Abstract

A STUDY ON THE DEACTIVATION OF USY ZEOLITES WITH DIFFERENT RARE EARTH CONTENTS

C.A. Henriques¹, J.o.J. Santos¹, C.M.S. polato², M.L. Murta Valle²,

E.F.S. Aguiar^2,3 and J.L.F. Monteiro⁴

¹Instituto de Química, UERJ, Rua São Francisco Xavier, 524, CEP: 20559-900,

Rio de Janeiro, RJ, Brazil, Fax: (021) 254-3687

²Escola de Química, UFRJ, Centro de Tecnologia, Bloco E, Rio de Janeiro, RJ, Brazil

³DICAT/GEAPRO - CENPES/PETROBRÁS, Cidade Universitária, Quadra 7, CEP: 21949-900,

Rio de Janeiro, RJ, Brazil

⁴NUCAT - Núcleo de Catálise, COPPE/UFRJ, CP68502, CEP: 21.945-970,

Rio de Janeiro, RJ, Brazil, Fax: (021) 590-7135

(Received: November 5, 1997; Accepted: February 3, 1998)

Abstract - The deactivation of USY zeolites with different rare earth contents due to the coke formed from n-heptane at 450^oC was studied. The results show that the presence of rare earth elements decreases the cracking and coking activities, increasing catalytic stability. However, reaction selectivity was not significantly influenced. The greater the rare earth content, the lower the coking rates and the coke contents. The TPO/DSC profiles suggested that the catalytic effect of the rare earth elements promoted coke oxidation.

Keywords: Coke, deactivation, REY zeolites.

INTRODUCTION

Y zeolites, whose ions have been exchanged with rare earth cations, are widely used as the active component for cracking catalysts in the petroleum industry. These cations improve catalytic activity and gasoline yield, lowering gas production and coke formation. They also promote the thermal and hydrothermal stability of the catalyst (Martins, 1993).

Coke formation and its retention inside the pores is the main cause of deactivation of zeolite-based catalysts in hydrocarbon processing. Both coking rate and coke composition depend on a variety of parameters such as characteristics of the zeolite pore structure, characteristics of the acid sites, nature of the reactant and operating conditions (temperature, pressure, etc.) (Guisnet and Magnoux, 1991a, 1991b, 1994).

In the literature there are several studies on zeolite deactivation by coking and on coke characterization using pure zeolites (USHY, HZSM-5, Mordenite, Erionite), but there are no similar studies in which the complete cracking catalysts were used.

As part of a research project aiming at systematically studying the influence of cracking catalyst composition (zeolite type and content, active matrix, inactive matrix, rare earth content, etc.) on coking, this work examined the deactivation of USY zeolites with different rare earth contents by the coke formed from n-heptane cracking. Coking rate and catalyst stability were evaluated. Since in most commercial processes the cost of catalyst deactivation is very high, regeneration of coked zeolites by coke removal is of particular interest. Therefore, the reactivity of the coke deposited on the various samples during its oxidation with air was also evaluated.

EXPERIMENTAL

Three samples of Y zeolites with different rare earth contents were studied. They were prepared from a parent NaY zeolite (SAR = 5.3), as depicted below:

• USY ® ion exchanged with NH₄Cl and hydrothermal treatment with 100% steam at 923 K.

• REUSY ® successively ion exchanged with NH₄Cl and RECl₃, followed by hydrothermal treatment with 100% steam at 923 K.

• CREY ® ion exchanged with RECl₃ and hydrothermal treatment with 100% steam at 923 K.

Zeolite composition was determined by X-ray fluorescence in a XRF Phillips PW 1407 spectrometer. The framework silica-to-alumina ratio (SAR) was determined by FTIR, by means of the frequency shift of a characteristic band in the structural vibration region around 800 cm^-1, with a Nicolet 60SXR spectrometer. Textural characteristics of the zeolites, such as BET specific area, micropore volume (t-plot) and mesopore area (BJH), were evaluated by physisorption of N₂ at -196^oC in a Micromeritics ASAP 2400 and their acid sites density by TPD of NH₃ (adsorption performed at 150^oC with a 4.0% NH₃/He mixture and TPD performed at 10^oC/min up to 550^oC).

n-Heptane cracking was carried out in a glass fixed-bed gas-phase reactor at 450^oC and 1 atm. The reactant was kept in a saturator at 60^oC, and N₂ was used as carrier gas with an n-heptane partial pressure equal to 0.30 bar. The reaction products were analyzed on line by gas chromatography using an Al₂O₃/KCL PLOT column. For those runs aiming at evaluating the activity and the selectivity of the samples, WHSV was adjusted to obtain initial conversions close to 25%, while for those aiming at coke characterization it was adjusted to allow significant coke formation (an initial conversion of about 60%).

Coke was characterized by TGA/DSC analysis. The coked samples were burnt in a stream of N₂ + O₂ (10% O₂) at a 10^oC/min heating rate up to 750^oC in a Rigaku Thermoanalyser TS-1000. The TGA/DSC data allowed the calculation of total coke content and the inspection of the TPO profiles provided information on coke reactivity.

RESULTS AND DISCUSSION

Table 1 shows the physicochemical characteristics of the samples. The results clearly indicate that all zeolites have the same global SAR and approximately the same sodium content. However, the SAR of the framework is much lower for the high rare earth content zeolite, showing that such a high content prevented dealumination. This result is confirmed by the higher microporous volume of this sample, as well as the lower mesoporous specific area. The beneficial effect of rare earth on zeolite stability is related to the formation of RE-O-RE bonds inside the zeolite cavities, which stabilizes the zeolite due to the formation of bridges with structural tetrahedra (Gianetto, 1990). The high resistance of RE-O-RE bonds to hydrothermal treatment at high temperatures could explain, according to Ward (1976), the increase in hydrothermal stability of REY zeolites. As expected, the samples containing rare earth elements have lower acid sites density. Indeed, a rare earth cation is trivalent and it neutralizes three acid sites. Some authors (Lemos et al., 1987) claim that some of this acidity can be regenerated after a hydrolytic reaction catalyzed by the rare earth cations themselves with the subsequent formation of protons. In any case, the initial activity is never restored.

Figure 1 shows the catalytic activity as a function of time on stream for the samples under study. Experimental points were adjusted according to Voorhies’ equation (1945) and the corresponding deactivation coefficients (n) are presented in Table 2.

It can be observed that a low rare earth content has little influence on the activity and the stability of sample REUSY when compared to USY, no matter how much acid sites density is reduced. However, the presence of a high rare earth content sharply decreases cracking activity, thus increasing the catalytic stability, as shown in Table 2. Indeed, the sample with the highest rare earth content presents the lowest deactivation coefficient. Murta Valle et al. (1996) also observed the greatest stability of the CREY zeolite for tri-isopropylbenzene cracking on the same samples as those studied in the present work. These results can be associated with the decrease in acidity (density of sites) in the samples as rare earth content increases, thus decreasing cracking by means of a bimolecular mechanism which predominates in large pore zeolites. A decrease in catalytic activity due to the presence of rare earth elements was reported by Camorim et al. (1993) for n-hexane cracking on USY zeolites and by Bittencourt et al. (1995) for the same reaction catalyzed by HZSM-5 and LaZSM-5. In both cases, the decrease in acid sites density was also used to explain the observed trends.

Figure 2 shows the initial selectivities at isoconversion (± 25%) and isocoking (» 0%). Reaction selectivity is not significantly influenced by the content of rare earth elements since the cracking product distribution is similar for the three samples. The main products are C3 and C4 (C3 + C4 ³ 85%), indicating that in the samples studied, n-heptane cracking proceeded predominantly by means of the classic bimolecular mechanism (involving the formation of carbenium ions and b -scission). This fact was reinforced by the low values (lower than 1) for the C.M.R parameter ("cracking mechanism ratio" Þ (C1+S C2)/iC4), as shown in Table 3, which was proposed by Wielers et al. (1991) to represent the relative contribution of the protolitic cracking monomolecular mechanism in relation to the bimolecular mechanism.

Figure 1: Catalytic activity as a function of time on stream (n-heptane cracking at 450^oC).

^oC).

The results presented in Table 3 also show that, independently of the range of conversion, a low rare earth content (REUSY sample) has no significant effect on reaction selectivity. However, the presence of high rare earth contents, which significantly reduces acid sites density, has a negative effect on the hydride transfer bimolecular reactions, thereby justifying the greater values for O/P (olefin to paraffin ratio) and for the C.M.R. parameter observed for the CREY sample. Lemos et al. (1988) reported similar results when comparing the performance of LaHY and HY zeolites in n-heptane cracking.

The high values for the iC4/nC4 ratio, particularly for samples USY and REUSY, makes evident the absence of steric constraints in the formation of terciary carbenium ions, the cracking intermediates for the bimolecular mechanism.

Figure 3 shows coke content as a function of time on stream (corresponding to an initial conversion close to 60% for all samples). The greater the rare earth cation content, the lower the coking rate and the total coke content, thus depicting the negative effect of rare earth on coke formation. This fact may be attributed to a reduction in both acid sites density and mesoporosity. A decrease in acid sites density hinders bi and polimolecular reactions involved in coking, while the absence of a significant secondary mesoporous system limits the room available for voluminous coke precursor molecules. Similar results, reported by Henriques et al. (1997) for tri-isopropylbenzene cracking on the same samples, confirm the negative effect of rare earth on coking.

TPO/DSC profiles of coke deposits obtained after 300min T. O. S. are shown in Figure 4. It can be seen that the TPO peak shifts toward higher temperatures with a decrease in rare earth content. This can be associated with both the greater rare earth content in the CREY sample, which could promote coke oxidation, and to the differences in coke composition since different coke contents are being compared (8.1% in USY and 2.8% in CREY, for example).

Moljord et al. (1995) have shown that the density of acid sites is the most important factor in determining the rate of coke oxidation for protonic Y zeolites and that the larger the number of Al atoms or protonic acid sites per unit cell, the easier the coke combustion. This was not observed in the present work, since coke formed in USY (high acid sites density) was more difficult to burn than that formed in CREY (low acid sites density). Since similar trends were reported for the coke formed from tri-isopropylbenzene (Henriques et al., 1997), the present results reinforce a previous suggestion that rare earth cations promote coke oxidation, a hypothesis that requires additional studies.

Figure 3: Coke formation as a function of time on stream. (n-heptane cracking at 450^oC)

CONCLUSIONS

The evaluation of the effects of the presence of rare earth on USY zeolite deactivation showed that the sample with the greatest content of these elements presented the lowest activity in n-heptane cracking and in coking, and was also the most resistant to deactivation by coke. This could be associated with a decrease in acid sites density as rare earth content increased, reducing the bimolecular reactions involved in n-heptane cracking and in coking.

The presence of rare earth cations also influenced the oxidation of coke, which apparently was favored by the presence of these elements. This result reinforced the supposition that rare earth elements probably promote coke oxidation.

ACKNOWLEDGEMENTS

The authors thank Ricardo da Silva Aderne (NUCAT/COPPE/UFRJ) for the analysis by TGA/DSC. J. O. J. Santos thanks Fábrica Carioca de Catalisadores S.A and C. M. S. Polato thanks CNPq/PIBIC/UFRJ for their scholarships, and C. A. Henriques expresses her gratitude to Fábrica Carioca de Catalisadores S.A for its financial support.

Publication Dates

History