Hydrogenation Conversion of Phenanthrene over Dispersed Mo-based Catalysts

Transcription

1 Scientific Research China Petroleum Processing and Petrochemical Technology 2015, Vol. 17, No. 3, pp 7-14 September 30, 2015 Hydrogenation Conversion of Phenanthrene over Dispersed Mo-based Catalysts Hu Yiwen; Da Zhijian; Wang Zijun (SINOPEC Research Institute of Petroleum Processing, Beijing ) Abstract: With oil-soluble molybdenum compound and sublimed sulfur serving as raw materials, two dispersed Mo-based catalysts were prepared, characterized and then applied to the hydrogenation conversion of phenanthrene. The test results showed that under the conditions specified by this study, the catalyst prepared in a higher sulfiding atmosphere was more catalytically active due to its higher content of MoS 2 and stronger intrinsic catalytic activity of MoS 2 unit, which demonstrated that the sulfiding atmosphere for the preparation of catalysts not only could influence the yield of MoS 2 but also the structure of MoS 2. The analysis on the selectivity of octahydrophenanthrene isomers revealed that the catalyst prepared in a lower sulfiding atmosphere had a relatively higher catalytic selectivity to the hydrogenation of outer aromatic ring and the structure of catalysts could be modified under the specific reaction conditions. Moreover, the selectivity between the isomers of as-octahydrophenanthrene at different reaction time and temperature was analyzed and, based on the results, a hydrogenation mechanism over dispersed Mo-based catalysts was suggested, with monatomic hydrogen transfer and catalytic surface desorption of the half-addition intermediates functioning as the key points. In addition, it is concluded that the catalyst prepared in a lower sulfiding atmosphere was more capable of adsorption than the other one. Key words: dispersed Mo-based catalysts; phenanthrene conversion; MoS 2 structure; product selectivity; hydrogenation mechanism 1 Introduction During the past decades, faced with the challenge of depleting petroleum resources and increasing consumption of transportation fuels, there had been an urgent demand for effective heavy-oil upgrading technologies, one of which was the study on hydrogenation of the non-substituted aromatic hydrocarbons into lighter constituents [1]. Recently, the slurry-phase hydrocracking process has been attracting more attention because of its special superiority in treating inferior quality feedstocks [2]. Likewise, it is very important for this developing process to hydrogenate the non-substituted aromatic hydrocarbons as far as possible. So, it is quite meaningful to investigate the catalytic hydrogenation in slurry phase where the dispersed micron/nanoscale particulate catalyst is currently being used [3]. As a typical non-substituted aromatic hydrocarbon in heavy oil, hydrogenation of phenanthrene (PH) has been extensively studied over different catalysts with its corresponding reaction networks proposed [4-7]. And the two proven to be most reasonable pathways are presented in Scheme 1 (the abbreviations therein are adopted in the following texts). The only difference between Network 1 and Network 2 was whether the path from DHP to as- OHP was really available. In this paper, two dispersed Mo-based solid particles, a typical catalyst popularly used in the slurry-phase hydrocracking process, were prepared by controlling the sulfiding atmosphere. Then, the hydrogenation conversion of PH in an autoclave reactor was carried out to evaluate the catalytic performance of the prepared catalysts. In order to simulate the operating conditions of slurryphase hydrocracking process, the catalyst particles have to be maintained in suspension during the reaction which means that a certain kind of solvent is required to keep the reaction mixture in liquid phase at the designed temperature and pressure. Therefore, decahydronaphthalene (DHN) was added as the reaction solvent. Received date: ; Accepted date: Corresponding Author: Wang Zijun, Telephone: ; wangzj.ripp@sinopec.com. 7

2 Scheme 1 Reaction networks of phenanthrene hydrogenation 2 Experimental 2.1 Materials All chemicals were used as received without further purification. Molybdenum 2-ethylhexanoate (containing 15.4% of Mo) and phenanthrene (with a purity of 98%) were purchased from the Alfa Aesar Co. Sublimed sulfur (chemically pure) and decahydronaphthalene (DHN, GR) were obtained from the Sinopharm Chemical Reagent Co., Ltd. The FCC cycle oil and carbon black were supplied by the Sinopec Beijing Yanshan Company. 2.2 Preparation of catalysts The dispersed Mo-based catalysts were prepared as follows: 50 g of molybdenum 2-ethylhexanoate, a certain amount of sulfur, 100 g of carbon black and 200 g of FCC cycle oil were added into a 1-L autoclave reactor. The reactor was purged with H 2 thrice before the final pressurization with H 2 to 7.0 MPa inside the reactor. Then the temperature-programmed heating from room temperature to 623 K at a temperature increase rate of around 6 K/min was started in a continuously stirred reactor. After 1 h of reaction at 623 K followed by cooling down, the mixture in the reactor was collected and filtered to obtain products with the filter residue washed by cyclohexane and dried at 393 K in vacuum overnight. Two kinds of catalysts were prepared by altering the addition of sulfur amount, where one sample with 3.85 g of sulfur added (S/Mo molar ratio=1.5) was labeled as DMoS1.5 and the other one with 6.42 g of sulfur added (S/Mo molar ratio=2.5) was labeled as DMoS2.5. Then the two catalyst samples were characterized by X-ray fluorescence spectrometry (XRF), X-ray photoelectron spectroscopy (XPS), X-ray diffractometry (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM), respectively. 2.3 Experimental details for hydrogenation conversion of PH In a typical run, g of PH dissolved in g of DHN (containing 5% of PH) and a certain amount of catalyst (with a Mo/PH molar ratio equating to 0.25) were added to a 100-mL autoclave reactor. The reactor was purged with hydrogen stream thrice before the final pressurization with H 2 to 9.0 MPa. Then the temperatureprogrammed heating from room temperature to the designed temperature at a temperature increase rate of 6 K/min 8

3 was started for the continuously stirred reactor. After the termination of reaction at the designed temperature and cooling down, the gas and liquid in the reactor were collected and analyzed by GC, respectively. 3 Results and Discussion 3.1 Characterization results of catalysts The composition analysis results of the prepared catalysts are listed in Table 1. The S/Mo ratio determined by XRF was quite similar to that of those elements initially added. It should be noted that there was obvious discrepancy between the results of bulk analysis (XRF) and surface analysis (XPS). For example, the C/Mo ratio determined by XRF was much less than that determined by XPS, while the case was opposite for the analysis of O/Mo ratio. The valence distribution of Mo and S showed that the MoS 2 amount in the sample DMoS2.5 was more than that in the sample DMoS1.5. Figure 1 shows the XRD patterns of two catalysts. It has previously been demonstrated that the peaks of the (100), (103) and (110) planes were attributed to the crystalline MoS 2 phase while the peak of the (002) plane reflected the stacking height of MoS 2 layers [8]. According to the peak intensity in Figure 1, on one hand, both catalysts contained the crystalline MoS 2 phase with the content of which in DMoS2.5 higher; and on the other hand, the MoS 2 layers in both catalysts only lightly stacked with Table 1 Composition analysis of the prepared catalysts Catalyst XRF (molar ratio) XPS (molar ratio) Mo valence distribution, % S valence distribution, % Mo, at.% S/Mo C/Mo O/Mo Mo, at.% S/Mo C/Mo O/Mo 1) Other DMoS DMoS ) The elements with trace amount were neglected. the stacking height in DMoS1.5 much smaller. The SEM and TEM images of two catalysts are summarized in Figure 2. The SEM images show that the plateshaped particles with different diameters (100 nm 1 μm) were randomly stacked and there was practically no difference between the two catalysts. Then, the TEM images clearly indicate that the MoS 2 slabs in DMoS1.5 were mainly comprised of single-layer species, about 10 nm in length, while the MoS 2 slabs in DMoS2.5 were mainly composed of two- or three-layer species, 10 nm 20 nm in length. Also, the above results measured by TEM were coincident with those of XRD analysis. Based on the obtained information, a structure mode for the prepared catalysts was proposed in Scheme 2. Figure 1 XRD patterns of two catalyst samples Scheme 2 The proposed model of prepared catalysts 9

4 Figure 2 SEM and TEM images of two catalyst samples 3.2 Effect of reaction time on hydrogenation reaction The molar fraction of main products versus reaction time over two catalysts is plotted in Figure 3. It is obvious that as the reaction time increased, the molar fraction of PH in the mixture gradually decreased coupled with the OHP increase. The profiles of DHP and THP revealed their kinetic stability as evidenced by the primary hydrogenation products of PH. In order to determine the reaction network of PH under these catalytic conditions, the molar fraction of as-ohp versus time is plotted in Figure 4. Since the formation rate of as-ohp almost did not change with time, the content of its reactants should slightly vary with an increasing time. Since the change curves in Figure 3 for DHP were sharp, DHP was not the reactant of as-ohp. Therefore, the path from DHP to as-ohp of Network 2 in Scheme 1 could be excluded and the hydrogenation reaction of PH at these catalytic conditions followed the pathway of Network 1. In comparison with Figure 3(a), Figure 3(b) indicates that the concentration of PH, DHP and THP was lower and OHP was higher, denoting that PH was hydrogenated deeper over DMoS2.5 than DMoS1.5. That is to say, DMoS2.5 showed a better catalytic hydrogenation activ- Figure 3 Effect of time on PH hydrogenation catalyzed by two catalysts PH; DHP; THP; OHP 10

5 Figure 4 Effect of time on the molar fraction of as-ohp over two catalysts DMoS2.5; DMoS1.5 ity than DMoS1.5. One clear reason for the better activity of DMoS2.5 was that DMoS2.5 had more active sites because of its relatively higher content of MoS 2 (as depicted in Table 1). However, the hydrogen-content increase per unit of MoS 2 showed by Figure 5 demonstrated that the intrinsic catalytic activity of MoS 2 in DMoS2.5 was also higher than that of DMoS1.5. Summarily, it was concluded that when the catalysts were prepared, the amount of added sulfur not only could influence their content of MoS 2 but also the structure of MoS 2. control to thermodynamic control in the conversion-temperature profile. In Figure 6, the content of PH and DHP both decreased with the increase in temperature, while the case was opposite for THP and OHP. On the other hand, the change curve was smooth for PH and THP while it was sharp for DHP and OHP. As for the whole conversion of PH, the reaction was still under the control of kinetics until 698 K although some of its steps had already moved to the thermodynamic control area like the transition from PH to DHP. Upon comparing the hydrogenation depth between Figure 6(a) and Figure 6(b), it is found out that the catalytic hydrogenation activity of DMoS2.5 was always higher than DMoS1.5 in the range of designed temperatures. However, the difference became quite small at 698 K. Figure 5 Effect of time on the hydrogen-content increase per unit of MoS 2 for the two catalysts DMoS2.5; DMoS Effect of temperature on hydrogenation reaction The molar fraction of main products versus temperature over two catalysts is plotted in Figure 6. Theoretically speaking, higher temperature can accelerate hydrogenation reaction but is disadvantageous to its thermodynamics which leads to the existence of knee point from kinetic Figure 6 Effect of temperature on PH hydrogenation catalyzed by PH; DHP; THP; OHP 3.4 The selectivity for the isomers of OHP The OHP mentioned above was a mixture of as-ohp and s-ohp and therefore the selectivity of s-ohp to as-ohp (s s-ohp/as-ohp for short) was especially concerned. The molar ratio of s-ohp to as-ohp versus time and temperature 11

6 the structure of active sites of catalyst [10]. So, two conclusions can be drawn from Figure 7, namely: (i) DMoS1.5 had a relatively higher catalytic selectivity to the hydrogenation of outer aromatic ring, and (ii) the structure of catalysts was somehow modified under the reaction conditions and the modification degree varied with temperature as well as the catalyst itself. 3.5 The selectivity for the isomers of as-ohp Just like DHN, there is also geometric isomerism in as- OHP. The isomers of as-ohp, cis-as-ohp and trans-as- OHP, were studied with the help of advanced GC-MS analysis. The molar ratio of cis-as-ohp to trans-as-ohp (s cis/trans for short) versus time and temperature over two catalysts is plotted in Figure 8(a) and Figure 8(b), respectively. Figure 7 Molar ratio of s-ohp to as-ohp versus (a) time; and (b) temperature over two catalysts (at 9.0 MPa of initial H 2 pressure, 623K for the upper graph and 2h for the lower one) DMoS2.5; DMoS1.5 over two catalysts is plotted in Figure 7(a) and Figure 7(b), respectively. It is interesting to see that for most samples, the content of s-ohp was higher than as-ohp and this difference was more obvious for DMoS1.5 in all cases. From the viewpoint of reaction time, the value of s s-ohp/as-ohp ratio in both catalytic systems varied slightly with 2.4 for DMoS1.5 and 1.1 for DMoS2.5. This observed selectivity was generally explained as the steric hindrance of middle ring in as-ohp which resulted in higher hydrogenation activation energy of as-ohp [7, 9]. Based on this explanation and the Arrhenius formula, the s s-ohp/as-ohp ratio would decrease as the temperature increased. Unexpectedly, the experimental results in Figure 7(b) were just on the contrary which implied that the pre-exponential factor or activation energy for the formation reaction of s-ohp/as-ohp also changed with the temperature. It is well known that the pre-exponential factor is related to the amount of active sites of catalyst while the activation energy is related to Figure 8 Molar ratio of cis-as-ohp to trans-as-ohp versus (a) time and (b) temperature over two catalysts (at 9.0MPa of initial H 2 pressure, 623K for upper graph and 2h for the lower one) DMoS2.5; DMoS1.5 With a more severe inner steric hindrance, cis-as-ohp was less energetically stable than trans-as-ohp. Nevertheless, it can be seen from Figure 8 that in most cases, 12

7 the value of s cis/trans was beyond 1.0 which implied that cisas-ohp kinetically predominated during the formation of as-ohp. In Figure 8(a), for both catalysts, the s cis/trans ratio decreased as time extended and the only possibility for this experimental phenomenon was the inter-conversion from cis-as-ohp to trans-as-ohp. In order to find out whether the inter-conversion was a catalytic process or only a pure heating reaction, the liquid collected from sample A in Figure 8(a) was further kept for 2 h under the same condition like A except the addition of DMoS2.5 to obtain the sample B. Since the s cis/trans ratio in sample B was much higher than sample C, the above-mentioned interconversion was definitely related to the catalyst. In Figure 8(b), the s cis/trans ratio decreased as temperature increased in both catalytic systems, which confirmed the view that the s cis/trans ratio was produced by the difference in activation energy between the isomers. Based on the information mentioned above, a hydrogenation mechanism over the dispersed Mo-based catalysts was proposed in Scheme 3, with its main points stated as follows: (i) the monatomic hydrogen transfer from catalyst surface to adsorbed reactant (Step A); (ii) the desorption of the half-addition intermediate from catalyst surface (Step D). If Step D did not occur, cis-as-ohp would be formed via Path I and if it did, the half-addition intermediate would have the chance to yield trans-as-ohp via Path II where the second active hydrogen was received from the opposite side (relative to the first one) of the ring plane. It is reasonable to state that Step D played a key role in the formation of s cis/trans. If the more difficult Step D took place, a higher s cis/trans ratio would be possibly obtained. Thus, the promotion to Step D by temperature was the main reason for the results referred to in Figure 8(b). Additionally, a higher s cis/trans ratio was always maintained in the DMoS1.5 catalytic system, which implied that it was more difficult for Step D to occur over DMoS1.5. This opinion was verified by the results depicted in Figure 8(b), indicating that there was a greater decrease in amplitude of s cis/trans ratio for DMoS1.5 as the temperature increased. Apparently, the adsorption ability of catalyst could directly influence the Step D and also the s cis/trans ratio. Furthermore, it could be inferred that DMoS1.5 could adsorb reactants stronger than DMoS2.5 which was actually confirmed by the blank experiments where less amount of PH was detected in the liquid sampled from the Scheme 3 The proposed mechanism of hydrogenation reactions over dispersed Mo-based catalysts DMoS1.5 system as compared with DMoS2.5 (in 2 h of stirring, in N 2 atmosphere, and at 473 K). 4 Conclusions (1) Under the conditions specified in this study on the two dispersed Mo-based catalysts used in the hydrogenation conversion of phenanthrene, the one prepared in higher sulfiding atmosphere showed better catalytic hydrogenation activity. (2) The sulfiding atmosphere adopted during the preparation of catalysts not only could influence the yield of MoS 2 but also the structure of MoS 2. The catalyst prepared in higher sulfiding atmosphere has higher content of MoS 2 and the MoS 2 unit therein exhibited higher catalytic hydrogenation activity. (3) The analysis on the selectivity of octahydrophenanthrene isomers revealed that the catalyst prepared in lower sulfiding atmosphere had a relatively higher catalytic selectivity to the hydrogenation of outer aromatic ring and a different degree of modification to the structure of catalysts occurring at different reaction temperatures. (4) According to the effect of reaction time and temperature on the selectivity between the isomers of as-octahydrophenanthrene, a hydrogenation mechanism for the dispersed Mo-based catalysts was proposed with monatomic hydrogen transfer and catalytic surface desorption of the half-addition intermediates serving as the main points. The catalyst prepared in higher sulfiding atmosphere would adsorb reactants much weaker as compared with the case implemented in lower sulfiding atmosphere. Acknowledgements: The authors are grateful to the financial support from the National Basic Research Program of China (Grant 2012CB224801). 13

8 References [1] Castaneda L C, Munoz J D. Combined process schemes for upgrading of heavy petroleum[j]. Fuel, 2012, 100(2): [2] Zhang S, Liu D, Deng W, et al. A review of slurry-phase hydrocracking heavy oil technology[j]. Energy Fuels, 2007, 21(6): [3] Liu D, Kong X, Li M, et al. Study on a water-soluble catalyst for slurry-phase hydrocracking of an atmospheric residue[j]. Energy Fuels, 2009, 23(2): [4] Korre S C, Klein M T, Quann R J. Polynuclear aromatic hydrocarbons hydrogenation: 1. Experimental reaction pathways and kinetics[j]. Ind Eng Chem Res, 1995, 34(1): [5] Qian W, Yoda Y, Hirai Y, et al. Hydrodesulfurization of dibenzothiophene and hydrogenation of phenanthrene on alumina-supported Pt and Pd catalysts[j]. Appl Catal, A, 1999, 184(1): [6] Beltramone A R, Resasco D E, Alvarez W E, et al. Simultaneous hydrogenation of multiring aromatic compounds over NiMo catalyst[j]. Ind Eng Chem Res, 2008, 47(19): [7] Qi S C, Zhang L, Wei X Y, et al. Deep hydrogenation of coal tar over a Ni/ZSM-5 catalyst[j]. RSC Advances 2014, 4 (33): [8] Delarosa M P, Texier S, Berhault G, et al. Structural studies of catalytically stabilized model and industrial-supported hydrodesulfurization catalysts[j]. J Catal, 2004, 225(2): [9] Liu C, Rong Z, Sun Z, et al. Quenched skeletal Ni as the effective catalyst for selective partial hydrogenation of polycyclic aromatic hydrocarbons[j]. RSC Advances, 2013, 3(46): [10] Dutta R P, Schobert H H. Hydrogenation/dehydrogenation of polycyclic aromatic hydrocarbons using ammonium tetrathiomolybdate as catalyst precursor[j]. Catal Today. 1996, 31(1): The Technology 30 kt/a Commercial Unit for Ethyleneamine Production via MEA Method Passed Appraisal The technology 30 kt/a unit for manufacture of ethyleneamine via MEA method developed by the CAS Dalian Institute of Chemical Physics with independent intellectual property rights has passed the achievement appraisal organized by the China Association of Petroleum & Chemical Industries. This MEA method for production of ethyleneamines consists of a whole set of proprietary technologies, which have created a series of technical inventions and innovative achievements in the area of development of catalysts, scaleup of commercial catalysts manufacture, R&D work on the trickle bed reactor internals, investigation of products separation process, design of a 10-kt/a-class production unit, and operation of the commercial unit. In August 2011 the first 10 kt/a unit for manufacture of ethyleneamines using the MEA method was put on stream in collaboration with the Shandong Union Chemical Company Limited (CUCCL). On the basis of the successful operation of the first unit, CUCCL then launched the construction of a new 30 kt/a unit for manufacture of ethyleneamines, which has become operational in March 2015 to achieve a more than 45% of MEA conversion, an over 80% of EDA selectivity, an over 99.7% of EDA purity, a 99.3% of DETA purity and a 99.9% of PIP purity. The Characterization Committee has recognized that this technology is quite innovative with its overall technical level reaching the internationally advanced level, among which the reaction pressure and the EDA selectivity have commanded a leading position in the international market. The expert group has unanimously agreed on the appraisal minutes and proposed to enhance the dissemination and application of this technology. 14