Journal of the Chilean Chemical Society
versão On-line ISSN 0717-9707
J. Chil. Chem. Soc. v.49 n.3 Concepción set. 2004
J. Chil. Chem. Soc., 49, N 3 (2004): 245-250
OBTENTION OF CAMPHENE WITH H3PW12O40 CATALYSTS SUPPORTED ON TiO2, SiO2 AND ZrO2nH2O.
Nora A. Comelli, Liliana M. Grzona, Omar Masini, Esther N. Ponzia and Marta I. Ponzi*
INTEQUI (CONICET- UNSL) 25 de Mayo N 384. V. Mercedes, San Luis. Argentina
(Received: December 23, 2003 - Accepted: July 7, 2004)
The -pinene isomerization reaction was studied on catalysts based on H3PW12O40 (HPW) supported on TiO2, SiO2 and ZrO2nH2O.
Catalysts with a 40% nominal charge of phosphotungstic acid (HPW) were prepared by impregnation of supports with a HPW solution. The study of the isomerization reaction was carried out in liquid phase by using a discontinuous reactor, magnetically stirred and maintaining a constant temperature.
The catalyst supported on SiO2 resulted to be the most active while the one supported on ZrO2nH2O was the less active. This behavior correlated with the number and strength of the acid sites. XRD studies revealed the formation of HPW crystals when it was supported on SiO2.
Key words: -pinene, camphene acid catalysts, phosphotungstic acid, isomerization
1 - INTRODUCTION
Camphene and its derivatives are widely used in numerous industrial processes and manufactures. A striking illustration of camphene industrial importance is shown in the scheme below. Camphene can be easily converted to these other compounds .
The camphene is used in the manufacture of camphor and its related compounds. The conventional form of camphene production is through the isomerization of -pinene catalyzed by TiO2 which has been activated by an acid treatment. The reaction is carried out at temperatures between 150 and 170 C and the major products obtained are camphene and limonene .
The -pinene isomerization in presence of acid catalysts occurs by a mechanism in parallel, where, on one side bicyclic compounds are obtained as camphene, tricyclene, and bornylene through a cycle rearrangement, and on the other side monocyclic compounds as limonene, terpinolene, and terpinene are obtained by means of the rupture of one of the rings (Fig.1).
Fig. 1: Reaction scheme.
The isomerization reaction is developed in presence of strong acid catalysts, where the catalyst acidity has a strong effect on the activity and the camphene yield.
Several studies of the isomerization reaction of -pinene are observed in the bibliography using different catalysts, some authors used zeolites and clays with different treatments as catalysts [2-9] while other authors used oxides treated with acids [10-13]. Catalysts based on sulfated zirconia  have provided good results but they deactivate easily. Heteropolyacids (HPA) appear within the acid catalysts offering interesting possibilities as heterogeneous catalysts . Its high acid strength makes them comparable with other acid catalysts.
Heteropolyacids (HPA) in solid state are formed by heteropolyanions, cations and crystallization water, H3PW12O40.xH2 O.
IR studies reveal the information about the primary structure, while XRD studies are related to the secondary structure, and for this reason, different diffractograms are obtained according to the hydration degree of the HPA. HPA, in general, have large crystallization water amounts. Most of the crystallization water is lost up to 100 °C. The HPA decomposition takes place at higher temperature between 350 and 600 °C and proceeds as follows .
H3PW12 O40 = 1/2 P2O5 + 12 WO3 + 3/2 H2O
HPA can replace to the conventional acids, thus obtaining cleaner processes from the environmental point of view since they produce less amount of harmful effluents and at the same time they are less corrosive. Anyway, they also present some disadvantages, the specific surface area is low around 5 m2/g and they are difficult to be separated from reaction mixtures. It is supposed that by impregnating the HPA on adequate supports, it is possible to solve these difficulties. These materials are able to act as acid solid catalysts. Different supports such as carbon, SiO2 and ZrO2nH2O have been used to prepare catalysts [15-19].
The aim of this work is to study the influence of the support, on the activity and selectivity of the a-pinene isomerization reaction. For this, catalysts of phosphotungstic acid were prepared (H3PW12O40.xH 2O ) on ZrO2nH2O, TiO2 and SiO2.
2.1. Catalyst preparation
The ZrO2nH2O was prepared in the laboratory by hydrolyzing zirconium oxychloride (ZrOCl2 . 6 H2O) (Fluka) with ammonia hydroxide (Tetrahedron 28%) up to pH 10. The gel was aged for 24 hours at room temperature, then it was washed up to the moment in which the identification reaction of the chloride gave a negative result and it was dried at 110 °C for 12 hours.
Catalysts were prepared by impregnation with phosphotungstic acid (HPW) on ZrO2nH2O, TiO2 and SiO2 (Aerosil 200). The impregnation was performed with the necessary amount of HPW dissolved in 50% ethanol-water to reach a 40% nominal charge in catalysts. Then, the material was calcined at 300 °C for 4 hours. Catalysts obtained are identified with the following denomination HPWZr, HPWTi and HPWSi and the materials without calcination are denominated precursors PHPWZr, PHPWTi and PHPWSi .
Transformations occurring during calcination of precursors were studied by using the following techniques, differential scanning calorimetry (DSC) and thermogravimetry (TGA). The surface area of the support and catalysts was determined by using the BET method, in a Micromeritics Accusorb 2100E equipment.
The crystalline structure of catalysts was determined by X ray diffraction studies (XRD) on a Rigaku D-Max III equipment with Cu K radiation (λ = 1.5378 Å, 40 K, 30 mA). The 2q range analyzed was the one between 10° and 60°.
FTIR spectra of HPW and catalysts were obtained in a Brucker IFS66 FTIR equipment using pellets with KBr. FTIR spectra of catalysts with adsorbed ammonia were used to determine the presence of Brönsted acidity. The adsorption was carried out at room temperature by using a pure ammonia flow 15 cm3/min for 30 min.
The acidity was determined with thermal programmed desorption (TPD) experiments of ammonia in a thermobalance Shimadzu model TGA 50. The catalyst was pre-treated at 300 °C in He flow for 1 hour, then, it was cooled up to 50°C and was exposed to a NH3-He flow (10%, 100 ml min-1), up to constant weight. The ammonia physically adsorbed was removed by passing a helium current.
2.3. Catalytic test
The reaction was carried out in a glass reactor with three openings. In one of them, a coolant to condense vapor was placed, the second one was used to remove samples with a micropipette and in the third one a thermocouple was placed to read the reaction temperature. The reactor was maintained at constant temperature with a glycerine bath. The reaction mixture was stirred with a magnetic stirrer at 500 rpm. In a typical experiment, a sample 20 ml of -pinene (98.7% purity of a mixture provided by Derivados de San Luis company) was placed in the reactor and heated up to the reaction temperature, then, the catalyst (0.2g) was added to start the reaction, 100 μl sample were extracted for analysis at determined time intervals.
The reaction products were analyzed in a chromatograph Shimadzu model GC-17A provided with a FID detector. For product separation, a capillary column J&W DB-1 60 m x 0.248 mm covered with a thickness of 0.25 μm was used. Reaction products were identified and quantified using standards and confirmed by mass spectrometry.
The reaction was carried out at 130 °C for all catalysts. For the HPWSi catalyst, experiments were performed at 130°C and also at 45 and 80°C.
3. RESULTS AND DISCUSSION
3.1. Surface area
Table 1 shows the surface area of the supports and catalysts. A decrease of the specific surface area about 70% is observed in catalysts supported on TiO2 and SiO2 which is expected due to the high charge of HPW supported on them.
3.2. Thermal behavior
DSC and TGA curves of HPW, of ZrO2nH2O and of precursors are shown in Figs. 2 and 3, respectively. Curve (a) of Fig. 2 corresponds to non-supported HPW and presents two endothermic signals at 65 and 185 C, the first one corresponding to the removal of adsorbed water by the solid and the second one to the crystallization water loss (H3PW12O40.6H 2O).
Peaks observed in the DSC diagrams are in agreement with the maximums of weight loss registered in the TGA experiment. In the curve (a) of Fig. 3, the weight loss in the second peak is 3.6%, corresponding to the loss of 6 molecules of H2O.
H3PW12O40 .6H2O →H3PW12O40 + 6H2O
In the TGA of HPW, a series of weak signals between 300 and 530 degrees is also observed which can be generated by the decomposition of the primary structure of HPW .
The DSC of ZrO2nH2O catalyst (Curve (b) of Fig. 2) presents an endothermic signal attributed to the adsorbed water loss and an exothermic signal at 448 °C assigned to the transformation of the amorphous phase to crystalline phase. The DSC of PHPWZr precursor (curve (c) of Fig. 2), presents an analogous behavior to the one of the support. The exothermic signal at 459 °C corresponds to a crystallization of the amorphous zirconia, and no other transformation is observed within the temperature range studied. The DSC of PHPWTi precursor (Curve (d) of Fig. 2) shows an endothermic signal corresponding to the adsorbed water loss and an exothermic signal at 532 °C assigned to the crystallization of the tungsten trioxide coming from the HPW decomposition.
The curve (e) of Fig. 2 (PHPWSi) shows two endothermic signals as the ones presented by the HPW and one exothermic signal at 600 °C that could be assigned to the formation of tungsten trioxide crystals. The temperature at which these crystals appear depends on the support . In curve (e) of Fig. 3, two signals are observed, corresponding to the water loss of the HPW, at 35 and 154° C.
Results indicate that ZrO2nH2O as well as TiO2 produce interactions that difficult transformations of the HPW. Also, it may be inferred that the tungsten trioxide species formed in the HPW decomposition presents a weaker interaction in TiO2 than in SiO2 since there exists a crystallization at lower temperature for the catalyst supported on TiO2, this interaction would be higher even for ZrO2nH2O whose signal is not observed in the temperature range studied, but according to Lopez-Salina et al.  it would be around 720 °C. This allows us to suppose that calcining at 300°C, the HPW decomposition is not produced in any of the three studied catalysts.
Fig. 2: DSC diagrams of precursors: (a) HPW, (b) ZrO2nH2O, (c) PHPWZr, (d) PHPWTi, (e) PHPWSi.
Fig. 3: DrTGA diagrams of precursors: (a) HPW, (b) ZrO2nH2O, (c) PHPWZr, (d) PHPWTi, (e) PHPWSi.
3.3. Evolution of the crystalline structure after supporting the HPW
Figures 4 and 5 show XRD diagrams of catalysts and supports, respectively. The curve (a) of Fig. 4 (HPWZr) shows a wide and flat signal characteristic of ZrO2nH2O, where no signals of crystalline phases are observed related with the HPW, indicating that if crystals exist they are very small. The curve (b) of Fig. 4 (HPWTi) is equal to the curve (b) of Fig. 5 belonging to the TiO2 support, in this case no signals are observed corresponding to crystallized HPW. Finally, curve (c) of Fig. 4 (HPWSi) shows the crystalline form of HPW .
The SiO2 surface (311 m2/g) is larger than the one of the ZrO2nH2O (170 m2/g ) and than of TiO2 (48 m2/g), so, in this way, it would be expected that the dispersion is higher in silica than in the other supports, but the result is exactly the contrary, this indicating that the interaction strengths play an important role in the formation of crystals. From DSC-TGA studies a very different behavior in these supports was observed. As catalysts used in this study were calcined up to 300 °C, the higher difference observed is that one referred to the loss of water molecules contained in the HPW.
Fig. 4: X ray diagrams of catalysts: (a) HPWZr, (b) HPWTi, (c) HPWSi.
Fig. 5: X ray diagrams of supports: (a) ZrO2nH2O, (b) TiO2, (c) SiO2.
3.4. Integrity of Keggin structure
The HPW structure in catalysts was studied by FTIR. Fig. 6 shows spectra of catalysts and of non-supported HPW for comparison.
The non-supported HPW, curve (a), presents signals at 1081 cm-1 nas (P-O), 981 cm-1 nas (W=O), 889 cm-1 nas (W-O-W) and 798 cm-1 nas (W-O-W) coinciding with those reported in the literature  and they are assigned to a Keggin type structure. The HPWZr catalyst, curve (b), presents a spectrum very similar to the one of non-supported HPW, but with less intense signals. The HPWTi catalyst (curve c) shows signals at 1104, 1058, 966, 894 and 812. The unfolding of the signal of 1080 cm-1 of the Keggin structure into two signals 1058 and 1104 cm-1 indicates disturbance in the P-O interaction, this unfolding is similar to the one reported for lacunar structure of Keggin anions [PW11O39]7- whose signals correspond to 1100, 1046, 958, 904, 812 and 742 cm-1 . The lacunar structure is formed by the loss of a WO unity from a Keggin structure.
Fig. 6: FTIR spectra. (a) HPW, (b) HPWZr, (c) HPWTi, (d) HPWSi.
3.5. Acidity of solids
To determine the nature of acid sites in catalysts, FTIR spectra were recorded using for identification the ammonia molecule. FTIR spectra performed on catalysts with adsorbed ammonia show a band placed at 1400 cm-1 assigned to an asymmetric vibration of the NH+4 ion which indicates the presence of Brönsted sites . This signal is more pronounced in HPWTi and HPWSi, and it is almost imperceptible in HPWZr.
The strength and amount of present sites were determined by ammonia programmed thermal desorption. The amounts of adsorbed ammonia (mmol NH3/g catalyst) are 0.56, 0.90 and 2.3 for HPWZr, HPWTi and HPWSi respectively.
In the HPWTi catalyst, the adsorbed ammonia amount is 1.6 times than the one corresponding to the amount adsorbed in HPWZr, while in the HPWSi catalysts, the absorption is increased in 4.1 times. According to results, the catalyst with lower acidity is the one supported on ZrO2nH2O and the one of higher acidity is that one supported on SiO2.
3.6. Study of isomerization reaction
In figure 7 the concentration of -pinene and of reaction products as a function of time are plotted, when the reaction is carried out at 80C and HPWSi is used as catalyst. The -pinene concentration decreases rapidly at the beginning. The principal products obtained are camphene, limonene, alpha and gamma terpinene, terpinolene and other non identified products.
Fig. 7: Isomerization of -pinene. Distribution of reaction products as a function of time, on HPWSi at 80 °C. (a) - pinene (b) camphene (c) others (d) limonene (e) and - terpinene (f) terpinolene.
Table 2 shows activity and selectivity results of different catalysts.
Comparing results obtained at 130 °C, it is observed that the HPWZr catalyst results to be less active, the HPWTi is moderately active and the HPWSi is very active reaching practically to a total conversion at 20 reaction minutes. For this reason, experiments were performed at lower temperatures (80 and 45 °C).
Selectivities in camphene for HPWZr and HPWTi catalysts are near 50%, these values are very similar to the ones published by other authors for this reaction [2-4] using other solid catalysts. Although the a-pinene conversions are lower than the ones reported by these authors, it is possible to increase them by adjusting the operation variables, reaction temperature and catalyst charge. For the HPWSi catalyst, selectivities around 50% were obtained when the reaction proceeded at 45 °C, while if the temperature was raised to 80 and 130 °C, the selectivity decreased to values near 40 and 10%, respectively.
Thus resulting the following activity order:
HPWZr < HPWTi < HPWSi
Since each catalyst has a different superficial area, this could be the reason of the different activities observed, however, activities expressed per area unity are not the same (Table 2), this indicating that the different activity is produced by other causes.
Figures 8 A and B show results of -pinene conversion and selectivity in camphene, respectively. The less active catalyst resulted to be the HPWZr, which is in agreement with the low acidity of this solid. The most active catalyst is the HPWSi which resulted to be a catalyst of high acidity. X ray studies indicated that only in the HPWSi catalyst the HPW crystals were formed. Although the HPWTi catalyst presents activity, this activity is lower than in the HPWSi because the amount of sites and the strength are considerably lower, in addition to a degradation observed in the Keggin structure with the appearance of lacunar phase.
In general, the camphene selectivity has not been affected except in those experiments performed at 80 °C with HPWSi, and in this case a conversion of the order of 95% is achieved in only 60 minutes. When the catalyst results highly active, products of high molecular weight are formed, thus decreasing the selectivity in camphene.
Fig. 8: A and B: Conversion of -pinene and selectivity in camphene: HPWSi (a) 80 °C, (b) HPWTi a 130 °C, (c) HPWSi at 45 °C and (d) HPWZr at 130 °C.
The isomerization reaction of -pinene is strongly affected by the catalyst acidity. DSC-TGA studies show that the support produces a marked influence in the interaction support-HPW, thus resulting solids with very different acidity and consequently with different catalytic behavior.
Catalysts prepared by impregnation of HPW on TiO2, SiO2 and ZrO2nH2O resulted to be active. The activity increases in the following order: HPWZr < HPW Ti < HPWSi, which is in correspondence with an increase in the acid strength and the number of sites present in the different solids.
XRD studies indicate that only the HPWSi catalyst presents HPW crystals on its surface visible by means of this technique while the HPW is found completely dispersed in HPWZr and in HPWTi.
The HPW structure maintains its Keggin structure after the HPW was supported on ZrO2nH2O and SiO2 supports while it deforms in lacunar phase when it is supported on TiO2
We thank to Tca. Graciela Valle, Tco. Néstor Bernava and Ing. Edgado Soto for technical assistance.
 Chemical Abstract 36: 7592 (1942). [ Links ]
 S. Findik and G. Gündüz, JAOCS 74 (1997) 1145. [ Links ]
 A. De Stefanis , G. Perez, O. Ursini, A.A.G. Tomlinson, Appl. Catal. A: General 132 (1995) 353. [ Links ]
 A. Severino, Esculsas, J. Rocha, J. Vital, L.S. Lobo, Appl. Catal. A: General 142 (1996) 255. [ Links ]
 A. Allahverdiev, S. Irandoust, D.Y. Murzin, J. Catal. 185 (1999) 352. [ Links ]
 C. Volzone, O. Masini, N. Comelli, L. Grzona, E. Ponzi, M. Ponzi, Appl. Catal. A: General 214 (2001) 213. [ Links ]
 C. Lopez, F. Machado, K. Rodríguez, D. Arias, B. Méndez, M. Hasegawa, Catal. Lett. 62 (1999) 221. [ Links ]
 F. Özkan, G. Gündüz, O. Akpolat, N. Besün, D. Murzin, Chem. Eng. J 91 (2003) 257. [ Links ]
 N. Besün, F. Özkan, G. Gündüz, Appl. Catal. A: General 224 (2002) 285. [ Links ]
 A. Severino, J. Vital, L. S. Lobo: Heterogeneous Catalysis and Fine Chemical. 3rd. International Symposium. Poitiers, France (1993) 445. [ Links ]
 O. Masini , L. Grzona, N. Comelli, R. Montenegro, A. Carrascull, M. Ponzi, Inf. Tec.18 (5) (1997) 163. [ Links ]
 L. Grzona, N. Comelli, O. Masini, E. Ponzi, M. Ponzi, React. Kinet. Catal. Lett. 69 (2) (2000) 271. [ Links ]
 E. Ponzi, O. Masini, N. Comelli, L. Grzona, A. Carrascull, M. Ponzi, Bol. Soc. Chil. Quim. 44 (1999) 271. [ Links ]
 M. Misono. Catal. Rev. Sci. Eng. 29 (1987) 269. [ Links ]
 B. I. Kozhevnikov, Catal. Rev.-Sci. Eng. 37 (2) (1995) 311. [ Links ]
 L.R. Pizzio, C.V. Cáseres, M.N. Blanco, Appl. Catal. A: General 167 (1998) 283. [ Links ]
 C. Hu, H. Zhang, L. Xu, G. Peng, Appl. Catal. A: General 177 (1999) 237. [ Links ]
 S. Patel, N. Purohit, A. Patel, J. Molec. Catal. A: Chemical 192 (2003) 195. [ Links ]
 E. Lopez-Salina, J. Hernández-Cortéz, F. Schifter, E. Torres- García, J. Navarrete, A. Gutiérrez-Carrillo, T. López, P. Lottici, D. Bersani, Appl. Catal. A: General 193 (2000) 215. [ Links ]
 B.K. Hodnett, J.B. Moffat, J. Catal. 88 (1984) 253. [ Links ]
 B.B. Bardin, R.J. Davis, Top. in Catal. 6 (1998) 77. [ Links ]
 C. Morterra, G. Cerrato, F. Pinna, M. Signoretto, J. Phys. Chem. 98 (1994) 12373. [ Links ]