Preparation of Thermalstable Pillared Clays

Preparation of Thermalstable Pillared Clays

O d ,L a d.(Editors), New Frontiers in Catalysis Proceedings of the 10th Intcmtional Congress on Catalysis, 19.24 July, 1992, Budapest, Hungary Q 1993...

291KB Sizes 0 Downloads 46 Views

O d ,L a d.(Editors), New Frontiers in Catalysis Proceedings of the 10th Intcmtional Congress on Catalysis, 19.24 July, 1992, Budapest, Hungary Q 1993 Elsevier Science Publishem B.V.All righB =wed


S.Mendorofl, F. Gonzalezb, C. Pesquerab, I. Benitob, C. Blancob and G. Poncelelc ahtituto de Catalisis y Petroleoquimica, Cantoblanco,28049 Madrid, Spain bDepartamentode Quimica, Universidad de Cantabria, Santander, Spain cGroupe de Physo-Chimie Minerale et de Catalyse, Place Croix du Sud 1, 1348 Louvain-la-Neuve, Belgium

Abstract RE-AlPILCs have been prepared by three different methods and the variations in properties with respect to the parent clay studied through various phisicochemical technics. The catalytic activity has been tested in n-heptane, h-isomerization and h-craking in the 200-400°C interval at atmospheric pressure. Some conclusions on the suitability of the proposed synthesis methods have been drawn. 1. INTRODUCTION

The main difficulty in dealing with pillared materials for the cracking of hydrocarbons arises from the need of using water vapour at high temperatures (steaming) to eliminate the coke generated during the reaction; PILCs have not shown sufficiently stable to withstand the necessary temperature (around 700°C) and they collapse with the corresponding loss of their interlamellar spaces and, therefore, of their surface area and activity (1). So, a series of attempts have been made in order to get that hydrothermal stability introducing modifications in the synthesis steps, all of which result in modifications of the physicochemical properties and catalitic activity of the resulting material. One of the possible ways to improve the hydrothermal resistance of the pillars would be to form mixed pillars with cations that would mutually stabilize, thus retarding dehydroxilation and sintering of the starting cations. When dealing with Aluminium pillars, it seems promising to introduce oxides of trivalent cations (RE: Ce, La, Ga, Nd, Sm, etc.), such as in zeolites (2), either in the structure of the pillaring cation or within the framework of the starting material, so as to prevent the breakdown of the pillars through the loss of structural water. In this work we intend to study how the method of preparation affects the catalytic performance and properties of mixed pillared clays, using aluminium as pillaring agent and Ce and La (RE) as stabilizing agents.



The starting material was a Spanish bentonite from Almeria, supplied by Minas de Gador, S.A.. Its percent chemical composition was: SO2, 42.9; A1,03, 12.6; FeO,, 4.58; MgO, 6.01; CaO, 1.37; K20, 1.09; PzOlr,0.03; TiO,, 0.27; MnO, 0.02;Na20, 1.74;I.L., 29. Mineralogically it is more than 95 % montmorillonite. The Na-montmorillonite (Na-M) has a surface area of 84 m2/g, the pore volume volume, as the takeup of N2 at 0.98 relative pressure, was 0.098 ml/g (0.0196 in micropores); its CEC 61.6 meq/100 g. Atomic absorption, N, adsorption, XRD, thermal analysis, ammonia adsorption, and I R spectroscopy of adsorbed pyridine have been used for samples characterization. Hisomerization and H-cracking of n-heptane in the 200-400°Cinterval, have been used as test reaction to evaluate the catalytic activity of the materials after incorporation of 1% Pt from ammonia solution (incipient wetness method) and "in situ" reduction, 400"C, 2 h. Three different methods were used to prepare Al-RE pillared clays: I) incorporation of the A1 polycation to materials previously exchanged with RE; 11) partial exchange of A1 for RE in a solution containing the Keggin cation (Al,304(OH),4(H,0),,)7+, marketed by Hoechst as Locron; 111) "in situ" cohydrolisis of A13+ and RE3+chlorides in order to get the corresponding Al-RE mixed cation. A OH/Me ratio of 2.0 was always used except in case I) at which 1.6 was used. The polycation solutions were suitably aged in order to get the intended cation. The final pHs altogether with the main synthesis parameters are gathered in table 1. Clay concentration in solution was 2.5 %. Following Plee (3), 20 meq Me/g clay were used in all three methods. In I1 and I11 various AURE ratios, 20/0,15/5, 10/10 , 5/15 and 0/20 were used. After the exchange reaction, the materials were washed by dyalisis until Cl- elimination and freeze dryed. Table 1 Main synthesis parameters cation



A1"IPILC mmo1/100g


20/0 20/0 10/10 10/10 10/10 10/10

4.87 4,02 5.05 7.60 5.00 7.60

294.4 322.2 235.2 333.3 221.5 329.3


"A1,Ce "'N,Ce "AI,La "'A1,La


Me'+/PIL,C mmol/lOOg

14.3 145.0 30.9 225.6




.21 .20 .25 .12 .24 .ll

100 94.3 69.7 87.8 59.2

'CEC/mmol Me3+incorporated 3. RESULTS AND DISCUSSION

In table 2 are gathered the basal spacings at room temperature and 500°C of all samples. Also the main textural parameters are included. As it can be observed method I is the most suitable with respect to microporosity creation, being followed by 11; method 111 has revealed as inadequate to prepare PILCs, the effect being more important in La than in Ce mixed pillars. Smaller losses in basal spacings at 500°C in samples becoming from method


I1 denote a greater thermal stability of these samples with respect to those from method I and 111; again a disordered structure from the latter is patent. Since the average charge of A1 in the Keggin ion is 0.54, and the CEC of the original material was 0.616 meq/g as previously said, several types of polyoxocations may have been exchanged and/or the (A11J7+ion hydrolyzed on the clay surface (4). Table 2 Textural parameters of the pillared clays Sample

Go,, A RT


11.9 18.9 17.3 18.0 20.6 19.8 20.0 16.7 17.2

9.89 17.5 16.7 16.4 17.5 17.1 17.7 13.2 13.0

Isotherm tYPe




84 392 242 266 180 176 280 162 51

316 224 246 128 93 192 73








Na-M 'Na-AlPM 'Ce-AlPM 'La-AlPM "Na-AlPM "Ce-AlPM "La-AlPM We-AlPM '"La-AlPM





.0196 .236 .175 .184 .184 .I58

.214 .161 .216

.124 .ll8 .130 .05 1 .120 .076 .031

pH as high as 7 or more are not compatible with the formation of the Keggin cation, thus explaining the obtained results. An unpillared clay with microporosity due to some kind of polycations of Al, RE or boths on the external surface of the clays, may result from method I11 whereas an ideal PILC with ordereded structure results from method I; method I1 gives rise to an intermediate material as seen by the types of N2 isotherm involved (table 2). Thermal analysis detects losses at around 450°C in 'Ce-AlPM and 'La-A1PM attributable to the transformation of the aluminium oligomer into alumina. The case is not the same in the rest of the samples, in which a continuous loss is seen all along the thermogram corresponding to not well developed species. IR spectra of adsorbed pyridine show important differences among samples in acidity. Pillaring, in general, creates Bronsted and increases Lewis acidity. The latter remains up to 500"C, although it slightly decreases above 200°C. As it is known the acidic properties of the pillared materials are by no means simple summationsof those properties of the pillar and the original clay material. However, Bronsted acidity is provided mainly by structural OH groups in 2:l layers. On the other hand, the pillaring agent, forming or not pillars, is converted in more or less extension, to metal oxide by calcination contributing to the Lewis acidity of the sample (5). Then, differences among samples can be easily related with the different synthesis methods involved; samples resulting from co-hydrolysispresent more Lewis acidity, whereas the use of Locron enables the creation of Bronsted centres. A certain disorder seems to occur in the structure of the original clay when mixed pillars are introduced, less in 10/10 samples. The evolution with temperature of the total conversion on various samples is given in fig.1. Only 10/10 samples from methods I1 and I11 have been included. As it can be seen, total conversions are quite similar in all three types of samples and, in no case larger than on "a-AlPM. Only sample '%e-AlPM-10/10 presents a dramatic change in activity reaching conversions of 100% above 350°C mainly in cracking. Also mLa-AIPM-lO/10


show an important decrease in activity which can be related with the loss in surface area previously detected. Product distribution at 350°C are given in table 3. Table 3. Conversions at 350°C on RE-A1 mixed PILCs Sample Conv.,96 h-ilh-c

M-Nat 'Na 19.1 ,810

57.2 ,204


39.9 .232


39.5 .219

"Na 44 1.267

"Ce 38.7 2.080


49.6 1.527


43 .441



99.2 .017

27.9 .366

The pillaring process I increases cracking with respect to the original clay, more on Ce than on La samples; in contrast, method I1 increases isomerization doubling the values reached on A1-PM with Ce inclusion.The effect is slightly lower on "La-AlPM. Finally, method I11 (cohydrolysis) is on line with method I, cracking surpassing by far isomerization, more on Ce than on La samples. Fig. 1 n-heptane conversion 4. CONCLUSIONS

Mixed pillars intended for preparing thermalstable PILCs have shown fairly succesful in hitting their target. From the three methods used, only the first one gave rise to props homogeneously distributed between layers. The remainig two with a q/n ratio below 0.36 in all cases, result in incompletely pillared materials, with microporosity created through excess material on the external surface of the clays, more through I11 than through I1 method. Acidity as well as activity have shown extremely different, the method 111, and specially on samples with Al/RE 1, being the most suitable for h-cracking and method I1 for h-isomerization.

5. REFERENCES 1. M.L. Occelli, I&EC Pro.Res & Dev., 22 (1983) 553. 2. D.Tichit, F. Fajula, F. Figueras, C. Gueguen and J. Busquet, Proc., 9 O ICC, M.J.Phillips and M. Ternan (eds), I (1988) 112. 3. D.Plee, F. Borg, L. Gatineau and J.J. Pripiat, Clays and Clay Min., 35 (1987) 81. 4. R. A. Schoonheydt, SSSR, 58 (1991) 201. 5. H.Ming-Yuan, L. Zhonghui and M. Eze. Catalysis Today, 2 (1988) 321.