pseudo-Homogeneous Irreversible Model (pH)
A homogeneous-like model was tested using the experimental data,
Where rA0 is the limonene (A) initial reaction rate, ρcat is the solid
for the purpose of evaluating a possible simplification of the heteroge­
catalyst density, Rcat is the average radius of the catalyst particle (taken
neous kinetics adapted to the homogeneous model. From the results
as an average of the particle size given for the mesh), DAB is the effective
shown Table 4 can be concluded that it appears that the reaction rate has
diffusivity of limonene into benzaldehyde (B) at 323 K, and CA,cat is the
a higher dependence on the limonene concentration than the benzal­
concentration of limonene at the catalyst surface.
dehyde due to the values obtained for the parameters α and β. A good
Diffusion coefficient can be calculated using the Wilke-Chang Eq. (2):
adjustment (R2 = 0.9897) was found with values of almost 0.4 and 0.05
1.173x10− 16 (∅MB )2 T (φσ ) for the reaction orders of limonene and benzaldehyde, respectively.
1

DAB = (2) Then, kinetics is government mainly for the concentration of limonene
μS VS 0.6 τ
in the solution instead of the concentration of benzaldehyde. Consid­
Where µS is the viscosity of the solution (0.001321 kg s− 1 m− 1), VS is ering these aspects as well as that the reaction is experimentally cata­
the molar volume of the solvent at the normal boiling point (or at least of lyzed by a heterogeneous system, the ER and LHHW models were further
the fluid where is dissolving the limiting reactant; 0.01 m3 kmol− 1), ∅ is taken into consideration.
the association parameter of benzaldehyde (2.26), MB is the molecular Heterogeneous Eley-Rideal (ER) Model
weight of benzaldehyde (112.16 kg kmol− 1) and T is the absolute tem­ A heterogeneous kinetic model was tested based on the Eley-Rideal
perature (323 K), φ is the catalyst porosity (0.01), σ is the constriction theory. In this type of reaction mechanism requires that only one of
factor (0.8) and τ is the tortuosity (3) [58]. Typical data of viscosity, the two reactants be adsorbed on the surface of the catalyst. Here we
molar volume and association parameters were obtained using Aspen fitted the experimental data of the initial reaction rate of limonene
Plus software. Table 5 shows the values of the Weisz-Prater criterion against the kinetic model obtained considering:
(CWP) obtained after evaluating different particles sizes, and it is
seeming that no internal mass transfer is presented at the particle sizes • Only adsorption of limonene on the surface of the catalyst (A + S =
tested in this research (the limit for avoiding internal mass transfer using A.S).
the Weisz-Prater criterion (CWP) is 0.6). Then, for further kinetic • The reaction of the adsorbed specie A.S with benzaldehyde assumed
studies, the catalyst was ground between 45 and 25 µm and tested in the as the limiting step to obtain (adsorbed or not) the main product 3-
reaction. oxabyciclo[3.3.1] nonane (A.S + B → C + S or A.S + B → C.S).
• The kinetic may not depend, i.e., CB almost constant, (ER-I) or may
3.3.2. Kinetic modeling depend (ER-II) on the concentration of benzaldehyde according with
The kinetic analysis of the condensation of limonene with benzal­ the low β parameter obtained from pH model.
dehyde, (Eq. (3)), was done with experimental data at several reaction
conditions collected from the liquid-phase solvent-free batch reaction; C The kinetic expression obtained from previous considerations and
is the main target of the reaction -3-oxabyciclo[3.3.1]nonane- and D, are evaluated at the initial conditions due to the implementation of the
the by-products obtained by the isomerization reaction of limonene over initial reaction rate method, lead to the kinetic models ER-I and ER-II
the acid sites of heteropoly acid material. The reaction rates calculated shown in Table 4. The best adjustment (R2 = 0.9901) of the experi­
by the initial reaction rate method in the different tests are compared mental data was found from the kinetic model ER-I with a similar value
with proposed kinetic models. of the kinetic constant k1 obtained for both pH and ER-I models and the
A + B→C + D (3) highest adjusted coefficient of determination (R2adj = 0.9876) between
all models. This model supports the idea that the kinetic of this reaction
To establish accurate kinetic models, all the data was collected under is government mainly by the concentration of limonene in the solution
the absence of internal mass transfer limitations. The absence of external instead of the concentration of benzaldehyde which is considered almost
mass transfer limitations is guaranteed at high stirring rate of around constant, adding to this narrative the suggestion of a contribution due to
750 rpm according with our preliminary studies involving similar the adsorption of limonene on the surface of the material. The evalua­
monoterpenes. The kinetic parameters obtained through an optimiza­ tion of the model considering the effect of the concentration of benzal­
tion of the sum of squared errors in the objective function OF in Eq. (4) dehyde was analyzed by the ER-II model which gave the lowest adjusted
fitting calculations of the experimental initial reaction rate r’A0,exp with coefficient of determination (R2adj = 0.8666) ant the highest root means
some kinetics expressions r’A0 are indicated in Table 6. The minimiza­ square deviation (Rmsd = 4.1573 × 10− 11).
tion of the OF was carried out incorporating the Genetic Algorithm in Langmuir-Hinshelwood-Hougen Watson (LHHW) Model
On the other hand, another typical heterogeneous kinetic model
based on the Langmuir-Hinshelwood-Hougen-Watson theory was
Table 5
Weisz-Prater criterion for different particles size tested in this research. considered for fitting the reaction data. This model requires that both
reactants be adsorbed on the surface of the catalyst. Here we fitted again
Range of particle size Average particle size Weisz-Prater Criterion
the experimental data of the initial reaction rate of limonene against the
(µm) (µm) (CWP)
75 < dp < 100 88 0.022 mathematical model obtained considering:
45 < dp < 75 60 0.009
45 < dp < 25 35 0.002

10
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

Table 6
Mathematical values of each variable in the kinetic modeling.
1
Model Rate law (mol g− min− 1) Constant Value ± CI95 OFx106 R2 (R2adj) Rmsd x1013 s2x105
5
*Value **CI95 x10−
pH − rA0 = k1 CαA0 CβB0
′
k1 0.11699 0.233 1.526 0.9897(0.9828) 3.8818 2.9596
α 0.39587 1.391
β 0.05478 1.460
ER-1 k1 CA0 k1 0.12073 0.1084 1.4653 0.9901(0.9876) 3.5786 2.95958
− rA0 =
′

1 + K1 CA0 K1 0.20503 0.2428

ER-1I k1 CA0 CB0 k1 0.01614 0.0865 1.4653 0.8933(0.8666) 415.76 2.95958

− rA0 =
′

1 + K1 CA0 K1 1.87485 10.34

LHHW-I k1 CA0 k1 0.38546 1391 1.554 0.9895(0.9825) 4.0272 2.95958

− rA0 =
′

(1 + K1 CA0 + K2 )2 K1 0.10695 386

K2 0.91779 6920
LHHW-II k1 CA0 CB0 k1 0.07335 0.0023 1.2618 0.9915(0.9858) 2.6537 2.95958
− rA0 =
′

(1 + K1 CA0 + K2 CB0 )2 K1 0.22131 0.0630

K2 0.13723 0.0169

*Simulation conditions by the GA (population size: 300, generations: 10,000, OF and parameters tolerance = 1 × 10− 8). ** CI95 calculated by coupling the GA to a
nonlinear least-squares regression function (nlinfit) using optimal GA values outputs as inputs (tolerance = 1 × 10− 8). k1 = pseudo-kinetic constant, K1, K2 =
Adsorption constants, α and β reaction order.

• A possible both adsorption of limonene on the surface of the catalyst temperature can affect the yields towards the formation of the desired
(A + S = A.S) and the adsorption of benzaldehyde as well (B + S = B. target. It was observed that a decrease until room temperature increases
S). the formation of isomers byproducts giving a change of the distribution
• The reaction of the adsorbed species A.S and B.S is assumed as the products and decreasing the synthesis of the desired 3-oxabyciclo[3.3.1]
limiting step to obtain (adsorbed or not) the main product 3-oxaby­ nonane. In addition, the effect of the reaction conditions was evaluated
ciclo[3.3.1] nonane (A.S + B.S → C + 2S or A.S + B.S → C.S + S). to establish whether solvent volume and time affect the performance of
• The kinetic may not depend, i.e., CB almost constant, (LHHW-I) or the reaction in course. It appears that an increase in the volume of the
may depend (LHHW-II) on the concentration of benzaldehyde as in solvent decreases the amount of the desired target, giving high selec­
the analysis for ER model. tivity to isomers of limonene.
The effect of HPA loading on TiO2-P25 was also explored, concluding
The parameters obtained from the kinetic expressions LHHW-I and that 20 wt% of HPA is enough to obtain the desired conversion and
LHHW-II evaluated at the initial conditions are shown in Table 6. selectivity to 3-oxabyciclo[3.3.1]nonane. With the aim to increase the
Although the adjusted coefficient of determination of LHHW models sustainability of the reaction, essential oils (Rosmarinus and orange oils),
were as good as the obtained from the ER-I, considering in this case CB extracted directly from biomass with limonene as one of the compo­
almost constant for LHHW-I model, results in an uncertain estimation of nents, were also tested obtaining different compositions where 3-oxaby­
the 95% confidence interval CI95 for all kinetic parameters and the ciclo[3.3.1]nonane is the major component. Evidently, this showed that
highest OF value. The lowest OF (1.2618 × 10− 6) and root mean square our catalyst (HPA/P-25) is selective, even, in samples of biomass.
deviation (Rmsd = 2.6537 × 10− 13) value were obtained from the LHHW- Kinetic analysis suggests a typical LHHW mechanism, within the
II model which considers the contribution effect of the concentration of adsorption as a critical stage. Finally, the laboratory scaled-up of the
benzaldehyde in the kinetic expression. This new perspective of the reaction (up to 10 g) can be performed obtaining the same yields that
reaction from LHHW-II model also seems in agreement with the when only 10 mg of sample was used. HPA/P25 is a heterogeneous
assumption of adsorption of limonene on the surface of the catalyst, in catalyst robust since can be used several times with a slight decrease in
this case higher than a possible adsorption of benzaldehyde because of the catalytic activity. This loss of activity was associated with fewer acid
the lower value of K2 obtained compared with K1. sites and with the adsorption of organic species on the surface of the
material.
4. Conclusions
CRediT authorship contribution statement
Several heterogeneous catalysts of heteropolyacids supported on
SBA-15, SiO2, TiO2 (Degussa P25, Al2O3, NP-TiO2, were tested in the Julián E. Sánchez-Velandia: Conceptualization, Data curation,
catalytic condensation of limonene with benzaldehyde. The synthesized Formal analysis, Investigation, Methodology, Validation, Visualization,
materials were widely characterized giving insights into the way as Writing – original draft, Writing – review & editing. Herme G. Baldoví:
heteropolyacids with a typical Keggin-like anion can be modified by the Formal analysis, Methodology, Resources, Writing – review & editing.
preparation methodology. XRD and physisorption evidenced a decrease A. Yu Sidorenko: Formal analysis, Methodology, Writing – review &
of the surface area after HPA impregnation and besides that, this can be editing. Jaime A. Becerra: Conceptualization, Data curation, Formal
well dispersed. Similarly, 31P NMR showed that apparently structure can analysis, Investigation, Methodology, Validation, Visualization, Writing
be whether like α- [PW12O4O]3- or such as well dispersed decomposed – review & editing. Fernando Martínez O: Formal analysis, Funding
HPA compound. Acidity by TPD- ammonia showed that almost all ma­ acquisition, Project administration, Resources, Supervision, Writing –
terials presented high content of acid sites and that evidently when in­ review & editing.
creases the amount of HPA in TiO2-P25, acidity increased until it was
constant. Good to excellent yields towards the formation of 3-oxabyciclo
Declaration of Competing Interest
[3.3.1]nonane at benign reaction conditions (50 ◦ C, free solvent, 1 h)
were achieved. The most promissory catalyst was HPA supported on
The authors declare the following financial interests/personal re­
titanium dioxide (Degussa P25) because of its textural and acidic
lationships which may be considered as potential competing interests:
properties (up to 187 µmol. gcat− 1). The reaction can be performed
Julian E Sanchez Velandia reports financial support was provided by
without solvent and under relatively soft conditions. Changes in the
Colombia Ministry of Science Technology and Innovation. Julian E.

11
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

Sanchez Velandia reports financial support was provided by Colombian cyclization reaction, Synlett 24 (2013) 1137–1141, https://doi.org/10.1055/s-
0032-1316915.
Ministry of Commerce Industry and Tourism. Julian E Sanchez Velandia
[15] I.V Il, N.S. Dyrkheeva, A.L. Zakharenko, A.Y. Sidorenko, N.S. Li-zhulanov, D.
reports financial support was provided by Industrial University of V Korchagina, R. Chand, D.M. Ayine-tora, A.A. Chepanova, O.D. Zakharova, E.
Santander. Jaime A Becerra Chala reports financial support was pro­ S. Ilina, J. Reynisson, A.A. Malakhova, S.P. Medvedev, S.M. Zakian, K.P. Volcho, N.
vided by Colombia Ministry of Science Technology and Innovation. F. Salakhutdinov, O.I. Lavrik, Design, synthesis, and biological investigation of
novel classes of 3-carene-derived potent inhibitors of TDP1, Molecules 25 (15)
(2020) 1–22, https://doi.org/10.3390/molecules25153496.
Data availability [16] A.Y. Sidorenko, Y.M. Kurban, A.V. Kravtsova, I.V. Il’ina, N.S. Li-Zhulanov, D.
V. Korchagina, J.E. Sánchez-Velandia, A. Aho, K.P. Volcho, N.F. Salakhutdinov, D.
Y. Murzin, V.E. Agabekov, Clays catalyzed cascade Prins and Prins-Friedel-Crafts
Data will be made available on request. reactions for synthesis of terpenoid-derived polycyclic compounds, Appl. Catal. A
Gen. (2022) 629, https://doi.org/10.1016/j.apcata.2021.118395.
[17] O.S. Patrusheva, K.P. Volcho, N.F. Salakhutdinov, Synthesis of oxygen-containing
heterocyclic compounds based on monoterpenoids, Russ. Chem. Rev. 87 (2018)
Acknowledgments 771–796, https://doi.org/10.1070/rcr4810.
[18] R.F. Cotta, R.A. Martins, K.A. da Silva Rocha, E.F. Kozhevnikova, I.V. Kozhevnikov,
E.V. Gusevskaya, Coupling of phenylacetaldehyde and styrene oxide with
This research was supported by Ministerio de Ciencia, Tecnología e
biorenewable alkenes in eco-friendly solvents, Catal. Today 381 (2021) 254–260,
Innovación, Ministerio de Educación Nacional; Ministerio de Industria, https://doi.org/10.1016/j.cattod.2020.05.068.
Comercio y Turismo, and ICETEX of Colombia through Programme [19] L.C.A. Barbosa, L.B. Nogueira, C.R.Á. Maltha, R.R. Teixeira, A.A. Silva, Synthesis
Ecosistema Científico-Colombia Científica from Fondo Francisco José de and phytogrowth properties of oxabicyclic analogues related to helminthosporin,
Molecules 14 (2009) 160–173, https://doi.org/10.3390/molecules14010160.
Caldas, Colombia; Grant RC-FP44842-212-2018. J.S.V is grateful to [20] S.V. Kumar, A. Yen, M. Lautens, P.J. Guiry, Catalytic asymmetric transformations
Universidad Industrial de Santander, Colombia for his postdoctoral of oxa- and azabicyclic alkenes, Chem. Soc. Rev. 50 (2021) 3013–3093.
fellowship: “Apoyo Estancias Postdoctorales”, 2020. J.A.B acknowl­ [21] A.Y. Sidorenko, A.V Kravtsova, P. Mäki-arvela, A. Aho, T. Sandberg, I.V Il, N.S. Li-
zhulanov, D.V Korchagina, K.P. Volcho, N.F. Salakhutdinov, D.Y. Murzin, Synthesis
edges Minciencias (before Colciencias), call 694, 2014. of isobenzofuran derivatives from renewable 2-carene over halloysite nanotubes,
Mol. Catal. 490 (2020), 110974, https://doi.org/10.1016/j.mcat.2020.110974.
[22] Y. Liu, Y. Wu, M. Su, W. Liu, X. Li, F. Liu, Developing Brønsted–Lewis acids
Supplementary materials
bifunctionalized ionic liquids based heteropolyacid hybrid as high-efficient solid
acids in esterification and biomass conversion, J. Ind. Eng. Chem. 92 (2020)
Supplementary material associated with this article can be found, in 200–209, https://doi.org/10.1016/j.jiec.2020.09.005.
the online version, at doi:10.1016/j.mcat.2022.112511. [23] M.B. Colombo Migliorero, V. Palermo, G.P. Romanelli, P.G. Vázquez, New niobium
heteropolyacid included in a silica/alumina matrix: application in selective
sulfoxidation, Catal. Today (2020), https://doi.org/10.1016/j.cattod.2020.10.034.
References [24] H. Zhang, J. Liu, C. Liu, T. Wang, W. Zhu, High dispersion of heteropolyacid
nanoparticles on hydrothermally Cs-modified three-dimensionally ordered
macroporous SiO2 with excellent selectivity in methacrolein oxidation, Chin. J.
[1] K.N. Mahmud, Z.A. Zakaria, Pyrolytic products from oil palm biomass and its
Chem. Eng. 28 (2020) 2785–2791, https://doi.org/10.1016/j.cjche.2020.07.017.
potential applications, valroization of Agro-industrial residues- volume II: non-
[25] L. Pizzio, P. Vázquez, C. Cáceres, M. Blanco, Tungstophosphoric and
biological approaches. Applied Environmentl Science and Engineering for a
molybdophosphoric acids supported on zirconia as esterification catalysts, Catal.
Sustainable Future, Springer, 2020.
Lett. 77 (2001) 233–239, https://doi.org/10.1023/A:1013218307792.
[2] M. Golets, S. Ajaikumar, J.P. Mikkola, Catalytic upgrading of extractives to
[26] E. Matijević. Fine Particles in Medicine and Pharmacy, Springer, New York, 2011,
chemicals: monoterpenes to “eXICALS, Chem. Rev. 115 (2015) 3141–3169,
ISBN 978-1-4614-0379-1, https://doi.org/10.1007/978-1-4614-0379-1.
https://doi.org/10.1021/cr500407m.
[27] V. Negro, G. Mancini, B. Ruggeri, D. Fino, Citrus waste as feedstock for bio-based
[3] M.J. Climent, A. Corma, S. Iborra, Heterogeneous catalysts for the one-pot
products recovery: review on limonene case study and energy valorization,
synthesis of chemicals and fine chemicals, Chem. Rev. 111 (2011) 1072–1133,
Bioresour. Technol. 214 (2016) 806–815, https://doi.org/10.1016/j.
https://doi.org/10.1021/cr1002084.
biortech.2016.05.006.
[4] N.F. Salakhutdinov, K.P. Volcho, O.I. Yarovaya, Monoterpenes as a renewable
[28] J.D. Tibbetts, S.D. Bull, p-menthadienes as biorenewable feedstocks for a
source of biologically active compounds, Pure Appl. Chem. 89 (2017) 1105–1117,
monoterpene-based biorefinery, Adv. Sustain. Syst. 5 (2021) 2000292, https://doi.
https://doi.org/10.1515/pac-2017-0109.
org/10.1002/adsu.202000292 (1-13).
[5] J.E. Sánchez-Velandia, J.F. Gelves, M.A. Márquez, L. Dorkis, A.L. Villa, Catalytic
[29] R.F. Cotta, K.A. da Silva Rocha, E.F. Kozhevnikova, I.V. Kozhevnikov, E.
isomerization of α-pinene epoxide over a natural zeolite, Catal. Lett. (2020),
V. Gusevskaya, Coupling of monoterpenic alkenes and alcohols with benzaldehyde
https://doi.org/10.1007/s10562-020-03225-9.
catalyzed by silica-supported tungstophosphoric heteropoly acid, Catal. Today 289
[6] N.D. Shcherban, R.Y. Barakov, P. Mäki-Arvela, S.A. Sergiienko, I. Bezverkhyy,
(2017) 14–19, https://doi.org/10.1016/j.cattod.2016.07.021.
K. Eränen, D.Y. Murzin, Isomerization of α-pinene oxide over ZSM-5 based micro-
[30] T. Pinto, K. Szeto, N. Oueslati, N. Essayem, V. Dufaud, F. Lefebvre, Comparison of
mesoporous materials, Appl. Catal. A Gen. 560 (2018) 236–247, https://doi.org/
the acidity of heteropolyacids encapsulated in or impregnated on SBA-15, Oil Gas
10.1016/j.apcata.2018.05.007.
Sci. Technol. (2016) 71, https://doi.org/10.2516/ogst/2016005.
[7] Y. Noma, Y. Asakawa, Biotransformation of monoterpenoids. Comprehensive
[31] G. Marcì, E. García-López, M. Bellardita, F. Parisi, C. Colbeau-Justin, S. Sorgues, L.
Natural Products II Chemistry and Biology 3, Elsevier, 2010, pp. 669–801, https://
F. Liotta, L. Palmisano, Keggin heteropolyacid H3PW12O40 supported on different
doi.org/10.1016/b978-008045382-8.00742-5.
oxides for catalytic and catalytic photo-assisted propene hydration, Phys. Chem.
[8] R.F. Cotta, K.A. da Silva Rocha, E.F. Kozhevnikova, I.V. Kozhevnikov, E.
Chem. Phys. 15 (2013) 13329–13342, https://doi.org/10.1039/c3cp51142a.
V. Gusevskaya, Heteropoly acid catalysts in upgrading of biorenewables:
[32] Q. Jin, F. Qu, J. Jiang, Y. Dong, W. Guo, H. Lin, A pH-sensitive controlled dual-drug
cycloaddition of aldehydes to monoterpenes in green solvents, Appl. Catal. B
release from meso-macroporous silica/multilayer-polyelectrolytes coated SBA-15
Environ. 217 (2017) 92–99, https://doi.org/10.1016/j.apcatb.2017.05.055.
composites, J. Sol Gel Sci. Technol. 66 (2013) 466–471, https://doi.org/10.1007/
[9] H. Martinez Q, Á.A. Amaya, E.A. Paez-mozo, F. Martinez O, S. Valange, Photo-
s10971-013-3033-6.
assisted O-atom transfer to monoterpenes with molecular oxygen and a dioxoMo
[33] J.E. Sánchez-Velandia, E.A. Páez-Mozo, F. Martínez-Ortega, A kinetic study of the
(VI) complex immobilized on TiO2 nanotubes, Catal. Today (2020) 1–17, https://
photoinduced oxo-transfer using a Mo complex anchored to TiO2, Rev. Fac. Ing.
doi.org/10.1016/j.cattod.2020.07.053.
Univ. Antioq. (2020) 83–93, https://doi.org/10.17533/udea.redin.20200477.
[10] C. Member, R.A.S.A.V. Kutchin, I.Y. Chukicheva, I.V Fedorova, O.A. Shumova,
[34] K. Djebaili, Z. Mekhalif, A. Boumaza, A. Djelloul, XPS, FTIR,EDX,XRD analysis of
Alkylation of aluminum phenolate with unsaturated monoterpenes, 437 (2011)
Al2O3 scales grown on Pm2000 alloy, J. Spectrosc. (2015) 1–16, https://doi.org/
776–777. 10.1134/S0012500811040124.
10.1155/2015/868109, 2013.
[11] M.J. Van Der Werf, K.M. Overkamp, J.A.M. De Bont, Limonene-1,2-epoxide
[35] Z. Wang, L. Liu, Mesoporous silica supported phosphotungstic acid catalyst for
hydrolase from Rhodococcus erythropolis DCL14 belongs to a novel class of
glycerol dehydration to acrolein, Catal. Today 376 (2021) 55–64, https://doi.org/
epoxide hydrolases, J. Bacteriol. 180 (1998) 5052–5057, https://doi.org/10.1128/
10.1016/j.cattod.2020.08.007.
jb.180.19.5052-5057.1998.
[36] J.E. Sánchez-Velandia, E. Pájaro, A.L. Villa, F. Martínez-O, Selective synthesis of
[12] M.B. Kolicheski, L.C. Cocco, D.A. Mitchell, M. Kaminski, Synthesis of myrcene by
camphene from isomerization of α- and β-pinene over heterogeneous catalysts,
pyrolysis of β-pinene: analysis of decomposition reactions, J. Anal. Appl. Pyrolysis
Microporous Mesoporous Mater. 324 (2021), https://doi.org/10.1016/j.
80 (2007) 92–100, https://doi.org/10.1016/j.jaap.2007.01.005.
micromeso.2021.111273.
[13] A.Y. Sidorenko, Y.M. Kurban, I.V. Il’ina, N.S. Li-Zhulanov, D.V. Korchagina, O.
[37] J. Wang, J. Yu, X. Zhu, X.Z. Kong, Preparation of hollow TiO2 nanoparticles
V. Ardashov, J. Wärnå, K.P. Volcho, N.F. Salakhutdinov, D.Y. Murzin, V.
through TiO2 deposition on polystyrene latex particles and characterizations of
E. Agabekov, Catalytic synthesis of terpenoid-derived hexahydro-2H-chromenes
their structure and photocatalytic activity, Nanoscale Res. Lett. 7 (2012) 1–8,
with analgesic activity over halloysite nanotubes, Appl. Catal. A Gen. (2021) 618,
https://doi.org/10.1186/1556-276X-7-646.
https://doi.org/10.1016/j.apcata.2021.118144.
[38] N. Xu, Z. Liu, Y. Dong, T. Hong, L. Dang, W. Li, Controllable synthesis of
[14] G. Baishya, B. Sarmah, N. Hazarika, An environmentally benign synthesis of
mesoporous alumina with large surface area for high and fast fluoride removal,
octahydro-2 H-chromen-4-ols via modified montmorillonite K10 catalyzed Prins

12
J.E. Sánchez-Velandia et al. Molecular Catalysis 528 (2022) 112511

ceramics international 42 (2016) 15253–15260, https://doi.org/10.1016/j. [49] R.I. Maksimovskaya, G.M. Maksimov, 31P NMR studies of hydrolytic conversions
ceramint.2016.06.164. of 12-tungstophosphoric heteropolyacid, Coord. Chem. Rev. 385 (2019) 81–99,
[39] V.S. Marakatti, D. Mumbaraddi, G.V. Shanbhag, A.B. Halgeri, S.P. Maradur, https://doi.org/10.1016/j.ccr.2019.01.014.
Molybdenum oxide/γ-alumina: an efficient solid acid catalyst for the synthesis of [50] A.Y. Sidorenko, Y.M. Kurban, A.V. Kravtsova, I.V. Il’ina, N.S. Li-Zhulanov, D.
nopol by Prins reaction, RSC Adv. 5 (2015) 93452–93462, https://doi.org/ V. Korchagina, J.E. Sánchez-Velandia, A. Aho, K.P. Volcho, N.F. Salakhutdinov, D.
10.1039/c5ra12106j. Y. Murzin, V.E. Agabekov, Clays catalyzed cascade Prins and Prins-Friedel-Crafts
[40] H. Zhang, J. Liu, C. Liu, T. Wang, W. Zhu, High dispersion of heteropolyacid reactions for synthesis of terpenoid-derived polycyclic compounds, Appl. Catal. A
nanoparticles on hydrothermally Cs-modified three-dimensionally ordered Gen. (2021), 118395, https://doi.org/10.1016/j.apcata.2021.118395.
macroporous SiO2 with excellent selectivity in methacrolein oxidation, Chin. J. [51] A.Y. Sidorenko, A.V Kravtsova, P. Mäki-arvela, A. Aho, T. Sandberg, I.V Il, N.S. Li-
Chem. Eng. 28 (2020) 2785–2791, https://doi.org/10.1016/j.cjche.2020.07.017. zhulanov, D.V Korchagina, K.P. Volcho, N.F. Salakhutdinov, D.Y. Murzin, Synthesis
[41] A. Micek-Ilnicka, N. Ogrodowicz, U. Filek, A. Kusior, The role of TiO2 polymorphs of isobenzofuran derivatives from renewable 2-carene over halloysite nanotubes,
as support for the Keggin-type tungstophosphoric heteropolyacid as catalysts for n- Mol. Catal. 490 (2020), 110974, https://doi.org/10.1016/j.mcat.2020.110974.
butanol dehydration, Catal. Today 380 (2021) 84–92, https://doi.org/10.1016/j. [52] S.R. Matkovic, L.E. Briand, M.Á. Bañares, Investigation of the thermal stability of
cattod.2021.04.021. phosphotungstic Wells-Dawson heteropoly-acid through in situ Raman
[42] A. Borodziński, M. Bonarowska, Relation between crystallite size and dispersion on spectroscopy, Mater. Res. Bull. 46 (2011) 1946–1948, https://doi.org/10.1016/j.
supported metal catalysts, Langmuir 13 (1997) 5613–5620, https://doi.org/ materresbull.2011.07.015.
10.1021/la962103u. [53] I. Pietro Oliveri, G. MacCarrone, S. Di Bella, A lewis basicity scale in
[43] A. Popa, V. Sasca, D. Bajuk-Bogdanović, I. Holclajtner-Antunović, Acidic nickel dichloromethane for amines and common nonprotogenic solvents using a zinc(II)
salts of Keggin type heteropolyacids supported on SBA-15 mesoporous silica, Schiff-base complex as reference lewis acid, J. Org. Chem. 76 (2011) 8879–8884,
J. Porous Mater. 23 (2016) 211–223, https://doi.org/10.1007/s10934-015-0072- https://doi.org/10.1021/jo2016218.
0. [54] S. Bhasker-Ranganath, M.S. Rahman, C. Zhao, F. Calaza, Z. Wu, Y. Xu, Elucidating
[44] P. Apopei, C. Catrinescu, C. Teodosiu, S. Royer, Mixed-phase TiO2 photocatalysts: the mechanism of ambient-temperature aldol condensation of acetaldehyde on
crystalline phase isolation and reconstruction, characterization and photocatalytic ceria, ACS Catal. 11 (2021) 8621–8634, https://doi.org/10.1021/
activity in the oxidation of 4-chlorophenol from aqueous effluents, Appl. Catal. B acscatal.1c01216.
Environ. 160–161 (2014) 374–382, https://doi.org/10.1016/j. [55] K. Dhahagani, J. Rajesh, R. Kannan, G. Rajagopal, Asymmetric Henry reaction of
apcatb.2014.05.030. aldehydes catalyzed by recyclable an MCM-41 supported copper(II) salen complex,
[45] B. Viswanadham, P. Jhansi, K.V.R. Chary, H.B. Friedrich, S. Singh, Efficient solvent Tetrahedron Asymmetry 22 (2011) 857–865, https://doi.org/10.1016/j.
free knoevenagel condensation over vanadium containing heteropolyacid catalysts, tetasy.2011.04.020.
Catal. Lett. 146 (2016) 364–372, https://doi.org/10.1007/s10562-015-1646-9. [56] S.A. Zavrazhnov, A.L. Esipovich, S.Y. Zlobin, A.S. Belousov, A.V. Vorotyntsev,
[46] Y. Liu, B. Cheng, K.K. Wang, G.P. Ling, J. Cai, C.L. Song, G.R. Han, Study of Raman Mechanism analysis and kinetic modelling of Cu NPs catalysed glycerol conversion
spectra for γ-Al2O3 models by using first-principles method, Solid State Commun. into lactic acid, Catalysts 9 (2019) 1–6, https://doi.org/10.3390/catal9030231.
178 (2014) 16–22, https://doi.org/10.1016/j.ssc.2013.09.030. [57] D. Casas-Orozco, E. Alarcón, A.L. Villa, Kinetic study of the nopol synthesis by the
[47] B. Reif, S.E. Ashbrook, L. Emsley, M. Hong, Solid-state NMR spectroscopy, Nat. Prins reaction over tin impregnated MCM-41 catalyst with ethyl acetate as solvent,
Rev. Methods Prim. 1 (2021), https://doi.org/10.1038/s43586-020-00002-1. Fuel 149 (2015) 130–137, https://doi.org/10.1016/j.fuel.2014.08.067.
[48] I.V. Kozhevnikov, Catalysis by heteropoly acids and multicomponent [58] H. Fogler. Elements of Chemical Reaction Engineering, Prentice Hall, 2010.
polyoxometalates in liquid-phase reactions, Chem. Rev. 98 (1998) 171–198,
https://doi.org/10.1021/cr960400y.

13