Preparation of a Nickel Molybdenum Carbide Catalyst and Its Activity in the Dry Reforming of Methane

CHINESE JOURNAL OF CATALYSIS Volume 32, Issue 5, 2011 Online English edition of the Chinese language journal RESEARCH PAPER

Cite this article as: Chin J Catal, 2011, 32: 771–776.

Preparation of a Nickel Molybdenum Carbide Catalyst and Its Activity in the Dry Reforming of Methane Taro HIROSE1, Yasushi OZAWA2, Masatoshi NAGAI1,* 1

Graduate School of Bio-Applications and Systems Engineering, Tokyo University of Agriculture and Technology, 2-24 Nakamachi, Koganei, Tokyo 184-8588, Japan

2

Energy Engineering Research Laboratory, Central Research Institute of Electric Power Industry, 2-6-1 Nagasaka, Yokosuka, Kanagawa 240-0196, Japan

Abstract: Nickel molybdenum carbide catalysts were prepared and their activities in the CO2 reforming of methane at a low CO2/CH4 reactant ratio were investigated using a microreactor at atmospheric pressure and at 973 K. The effect of the catalyst preparation method and the Ni/Mo ratio on the increase in catalyst life and the promotion of catalytic activity were investigated using N2 adsorption, X-ray diffraction, temperature-programmed carburization, temperature-programmed reaction, and a reforming reaction. The 25Ni75Mo catalyst that was carburized at 813 K exhibited the highest hydrogen formation ability and gave the least carbon deposition. The incomplete carburization of the Mo oxide species in the catalyst that was carburized at a lower temperature gradually gave a more active carburized species. The NiMoOxCy in the catalyst was more active in hydrogen formation during the dry reforming of methane while ȕ-Mo2C and Ș-Mo3C2 were less active. Key words: catalyst activation; catalyst selectivity; fuel; dry reforming; methane

The methane reforming process produces a synthetic gas that can be used to make hydrogen for fuel cells and industrial chemicals. However, this conventional process produces large amounts of carbon dioxide, which is a cause of global warming. The dry reforming of methane (CH4 + CO2 ĺ 2CO + 2H2) is applicable to the Fischer-Tropsch reaction, to diethyl ether synthesis, and to the reduction of carbon dioxide. Although noble metal catalysts and nickel catalysts have been used for synthetic gas to obtain a high degree of conversion during CO2 reforming [1–6], few studies have been conducted on methane reforming with and without CO2 on carbide or oxycarbide catalysts [7–11]. In these studies, the catalysts are readily deactivated by coke deposition during methane reforming and they transform into carbides in the presence of graphite. Therefore, the addition of cobalt to molybdenum carbides was studied during CO2 reforming to reduce carbon formation [12–14]. In addition, Shu et al. [15] reported that 5% CO2 among reactions of 1%–6% CO2 that were added to a 10% Ar-90% CH4 feed was effective for the CO2 reforming of CH4

on Mo carbide. In this study, we investigated the CO2 reforming of methane at a low CO2/CH4 reactant ratio on a nickel molybdenum carbide catalyst at atmospheric pressure at 973 K using a microreactor. We determined the effects of catalyst preparation and the Ni/Mo ratio on the extension of catalyst life and the promotion of catalytic activity. We also discuss the active species on the basis of X-ray diffraction (XRD), temperature-programmed carburization (TPC), and temperature-programmed reduction (TPR) results.

1

Experimental

Aqueous solutions of the Ni and Mo precursors of Ni(NO3)2·6H2O (Kishida Chemical Co., 98%) and (NH4)6Mo7O24·4H2O (Kishida Chemical Co., 99%) were prepared in water, dried at 393 K overnight, and calcined at 773 K for 5 h in air. Catalyst carburization and the experimental reactions were performed in a quartz microreactor at atmospheric

Received 14 October 2010. Accepted 19 November 2010. *Corresponding author. Tel: +81-42-388-7060; Fax: +81-42-381-4201; E-mail: [email protected] Copyright © 2011, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. Published by Elsevier BV. All rights reserved. DOI: 10.1016/S1872-2067(10)60185-4

Taro HIROSE et al. / Chinese Journal of Catalysis, 2011, 32: 771–776

2 Results and discussion 2.1 Dry reforming reaction All the catalysts gave a CO2 conversion of approximately 100% for the dry reforming reaction at 973 K. The H2 formation rate and CH4 conversion over each catalyst rapidly dropped after the start of the reaction and soon reached the steady state or gradually increased. The H2 formation rate and CH4 conversion in the dry reforming reaction over the 50NiMo catalyst that was carburized from 773 to 873 K are shown in Fig. 1. For the catalysts carburized at 773–798 K, the rate and conversion at the start of the reaction was high and was followed by a rapid decrease and then a gradual increase until the end of the experiment. In contrast, for the 810–873 K carburized catalysts the H2 formation rate at the start of the reaction was low followed by slight change and then a steady state was

773 K 786 K 798 K 810 K 823 K 873 K

H2 formation rate (mmol/min)

1.0

0.8

0.6

0.4 100 773 K 786 K 798 K 810 K 823 K 873 K

90 CH4 conversion (%)

pressure. Ni and Mo oxides (0.2 g) were placed in the microreactor and pretreated with a stream of air (66.7 ml/min) at 773 K for 1 h and carburized with a stream of a 20% CH4-80% H2 mixture (66.7 ml/min) at 773–873 K for 1 h. The gas was then purged with He (15 ml/min) and changed to a mixed gas (18 ml/min) containing 33% CH4 and 1.1% CO2 in He, which was supplied to the catalyst for the reforming reaction at 973 K for 6 h. Three kinds of nickel molybdenum carbide catalysts were prepared: 25NiMo, 50NiMo, and 75NiMo and these represent catalysts with Ni/Mo mole ratios of 25/75, 50/50, and 75/25, respectively. The reaction products and feed were analyzed using quadrupole mass spectrometry (Quadstar 422, Balzers Co.), which was connected on-line. H2, CH4, H2O, C2H2, C2H4, CO (or C2H6 or C2H4), C2H6, C3H8, CO2, and C6H6 were determined by their mass-to-charge ratios (m/z) of 2, 15, 18, 26, 27, 28, 29, 30, 44, and 78, respectively. The CH4 conversion and H2 formation rate were then obtained from this mass spectrometric data using He as the internal standard. The BET surface area of the catalysts was then measured using an Omnisorp 100CX (Beckman Coulter Co.) before and after the reaction. The XRD patterns were also measured using a RINT2100 PC/V (Rigaku Co.) with Cu KĮ radiation. The peaks were identified using ASTM cards. The carbide structure of the catalysts was then investigated by TPC with the same experimental setup used for the reforming reaction. The Ni and Mo oxides (0.2 g) were pretreated with a stream of air (66.7 ml/min) at 773 K for 1 h, cooled to 573 K with He (15 ml/min), and then heated by TPC from 573 to 1073 K at 1 K/min using a mixed gas stream (30 ml/min) composed of 6.7% CH4 and 26.7% H2 in He. The active catalyst species was then investigated using TPR. After catalyst carburization or the reforming reaction, the catalyst was cooled to room temperature in situ using He (15 ml/min) and then heated by TPR in situ from room temperature to 1273 K at 5 K/min with a stream of 10% H2 in He (15 ml/min).

80 70 60 50 0

1

2 3 4 Time on stream (h)

5

6

Fig. 1. H2 formation rate and CH4 conversion over 50NiMo carburized at different temperatures.

reached, while the CH4 conversion at the start of the reaction was high followed by a rapid decrease and then finally reaching a steady state. The 786 K carburized 50NiMo catalyst exhibited the highest H2 formation rate and CH4 conversion at 6 h on stream while the rate and the conversion roughly decreased with an increase in carburizing temperature. The H2 formation rate and CH4 conversion in the dry reforming reaction at 6 h on stream for 25NiMo, 50NiMo, and 75NiMo are shown in Fig. 2. At the same carburizing temperature (786 and 798 K, respectively) the H2 formation rate decreased and the CH4 conversion increased as the Ni/Mo ratio increased while the CO2 conversion for all the catalysts remained at approximately 100%. The H2 formation rate for the 75NiMo catalyst decreased as the carburizing temperature increased while the rate for the 25NiMo catalyst increased. CH4 conversion decreased as the carburizing temperature increased for all the catalysts. Low conversion of CH4 leads to low carbon deposition because almost all the CO2 reacts with CH4 for all the catalysts and further CH4 conversion corresponds to the decomposition to carbonaceous compounds. Small amounts of C2H6, C2H4, C2H2, and C6H6 were observed for all the catalysts at 6 h after the reforming reaction but no relation was found between these amounts and the Ni/Mo ratio. As a result, the 813 K carburized 25NiMo catalyst exhibited the highest H2 formation rate with the lowest carbon deposition.

Taro HIROSE et al. / Chinese Journal of Catalysis, 2011, 32: 771–776

CH4 conversion (%)

H2 formation rate (mmol/min)

1.0 0.8 0.6 0.4 0.2

75NiMo-773 K 50NiMo-786 K 25NiMo-798 K

75NiMo-786 K 50NiMo-798 K 25NiMo-813 K

75NiMo-773 K 50NiMo-786 K 25NiMo-798 K

75NiMo-786 K 50NiMo-798 K 25NiMo-813 K

0.0 90 80 70 60

Fig. 4. XRD patterns of the 50NiMo catalysts before the reaction and carburized at 773 (1), 786 (2), 823 (3), and 873 K (4).

50 40

770

780 790 800 Carburizing temperature (K)

810

820

Fig. 2. H2 formation rate and CH4 conversion as a function of carburizing temperature at 6 h after the reaction over different catalysts.

2.2 TPC, TPR, XRD, and BET analyses The TPC profiles of the catalysts are shown in Fig. 3. Water formation peaks were observed at 677, 719, and 886 K while the CO (or C2H6 or C2H4) formation peak was at 906 K in the TPC measurement for 25NiMo. For 50NiMo and 75NiMo the water peak at 677 K became broader. An increase in the Ni/Mo ratio lowered the water formation temperature from 886 to 870 719

677

886

(a)

906

Intensity

731 (b)

878 881 720

(c) 870 875 600

650

700 750 800 Temperature (K)

850

900

950

Fig. 3. TPC profiles of the NiMo catalysts. (a) 25NiMo; (b) 50NiMo; (c) 75NiMo. Upper: H2O; lower: CO (or C2H6 or C2H4).

K and the CO formation temperature from 906 to 875 K during TPC. XRD patterns of the 50NiMo catalyst after TPC are shown in Fig. 4. Nickel molybdenum oxycarbide (NiMoOxCy) and molybdenum oxides (MoO2) were observed at a carburization temperature of 773 K but NiMoOxCy and Ni metal, without MoO2 formation, were detected at 786 K. At 873 K, ȕ-Mo2C and Ni metal were observed and NiMoOxCy was absent. Therefore, the oxidic catalysts were insufficiently carburized below 780 K and NiMoOxCy was formed below 770 K. Combining the TPC was evident because of the decomposed oxygen from the transformation of MoO3 to MoO2. Additionally water formation at approximately 720 K was due to the transformation of NiMoO4 to NiMoOxCy, and water and CO (or C2H6 or C2H4) at approximately 880 K was formed by the oxygen and carbon from the transformation of NiMoOxCy to ȕ-Mo2C and Ni metal. Although NiMoOxCy was not detected at 873 K in the XRD of 50NiMo, TPC data shows that water and CO continued to evolve up to approximately 900 K (Fig. 3). From these results, it is evident that NiMoOxCy is present on the surface. Consequently, the increase in Ni content lowered the transformation temperature of NiMoOxCy as it produced ȕ-Mo2C and Ni metal. The TPR profiles of the 50NiMo catalyst that was carburized at 786 and 873 K before the reaction are shown in Fig. 5. Methane was created in the reaction between NiMoOxCy and H2 at 715 K and after carburizing as shown by the XRD of the catalyst that was carburized at 786 K (Fig. 4). For the catalyst that was carburized at 873 K, the methane formation peak attributed to NiMoOxCy decreased and a new methane formation peak was present at 920 K. These were deconvoluted into peaks of ȕ-Mo2C (860 K) and pyrolytic carbon (960 K) and the identification was based on a previous report [13]. By combining the results obtained at the start of the dry reforming reaction (Fig. 1), NiMoOxCy is shown to be a more active

Taro HIROSE et al. / Chinese Journal of Catalysis, 2011, 32: 771–776

715

CH4 intensity

860 960 (1)

(2)

400

600

800 1000 Temperature (K)

1200

Fig. 5. TPR profiles of the 50NiMo catalyst carburized at 786 (1) and 873 K (2) before the reaction.

species than ȕ-Mo2C during H2 formation in the dry reforming of CH4. High H2 formation and high CH4 conversion with a low CO2/CH4 ratio at the start of the reaction for the catalyst carburized at 786 K indicates that NiMoOxCy promotes the dry reforming reaction of CH4 as well as the direct decomposition reaction of CH4. Additionally, low H2 formation with a high CH4 conversion at the start of the reaction for the catalyst that was carburized at 873 K suggests that ȕ-Mo2C can promote a reverse water gas shift reaction (CO2 + H2 ĺ H2O + CO). The TPR profiles of the 50NiMo catalyst that was carburized at 786 K before the reaction and at 1 h after the reaction are shown in Fig. 6. After a 1 h reaction the peak at 715 K disappeared and the broad CH4 formation peak was present at 1039 K, which was deconvoluted into ȕ-Mo2C (860 K), pyrolytic carbon (960 K), Ș-Mo3C2 (1039 K), and graphitic carbon (1149 K) peaks. These identifications were also based on a previous report [13]. This result indicates that the rapid drop in H2 formation rate and CH4 conversion at the start of the reaction (Fig. 1) is related to the transformation of NiMoOxCy to ȕ-Mo2C and

Ș-Mo3C2. This proves that NiMoOxCy is more active than ȕ-Mo2C and Ș-Mo3C2 in the dry reforming of CH4. XRD patterns of the 50NiMo catalysts obtained after the reforming reaction are shown in Fig. 7. After the reaction, Ni metal was observed for each catalyst carburized from 773 to 873 K. Graphitic carbon was found to have a higher peak for the catalysts that were carburized at lower temperatures, e.g., the incomplete carburization at lower temperature promoted H2 formation together with carbon deposition during reforming. Moreover, MoO2 has very low activity during methane dry reforming [16–18]. The oxidized catalyst of the NiO-MgO solid solution showed no reforming activity during the autothermal CO2 reforming of methane whereas the catalyst that was reduced to Ni0 on the surface exhibited reforming activity [19]. Although MoO2 was not identified in the XRD of the 50NiMo catalyst that was carburized at 786 K (neither was NiO between 773 and 786 K, as shown in Fig. 4), the oxidized Mo and/or Ni remained on the catalyst that was carburized incompletely at 773–798 K. It gradually transformed into the more active carburized species or metallic Ni during the reforming reactions leading to a gradual increase in the H2 formation rate and in CH4 conversion (Fig. 1). Brungs et al. [17] reported that the main route for catalyst deactivation in the dry reforming of methane over supported molybdenum carbide was likely to be the oxidation of Mo2C to MoO2. However, Lacheen et al. [18] showed that CH4 conversion over MoCx-ZSM-5 does not decrease after 15 ks at a CO2/CH4 reactant ratio of 0.05 but it does decrease at ratios above 0.1 and this is caused by the transformation of MoCx to MoOx. Because the ratio used in our experiment was 0.033, which is lower than 0.05, such an oxidation does not occur during the CO2 reforming of methane. Graphite carbon K-Mo3C2

Ni-metal E-Mo2C

773 K

Intensity

CH4 intensity

786 K

(1)

810 K 860

960 1039

1149

823 K

(2)

400

798 K

873 K 600

800 1000 Temperature (K)

20

1200

Fig. 6. TPR profiles of the 50NiMo catalyst carburized at 786 K. (1) Before the reaction; (2) 1 h after the reaction.

30

40

50

60

2T/( o ) Fig. 7.

XRD patterns of the 50NiMo catalysts after the reaction and

carburized at different temperatures.

Taro HIROSE et al. / Chinese Journal of Catalysis, 2011, 32: 771–776

Graphitic carbon Ni-metal K-Mo3C2 E-Mo2C

Specific surface area (m2/g)

120

(1)

Intensity

(2)

(3)

After reaction for 6 h Before reaction

80 60 40 20 760

780

800 820 840 860 Carburizing temperature (K)

880

(4)

Fig. 9. Relationship between the surface area and the carburizing temperature of the 50NiMo catalysts.

(5)

catalyst that was carburized at 773 K accorded with that at 786 K. As a result, the high carburizing temperature for the low Ni/Mo ratio catalyst increased H2 formation and low amounts of water and carbon were formed. The BET surface area of the 50NiMo catalyst before and after the reforming reaction is shown in Fig. 9. The surface area decreased after the reaction because of the transformation of the oxide to a carbide or metal and because of catalyst sintering. In addition, the surface area roughly decreased with an increase in the carburizing temperature. This is one of the reasons for the H2 formation rate and the decrease in CH4 conversion with an increase in the carburizing temperature.

20 Fig. 8.

100

30

40 2T/( o )

50

60

XRD patterns of the NiMo catalysts after the reaction. (1)

25NiMo-798 K; (2) 25NiMo-813 K; (3) 50NiMo-786 K; (4) 50NiMo-798 K; (5) 75NiMo-773 and 786 K.

XRD patterns for the 25NiMo, 50NiMo, and 75NiMo catalysts were obtained after the reforming reaction and are shown in Fig. 8. Nickel and graphitic carbon were formed as the Ni/Mo ratio increased. This type of carbon formation is in agreement with the calculations of Sehested et al. [20] who found that carbon formation on a Mo2C catalyst requires more Gibbs-free energy than a nickel catalyst. This result means that a molybdenum carbide catalyst is more resistant to carbon formation than nickel catalysts. Additionally, an increase in the Ni/Mo ratio decreased the transformation temperature of NiMoOxCy to ȕ-Mo2C and Ni metal, as determined by the TPC measurement. Therefore, the high Ni/Mo ratio enhances the decomposition of the more active NiMoOxCy to the less active ȕ-Mo2C and Ni metal. Furthermore, the high Ni/Mo ratio reduced H2 formation because of an increase in water formation. Xiao et al. [21] reported that pretreating NiMoAl catalysts using carburization can increase catalyst stability while large amounts of Mo dopants have a damaging effect and this is probably because of carbon deposition caused by CH4 decomposition. They proposed that the formation of NiMoO4 is probably responsible for catalyst deactivation. In contrast, we found that NiMoO4 transformed into NiMoOxCy during carburizing and it was not identified before and after the reforming reaction because of the low CO2/CH4 ratio, which was much lower than the ratio of 1 that was the mentioned in the study. XRD patterns of the catalysts that formed after the reforming reaction and that were carburized at different temperatures are also shown in Fig. 8. The 25NiMo catalyst that was carburized at 798 K produced more graphitic carbon than the catalyst carburized at 813 K while the XRD pattern of the 75NiMo

3 Conclusions The CO2 reforming of methane at a low CO2/CH4 reactant ratio on nickel molybdenum carbide catalysts prepared at various carburizing temperatures and Ni/Mo ratios was investigated using XRD, TPC, and TPR. For the 50NiMo catalysts carburized at 810–873 K the H2 formation rate and CH4 conversion reached a steady state after rapidly dropping initially during the reaction while they gradually increased for the catalysts carburized at 773–798 K. At the same carburizing temperature, the H2 formation rate decreased and the CH4 conversion increased as the Ni/Mo ratio increased. NiMoOxCy was more active for H2 formation while ȕ-Mo2C and Ș-Mo3C2 were less active. The incomplete carburization of the Mo oxide species and/or Ni oxide species in the catalyst carburized at a lower temperature gradually led to their transformation into a more active carburized species or metallic Ni species among the studied reforming reactions leading to a gradual increase in the H2 formation rate and CH4 conversion. At the same carburizing temperature the H2 formation rate increased and CH4 conversion decreased as the Ni/Mo ratio decreased. The decrease in the Ni/Mo ratio retarded the decomposition of NiMoOxCy into ȕ-Mo2C and Ni metal. In addition, the high carburizing temperature used for the low Ni/Mo ratio catalyst

Taro HIROSE et al. / Chinese Journal of Catalysis, 2011, 32: 771–776

increased the H2 formation rate and low amounts of water and carbon were formed. Consequently, the 25NiMo catalyst that was carburized at 813 K gave the best H2 formation behavior and it deposited the lowest amount of carbon.

References 1 Ashcroft A T, Cheetham A K, Green M L H, Vernon P D F. Nature, 1991, 352: 225 2 Nagai M, Nakahira K, Ozawa Y, Namiki Y, Suzuki Y. Chem Eng Sci, 2007, 62: 4998 3 Chen X, Honda K, Zhang Z-G. Appl Catal A, 2005, 288: 86 4 San-José-Alonso D, Juan-Juan J, Illán-Gómez M J, RománMartínez M C. Appl Catal A, 2009, 371: 54 5 Roh H, Jung Y, Koo K Y, Jung U H, Seo Y, Yoon W L. Appl Catal A, 2009, 383: 156 6 Liu D P, Quek X Y, Cheo W N E, Lau R, Borgna A, Yang Y H. J Catal, 2009, 266: 380 7 Solymosi F, Cserényi J, Szöke A, Bánsági T, Oszkó A. J Catal, 1997, 165: 150

8 9 10 11 12 13 14 15 16 17 18 19 20 21

Oshikawa K, Nagai M, Omi S. J Phys Chem B, 2001, 105: 9124 Tominaga H, Nagai M. Appl Catal A, 2007, 328: 35 Oshikawa K, Nagai M, Omi S. Jpn J Appl Phys, 2000, 39: 4456 LaMont D C, Thomson W J. Appl Catal A, 2004, 274: 173 Nagai M, Kanesugi H. J Chem Eng Jpn, 2006, 39: 569 Izhar S, Kanesugi H, Tominaga H, Nagai M. Appl Catal A, 2007, 317: 82 Nagai M, Matsuda K. J Chem Eng Jpn, 2006, 39: 575 Shu Y, Ohnishi R, Ichikawa M. J Catal, 2002, 206: 134 York A P E, Claridge J B, Brungs A J, Tsang S C, Green M L H. Chem Commun, 1997: 39 Brungs A J, York A P E, Claridge J B, Márquez-Alvarez C, Green M L H. Catal Lett, 2000, 70: 117 Lacheen H S, Iglesia E. J Catal, 2005, 230: 173 Matsuo Y, Yoshinaga Y, Sekine Y, Tomishige K, Fujimoto K. Catal Today, 2000, 63: 439 Sehested J, Jacobsen C J H, Rokni S, Rostrup-Nielsen J R. J Catal, 2001, 201: 206 Xiao T, Suhartanto T, York A P E, Sloan J, Green M L H. Appl Catal A, 2003, 253: 225