Molybdenum oxide/γ-alumina: an efficient solid acid catalyst for the synthesis of nopol by Prins reaction

Abstract

Prins condensation of β-pinene with paraformaldehyde was carried out over MoO_x/γ-Al₂O₃ catalyst in liquid phase. MoO_x/γ-Al₂O₃ catalysts with different loadings were synthesized by impregnation method and characterized by XRD, N₂ sorption, NH₃-TPD, UV-DRS, TEM and SEM. The surface area, TEM and UV-DRS measurements reveal complete coverage of alumina support by MoO_x was achieved between 9 and 11 wt% loading in which the molybdenum is present as octahedral and tetrahedral species in the form of a monolayer domain and small clusters of MoO_x. The effect of MoO_x loading and calcination temperature of the catalysts on total acidity and catalytic activity in the Prins reaction were correlated. The total surface acid density of MoO_x/γ-Al₂O₃ increased with an increase in MoO_x loading until 11 wt% and further loading decreased the acidity. Among the different supports, ZrO₂, TiO₂, SiO₂ and Nb₂O₅, γ-Al₂O₃ exhibited the best performance. The influence of solvent on the Prins reaction in terms of dielectric constant, acceptor and donor number was investigated. Solvents with both AN and DN in the range of 10 to 20 were suitable to facilitate the reaction. The effect of reaction temperature, catalyst amount, reactant mole ratio and reusability of the catalyst for the Prins reaction were investigated. The 11 wt% MoO_x loaded on γ-Al₂O₃ showed the best performance with 96% β-pinene conversion with 86% nopol selectivity.

---

1. Introduction

The reaction of alkene with an aldehyde to produce 1,3-dioxanes, 1,3-diols and unsaturated alcohols is known as the Prins reaction. Nopol is an unsaturated, bicyclic, monoterpinic primary alcohol obtained from naturally occurring β-pinene and formaldehyde. Nopol has wide application in the perfumery industry as a fragrance, in detergent industries as n aroma and as a pesticide in agrochemical industries.¹ The Prins reaction is an acid catalyzed reaction and proceeds with electrophillic addition of an activated aldehyde to an olefin. Traditionally, nopol was produced from Prins condensation of β-pinene and paraformaldehyde under anhydrous conditions or using the homogenous catalyst ZnCl₂.^2,3 The disadvantages associated with these process are the use of unfavorable conditions like high temperature and pressure, with low yield and formation of unwanted side products. To overcome these problems, different heterogeneous catalysts have been designed and studied for the Prins reaction. The most widely explored catalysts for Prins reactions are Sn-beta, Sn-MCM-41, Sn-SBA-15, Zr-SBA-15 and Zn-MCM-41, where metal ions are introduced by isomorphous substitution, wet impregnation and chemical vapor deposition methods.^4–9 A good to moderate yield for nopol has been reported for the aforementioned catalysts. However, the use of these catalysts may be limited due to the tedious synthesis procedures, leaching of metal ions and use of toxic metal tin. In accordance with the insight of green chemistry, some of the catalysts like sulfated zirconia, mesoporous iron phosphate, Fe–Zn double cyanide complex, metal organic framework and MWW type zeolite have been applied for the Prins reaction to produce nopol.^10–13 Most of the catalysts studied, indicate the requirement of either Lewis acid sites or weak to moderate Brönsted acid sites for the selective synthesis of nopol.^4–13

Supported oxides of transition metals are widely used as catalysts for various organic transformations.¹⁴ Due to the dispersion of one metal oxide phase over the other; these are known as “monolayer type” of catalysts. The molybdenum oxide is impregnated on to inorganic supports like TiO₂, SiO₂, Al₂O₃ and ZrO₂ to improve catalytic activity, life and mechanical strength of catalyst. The MoO_x supported on silica has been studied as an acid catalyst for the Prins reaction for the synthesis 1,3-dioxolanes and 1,3 dioxanes.¹⁵ MoO₃/TiO₂ supported catalyst has been widely studied as desulfurization catalytic process; MoO₃/ZrO₂ supported catalyst was studied for several reactions such as alkane isomerization, synthesis of furfural etc.^14–20 Among the different support, γ-alumina has attracted much attention because of its large surface area, high stability and porosity. Molybdenum oxide supported on γ-alumina has been studied for the catalytic processes such as epoxidation of allyl alcohol, oxidative dehydrogenation of propane, hydrodesulphurization of thiophene.^21–26 However, there are only few reports discussing the relation between the acidic properties of MoO_x/γ-Al₂O₃ with catalytic activity of organic reactions.¹⁴ To further explore on the acidic properties, MoO_x/γ-Al₂O₃ catalyst has been chosen to study for the Prins reaction.

Recently, we reported Sn(OH)Cl catalyst containing weak Brönsted acidic sites for the Prins reaction of β-pinene with paraformaldehyde to produce nopol.¹² There are only few Brönsted acidic catalyst applied for the Prins reaction of β-pinene to produce selectively nopol. Since, the acidic properties of the MoO_x/γ-Al₂O₃ catalyst differ from weak to moderate upon variation of loading and calcination temperature, study of the following acid catalyst for Prins reaction is interesting. To the best of our knowledge, Prins reaction of β-pinene with paraformaldehyde to produce nopol has not been studied over MoO_x/γ-Al₂O₃ catalysts. In the present work, a series of MoO_x/γ-Al₂O₃ was prepared to study the effect of MoO_x loading and calcination temperature of catalysts on the catalytic performance in Prins reaction. In order to study the effect of support on catalytic activity of molybdenum oxide catalyst, MoO_x was loaded on different supports like TiO₂, SiO₂, ZrO₂ and Nb₂O₅. Furthermore, the influence of solvent in Prins reaction in terms of their acceptor number, donor number and dielectric constant was studied. The effects of reaction parameters like reaction temperature, catalyst amount and reactant mole ratio were studied. The physiochemical properties of the catalyst were correlated with catalytic activity.

2. Experimental

2.1 Materials

Paraformaldehyde (hereafter PF) was obtained from the Loba Chemie, India. β-Pinene was purchased from Sigma-Aldrich, USA. Nopol, α-pinene, camphene and limonene were purchased from Alfa Aesar, USA. Pseudoboehmite was a gift sample by Süd-Chemie India Pvt Ltd. All the chemicals were used without further purification.

2.2 Catalyst preparation

The support γ-alumina was prepared by calcining pseudoboehmite at 550 °C for 4 h. MoO_x/γ-Al₂O₃ was prepared by the impregnation method using γ-Al₂O₃ and a solution of ammonium molybdate tetrahydrate [(NH₄)₆Mo₇O₂₄·4H₂O] as reported in the literature method.¹⁴ In a typical procedure, 10 g of γ-Al₂O₃ was dissolved in 50 ml of solution containing required amount of ammonium molybdate. The solution was stirred at 80 °C for 2 h followed by water evaporation. Finally, the catalyst was dried at 120 °C for 4 h followed by calcination at 800 °C. A series of MoO_x/γ-Al₂O₃ catalysts with MoO_x ranging from 0.2 to 25 wt% on γ-Al₂O₃ were prepared. Similarly, MoO_x loaded onto different supports like TiO₂, SiO₂, ZrO₂, Nb₂O₅ and pseudoboehmite were prepared. For the comparison, Al₂Mo₃O₁₂ was prepared by sol–gel method as reported elsewhere.³⁰

2.3 Catalyst characterization

The crystallinity and phase purity of all the catalysts after calcination were measured with an X-ray powder diffractometer instrument (Bruker D-2 phaser) using Cu-Kα radiation (λ – 1.542 Å). The samples were scanned in the 2θ range of 10–80 with scanning rate of 0.02 degrees per second.

The temperature programmed desorption of ammonia was performed in the temperature range of 100–600 °C with ramp rate of 12.5 °C min⁻¹. Before the TPD measurements, samples were pretreated at 550 °C for 1 h in flow of helium gas. Then samples were saturated with anhydrous ammonia gas (10% NH₃ + 90% He) at 100 °C for 45 min and finally flushed with He gas at 100 °C for 1 h to remove physisorbed ammonia. Nitrogen sorption measurements were carried out at 77 K in a Nova 1000 Quantachrome instrument. Prior to analysis, the samples were heated to 200 °C for 1 h. Specific surface areas of catalysts were determined by Brunauer–Emmett–Teller (BET) equation. Average pore diameter and total pore volume were measured by BJH method. The TEM images of the catalysts were analyzed by transmission electron microscope (TEM-JEOL-2010) instrument with SAED.

2.4 Catalytic activity studies

The liquid phase Prins condensation reaction of β-pinene with paraformaldehyde was carried out in the batch reactor in the presence of MoO_x/γ-Al₂O₃ solid acid catalyst. In a typical experiment, required amounts of the reactants (β-pinene and paraformaldehyde) in the presence of a solvent (benzonitrile) with an appropriate amount of catalyst (20 wt% with respect to total weight of reactant) were taken in a round bottom flask fitted with a condenser with continuous circulation of cold water and heated over oil bath with continuous stirring. The required temperature of oil bath was controlled and monitored by the PID controller and thermometer respectively.

During the course of reaction, samples were collected at certain intervals of time, the reaction mixture was cooled and the product was analyzed by gas chromatograph (Shimadzu-2014, FID detector) equipped with RTX-5 column (0.25 mm I.D and 30 m length). The oven temperature programme was 80 °C (5 min), 80–240 °C (15 °C min⁻¹), 240 °C (5 min). Reaction products were quantified using a multi-point calibration curve through the external standard method. The yield was calculated by the GC analysis using the formula, product (mol%) = [conversion (mol%) × selectivity (mol%)]/100.

3. Results and discussion

3.1 Characterization of MoO_x/γ-Al₂O₃ catalyst

The XRD pattern of MoO_x/γ-Al₂O₃ catalysts calcined at 800 °C with different MoO_x loadings are as shown in Fig. 1. There were no intense diffraction peaks corresponding to the crystalline MoO_x and Al₂Mo₃O₁₂ phase detected for the catalyst loading less than 11 wt% of MoO_x. This clearly indicates that MoO_x is highly dispersed on the support as amorphous MoO_x species. Further, increase in loading of MoO_x up to 25 wt%, the diffraction peaks due to formation of crystalline MoO_x and Al₂Mo₃O₁₂ were observed.³¹ The XRD patterns of 11 wt% MoO_x/γ-Al₂O₃ catalyst calcined at various temperatures are as shown in Fig. 2. Calcination of MoO_x/γ-Al₂O₃ catalyst below 700 °C retained the γ phase of alumina, whereas further calcination at higher temperature resulted in the formation of α-alumina. However the α-phase was found to be more predominant for the catalyst calcined above 900 °C. Typically, the γ phase of alumina turns to α-phase at 1050 °C, but in the present case, α-alumina was found to appear at 900 °C, which could be attributed to stabilization of alpha phase by MoO_x present on γ-alumina at lower temperature.^15,27,28

Fig. 1 XRD patterns of MoO_x loading on γ-Al₂O₃ calcined at 800 °C (a) γ-Al₂O₃ (b) 1 wt% (c) 7 wt% (d) 11 wt% (e) 15 wt% (f) 20 wt%.

Fig. 2 XRD patterns of MoO_x loading on γ-Al₂O₃ calcined at different temperatures.

The BET surface area, pore volume and average pore diameter of MoO_x/γ-Al₂O₃ catalyst are tabulated in Table 1. The γ-Al₂O₃ support showed a high surface area of 200 m² g⁻¹. The N₂ adsorption–desorption isotherm of γ-Al₂O₃ and MoO_x/γ-Al₂O₃ catalysts showed type-IV isotherm with H1 type hysteresis loop confirming the presence of mesopores (Fig. 3). As the MoO_x loading increased on γ-Al₂O₃ support, the surface area of the catalysts decreased due to shrinkage of alumina support and at 9 wt% of loading sudden decrease of the surface area was observed which is in concurrence with the previous report.¹⁴ The average pore diameter and pore volume showed reverse trend to that of surface area. With increasing MoO_x loading, the pore diameter and pore volume initially decreased up to 9 wt% and subsequently increased which could be associated with the formation of crystalline MoO_x.³¹ Based on the higher catalytic activity of 11 wt% of MoO_x in Prins reaction, the effect of calcination temperature was studied on the same catalyst system. The calcination of 11 wt% MoO_x/γ-Al₂O₃ catalyst at higher temperatures decreased the surface area of the catalysts as shown in Table 2. The decrease of surface area below 800 °C is due to the shrinkage of alumina support, whereas abrupt decrease in surface area above 800 °C is due to formation of α-phase of alumina and Al₂Mo₃O₁₂ as inferred by XRD. The decrease of surface area followed decrease in pore volume and pore diameter.

Table 1 Physicochemical properties of MoO_x loaded γ-Al₂O₃ catalysts calcined at 800 °C and their catalytic activity in Prins reaction^a

Amount of MoOx loading	Surface area (m^2 g^−1)	Pore diameter (nm)	Pore volume (cm^3 g^−1)	Acidity (mmol NH3 per g)	Surface acid density (μmol m^−2)	Conversion of β-pinene (mol%)	Selectivity (mol%)
							A	B	C	D	E
a Reaction conditions: β-pinene = 10 mmol, PF = 20 mmol, catalyst amount = 20 wt%, benzonitrile = 5 ml, reaction temperature = 90 °C, reaction time = 10 h. A = nopol, B = α-pinene, C = camphene, D = limonene, E = mixture of β-pinene isomerized products such as terpinenes and terpinolenes.
0.4	163	40.1	0.54	1.40	8.6	60.0	83.4	0.40	0.5	2.10	13.5
1	162	39.0	0.54	1.30	8.0	70.2	86.5	0.48	0.6	0.95	11.4
3	138	35.0	0.52	1.30	9.4	84.0	86.4	0.85	1.1	0.37	11.3
7	122	28.5	0.44	1.20	9.8	93.0	83.6	1.60	2.3	0.30	12.2
9	79	24.1	0.46	1.00	12.6	94.0	84.0	1.71	2.5	0.25	11.5
11	88	30.5	0.44	0.82	9.3	96.0	86.0	1.68	2.3	0.20	9.80
15	65	33.6	0.47	0.73	11.2	87.9	88.3	1.36	1.9	0.18	8.20
20	62	34.0	0.46	0.59	9.5	83.5	90.0	0.85	1.6	0.20	7.30

---

Fig. 3 Nitrogen adsorption–desorption isotherms and pore size distribution of 11 wt% MoO_x/γ-Al₂O₃ catalyst.

Table 2 Physical properties of MoO₃/Al₂O₃ calcined at 800 °C

Amount of Mo loading (mmol g(Al2O3)^−1)	Amount of MoOx loading (wt%)	Surface area (m^2 g^−1)	S0^a (m^2)	Soccupied^b (m^2 g(Al2O3)^−1)	Mo surface density^c (Mo per nm^2)
a Surface area of g(Al2O3)^−1 carrier.b Occupied area by MoO3 unit.c Number of Mo atoms per square nanometer BET surface area.
0.03	0.4	163	163.6	3.5	0.1
0.07	1.0	162	163.6	9.2	0.2
0.20	3.0	138	142.2	26.5	0.9
0.47	7.0	122	131.1	62.2	2.4
0.60	9.0	79	86.8	79.5	4.7
0.74	11.0	88	98.8	98.0	5.2
1.01	15.0	65	76.4	133.8	9.6
1.35	20.0	62	77.5	198.7	13.5

---

Physical property of MoO_x/Al₂O₃ catalyst calcined at 800 °C is as given in Table 2. S₀ is the surface area g(Al₂O₃)⁻¹ of a carrier, and was calculated from the BET specific surface area and MoO₃ content. The surface area of MoO_x/Al₂O₃ decreased with loading amount of MoO_x due to a shrink of alumina support and sudden decrease of surface area was observed at 11 wt%. The area of alumina support (S₀) decrease and the area occupied by MoO_x units (S_occupied) increase with loading amount of MoO_x and S_occupied become almost similar and larger than S₀ at 11 wt% and above loadings. The surface density of Mo atom on 8 and 10 wt% MoO_x/Al₂O₃ calcined at 800 °C catalyst is found to be 3.4 and 5.6 atoms nm⁻², respectively.³² The surface density of Mo atoms in monolayer coverage with MoO_x was 4.5 atom nm⁻² and the value was larger than the surface density of Mo atom on 7 wt% and a little smaller than 11 wt%. These results indicate that MoO_x was loaded as two-dimensional molybdenum oxide until 7 wt% and formation of three-dimensional MoO_x just started on 9 wt% MoO_x/γ-Al₂O₃. A detailed study on the transmission electron microscopy (TEM) analysis of 11 wt% MoO₃/γ-Al₂O₃ catalyst was carried out. Fig. 4a and b shows a large bright substrate with a dark shadow covering them with the amorphous mass. The bright substrate was assigned to be γ-Al₂O₃ as determined by their d spacing values (d(311) = 0.258 nm) of the respective plane and EDAX (Fig. 4c). The dark shadow was amorphous MoO₃ as confirmed by the EDAX measurements. Furthermore, no d spacing values are observed indicating amorphous nature of MoO₃ as shown in Fig. 4c. The selected area electron diffraction (SAED) pattern of catalyst exhibited the concentric rings indicating the presence of crystalline Al₂O₃ with well-defined lattice planes (311), (400), (440) of cubic γ-Al₂O₃. No concentric rings due to MoO₃ was observed in SAED pattern due to amorphous nature MoO₃ (Fig. 4d). The TEM image of 11 wt% of MoO₃/Al₂O₃ additionally indicates the MoO_x is uniformly loaded on alumina in the form amorphous monolayer domains and small clusters of MoO_x (Fig. 5).

Fig. 4 TEM images of 11 wt% MoO_x/Al₂O₃.

Fig. 5 EDAX pattern of 11 wt% MoO_x/Al₂O₃ catalyst.

The solid UV-DRS spectroscopic measurements were carried out to study the presence of molybdenum species in the supported catalyst. The most of the supported catalysts showed the absorption bands between 200 and 400 nm (Fig. 6), which is assigned to the ligand to metal charge transfer transitions.³³ Furthermore, the loading of 5 wt% exhibited two bands at 230 and 300 nm, which could be assigned to isolated monomeric species of tetrahedral and octahedral species, respectively. Whereas, above 5 wt% loading, only one broad peak was observed due to the overlap of tetrahedral and octahedral molybdenum species. In summary, surface area, TEM and UV-DRS measurements indicated that complete coverage of alumina support by molybdenum was achieved between the 9 and 11 wt% loading in which the molybdenum is present in octahedral and tetrahedral species in the form monolayer domain and small clusters of MoO₃.

Fig. 6 UV-DRS spectrum of different loaded MoO_x/Al₂O₃ catalysts.

To determine the amount and strength of acid sites, a temperature programmed desorption (TPD) profile of all the MoO_x/γ-Al₂O₃ samples were measured (Fig. 7). The bare γ-Al₂O₃ catalyst exhibited around 0.76 mmol of acidity. The total acidity decreased with increase of Mo loading and weaker to moderate new acid sites of Brönsted type are generated as proposed in the literature.^14,29,30 The decrease in total acidity could be due to the decrease of surface area. After loading of MoO_x on γ-Al₂O₃, high temperature desorption peak (450 °C) reduced in intensity and the low temperature desorption peak (250 °C) was found major in all loaded catalysts, which could be due to the formation of weak to moderate Brönsted acid sites.³⁰ The effect of different calcination temperature on acidity was also determined as shown in Fig. 8. It is observed that, the increase of calcination temperature from 450 to 1000 °C decreased the total acidity, which could be due to decrease in surface area and change of γ-Al₂O₃ to α-Al₂O₃ as shown in Table 3.

Fig. 7 TPD-NH₃ profile of MoO_x/γ-Al₂O₃ catalyst with different Mo loadings.

Fig. 8 TPD-NH₃ profile of 11% MoO_x/γ-Al₂O₃ calcined at different temperature.

Table 3 Effect of different calcination temperatures on physicochemical properties of MoO_x/γ-Al₂O₃ catalyst and their effect on catalytic activity^a

Calcination temperature	Surface area (m^2 g^−1)	Pore diameter (nm)	Pore volume (cm^3 g^−1)	Acidity (mmol of NH3 per g)	Surface acid density (μmol g^−1)	Conversion (mol%)	Selectivity (mol%)
a Reaction conditions: β-pinene = 10 mmol, PF = 20 mmol, catalyst amount = 20 wt%, benzonitrile = 5 ml, reaction temperature = 90 °C, reaction time = 10 h.
450	118	37.6	0.38	1.81	15.3	83.8	82.9
550	111	36.0	0.39	1.15	10.3	91.7	82.4
700	90	30.7	0.44	0.91	10.1	94.4	84.0
800	88	30.5	0.44	0.82	9.31	96.0	86.0
900	35	26.0	0.076	0.67	19.1	73.6	90.1
1000	36	25.5	0.054	0.60	16.6	59.4	91.2

---

To study the morphological properties of catalyst, SEM images of γ-Al₂O₃ and 11 wt% MoO_x/γ-Al₂O₃, were taken (Fig. 9). Both γ-Al₂O₃ and 11 wt% MoO_x/γ-Al₂O₃ contained spherical shaped particles with a particle size ranging from 5–10 μm. The morphology of γ-alumina remained same even after 11 wt% MoO_x loading.

Fig. 9 SEM images of (a) γ-Al₂O₃ (b) 11 wt% MoO_x/γ-Al₂O₃.

3.2 Catalytic activity by MoO_x/Al₂O₃ catalyst

The Prins reaction of β-pinene with paraformaldehyde was carried over different Mo loaded γ-Al₂O₃ catalysts. The γ-Al₂O₃ exhibited β-pinene conversion of 52% with 86% nopol selectivity. The lower catalytic activity of γ-Al₂O₃ could be due to the less acidity (0.7 mmol g⁻¹). As the amount of Mo loading on γ-Al₂O₃ increased up to 11 wt%, the conversion of β-pinene increased, and further increasing of MoO_x loading to 25 wt% decreased the β-pinene conversion (Fig. 10). This could be explained on the fact that surface acid density (μmol m⁻²) increases with increase of MoO_x loading up to 11 wt% and above 11 wt%, the surface acidity decreased as shown in the Fig. 10. Similar kind of correlation of Brönsted acidity with the MoO_x loading were reported earlier, where Brönsted acidity increases with increase in loading of MoO_x up to 11 wt% and above 11 wt%, the Bronsted acidity decreases due to the formation of crystalline MoO₃ and Al₂Mo₃O₁₂.¹⁴ Similar, trend of correlation between the acidity and the catalytic activity was observed in case of benzylation of anisole and isomerization of α-pinene.¹⁴ These above findings clearly indicates the formation of new Brönsted acid sites on the surface of the γ-Al₂O₃ which can be noticed and correlated with the change of surface acid density after MoO_x loading. From literature reports it is concluded that, at lower Mo loading tetrahedral MoO_x species are formed and higher Mo density results in octahedral MoO_x structures.²³ The catalytic activity increased with increase of Mo loading, and once the formation of Al₂(MoO₄)₃ (mainly, contains tetrahedral MoO_x species) started the catalytic activity decreased due to the presence of tetrahedral Mo species in the Al₂(MoO₄)₃. These results clearly indicate that octahedral MoO_x species are more active in the Prins reaction. The decrease of catalytic activity for alpha alumina could be due to its lower surface area. The nopol selectivity was found to be in the range of 82–92% irrespective of the MoO_x loading. The side products mainly consisting of α-pinene, camphene, limonene and mixture of terpinenes and terpinolenes were formed due to the isomerisation of β-pinene (Table 4). The 11 wt% of MoO_x on γ-Al₂O₃ exhibited a high β-pinene conversion of 96% with nopol selectivity of 86%. For comparison, MoO_x, Al₂Mo₃O₁₂ and (NH₄)₆Mo₇O₂₄. 4H₂O were also screened for the Prins reaction, which exhibited lower activity as shown in Table 4. The above result clearly indicates, loading of MoO_x on γ-Al₂O₃ support generated new acid sites which are catalytically active in Prins reaction rather than MoO_x, Al₂Mo₃O₁₂ or ammonium molybdate alone.

Fig. 10 Activity and selectivity of MoO_x/γ-Al₂O₃ catalyst calcined at 800 °C with various loadings on Prins reaction.

Table 4 Effect MoO_x loading on different supports and their catalytic activity on Prins reaction^a

Different support	Surface area (m^2 g^−1)	Acidity (mmol of NH3 per g)	Surface acid density (μmol m^−2)	Conversion of β-pinene (mol%)	Selectivity for nopol (mol%)	Nopol yield (mol%)	TON^b
a Reaction conditions: β-pinene = 10 mmol, PF = 20 mmol, catalyst amount = 20 wt%, benzonitrile = 5 ml, reaction temperature = 90 °C, reaction time = 10 h.b TON determined by mmol of nopol formed per mmol of active sites of the catalyst.
11 wt% MoOx/γ-Al2O3	88	0.91	10.3	96.0	86.0	82.5	22.6
11 wt% MoOx/ZrO2	77	0.98	12.7	99.5	1.90	1.9	4.8
11 wt% MoOx/Nb2O5	12	0.25	20.8	21.5	34.9	7.5	7.5
11 wt% MoOx/SiO2	45	0.26	5.7	38.5	89.0	34.2	32.8
11 wt% MoOx/TiO2	23	0.27	11.7	19.7	60.8	12.0	11.1
Al2Mo3O12	—	—	—	0.0	0.0	—	—
MoOx	—	—	—	12.23	24.2	—	—
Ammonium molybdate	—	—	—	8.0	12.0	—	—

---

In order to study the effect of calcination temperature of MoO_x/γ-Al₂O₃ catalyst on the catalytic activity, the 11 wt% MoO_x/γ-Al₂O₃ catalyst was calcined at different temperature ranging from 450 to 1000 °C. Increase of calcination temperature from 450 °C to 800 °C increased the β-pinene conversion from 83 to 96%. Further, increase of calcinations temperature of catalysts to 900 and 1000 °C resulted in lower catalytic activity which could be due to the formation of α-Al₂O₃ phase with lower surface area (Table 3). The calcination temperature was well studied by T. Kitanoa et al. Indicating Brönsted acid generation of alumina-supported molybdenum oxide is at high temperatures in the range of 700–800 °C.¹⁴ Therefore the catalyst with lower calcination temperature has shown to be less active even with high surface acid density. Whereas, the calcination of catalyst above 800 °C exhibited lower catalytic activity due to the phase change of Al₂O₃. The catalyst with 11 wt% MoO_x/γ-Al₂O₃ calcined at 800 °C showed higher catalytic activity with lower surface acid density value.

The Prins reaction was also carried over MoO_x loaded on different supports such as TiO₂, SiO₂, ZrO₂ and Nb₂O₅ as shown in the Table 3. Among the different supported catalyst, SiO₂ and γ-Al₂O₃ support showed high TON, whereas other supports like TiO₂, pseudoboehmite, ZrO₂ and Nb₂O₅ showed moderate to low TON values. Since most of the supports studied have different surface areas, the surface acid densities of the MoO_x loaded on different support (μmol m⁻²) were determined for the catalytic activity correlation and are tabulated in the Table 4. In spite of high surface acid density of Nb₂O₅, TiO₂, ZrO₂ supports, exhibited lower catalytic activity and TON than γ-Al₂O₃. Conversely, the SiO₂ support with lowest surface acid density exhibited lower catalytic activity. The MoO_x loaded on γ-Al₂O₃ with high TON and low surface acid density was found to be the best support among the others, wherein high β-pinene conversion of 96% and 86% nopol selectivity was obtained for the Prins reaction.

3.3 Effect of solvent

The reaction of β-pinene with paraformaldehyde in absence of solvent resulted in low yield of nopol (30.2%). This indicates that appropriate solvent is required for the generation of formaldehyde from PF. Hence, the effect of different solvents on the Prins reaction of β-pinene over a MoO_x/γ-Al₂O₃ was studied in detail. A range of solvents were screened for the Prins reaction and the results are represented in Table 5.

Table 5 Effect of nature of solvents in Prins reaction^a

Solvent	Dielectric constant (DC)	Acceptor number (AN)	Donor number (DN)	β-Pinene conversion (mol%)	Nopol selectivity (mol%)
a Reaction conditions: β-pinene = 10 mmol, PF = 20 mmol, catalyst amount = 20 wt%, solvent = 5 ml, reaction temperature = 80 °C, reaction time = 10 h.
No solvent	—	—	—	89.0	34.0
Cyclohexane	2.0	0	0	81.2	33.0
Toluene	2.4	8.2	0	92.4	42.4
Dichloromethane	10.4	16.7	0	100	40.0
Nitrobenzene	34.8	14.8	4.4	89.9	52.9
Benzonitrile	26.0	15.5	11.9	81.5	88.6
Acetonitrile	37.5	18.9	14.1	53.0	90.0
DMSO	46.7	19.3	29.8	20.6	10.3
Triethylamine	2.4	1.4	61.0	0.0	0.0

---

The influence of dielectric constant (DC), acceptor number (AN) and donor number (DN) of the solvents on the Prins reaction was studied. In the beginning, the effect of DC on the activity was studied by choosing solvents with similar dielectric constants. Solvents such as acetonitrile (DC = 37.5) and nitrobenzene (DC = 34.8) with comparable DC showed different yields for nopol. The solvents, toluene (DC = 2.4), cyclohexane (DC = 2.02) and triethylamine (DC = 2.42) also gave different activity profiles. This clearly indicates that dielectric constant of a solvent has no influence on the catalytic activity of MoO_x/γ-Al₂O₃ for Prins reaction.

Further, study on the solvent parameters like AN and DN gave better understanding on the role of solvent on the reaction. The AN and DN were first measured by Gutmann, which assess the acidic and basic properties of solvents, respectively.³⁴ The acceptor (AN) and donor number (DN) are the measure of strength of the solvents as Lewis acids or bases. The AN is based on the ³¹P-NMR chemical shift of triethylphosphine oxide in the solvent and DN is based on the heat of reaction between the ‘solvent’ and SbCl₅ in 1,2-C₂H₄Cl₂.^34,35 These values have been well established and used by several researchers to explain the role of solvent in chemical process.^34–38 As the AN of the solvents increased (DN = 0) from cyclohexane to dichloroethane, the β-pinene conversion improved from 81.2 to 100%, whereas selectivity was low (∼40%). Further increasing the DN of the solvents (having similar AN) from dichloroethane to acetonitrile, conversion dropped from 100 to 53% and selectivity increased substantially from 40 to 90%. However, increase of DN from acetonitrile to DMSO decreased both conversion and selectivity. Finally, triethylamine with highest DN (AN = negligible) showed no activity for Prins reaction.

From the above results, it is clear that solvent with only AN or DN like toluene, triethylamine, results in low yield for nopol. Solvents with both AN and DN values in the range of 10 to 20 (acetonitrile and benzonitrile) are the best to give high yield for nopol. The AN of a solvent represents acidic property, which facilitates fast decomposition of PF to formaldehyde, as a result, high conversion with low selectivity are obtained over the MoO_x/γ-Al₂O₃. The low selectivity could be due to the activation of Brönsted acid sites by the acidic solvents leading to the formation of isomerization products such as limonene, α-pinene and camphene etc. The DN of solvent represents the basic property of the solvents. Therefore, solvents with DN in the range of 10 to 20 could bind to the strong Brønsted acidic sites of the catalyst resulting in low conversions and high selectivity. Further increase of DN (>14), solvent may preferentially bind to most of the acid sites of the MoO_x/γ-Al₂O₃ catalyst leading to the low conversion. The low activity in presence of solvents with only high DN (triethylamine) could be due to the preferential adsorption on the acid sites leading to low generation of formaldehyde from PF. Hence, solvents with both AN and DN in the range 10 to 20 are suitable to facilitate reaction by balancing formation of side products and decomposition of PF. Among all the solvents taken in this study, benzonitrile with AN (15.5) and DN (11.9) exhibited good conversion and selectivity.

3.4 Influence of reaction conditions

In order to optimize the reaction condition, influence of reaction temperature on the conversion of β-pinene and selectivity towards nopol was studied using MoO_x/γ-Al₂O₃ with β-pinene : PF mole ratio of 1 : 2 using 20 wt% of catalyst. The β-pinene conversion increased with increase in temperature as shown in Fig. 11a. The low conversion of β-pinene at lower temperature (60 °C) is due to less in situ generation of formaldehyde from PF. At 100 °C, the formation of formaldehyde was faster and as a result higher conversion was achieved. The reaction temperature also had an influence on selectivity of nopol. With increase in reaction temperature from 80 to 100 °C, nopol selectivity decreased due to formation of side products such as camphene, limonene, α-pinene. The reaction carried out at 90 °C gave better conversion of β-pinene as well as high selectivity to nopol.

Fig. 11 Influence of reaction conditions. (a) Effect of temperature: reaction conditions: β-pinene = 10 mmol, PF = 20 mmol, catalyst amount = 20 wt%, benzonitrile = 5 ml, reaction time = 10 h. (b) Effect of β-pinene : PF mole ratio: reaction conditions: catalyst amount = 20 wt%, benzonitrile = 5 ml, reaction temperature = 90 °C, reaction time = 10 h. (c) Effect of catalyst amount: reaction conditions: β-pinene = 10 mmol, PF = 20 mmol, benzonitrile = 5 ml, reaction temperature = 90 °C, reaction time = 10 h. (d) Effect of reaction time: β-pinene = 10 mmol, PF = 20 mmol, catalyst amount = 20 wt%, benzonitrile = 5 ml, reaction temperature = 90 °C.

Further, to study the effect of different concentrations of reactants, Prins reaction was carried out with β-pinene : PF mole ratios of 1 : 1 to 1 : 4 as shown in Fig. 11b. The conversion and selectivity improved with increase in β-pinene : PF ratio from 1 : 1 to 1 : 2 and further increase of mole ratio did not have any effect on the catalyst performance. The less conversion of β-pinene at low PF concentration is due to less availability of active reactant, formaldehyde. The β-pinene : PF molar ratio of 1 : 2 showed high β-pinene conversion (92%) and nopol selectivity (93%).

The effect of catalyst amount on β-pinene conversion and selectivity for nopol was examined for Prins condensation reaction with different amount of MoO_x/γ-Al₂O₃ catalyst with 1 : 2 mole ratios of β-pinene : paraformaldehyde in the presence of benzonitrile solvent at 90 °C. As the catalyst amount was increased from 15 to 30 wt%, the β-pinene conversion increased from 88 to 98% (Fig. 11c). Increase of catalyst amount above 20 wt% decreased the selectivity for nopol due to the formation of side products. Hence, catalyst with 20 wt% was sufficient to produce a good conversion of β-pinene and nopol selectivity.

Effect of reaction time for Prins condensation of β-pinene with PF was carried out under optimized reaction conditions (Fig. 11d). The conversion of β-pinene after 1 h was 35% and further increase in reaction time, increased the conversion and reached to 96% after 10 h. But the conversion of β-pinene/hour decreased as reaction time increased. With increase of reaction time from 1 to 10 h decreased the selectivity for nopol from 91 to 86%.

3.5 Catalyst recyclability

The reusability studies were carried out by using 11 wt% MoO_x/γ-Al₂O₃ catalyst calcined at 700 °C. The used catalyst was filtered after reaction, washed with acetone and dried for 2 h at 120 °C. In order to remove the excess PF adsorbed on the catalyst and then catalyst was regenerated at 500 °C for 1 h. Catalyst gave good catalytic activity up to 3 cycles with slight decrease in β-pinene conversion and nopol selectivity (Table 6). The XRD patterns of catalyst after 3 recycle showed the intense peak of γ-Al₂O₃ phase indicating the structural integrity of the catalyst (Fig. 12). The recycled catalyst and reaction filtrate was carefully analyzed for the ICP-OES. The reaction filtrate showed the absence of Mo under the detection limit of ≥0.1 ppm and there was no decrease in the Mo content of fresh and recycled catalyst, indicating catalyst is truly heterogeneous at reaction conditions.

Table 6 Catalysts recycle study^a

Cycle	Conversion of β-pinene (mol%)	Selectivity for nopol (mol%)
a Reaction conditions: β-pinene = 40 mmol, PF = 80 mmol, catalyst amount = 20 wt%, benzonitrile = 20 ml, reaction temperature = 90 °C, reaction time = 10 h.
Fresh	94.4	84.0
Cycle-1	93.4	84.5
Cycle-2	91.4	84.1
Cycle-3	93.7	80.0

---

Fig. 12 XRD pattern of 11 wt% MoO_x/γ-Al₂O₃ catalyst (a) fresh catalyst (b) after 3 recycle.

3.6 Reaction mechanism

Plausible reaction mechanism for Prins condensation of β-pinene with PF over MoO_x/γ-Al₂O₃ is as shown in Scheme 1. The loading of MoO_x onto γ-Al₂O₃ generates weak to moderate acidic sites which are due to the molybdenum oxide domain (Mo–OH), three dimensional MoO₃ (Mo–(OH)–Mo) and at the interface by MoO₃ domain with γ-Al₂O₃ support (Mo–(OH)–Al). These could be the main catalytically active sites for the Prins reaction over MoO_x/γ-Al₂O₃ catalyst.¹⁴ The formaldehyde generated from PF interacts with acid sites to form a carbocation via polarization of carbonyl group, thereby generating an electrophile. The β-pinene attacks the electrophilic center followed by the allylic proton transfer completing the catalytic cycle with formation of nopol. The presence of strong Brönsted acidity activates the C=C of β-pinene to a greater extent resulting in isomerized products such as camphene, limonene, α-pinene etc. leading to lower selectivity for nopol.

Scheme 1 Plausible reaction mechanism for Prins condensation of β-pinene with PF over MoO_x/γ-Al₂O₃.

3.7 Comparison of MoO_x/γ-Al₂O₃ with reported solid acid catalysts

Table 7 shows the comparison of catalytic activity of MoO_x/γ-Al₂O₃ with other reported solid acid catalysts. The tin based catalysts like Sn-SBA-15, Sn-MCM-41 and Sn(OH)Cl showed high conversion and high selectivity for nopol. The other Lewis acidic metal ions such as Zn²⁺ is also found to be active sites for the Prins reaction. Zn–montmorillonite catalyst showed slightly higher nopol yield (87%) however, the leaching of the Zn was observed during the reaction. Prins reaction with sulfated zirconia was reported to give high nopol yield (98%). Zn-beta catalyst showed nopol yield of 85.5%. MoO_x/γ-Al₂O₃ catalyst in present work exhibited higher nopol yield with lower amount of catalyst than previously reported catalysts such as Sn-MCM-41, Zn-MCM-41, Zr-SBA-15, Na-ITQ, Sn-kenyaite. These results clearly indicate the potential application of MoO_x/γ-Al₂O₃ as solid acid catalyst in Prins reaction.

Table 7 Comparison of MoO_x/γ-Al₂O₃ with reported solid acid catalysts

Catalyst	Catalyst amount^a (wt%)	β-Pinene conversion (%)	Nopol yield	References
a Catalyst amount is calculated on the basis of total weight of catalyst with respect to total weight of the reactants.
Sn-SBA-15	13	99.8	98.7	6
Sn-MCM-41	51	99.3	98.0	5
Sn-kenyaite	50	50.8	49.8	5
Sn(OH)Cl	12	98.0	97.0	12
Sulfated zirconia	11	99.0	98.0	10
Na-ITQ	25	60.0	52.2	2
Zr-SBA-15	25	74.0	74.0	7
Zn–montmorillonite	41	90.0	87.3	39
Fe–Zn metal cyanide	10	52.0	49.9	11
Zn-MCM-41	50	91.0	75.5	8
MOF	12	—	82.0	13
Zn-beta	20	92.0	85.5	38
MoOx/γ-Al2O3	20	96.0	82.5	Present work

---

4. Conclusions

MoO_x/γ-Al₂O₃ exhibited to be highly selective and reusable catalyst for the synthesis of nopol from β-pinene and paraformaldehyde by Prins condensation. The XRD shows the phase transformation from MoO_x/γ-Al₂O₃ to MoO_x/α-Al₂O₃ and Al₂Mo₃O₁₂ at high calcination temperature and high MoO_x loading respectively. All the MoO_x/γ-Al₂O₃ catalysts showed type-IV isotherm with H1 type hysteresis loop confirming the presence of mesopores. The total acidity of the MoO_x/γ-Al₂O₃ decreased with increase of loading and calcinations temperature due to decrease of surface area. The MoO_x supported on γ-Al₂O₃ showed best performance among different types of supports used. As the loading of MoO_x increased, the β-pinene conversion increased up to 11 wt%, substantial loading decreased the conversion which is well correlated with surface acid density of the loaded catalysts. The high selectivity for nopol in the range of 82–86% was observed over the MoO_x/γ-Al₂O₃ catalyst. The solvents like acetonitrile and benzonitrile with both acceptor and donor numbers in the range of 10 to 20 are necessary to enhance the performance of the catalyst. The 11 wt% MoO_x loaded on γ-Al₂O₃ showed a better performance of 96% β-pinene conversion and 86% nopol selectivity. The catalyst was truly heterogeneous and can be recycled thrice with negligible decrease in the yield of nopol. Comparison of MoO_x/γ-Al₂O₃ with reported catalyst showed high remarkable activity indicating its potential application as solid acid catalyst for Prins reaction.

Acknowledgements

Vijaykumar S. M. and Dundappa Mumbaraddi acknowledge Admar Mutt Education Foundation (AMEF) for a fellowship. The authors are thankful to BIT, Bangalore for the BET measurements.

References

1. U. R. Pillai and E. S. Demessie, Chem. Commun., 2004, 826.
2. J. Wang, S. Jaenicke, G. K. Chuah, W. Hua and Z. Gao, Catal. Commun., 2011, 12, 1131.
3. J. P. Bain, J. Am. Chem. Soc., 1946, 68, 638.
4. A. Corma, S. Iborra, M. Mifsud and M. Renz, ARKIVOC, 2005, 9, 124.
5. E. A. Alarcón, L. Correa, C. Montes and A. L. Villa, Microporous Mesoporous Mater., 2010, 136, 59.
6. M. Selvaraj and Y. Choe, Appl. Catal., A, 2010, 373, 186.
7. D. M. Do, S. Jaenicke and G. Chuah, Catal. Sci. Technol., 2012, 2, 1417.
8. E. A. Alarcón, A. L. Villa and C. M. Correa, Microporous Mesoporous Mater., 2009, 122, 208.
9. A. Villa de, E. Alarcón and C. Montes de, Catal. Today, 2005, 942, 107.
10. S. V. Jadhav, K. M. Jinka and H. C. Bajaj, Appl. Catal., A, 2010, 390, 158.
11. M. V. Patil, M. K. Yadav and R. V. Jasra, J. Mol. Catal. A: Chem., 2007, 273, 39.
12. V. S. Marakatti, G. V. Shanbhag and A. B. Halgeri, RSC Adv., 2013, 3, 10795.
13. M. Opanasenko, A. Dhakshinamoorthy, Y. Hwang, J. Chang, H. Garcia and J. Čejka, ChemSusChem, 2013, 6, 865.
14. T. Kitanoa, S. Okazakia, T. Shishidoa, K. Teramura and T. Tanakaa, J. Mol. Catal. A: Chem., 2013, 371, 21.
15. A. P. Amrute, S. Sahoo, A. Bordoloi, Y. K. Hwang, J. Hwang and S. B. Halligudi, Catal. Commun., 2009, 10, 1404.
16. J. C. Yori, C. L. Pieck and J. M. Parera, Catal. Lett., 2000, 64, 141.
17. P. Bhaumik, T. Kane and P. L. Dhepe, Catal. Sci. Technol., 2014, 4, 2904.
18. K. Y. S. Ng and E. Gulari, J. Catal., 1985, 92, 340.
19. K. Arata, Appl. Catal., A, 1996, 3, 146.
20. C. C. Williams, J. G. Ekerdt, J. M. Jehng, F. D. Hardcastle, A. M. Turek and I. E. Wachs, J. Phys. Chem., 1991, 95, 8781.
21. S. Ito, N. Umehara, H. Takata and T. Fujii, Solid State Ionics, 2004, 172, 403.
22. S. Imamura, H. Sasaki, M. Shono and H. Kanai, J. Catal., 1998, 177, 72.
23. K. Chen, S. Xie, A. T. Bell and E. Iglesia, J. Catal., 2001, 198, 232.
24. M. C. Abello, M. F. Gomez and O. Ferretti, Appl. Catal., A, 2001, 207, 421.
25. Y. Okamoto, H. Tomioka, Y. Katoh, T. Imanaka and S. Teranishi, J. Phys. Chem., 1980, 8, 1833.
26. M. Ari, K. J. Miller, B. A. Marinkovic, P. M. Jardim, R. Avillez, F. Rizzo and M. A. White, J. Sol-Gel Sci. Technol., 2011, 58, 121.
27. B. Wang, G. Ding, Y. Shang, J. Lv, H. Wang, E. Wang, Z. Li, X. Ma and S. Qi Sun, Appl. Catal., A, 2012, 432, 144.
28. C. J. P. Steiner, D. P. H. Hasselman and R. M. Spriggs, J. Am. Ceram. Soc., 1971, 54, 412.
29. S. Cava, S. M. Tebcherani, I. A. Souza, S. A. Pianaro, C. A. Paskocimas, E. Longo and J. A. Varela, Mater. Chem. Phys., 2007, 103, 394.
30. S. Rajagoapal, J. A. Mazari and R. Miranda, J. Catal., 1995, 151, 192.
31. M. C. Abello, M. F. Gomez and O. Ferretti, Appl. Catal., A, 2001, 207, 421.
32. K. Chen, S. Xie, A. T. Bell and E. Iglesia, J. Catal., 2001, 198, 232.
33. J. P. Thielemanna, T. Resslerc, A. Walterc, G. Tzolova-Müllera and C. Hessb, Appl. Catal., A, 2011, 399, 28.
34. V. Gutmann, Coord. Chem. Rev., 1976, 18, 225.
35. J. Deutsch, H. A. Prescott, D. Muller, E. Kemnitz and H. Lieske, J. Catal., 2005, 231, 269.
36. Y. Chung and H. Rhee, J. Mol. Catal. A: Chem., 2001, 175, 249.
37. V. S. Marakatti, G. V. Shanbhag and A. B. Halgeri, Appl. Catal., A, 2013, 451, 71.
38. V. S. Marakatti, G. V. Shanbhag and A. B. Halgeri, Catal. Sci. Technol., 2014, 4, 4065.
39. M. K. Yadav and R. V. Jasra, Catal. Commun., 2006, 7, 889.

---