The Effect of Preparation Method on Pt/Nb2O5 Catalysts

The Effect of Preparation Method on Pt/Nb₂O₅ Catalysts

A. Eleutério¹, J.F. Santos, F.B. Passos1,^*, D.A.G. Aranda^2,3 and M. Schmal^2,3

¹Departamento de Engenharia Química, Universidade Federal Fluminense,

Rua Passo da Pátria, 156, São Domingos, Niterói, RJ, 24210-240, Brazil, FAX [phone],

E-mail: [e-mail]

²Departamento de Engenharia Química, Escola de Química,

Universidade Federal do Rio de Janeiro, Caixa Postal 68542, Rio de Janeiro, 21945-970, Brazil,

FAX [phone], E-mail: [e-mail]

³NUCAT - Programa de Engenharia Química - COPPE, Universidade Federal do Rio de Janeiro,

Caixa Postal 68502, Rio de Janeiro, RJ, 21945-970, Brazil, FAX [phone],

Email: [e-mail]

(Received: November 5, 1997; Accepted: March 26, 1998)

Abstract - The use of the ion-exchange method and the addition of lithium to Pt/Nb₂O₅ catalysts were investigated in this work, using techniques of temperature-programmed reduction, H₂ and CO chemisorption, UV-Vis diffuse reflectance spectroscopy and the conversion of n-heptane as a catalytic test. The superficial precursor present after the calcination step is platinum oxide, as previously observed for Pt/Nb₂O₅ catalysts prepared by the incipient wetness method. For some of the samples, autoreduction was observed during the calcination step, with the formation of low dispersion metallic platinum. The Pt/Nb₂O₅ catalysts prepared by ion exchange showed a high yield of olefins, as compared to aromatics, in the conversion of n-heptane. However, a high yield of hydrogenolysis products was also observed. For some of the lithium-containing samples, there was a suppression of dehydrogenation and aromatization reactions, with an increase in central C-C bond hydrogenolysis.

Keywords: Preparation method, catalysts, n-heptane.

INTRODUCTION

Pt/Nb₂O₅ catalysts were recently investigated (Aranda et al., 1993a), and promising results were obtained for the use of these catalysts in alkane dehydrogenation reactions. In the case of n-heptane dehydrogenation, these catalysts showed a higher yileld of olefins, with a decrease in selectivities for the undesired side reactions: hydrogenolysis, cracking, isomerization and aromatization. However, these catalysts showed low stability and deactivation due to coke formation. Furthermore, even when low reduction temperatures (300^° C) were employed, these catalysts showed lower metal-exposed fractions than the values found for similar Pt loading on Pt/Al₂O₃ catalysts (Aranda et al., 1993b).

In this paper, the ion-exchange method will be used to prepare Pt/Nb₂O₅ catalysts. In the case of Pt/SiO₂ catalysts, preparation by ion exchange, using Pt(NH₃)Cl₂ as the precursor, resulted in catalysts which are better dispersed than those prepared by incipient wetness impregnation (Benesi et al., 1968).

The effect of lithium addition to these catalysts will also be investigated. Lithium is a promoter of commercial dehydrogenation catalysts, and interest exists in discovering its effect on Nb₂O₅- based catalysts.

The prepared catalysts were examined by temperature-programmed reduction (TPR), H₂ and CO chemisorption, and UV-Vis absorption diffuse reflectance spectroscopy (DRS), and were evaluated for conversion of n-heptane.

EXPERIMENTAL

Pt/Nb₂O₅ catalysts were prepared by the ion-exchange method, using the procedure described by Benesi et al. (1968). Five grams of the support were continuously stirred in an aqueous solution containing Pt(NH₃)₄Cl₂. The solution was bufffered at a pH of 10 by adding NH₄OH (1N). The reaction mixture was vigorously stirred during four hours, followed by filtration and washing with deionized distilled water. The resulting solid was dried at 120° C for 16h. After drying, the resulting solid was calcinated at 500° C for 2h. In order to evaluate the effect of calcination temperature, Pt/Nb₂O₅ was calcinated at 300^oC.

Lithium was added by the incipient wetness method. An aqueous solution of LiCl was added to Pt/Nb₂O₅ samples. After impregnation, the mixture was dried at 120^oC and calcined at 500^oC.

Platinum and lithium contents were determined by atomic absorption. The prepared samples contained 1%w/w of platinum and the lithium contents were 1%, 0.5% and 1% on a weight basis.

TPR profiles were monitored by a thermal conductivity detector. The precursors were dehydrated at 393 K in a flow of Ar prior to reduction. A mixture of 1.5% hydrogen in argon flowed through the sample at 30ml/min, raising the temperature at a heating rate of 10 K/min up to 773 K.

For UV-Vis diffuse reflectance spectroscopy, experiments were performed at room temperature in a VARIAN model Cary 5 UV-Vis-NIR spectrophotometer equipped with a diffuse reflectance accessory (HARRICK). In order to eliminate the contribution of the support, the reflectance r (l ) was rated for reflectance of the support and results were calculated as a function of the Shultz Kubelka-Munk equation.

Carbon monoxide and hydrogen uptakes were obtained using an automatic adsorption system (ASAP 2900, Micromeritcs). After reduction at 573 K or 773K in a H₂ flow, the samples were evacuated at 10^-6Torr for 30min at the reduction temperature and cooled to room temperature. Irreversible H₂ uptakes were obtained from the total and reversible adsorption isotherms.

Dehydrogenation of n-heptane was carried out in an atmospheric glass microreactor. The sample was dried at 393 K for 30 min in a N₂ flow, followed by reduction in a H₂ flow at 773 K for 30 min. The reaction mixture was obtained by sending H₂ through a saturator kept at 298 K, which led to an H₂/n-heptane ratio equal to 16. Reaction temperature was 773 K and products were analyzed by using a gas chromatograph equipped with a flame ionization detector and a 50 m KCl-Al₂O₃ capillary column. After 4 hours on stream, the flow of the mixture was changed and product distribution was compared within the same range of conversion (10-15%).

RESULTS AND DISCUSSION

TPR profiles of the catalysts are presented on Figure 1. The profile of the Pt/Nb₂O₅ catalysts at 500⁰C showed a maximum at 110⁰C, ascribed to the reduction of platinum oxide, and a broad shoulder around 200⁰C. In addition, an H₂ uptake, was observed around 380⁰C, and it is ascribed to the reduction of niobia (Aranda et al., 1993b). Pt-Li/Nb₂O₅ catalysts showed similar profiles, indicating that the presence of lithium had not changed the surface precursor.

Table 1 lists the H₂ uptakes observed during reduction. It can be noted that the presence of lithium caused a decrease in H₂ consumption. This decrease was probably caused by the formation of reduced Pt particles during the calcination step. During the calcination procedure, NH₃ was produced, forming a reducing atmosphere which was able to reduce Pt. Moreover, reduction of the support also decreased for the samples containing lithium.

For the Pt/Nb₂O₅ catalyst calcined at 300⁰C, the fraction of H₂ consumed due to reduction of the metal was higher than that observed for the catalyst calcined at 500⁰C. This is consistent with the hypothesis that platinum is reduced during the calcination step, and this reduction is lower at low temperatures.

DRS results are shown in Figure 2. The Pt/Nb₂O₅ catalyst presented a band at 410 mm due to the presence of platinum oxide. (Lieske et al., 1982). Li-containing catalysts displayed an additional band at 340 mm, which is commonly ascribed to a oxychloroplatinum complex (Lieske et al., 1982). A small amount of this complex was probably formed due to impregnation with an aqueous solution of lithium chloride. Furthermore, the spectra of the Li-containing catalysts did not return to baseline. This behavior is consistent with platinum reduction after the calcination step.

H₂ and CO chemisorption results are presented in Table 2. The catalysts presented low H/Pt and CO/Pt values, even after reduction at 300° C. Results are consistent with the hypothesis that reduction occurs during the calcination step, causing the formation of large particles. Other possible reasons for this low dispersion are the high mobility of platinum particles during reduction and reduction of the support at 300° C. Unfortunately, due to contrast problems, it was not possible to carry out TEM (transmission electronic microscopy) experiments for these samples.

Reduction at high temperatures caused a decrease in H/Pt and CO/Pt values. This is generally observed when reducible supports are employed. In the case of Pt/Nb₂O₅, there was a complete suppression of H₂ adsorption capacity.

Li addition caused different effects on the catalysts reduced at 300° C; for the samples containing 0.1% and 0.5% Li, there was a decrease in platinum adsorption capacity, while for the 1%Li catalyst, there were not any significant changes in H/Pt and CO/Pt. In addition, the decrease in platinum adsorption capacity due to reduction at the higher temperature, was lower for the 0.1% Li and 0.5% Li catalysts, while a higher decrease was obtained in the case of the 1% Li catalyst.

Figure 1: TPR profiles for Pt/Nb₂O₅ catalysts.

Figure 2: DRS UV-Vis spectra of calcined Pt/Nb₂O₅ catalysts.

Table 1:
Temperature-Programmed Reduction Results

Catalyst	H2 uptake (m mol/gcat)	Fraction of the uptake due to metal precursor (%)
1% Pt/Nb2O5 calc. 500oC.	84.9	56.0
1% Pt/Nb2O5 calc. 300oC.	85.4	68.7
1%Pt-0.1%Li/Nb2O5	82.7	55.3
1%Pt-0.5%Li/Nb2O5	67.5	49.0
1%Pt-1 %Li/Nb2O5	46.4	42.7

Table 2:
H

Catalyst	H2 Uptakea (m mol/gcat)	H/Pt	CO Uptakea (m mol/gcat)	CO/Pt
1% Pt/Nb2O5 red. 300oC.	3.2	0.124	6.2	0.12
1% Pt/Nb2O5 red. 500oC.	-	-	0.49	0.007
1%Pt-0.1%Li/Nb2O5 red. 300oC	0.67	0.026	2.9	0.057
1%Pt-0.1%Li/Nb2O5 red. 500 oC	-	-	1.3	0.026
1%Pt-0.5%Li/Nb2O5 red. 300oC	0.31	0.012	0.94	0.018
1%Pt-0.5%Li/Nb2O5 red. 500oC	-	-	0.71	0.014
1%Pt-1%Li/Nb2O5 red. 300oC	2.4	0.122	5.5	0.107
1%Pt-1%Li/Nb2O5 red. 500oC	1.5	0.059	0.80	0.0016

₂ and CO Chemisorption on Pt/Nb₂O₅

Figure 3 displays the time dependence of n-heptane conversion at 500° C. All samples suffered a high initial deactivation process. Stationary activity was greater in the case of 1%Pt-0.1%Li/Nb₂O₅ followed by 1%Pt/Nb₂O₅ calcined at 500° C and 1%Pt/Nb₂O₅ calcined at 300° C. A higher lithium addition caused a small decrease in stationary conversion.

The results for product distribution did not follow the same trend as those for conversion (Table 3). The 1% Pt/Nb₂O₅, calcined at 500° C, presented a high selectivity for the formation of olefins, with a ratio of olefins to toluene (R) equal to 23.5. This result is better than observed for Pt/Nb₂O₅ prepared by incipient wetness, for which ratios between 7 and 17 were obtained (Aranda, 1995). Ion exchanged Pt/Nb₂O₅ catalysts are also more selective for olefins than Pt-Sn/Al₂O₃ catalysts, which presented values of R between 0.5 and 2 under similar conditions (Passos, 1994).

For all catalysts tested, the yield of hydrogenolysis products was higher than previously anticipated for Pt/Nb₂O₅ catalysts. This is indicative of the presence of large particles on the surface of the catalysts. Thus, the low H/Pt and CO/Pt values obtained, even after reduction at 300° C, could be at least partially due to the presence of large Pt particles. A high yield for central bond hydrogenolysis reactions was observed for the 1%Pt-0,1%Li/Nb₂O₅ catalyst. This reaction, which can occur by a mechanism involving carbonium ions (Gates et al., 1979), is catalyzed on acid sites. Thus, it should not be favored on Nb₂O₅-supported catalysts promoted by lithium. A metal catalyzed central-bond breaking hydrogenolysis probably became important for these catalysts.

The metal-catalyzed hydrogenolysis and isomerization reaction sites consist of a critical reaction site and a secondary site. The critical site is related to hydrocarbon chemisorption mode, while the secondary site is related to the hydrogen chemisorption to dehydrogenate the hydrocarbon (Foger and Anderson, 1980). Two routes using critical sites were identified: the C₂ mode requires at least two surface atoms and leads preferentially to central bond breaking. The iso-unit mode requires a smaller number of surface atoms (probably one). The ratio between central bonding and terminal bond rates was altered for several catalysts, indicating a modification of the metallic phase.

Figure 3: Conversion of n-heptane at 500° C on Pt/Nb₂O₅ catalysts.

Table 3:
Product Distribution for n-Heptane Conversion at 500

Catalyst	Term. Hydrog.a	Central Hidrog. b	C7Olef. c	Aro.d	Isom.e	Rf
1% Pt/Nb2O5 calc. 500oC.	9.5	1.5	75.6	3.2	10.2	23.6
1%Pt-0.1%Li/Nb2O5	34.3	24.1	26.4	7.4	7.8	3.6
1%Pt-0.5%Li/Nb2O5	9.2	4.9	71.3	5.1	9.5	14.0
1%Pt-1.0%Li/Nb2O5	9.9	1.1	72.2	11.1	5.7	6.5

^oC

^bCentral hydrogenolysis products yield (propane and butanes);

^c C

₇ olefin yield;

^daromatic yield;

^e n-Heptane isomers yield;

^fRatio of olefin yield to aromatic yield.

CONCLUSION

After calcination, the surface precursor present on Pt/Nb₂O₅ catalysts prepared by the ion-exchange method is platinum oxide. For some samples, autoreduction was observed during the calcination step, with the formation of low dispersion platinum particles. The Pt/Nb₂O₅ catalyst, prepared by ion-exchange, showed a high yield of olefins in the conversion of n-heptane. However, a high yield for hydrogenolysis products was also observed. For some lithium-containing samples, there was a suppression of dehydrogenation and aromatization reactions, with an increase in central-bond hydrogenolysis reaction.

ACKNOWLEDGMENTS

The authors thank FAPERJ (Fundação de Amparo a Pesquisa do Estado do Rio de Janeiro) and CNPq (Conselho Nacional de Pesquisa Científica).

• Aranda, D. A. G., Passos, F. B., Noronha, F. B. and Schmal, M., Characterization of Pt-Sn Bimetallic Catalysts Supported on Alumina and Niobia, Applied Catalysis A:General, vol. 100, p. 77 (1993a).
• Aranda, D. A. G., Passos, F. B., Noronha, F. B. and Schmal, M., Activity and Selectivity of Pt-Sn/Nb₂O₅ in n-Heptane Conversion, Catalysis Today, v. 16, p. 397, (1993b).
• Aranda, D. A. G., Influęncia da interaçăo Pt-Sn e do estado SMSI na superfície metálica de catalisadores de platina suportados. Ph. D. diss., Rio de Janeiro, COPPE-UFRJ (1995).
• Benesi, H. A, Curtis, R. M. and Studer, H. P., Preparation of Highly Dispersed Catalytic Metals , Journal of Catalysis, vol. 10, p. 328 (1968).
• Foger, K. and Anderson, J. R., Skeletal Reactions of Hydrocarbons over Supported Iridium-Gold Catalysts., Journal of Catalysis, vol.64, p. 448 (1980).
• Gates, B. C., Katzer, J. R. and Schuit, G. C.A., Chemistry of Catalytic Processes, Mc-Graw-Hill, New York (1989).
• Lieske, H., Lietz, G., Spindler, H. and Volter, J., Reactions of Platinum in Oxygen and Hydrogen-Treated Pt/g -Al₂O₃ Catalysts, Journal of Catalysis,vol. 81, p. 17 (1983).
• Passos, F. B., Estudo do efeito do lítio sobre as propriedades de catalisadores Pt/Al₂O₃ M.Sc. thesis, Rio de Janeiro, COPPE-UFRJ (1990).

* To whom correspondence should be addressed.

To whom correspondence should be addressed.

Publication Dates

• Publication in this collection
  09 Oct 1998
• Date of issue
  June 1998

History

• Accepted
  28 Mar 1998
• Received
  05 Nov 1997