ZSM-5 for propane dehydrogenation

ZSM-5 for propane dehydrogenation

Catalysis Communications 10 (2009) 2013–2017 Contents lists available at ScienceDirect Catalysis Communications journal homepage: www.elsevier.com/l...

422KB Sizes 2 Downloads 49 Views

Catalysis Communications 10 (2009) 2013–2017

Contents lists available at ScienceDirect

Catalysis Communications journal homepage: www.elsevier.com/locate/catcom

Effect of calcination atmosphere on the catalytic properties of PtSnNaMg/ZSM-5 for propane dehydrogenation Linyang Bai, Yuming Zhou *, Yiwei Zhang, Hui Liu, Xiaoli Sheng, Yongzheng Duan School of Chemistry and Chemical Engineering, Southeast University, Nanjing 211189, PR China

a r t i c l e

i n f o

Article history: Received 11 June 2009 Received in revised form 18 July 2009 Accepted 21 July 2009 Available online 25 July 2009 Keywords: Pt–Sn catalyst ZSM-5 Propane dehydrogenation Propylene Calcination atmosphere

a b s t r a c t The influence of calcination atmosphere (flowing air, static air and flowing N2) on the catalytic properties of PtSnNaMg/ZSM-5 catalysts for propane dehydrogenation was investigated. The catalysts were characterized by XRD, BET, TEM, and TPR. Results showed that the calcination atmosphere affected greatly the catalysts in many ways including surface area, the structure of the metallic phase and the interaction between platinum and tin, which are related to their catalytic behaviors. Among the catalysts studied, the PtSnNaMg/ZSM-5 catalyst calcined in flowing air exhibited the best catalytic performance. Ó 2009 Elsevier B.V. All rights reserved.

1. Introduction Platinum–tin catalysts have been intensively studied due to their applications in the hydrocarbon dehydrogenation and naphtha reforming processes [1–4] due to high catalytic properties. However, it has been reported that the catalytic performances of the catalysts would be dependent on the catalyst preparation methods [5–12]. Among these, calcination is an important factor to influence the catalytic properties [13,14]. In our previous work [15], we investigated the effect of calcination temperature on catalytic properties of PtSnNa/ZSM-5 for propane dehydrogenation. It was demonstrated that proper calcination temperature could enhance the catalytic activity of propane dehydrogenation. This phenomenon could be attributed to the strong interaction between Pt and Sn, whereas high temperature calcination could lead to the Pt sintering and suppress the dehydrogenation reaction. Huang et al. [16] investigated the effect of calcination atmosphere on Cu/Al2O3 catalyst for carbon monoxide oxidation and found that high temperature calcination in oxidizing atmosphere led to a redispersed copper surface and loss of activity, while in reducing atmosphere produced sintering of copper surface and increase of activity. However, to the best of our knowledge, the studies on calcination atmosphere have been scarcely made for the supported Pt–Sn catalysts. The aim of the current work was to investigate the effect of calcination atmosphere on the structural characteristics and the catalytic properties of PtSnNaMg/ZSM-5 catalysts used in the * Corresponding author. Tel./fax: +86 25 52090617. E-mail address: [email protected] (Y. Zhou). 1566-7367/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.catcom.2009.07.020

dehydrogenation of propane. The characterization of the catalysts was carried out by using XRD, BET, TEM, and TPR. 2. Experiment 2.1. Catalyst preparation PtSnNaMg/ZSM-5 catalysts were prepared by sequential impregnation method. The powder H-ZSM-5 was impregnated in an aqueous solution of 0.587 M Mg(NO3)2 at 80 °C for 4 h, then dried at 80 °C for 3 h. Afterwards, the sample was impregnated at 80 °C for 4 h in solutions mixture of 0.033 M H2PtCl6, 0.153 M SnCl4 and 0.427 M NaCl. Finally the prepared sample was dried at 80 °C for 3 h. The metal content in the sample was 0.5 wt.% Pt, 1.0 wt.% Sn, and 1.0 wt.% Na, 0.5% Mg, respectively. Afterwards, the prepared sample was fully agglomerated with 5.0 wt.% alumina during the process of pelletization. After heating at 120 °C overnight, the sample was divided into three portions. They were calcined at 500 °C in flowing air (a flow rate of 100 ml/min), static air in a muffle furnace and flowing N2 (a flow rate of 100 ml/min) for 4 h, respectively (part of each sample was taken out after cooling to room temperature for TPR experiment), followed by reduction for 8 h in flowing H2 at 500 °C. The different catalysts are denoted as CatA, Cat-B, and Cat-C, respectively (see Table 1). 2.2. Catalyst characterization The BET surface area and pore volume were obtained from nitrogen isotherms determined at liquid nitrogen temperature on

2014

L. Bai et al. / Catalysis Communications 10 (2009) 2013–2017

Table 1 Textural properties of the various samples. Catalysts

Atmosphere

Color

SBET (m2 g

ZSM-5 Cat-A Cat-B Cat-C

Static air Flowing air Static air Flowing N2

White White White Gray

332.1 309.2 298.4 284.3

a

1

)

Vp (ml g 0.2786 0.2732 0.2517 0.2467

1

)

Relative crystallinitya (%) 100 36 36 37

Relative crystallinity considering the ZSM-5 sample as 100% crystallinity.

an automatic analyzer ASAP2020 from Micromeritics. Samples were degassed at 350 °C for 4 h prior to the analysis. The powder X-ray diffraction patterns of the samples were obtained on a XD3A diffractometer using Cu Ka radiation in the scan 2h range of 5°–40°. The samples were checked for crystallinity by comparing the intensity of peak of the samples with that of parent ZSM-5 zeolites which was considered to be 100% crystalline. The Temperature-programmed reduction (TPR) was measured in TP-5000 apparatus. Prior to the TPR experiments, 0.15 g catalyst was dried in flowing N2 at 100 °C for 1 h. Five percentage H2/N2 was used as the reducing gas at a flow rate of 40 ml/min. The rate of temperature rise was 10 °C/min up to 800 °C. Conventional transmission electron microscopy (TEM) studies were conducted using JEOL2000 instrument. High-resolution TEM was performed using JEOL-JEM2010 apparatus. Samples were prepared by grinding, suspending and sonicating them in ethanol and placing a drop of the suspension on a copper grid with a perforated carbon film. 2.3. Propane dehydrogenation The catalytic experiments were performed in a stainless fixed bed tubular reactor, at 590 °C and atmospheric pressure using a catalyst charge of 1.0 g. The propane weight hourly space velocity (WHSV) was 3 h 1 and the mole ratio of H2/C3H8 was 0.25. The reaction products were analyzed with an online GC-14C gas chromatography. 3. Result and discussion

the zeolite, whereas part of tin, Na and Mg species could enter the channels of the zeolite. The absence of peaks for Pt should be due to the small loading amount. No phases containing different promoters could be detected on all catalysts, suggesting these species are incorporated in the network of the ZSM-5. Relative crystallinity data for all samples is shown in Table 1. It can be observed that the intensities of the (0 5 1) peak (at 23° 2h) decrease, implying these promoters can degrade the crystallinity of ZSM-5. Fig. 2 shows the TEM images of the different catalysts. No agglomerated particles on Cat-A can be seen. As for cat-B, some agglomeration of metal particles on the surface of the support can be observed, which means some particles are heterogeneously distributed. It is interesting to note that the population of large particles for Cat-C increases drastically owing to their strong sintering of metal particles. Thus, it can be inferred that the metal particles distribution decreases in the order: Cat-A > Cat-B > Cat-C. In order to clarify the possibility of Pt sintering, selected area electron diffraction analysis is measured, as shown in Fig. 2(4), SADP taken from the agglomerated particles. By measuring three different groups of ring radius, the diffraction rings can be indexed as (1 1 1), (2 0 0), and (2 2 0) crystal faces of Pt. Therefore, these particles can be identified as Pt phase by such diffraction analysis. These findings suggest that the calcination atmosphere can affect the dispersion of Pt particles on the external surface of the catalysts effectively, which would have great impact on catalytic performance. Despite high dispersion of Pt particles on Cat-A, we can’t rule out the possibility of metal particles agglomeration, thus a close inspection of Cat-A is necessary. High-resolution TEM image of

3.1. Catalyst characterization XRD patterns of the samples are depicted in Fig. 1. Regardless of the different calcination atmosphere treatments, the XRD patterns of the different catalysts are very similar to ZSM-5 structure, which suggests that the catalysts still maintain the well-ordered microstructure of ZSM-5. According to our previous results [17,18], the platinum particles were located mainly on the external surface of

(4) (3) (2)

(1)

5

10

15

20

25

30

35

40

0

2 THETA( ) Fig. 1. XRD patterns of the different samples: (1) ZSM-5, (2) Cat-A, (3) Cat-B, and (4) Cat-C.

Fig. 2. TEM micrographs of the different catalysts: (1) Cat-A, (2) Cat-B, (3) Cat-C, and (4) SADP of the agglomerated particles.

2015

L. Bai et al. / Catalysis Communications 10 (2009) 2013–2017

and Cat-C, small part of platinum particles agglomeration does occur. Data on the specific surface area and pore volume of different samples are shown in Table 1. ZSM-5 as a reference is also included. It is seen that the surface area and pore volume of ZSM-5 decrease, when Pt and other promoters were incorporated. The specific surface areas of Cat-A, Cat-B, and Cat-C are 309.2, 298.4, and 284.3 m2/g, respectively. Among the three catalysts, the CatA catalyst shows the highest surface area in that steam could be drained off in time and the micropore structure is preserved very well (some strong adsorbed water remains on the support before calcination treatment). For Cat-B, there is a decrease in surface area compared with Cat-A. This is because steam couldn’t be removed quickly in static air; some of micropore must be plugged. Accordingly, a diminution of the surface area occurs. On the Cat-C catalyst, water vapor could be removed quickly in flowing N2, the pore blocking could be avoided, and however, the strong sintering of platinum particles results in the decrease of specific surface area. 3.2. Temperature-programmed reduction (TPR)

Fig. 3. High-resolution TEM image of Cat-A.

TCD signal (a.u.)

(3)

(2)

(1)

200

300

400

500

600

700

0

Temperature ( C) Fig. 4. TPR profiles of the different catalysts: (1) Cat-A, (2) Cat-B, and (3) Cat-C.

Cat-A is shown in Fig. 3. The Pt particle size is in the range of 1–5 nm and the average particle size is 2.4 nm. This result suggests that although the Pt dispersion on Cat-A looks better than Cat-B

a

The TPR profiles of different catalysts are presented in Fig. 4. For Cat-A and Cat-B, the first reduction peak around 270 °C corresponds to the reduction of platinum species. The second hydrogen consumption peak at ca. 450 °C is assigned to the reduction of Sn4+–Sn2+. And the third peak at about 580 °C is attributed to the reduction of Sn2+–Sn0, which is in good agreement with the reported literature data [17]. It is worthy of remark that the first peak of Cat-C shifts towards high temperature overlapping with the reduction peak of tin species, which implies that the non-sintering Pt species having high reduction temperature is different from that of Cat-A and Cat-B. The difference between these two types of species may be attributed to their agglomeration state. The bulk Pt species has high reduction temperature, while the dispersed Pt species has low reduction temperature. A similar result was also reported by Shen and Kwai [19] as they suggested that two types of Pt species have different reduction temperatures on Pt/MCM-41. Moreover, the hydrogen consumption corresponding to the Pt species upon N2 calcination treatment decreases, whereas the area of the third peak increases in comparison to Cat-A and Cat-B. The decrease of hydrogen consumption can be explained by considering that most of platinum species upon calcination in N2 are decomposed to the metallic Pt. (In fact, the study of the catalyst color after N2 calcination supports this idea. The fresh catalyst upon calcination in N2 becomes gray, which is similar to that of the reduced sample). The

b

35

25

(2)

20 15 10 (3)

5 0

2

4

6

8

Time on stream (h)

10

Propylene selectivity (%)

Propane conversion (%)

30

(1) (2)

95

(1)

90

85

80

(3)

75 0

2

4

6

8

10

Time on stream (h)

Fig. 5. Propane conversion (a) and propylene selectivity (b) versus reaction time of the different catalysts: (1) Cat-A, (2) Cat-B, and (3) Cat-C.

2016

L. Bai et al. / Catalysis Communications 10 (2009) 2013–2017

Table 2 Reaction data of different catalysts after reaction for 6 h. Catalysts

Conversion (%)

Cat-A Cat-B Cat-C

34.05 28.27 10.14

Selectivity (%) Propylene

Methane

Ethane

Ethylene

C1 + C2

95.18 93.82 84.24

1.74 1.89 5.26

2.22 3.31 4.69

0.86 0.98 5.81

4.82 6.18 15.76

increase of hydrogen consumption is due to the reduction of more amounts of tin species, implying the tin species is readily reduced. This phenomenon can also account for the tin species segregation effect produced by the calcination during N2 treatment, which results in weakening the interaction between Pt and Sn greatly. However, for Cat-B, the area of the third peak is larger than that of Cat-A, indicating that the presence of water vapor could weaken the interaction between the tin species and the support, leading to more amounts of zerovalent tin formation. Therefore, the calcination atmosphere has an important influence on the interactions between Pt and Sn as well as them and the support. 3.3. Catalytic test The effect of calcination atmosphere of PtSnNaMg/ZSM-5 catalysts on the dehydrogenation of propane was investigated

and results are depicted in Fig. 5a and b. Among the three PtSnNaMg/ZSM-5 catalysts, Cat-A exhibits the highest propane conversion and propylene selectivity, whereas Cat-C displays the lowest propane conversion and propylene selectivity. It is well known from the literature that the active sites for the dehydrogenation catalysts are the ensembles of surface platinum [20]. The best dehydrogenation reaction performance of Cat-A can be attributed to the following reasons: one reason is related to the highest dispersion of platinum. The other is ascribed to the strong interaction between Pt with SnOx. As for Cat-B, the catalytic activity is lower than that of Cat-A. The relatively low activity can be due to lower dispersion of Pt ensembles. Just as mentioned above, the water vapor produced was not removed quickly enough in static air. The presence of water vapor has been reported to induce the agglomeration and sintering of the Pt particles during calcination and reduction [21–23]. Thus, treatment in static air results in a lower

Pt SnOx (1) M1 site

M2 site

ZSM-5

(2)

M1 site

M2 site

M1 site

M2 site

ZSM-5

ZSM-5

(3) M1 site

M1 site

ZSM-5

ZSM-5

Fig. 6. Schematic models for the influence of calcination atmosphere on the catalytic properties: (1) Cat-A, (2) Cat-B, and (3) Cat-C.

L. Bai et al. / Catalysis Communications 10 (2009) 2013–2017

Pt dispersion than treatment in flowing air. On the other hand, calcination in flowing N2 atmosphere leads to the lowest catalytic performance, which suggests that the bimetallic Pt–Sn catalyst changes into the monometallic Pt catalyst to some extent. The segregation of Pt and Sn results in disruption of the interaction between Pt and Sn. Moreover, the strong sintering of Pt particles on the support may be responsible for the lowest activity of Cat-C. Hence, it can be concluded that calcination atmosphere significantly affects the catalytic activity. In Fig. 5b, the initial propylene selectivity for the three catalysts decreases in the following order: Cat-A > Cat-B > Cat-C. This behavior could be assigned to a very different dehydrogenation capacity of these catalysts, probably related to a different structure of metallic phase when the calcination atmosphere is changed. It is generally accepted that the hydrogenolysis, isomeration and coking reactions demand large Pt ensembles [24,25], whereas the desired dehydrogenation reaction can proceed on small ensembles of surface platinum atoms [26]. As a consequence of the reduction of the size of platinum ensembles by calcination in flowing air, hydrogenolysis reactions are suppressed, which leads to enhancing the propylene selectivity. Table 2 contains results corresponding to the propane dehydrogenation test reaction after 6 h of reaction. The table includes values of conversion and selectivity to propylene, methane, ethane, ethylene, and methane + ethane + ethylene (C1 + C2). It is known that the Pt based catalysts have an important density of hydrogenolytic sites, which produces the C–C bound breaking of propane, thus giving (C1 + C2) hydrocarbons [27]. The Cat-A exhibits the lowest selectivity to side reaction products (C1 + C2), which is attributed to small particle size of Pt ensembles. However, on Cat-B the percentage of Pt agglomeration is increased a little, the selectivity to byproducts increases as compared to Cat-A. Concerning Cat-C, strong sintering of Pt ensembles occurs, thus the side reactions take place evidently. 3.4. Models for the influence of calcination atmosphere on the catalytic properties of PtSnNaMg/ZSM-5 catalysts Based on the obtained results, schematic models proposed for the influence of calcination atmosphere are depicted in Fig. 6. For Cat-A, the Sn species are mainly deposited on the external surface of the ZSM-5, and large fractions of Pt particles are relatively well-dispersed on the interface of SnOx, and small fractions of Pt are deposited on the support. It is reported that two kinds of active Pt species may coexist on the surface of the dehydrogenation catalyst, named M1 sites and M2 sites [28]. M1 sites are the sites in which Pt directly anchors on the carrier surface, while M2 sites corresponds to the sites in which Pt anchors on tin oxides surface, resulting in the formation of ‘‘sandwich structure’’. Generally, M1 sites are responsible for the side reactions, while M2 sites are the main reaction active sites for the dehydrogenation of propane. We can assume that the Sn could be found on the interface of platinum and support and therefore a close contact between platinum and Sn oxides could be achieved for Cat-A. In the case of Cat-B, M1 and M2 sites still remain, and the ‘‘sandwich structure” is preserved well. However the presence of water vapor could weaken the interaction between the metals and the support to some extent, which leads to more amounts of platinum particles agglomeration. For Cat-C, the Pt–Sn catalyst nearly changed into monometallic Pt catalyst. The segregation of Pt and SnOx occurs, which leads to large fractions of Pt atoms into large clusters. Moreover

2017

M2 sites are not formed, Pt directly anchors on the support. In this sense, the close contact between Pt and Sn is largely minimized. 4. Conclusions From the above results, it is clear that the effect of calcination atmosphere has an important impact on the physico-chemical properties including surface area and the structure of the metallic phase, which results in different catalytic behavior of PtSnNaMg/ ZSM-5. When the PtSnNaMg/ZSM-5 catalyst is calcined in flowing air, it exhibits the highest activity and selectivity for converting propane to propylene. This behavior is attributed to the strong interactions between Pt and Sn as well as them and the support. On the other hand, calcination in static air or flowing N2 is detrimental for the dispersion of Platinum. TEM results show that the PtSnNaMg/ZSM-5 catalyst calcined in flowing N2 results in the strong sintering of Pt ensembles, which severely modifies the metallic phase. Accordingly, the poorest catalytic performance is displayed. Acknowledgments The authors are grateful to the National Nature Science Foundation of China (50873026), Science and Technology Support Program of Jiangsu Province of China (BE2008129) and ‘‘Six Talents Pinnacle Program” of Jiangsu Province of China (06-A-033) for financial supports. References [1] J. Llorca, N. Homs, J.G. Fierro, J. Sales, P.R. de la Piscina, J. Catal. 166 (1997) 44. [2] S.A. Bocanegra, A.A. Castro, O.A. Scelza, S.R. de Miguel, Appl. Catal. A 333 (2007) 49. [3] O.A. Bariås, A. Holemn, E.A. Blekkan, J. Catal. 158 (1996) 1. [4] F. Epron, C. Carnevillier, P. Marécot, Appl. Catal. A 295 (2005) 157. [5] C.M. M’Komble, M.E. Dry, C.T. O’Connor, Zeolites 19 (1997) 175. [6] G.J. Siri, M.L. Casella, G.F. Santori, O.A. Ferretti, Ind. Eng. Chem. Res. 36 (1997) 4821. [7] C.L. Pieck, P. Marecot, C.A. Querini, J.M. Parera, J. Barbier, Appl. Catal. A 133 (1995) 281. [8] L.R. Raddi de Araujo, M. Schmal, Appl. Catal. A: Gen. 203 (2000) 275. [9] C.S. Triantafillidis, A.G. Vlessidis, N.P. Evmiridis, Ind. Eng. Chem. Res. 39 (2000) 307. [10] C.S. Triantafillidis, A.G. Vlessidis, L. Nalbandian, N.P. Evmiridis, Micropor. Mesopor. Mater. 47 (2001) 369. [11] V.A. Tsiatouras, T.K. Katranas, C.S. Triantafillidis, A.G. Vlessidis, E.G. Paulidou, N.P. Evmiridis, Stud. Surf. Sci. Catal. 142 (2002) 839. [12] T.K. Katranas, K.S. Triantafyllids, A.G. Vlessidis, N.P. Evmiridis, Catal. Lett. 118 (2007) 79. [13] R.P. Wang, X.W. Guo, X.S. Wang, J.Q. Hao, Catal. Today 93–95 (2004) 217. [14] S. Ku´s, M. Otremba, A. Tórz, M. Taniewski, Appl. Catal. A 230 (2002) 263. [15] Y.W. Zhang, Y.M. Zhou, Y.A. Li, Y. Wang, Y. Xu, P. Ch Wu, Catal. Commun. 8 (2007) 1009. [16] T.J. Huang, T.C. Yu, S.H. Chang, Appl. Catal. 52 (1989) 157. [17] Y.W. Zhang, Y.M. Zhou, H. Liu, Y. Wang, Y. Xu, P.C. Wu, Appl. Catal. A 333 (2007) 202. [18] Y.W. Zhang, Y.M. Zhou, A.D. Qiu, Y. Wang, Y. Xu, P.C. Wu, Acta Phys. Chim. Sin. 22 (2006) 672. [19] S.-C. Shen, S. Kawi, Appl. Catal. B 45 (2003) 63. [20] P. Biloen, F.M. Dautzenberg, W.M.H. Sachtler, J. Catal. 50 (1977) 77. [21] T.F. Garetto, C.R. Apesteguia, Catal. Today 62 (2000) 189. [22] T.F. Garetto, C.R. Apesteguia, Appl. Catal. B: Environ. 32 (2001) 83. [23] L.J. Hu, K.A. Boateng, J.M. Hill, J. Mol. Catal. 259 (2006) 51. [24] R.D. Cortright, J.M. Hill, J.A. Dumesic, Catal. Today 55 (2000) 213. [25] R.D. Cortright, J.A. Dumesic, J. Catal. 148 (1994) 771. [26] P. Biloen, F.M. Dauzenberg, W.M.H. Sachtler, J. Catal. 50 (1997) 77. [27] E.L. Jablonski, A.A. Castro, O.A. Scelza, S.R. de Miguel, Appl. Catal. A 183 (1999) 189. [28] L.W. Lin, W.S. Yang, J.F. Jia, Z.S. Xu, T. Zhang, Y.N. Fan, Y. Kou, J.Y. Shen, Sci. China (Series B) 42 (1999) 571.