Subscriber access provided by Kaohsiung Medical University
Article
Combining CO2 Reduction with Ethane Oxidative Dehydrogenation by Oxygen-modification of Molybdenum Carbide Siyu Yao, Binhang Yan, Zhao Jiang, Zongyuan Liu, Qiyuan Wu, Jihoon Lee, and Jingguang G Chen ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b00541 • Publication Date (Web): 02 May 2018 Downloaded from http://pubs.acs.org on May 2, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
Combining CO2 Reduction with Ethane Oxidative Dehydrogenation by Oxygen-modification of Molybdenum Carbide Siyu Yao 1, Binhang Yan 1, Zhao Jiang 2, 3, Zongyuan Liu 1, Qiyuan Wu 1, 4, Jihoon Lee 2, Jingguang G. Chen 1,2* 1. Chemistry Division, Brookhaven National Laboratory, Upton, NY, 11973 2. Department of Chemical Engineering, Columbia University, New York, NY, 10027 3. School of Chemical Engineering and Technology, Xi'an Jiaotong University, Xi’an, China, 710049 4. Department of Chemistry, SUNY Stony Brook, Stony Brook, NY, 11790 ABSTRACT: The surface properties that determine the selectivity of Mo2C catalysts in ethane oxidative dehydrogenation with CO2 as a soft oxidant were investigated using a combination of pulse experiments and in-situ spectroscopic methods. Oxygen modification was discovered to be crucial for inhibiting the cleavage of C-C bond in ethane and enhancing the production of ethylene. The addition of the Fe promoter accelerated the formation of surface oxygen species and stabilized them from reduction by ethane, leading to a shorter induction period, higher ethylene yield and improved stability.
Keywords: oxidative dehydrogenation, ethane, carbon dioxide, oxidation-modified molybdenum carbide, pulse reactor system
Introduction The catalytic conversion of light alkanes into alkenes via the oxidative1,2 dehydrogenation or other pathways is of increasing importance because of the growing demand for light olefins, which are the major building blocks in the petrochemical industry3,4. Currently, ethane (C2H6), an abundant component of shale gas (up to 10%), is typically converted into ethylene via thermal dehydrogenation at high temperatures (973-1123 K)5. The high energy consumption and severe catalyst deactivation caused by coke deposition involved in the process have motivated efforts in developing catalytic oxidative dehydrogenation (ODH) process that operates at lower temperatures and reduces rates of carbon deposition6. However, it is still a challenge to prevent the total oxidation of ethane to CO2 when molecular oxygen is utilized as an oxidant. The difficulty in heat regulation also limits the scale of industrial ethane ODH reactors1. Employing CO2 as a mild oxidant for the ODH process of ethane (C2H6 + CO2 = C2H4 + CO + H2O) has attracted recent interests, as it is able to reduce the working temperature and prevent the over oxidation of C2H65,7,8. Meanwhile, the ODH process also converts CO2 into CO that could be used as precursor for producing valuable chemicals, mitigating CO2 emission while simultaneously achieving ethane dehydrogenation 9. The critical factor that determines the applicability of ethane ODH with CO2 is the ethylene selectivity. The major side reactions include dry reforming reaction (C2H6 + 2CO2 = 4CO + 3H2) and the cracking of ethane to car-
bon and methane (C2H6 = 2C + 3H2, C2H6 + H2 = 2CH4). In order to develop efficient catalytic systems for ODH using CO2 as oxidant, the catalyst needs to be capable of activating CO2 to produce surface oxygen, cleaving C-H bonds of ethane, while retaining the C-C bond simultaneously. Controlling the subtle balance of activity of the three reactions remains to be a key challenge in catalyst development. Molybdenum carbides are among the possible potential candidates for ethane ODH with CO2, as it has been reported that Mo2C is a highly active catalyst for CO2 activation10,11. Meanwhile, as a potential alternative of platinumgroup metals, Mo2C was reported as an effective catalyst for alkane dehydrogenation12,13, demonstrating its capability to activate C-H bonds of hydrocarbons. Previous attempts14,15 have proven that oxide (SiO2 and Al2O3) supported Mo2C catalysts were able to selectively catalyze C2H6 oxidative dehydrogenation to ethylene. However, the catalysts showed relatively poor stability under reaction conditions. Although the oxygen modification effect has been discovered to be related with the preferential conversion of ethane to ethylene from theoretical15 and kinetic approaches16, the knowledge of structure and catalytic properties of active sites on Mo2C is still relatively limited. Understandings of the active species over Mo2C and the selective activation of ethane and CO2 are required for the rational design and promotion of carbidebased catalysts in pursuing the optimal performance of ethane ODH with CO2.
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Herein, we present detailed catalytic evaluation and structural characterization to investigate active sites over hexagonal β-Mo2C for ethane ODH with CO2. Based on the results of transient pulse experiments, the reactivity of Mo2C toward ethylene from C2H6 and CO2 oxidative dehydrogenation is controlled by the surface oxidation state of Mo. An oxygen-rich Mo2C surface is essential for the selective oxidative dehydrogenation of ethane and prevents side reactions of dry reforming and cracking. Ambient-pressure X-ray photoelectron spectroscopy (APXPS) is used to confirm that the active sites for C2H6 ODH is Mo (II/IV) oxides on the surface of Mo2C. Fe promoter is discovered to be able to accelerate the formation and stabilization of surface oxygen species, rendering a much shorter induction period and significantly higher ethylene selectivity and yield. In-situ X-ray diffraction and absorption techniques and density functional theory (DFT) calculations are also employed to further understand the active sites of the catalysts.
Experimental and Computational Methods Synthesis of Mo2C and Fe/Mo2C catalysts. The Mo2C catalyst was synthesized via a temperature program carburization method.17 Ammonium heptamolybdate (Sigma-Aldrich) powder was ground into fine powder (< 80 mesh) and calcined at 773 K for 4 h to obtain MoO3. Typically, 0.6 g MoO3 was placed into a quartz tube and carburized in the mixture of CH4/H2 (15/85, v/v, 50 ml/min). The sample was heated to 573 K at 5 K/min and held for 2 hours and then ramped to 973 K at a rate of 0.5 K/min and held at 973 K for another 2 hours. After cooling, the samples were passivated with 20 ml/min CO2 for 8 hrs. The Fe/Mo2C catalysts were synthesized by mixing the appropriate amount of iron nitrate into ammonium heptamolybdate (MA) aqueous solution (3 g in 3 ml deionized water). The generated suspension was stirred vigorously at 323 K until the solvent was completely evaporated. The generated precipitate was dried in the oven overnight at 393 K and then calcined at 773 K for two hours in order to obtain the precursor FeOx/MoO3, which was carburized using the same temperature program procedures described for Mo2C. Physical characterization. The ambient pressure X-ray photo-electron spectroscopy (AP-XPS) spectra were collected using a SPECS AP-XPS chamber equipped with a PHOIBOS 150 EP MCD-9 analyzer with an energy resolution of 0.4 eV. The C1s photoemission line with the surface carbon feature (284.5 eV) was used for the energy calibration. Before the measurements, powder catalysts of Mo2C and 1% Fe/Mo2C were activated in the reactor using the same procedures as the catalytic reaction. After cooling down, the samples were passivated in the flow of pure CO2 at room temperature for 4 hours. The passivated samples were then pressed into clean aluminum plates and used as XPS samples. Typically, the fresh sample was carburized in the AP-XPS chamber by 100 mTorr CH4/H2 mixture at 823 K (the upper limit of the instrument) for 2 hours. After carburization, the XPS spectra of activated sample was collected under 10-6 mTorr at 823 K. Then CO2
Page 2 of 9
with different pressure from 1 to 100 mTorr was introduced into the chamber subsequently. The spectra at Mo 3d, O 1s and C 1s regions were collected at 823 K after 30 min treatment with the same scan range and scan rate. The X-ray powder diffraction patterns of Mo2C and Fe/Mo2C were collected using an Ultima III X-ray powder diffractometer with Cu Kα1 radiation. The high resolution Scanning transmission electron microscope (STEM) images were collected using a JEM-2100F field emission electron microscope. The in-situ X-ray diffraction patterns of Mo2C and Fe/Mo2C under reaction condition were collected at 17 BM of Advanced photon source (APS) in Argonne National Laboratory with the incident X-ray wavelength of 0.24128 Å. The in-situ Clausen cell with quartz tube (1.1 mm OD and 0.9 mm ID) was used for the measurements. Before the reaction, 5 mg of catalyst was activated in a mixture of CH4 and H2 (15/85 v/v) at 873 K for 1 hour and pure H2 for 0.5 hour. The in-situ XRD measurements under reaction conditions (25% C2H6/ 25% CO2/ He, 20 ml/min gas flow rate) were performed at 873 K after 0.5 hour treatment. The data refinement was done using the GSAS-II packages. The in-situ X-ray absorption fine structure (XAFS) XAFS characterization was performed at the 2-2 beamline of Stanford Synchrotron Radiation Laboratory (SSRL). The Fe K edge (7112 eV) XAFS spectra were collected in the fluorescence mode using a 13-channel Ge detector. The Mo K edge (20000 eV) XAFS spectra were collected in the transmission mode using an ion chamber as detector. Home-made micro-channel in-situ cell using graphite paper as window was used for the measurements. Before the reaction, 0.2 g of catalyst was activated in a mixture of CH4 and H2 (15/85 v/v) at 873 K for 1 hour and pure H2 for 0.5 hour. The in-situ XAS measurements under reaction conditions (25% C2H6/ 25% CO2/ He, 20 ml/min gas flow rate) were performed at 873 K after 0.5 hour treatment. The data pretreatment and spectra fitting were performed using the Ifeffit packages.18 Catalytic evaluation. For each experiment, 50 mg catalyst was loaded in a 1/8 inch quartz tube and activated in the flow of CH4/H2 mixture (15/85, v/v, 50 ml/min) in a fixed bed flow reactor for 1.5 hour at 873 K. The CH4 flow was then switched off in order to remove surface carbon deposition. After the activation, the reactant (C2H6/CO2/N2=1:1:2, WHSV= 24,000 ml/(gcat*h)) was introduced into the reactor. The products were analyzed by an Agilent 7890B gas chromatograph (GC) equipped with TCD and FID detectors. The conversion of the reactant is defined as: − Conversion a = ∗ 100% The selectivity of product CO is defined as: Selectivity CO 0.5 ∗ $ − %2$ ∑ ()*+, -, ℎ/0)+()*+, 1)+02(34 + 0.5 ∗ $ − %2$ ∗ 100%
=
ACS Paragon Plus Environment
Page 3 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis Selectivity hydrocarbon product a Where Eadsorbate + surface is the total energy of the adsorbate ∑ ()*+, -, 1)+02(3 together with the surface, Eadsorbate is the total energy of = the free adsorbate in the gas phase, and Esurface is the total ∑ ()*+, -, ℎ/0)+()*+, 1)+02(34 + 0.5 ∗ $ − %2$ energy of the surface. ∗ 100% The yield of product a is defined as: Yield a = Selectivity a ∗ conversion ethane Pulse reactor experiments. The pulse reactor experiments were carried out in a home-designed system with a micro-fixed bed reactor and GC, as shown in scheme S1. In these experiments, 20 mg catalyst was first activated in the reactor and treated under certain pretreatment program (Table S2), followed by changing the system to purging mode using Ar, followed by injecting pulse of reactant gas. The loop of the pulse system was kept at 110 °C with a volume of 1.00 ml. The analysis of the products was achieved using the same method as the steady-state experiments mentioned above. The amounts of carbon deposition and oxygen consumption were calculated based on mass balance. Density functional theory (DFT) calculations. DFT calculations were carried out using the Vienna ab initio simulation package (VASP)19,20. The electronic-ion interaction was modeled by the projector augmented wave (PAW) method21. The Perdew-Wang-91 (PW91) functional22 with the generalized gradient approximation was employed to deal with the electronic exchange and correlation. The kinetic wave cutoff energy was set at 400 eV to describe the electronic wave functions. The Brillouinzone integration was sampled using a of 4×4×1 Monkhorst-Pack k-points grid with a Gaussian smearing of 0.1 eV. Geometry optimization was converged until the forces acting on the atoms were smaller than 0.01 eV/Å, whereas the energy threshold-defining self-consistency of the electron density was set to 10−5 eV. The hexagonal Mo2C phase was used with an eclipsed configuration as the unit cell23,24. The calculated lattice parameter of the cell is 2a = 6.079 Å, 2b = 6.073 Å and c = 4.722 Å, in good agreement with the experimental values (a = b = 3.011 Å c = 4.771 Å25. Among all the surfaces of hexagonal Mo2C, it was reported that the (101) surface with Mo/C = 1/1 ratio and surface energy of 2.19 J/m2 was most stable. Therefore, the (101) surface was selected in this study, which was modeled by a four bilayer 3×3 surface slab (a bilayer contains a unit of one Mo layer and one C layer). A vacuum layer of 12 Å was added perpendicular to the slab to avoid artificial interactions between the slab and its periodic images. During optimization, the upper two layers together with the adsorbed species were allowed to relax, whereas the bottom two layers were fixed. The binding energies (BE) for all intermediates are calculated as follows: BE = >?@ABC?DABE?F + >?@ABC? − >ABE?F
Results and Discussion Correlating surface state of Mo2C with catalytic performance. Previous studies have demonstrated that the surface of molybdenum carbide is under complicated dynamic state under different hydrogen, oxygen and carbon chemical potentials in the atmosphere26,27. The irreversible oxidation or carbon deposition of the Mo2C surface with reactants under working condition of methane dry reforming (CH4 + CO2 = CO + H2), alkane dehydrogenation and isomerization are believed to be the main reasons for the poor stability of the Mo2C catalysts28,29. As the catalytic activity of Mo2C is most likely determined by its surface state, it is necessary to evaluate the influence of pretreatment atmosphere on the catalytic performance of ethane ODH with CO2 to identify the optimal surface state for ethylene production. Similar to dry reforming reaction, three different surface states may exist on the Mo2C catalyst, namely O-rich surface from CO2 oxidation, C-rich surface from alkane decomposition followed by carbon deposition, and Mo-rich surface generated from O and C removal via reduction/methanation by H2 generated from alkane decomposition. As a comparison, the activated Mo2C was treated with CO2, pure C2H6 or H2 for 30 min before catalytic performance evaluation. As shown in Figure 1, H2 pretreated Mo2C exhibited highest ethane conversion (~14%), while both CO2 and C2H6 treatment showed negative effect on C2H6 conversion. The C2H6 treated Mo2C was barely able to catalyze the ethane conversion, which was probably due to the block of active sites by carbon deposition. Judging on the ethylene selectivity, the O-rich Mo2C surface from CO2 treatment exhibited significant advantage in producing ethylene, with over 70% initial C2H4 selectivity comparing to the 10% on H2 treated catalyst. With increasing time on stream, all three catalysts deactivated and the ethylene selectivity increased simultaneously. Consequently, the Mo-rich Mo2C catalyst was gradually transformed into an ODH preferential catalyst, with the yield of C2H4 reached a maximum of 5.2% at 8.5 hour and the ethylene formation rate reached 4000 nmol/(gcat*s), 20 times higher than the literature report14, indicating that Mo2C was a potentially high performance ethylene production catalyst via ethane ODH after proper treatment. On the contrary, the increase of SC2H4 was unable to compensate the dramatically decreasing C2H6 conversion on O-rich Mo2C, causing the monotonic decrease of C2H4 yield. However, these results suggest that the modification of Mo2C by oxygen could effectively inhibit the C-C bond dissociation and promote the formation of ethylene.
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 1. Catalytic performance of C2H6 ODH reaction on Mo2C with different pretreatments. (A) ethane conversion; (B) ethylene selectivity and (C) ethylene yield.
Pulse reactor experiments. Although the flow reactor evaluation of the three catalysts revealed the importance of oxygen, the quantitative relationship of the intrinsic activity and selectivity of Mo2C with specific surface state was not clear. As shown in Figure 1, the performance of Mo2C never reached steady state, which suggested that the surface state of the catalyst underwent continuous
change under reaction conditions. Pulsing tests were designed to probe the catalytic behavior of the surface state with trace amount of reactants and to further understand the structural change of the Mo2C surface under the redox atmosphere (the details of pulse experiments were listed in Table S2). The reaction of CO2 with Mo-rich and C-rich surfaces of Mo2C catalysts are shown in Figure 2a.
Figure 2. (A) Reaction of CO2 pulses with Mo-rich and C-rich Mo2C; (B) Reaction of C2H6 pulses with Mo-rich and O-rich Mo2C; (C) The determination of surface O* coverage by CO2 pulses titration; (D) The relationship of ethane conversion, ethylene selectivity and yield with various surface oxygen coverage.
Based on the pulse experiments, CO2 was able to transform into CO via two pathways: (1) direct dissociation into CO and surface O* (CO2 + *= CO + O*)10,11 and (2) reverse-Boudouard reaction with carbon in the catalyst and generate 2 CO (CO2 + C = 2CO)30,31. Based on mass balance calculations, the Mo-rich surface was able to effectively catalyze the CO2 dissociation reaction and gen-
erate large amount of surface O* in the first pulse (over 70% selectivity through CO2 dissociation). The CO2 dissociation activity decreased with increasing O* coverage, and the Mo-rich surface was gradually converted to Orich surface that was unable to further activate CO2. On the contrary, the major CO2 conversion route on the Crich Mo2C catalyst was the reverse-Boudouard reaction, as
ACS Paragon Plus Environment
Page 4 of 9
Page 5 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis the ratio of CO production and C consumption was around 2 in the first 4 pulses. Only after the reactive surface carbon was removed by the CO2, the generation of O* could be observed. In other word, on C-rich surface, CO2 could be effectively activated to form active O* species that was important for oxidative dehydrogenation. The reaction results of C2H6 with Mo-rich surface and Orich surface are listed in Figure 2b. On the Mo-rich surface, the complete ethane decomposition to form carbon deposition and ethane hydrogenolysis to methane were the main surface reactions (55% and 40% selectivity, respectively, in the first pulse). The pulse experiments also suggested that the accumulated carbon tended to block the active site on Mo2C, reducing the ability of C-C and CH cleavage of Mo2C significantly in the subsequent pulses. In comparison, although the initial C2H6 conversion was relatively low over CO2 treated Mo2C, the dehydrogenation product ethylene was observed in the first 3 pulses of ethane injections (over 54% ethane convert into ethylene in the first injection). The selectivity of methane in these pulses was suppressed at the same time, indicating that oxygen covered Mo2C could effectively inhibit the C-C dissociation of ethane. With the gradual removal of the surface oxygen as the number of pulses increased, the conversion of ethane increased significantly and the yield of ethylene decreased. Therefore, it is highly probable that the active site of partial dehydrogenation of ethane was related with the surface O* species from CO2 dissociation. Furthermore, to reveal the relationship between ethane ODH catalytic performance and surface oxygen coverage, the density of O* adsorption sites on the β-Mo2C was determined using the CO2 titration method32,33. As shown in Figure 2c, the oxygen deposition curve dropped quickly with the increasing concentration of surface O* species. After 21 pulses the oxygen deposition on the surface was below 15nmol/g per pulse and the deposition rate was relatively stable, indicating that further CO2 injection mainly resulted in the slow oxidation of the bulk phase of Mo2C. The O/Mo ratio was estimated to be 1.5 based on the BET surface area and the utilization of the Mo2C(101) surface as the preferential exposing face23.
Figure 3. Products distribution of C2H6 pulses reduction on CO2 pretreated Mo2C and Fe/Mo2C catalysts.
The catalytic performance of ethane ODH with CO2 over Mo2C was measured at different surface O* coverages (Table S3). Without O* modification, the C2H4 selectivity was only 5% and the majority of C2H6 was converted via the ethane hydrolysis and dry reforming side reactions. With increasing O* coverage, the selectivity of ethylene exhibited a dramatic increase, indicating that high O* coverage was able to effectively suppress the side reactions. At 60% coverage, the selectivity and yield of ethylene reached 43% and 2.8%, respectively. This phenomenon also explains the activity and selectivity evolution of H2 pretreated Mo2C in ethane ODH with CO2, which is related with the gradual surface oxidation of Mo2C under reaction condition. Therefore, in order to enhance the ethylene production, the control of proper O* surface coverage is of great importance. Accelerating the oxidation of the catalyst surface should reduce the induction period of ethane ODH reaction over the Mo2C catalyst. Enhancing ethylene production via Fe modification. In order to stabilize the surface oxygen species, Fe as a promoter was introduced on the Mo2C catalyst. A comparative pulse study of the reduction of oxygen-covered Mo2C and Fe/Mo2C by ethane was performed (Table S4). As shown in Figure 3, the removal of oxygen species on the Fe-Mo2C catalyst by ethane pulses was much slower than bare Mo2C. The formation of ethylene lasted for six pulses on Fe/Mo2C compared with four pulses on Mo2C. In the meantime, the total amount of ethylene generated on Fe/Mo2C was twice that of bare Mo2C. In addition, the formation of methane over Fe/Mo2C was almost completely inhibited. Therefore, the Fe promoter could effectively stabilize the oxygen on the Mo2C surface, enhance the ethylene production and reduce the formation of methane byproduct.
Figure 4. The influence of Fe loading on the catalytic performance of Fe/Mo2C catalysts. (A) C2H6 conversion; (B)
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ethylene selectivity (C) ethylene yield and (D) the summary of the influence of loading. The catalytic performance of a series of Fe-modified Mo2C catalysts with different Fe loadings was evaluated under the same reaction conditions as Mo2C. The 0.5 wt% Fe/Mo2C catalyst showed a slight improvement on the initial CO2 and C2H6 conversion, and the initial C2H4 selectivity and yield were slightly smaller than Mo2C. When the Fe loading increased above 1 wt%, all tested samples exhibited preferential selectivity to ethylene. The higher Fe loading renders higher initial ethylene selectivity and shorter induction period. However, the conversion of ethane decreased with increasing Fe loading, resulting in an optimal C2H4 yield over the 1% Fe/Mo2C catalyst. The C2H4 yield retained at above 5.5% for over four hours and the formation rate of ethylene was around 4,000 nmol/(gcat*s) during this period, approximately 20 times higher than the previous reported Mo2C/SiO2 catalyst14 (Figure 4).
respondingly, the intensities of the O 1s spectra also suggested that the oxygen species gradually accumulated on the surface in CO2 atmosphere and Fe/Mo2C was oxidized much faster than Mo2C. The binding energy of the major peaks in the O 1s region in both catalysts at 530.1 eV, consistent with the formation of metal oxides.
Detailed characterization was performed to understand the structural properties of Fe over Mo2C (Figure S2). The powder XRD profiles revealed that Fe formed highly dispersed morphology over Mo2C, with no signals related with Fe being observed until the loading increased to 10%. The in-situ Fe K edge XANES spectra suggested that, even after the carburization activation, Fe promoters were not fully reduced34 (Table S5). Under reaction conditions, Fe exhibited pre-edge features similar to those of Fe3O4. Detailed EXAFS fitting (Table S6) on the radial distribution profiles of 1% Fe/Mo2C under reaction conditions suggested that the first nearest neighbors of Fe centers were oxygen with an average C.N.Fe-O (coordination number) near 4 at 1.95 Å, indicating that Fe formed oxide domains on the Mo2C support. The remote Fe-Mo coordination at 2.86 Å was related with the Fe-O-Mo species with a coordination number around 1.2. Only small amount of Fe-Fe coordination (about 0.6) was observed near 2.43 Å, indicating that the average diameter of the Fe domain was relatively small and suggesting a fine dispersion of loaded Fe. HR-STEM elemental mapping in Figure S2d showed that Fe formed homogeneous dispersion on Mo2C with no large particles observed, in agreement with the EXAFS fittings. The structural characterization showed that Fe formed stable small oxide clusters over the Mo2C surface, which enhanced the oxidation of Mo2C by CO2. The AP-XPS method was utilized to identify the surface chemical state of Mo2C and Fe/Mo2C to further verify the importance of surface oxygen species after CO2 treatment (Figure 5). The Mo 3d spectrum of the activated catalyst suggested that there were three different forms of Mo on the surface, including Mo carbide at 228.2eV, Mo oxycarbide at 228.8 eV and Mo(IV) at 229.3 eV. Both Mo2C and Fe/Mo2C catalysts shared similar Mo oxidation state after carburization. With increasing CO2 pressure, the intensity of Mo carbide decreased and transformed into Mo(II) and Mo(IV) subsequently35. Fe-modified Mo2C exhibited higher inclination for oxidation at the same CO2 partial pressure. At 100 mTorr, part of Mo on the surface of Fe/Mo2C was even oxidized to Mo(V) and Mo(VI). Cor-
Figure 5. The ambient pressure X-ray photoelectron spectroscopy (AP-XPS) characterization of Mo2C and Fe/Mo2C catalysts under various CO2 pressure (1-100 mTorr). (A) Mo 3d region; (B) C 1s region and (C) O 1s region spectra.
Furthermore, the in-situ XRD patterns of Mo2C and 1% Fe/Mo2C demonstrated the formation of MoOx species not only formed over the catalyst surface but also in the bulk phase (Figure S3). The bulk phase oxidation of Mo2C was probably the reason for the deactivation of the Mo2C based catalysts. Detailed Rietveld refinement results revealed that the oxidation rate of the support in 1%
ACS Paragon Plus Environment
Page 6 of 9
Page 7 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis Fe/Mo2C was much faster than pure Mo2C, which could be used to explain the catalytic behavior of Fe promoted Mo2C catalysts.
of ethane conversion when the surface is covered by oxygen species. The binding energy of oxygen at the Fe-Mo interface is also calculated (Figure S6), revealing that the binding of the O* atoms at the Fe-Mo interface is 0.53 eV higher than on bare Mo2C. This result support the experimental observation that the addition of Fe promoters would enhance oxide formation than unmodified Mo2C under reaction conditions, leading to the accelerated formation of the necessary MoOx species for ethylene production.
Conclusions
Figure 6. The DFT calculation of the Gibbs free energy change (∆G) of three different ethane conversion routes under different surface oxygen coverage. C2H6* = 2C* + 6H* (blue bar); C2H6* = C2H4* + 2H* (red bar); C2H6* = 2CH3* (yellow bar).
In summary, we have demonstrated that the MoOx modified Mo2C surface is related to the ethylene formation in ethane ODH with CO2. The existence of oxygen is able to prevent the C-C bond cleavage side reactions and reduce the formation of coke. Fe promoter is discovered to be able to accelerate the formation of the MoOx layer, which significantly shorten the induction period and stabilize surface MoOx. The rate for selective ethylene production on a 1wt% Fe/Mo2C is 20 times than previously reported Mo2C catalysts. The discovery of the promoting role of surface oxides should provide guidance on developing other carbide-based catalysts for ethane ODH with CO2.
ASSOCIATED CONTENT DFT results. DFT calculations were performed to verify the role of surface MoOx species in the ODH reaction36-38 (Table S7). The reaction Gibbs free energy change of complete ethane decomposition (C2H6* = 2C* + 6H*), direct ethane dehydrogenation to ethylene (C2H6* = C2H4* + 2H*) and the C-C dissociation to surface methyl (C2H6* = 2CH3*, which could considered as the precursor of methane) were calculated on Mo2C(101) and O* modified Mo2C(101) surfaces (Figure 6). On the clean Mo2C (101) face23, the complete decomposition of ethane into C* and H* is the thermodynamically favored route, with the ΔG of -3.18 eV. The C-C bond cleavage to -CH3* exhibits a ΔG of -2.16 eV, while the partial dehydrogenation route is the least favorable route, with a ΔG value of -2.02 eV. When the surface is modified by 2 oxygen atoms (at the Mo-Mo bridge site, namely 2/9 oxygen coverage), the ΔG of ethylene formation is almost unchanged (-1.98 eV). On the contrary, the ΔG values of both C* and -CH3* formation are significantly reduced (to -2.35 and -1.42 eV respectively), making the dehydrogenation route the second favorable pathway for ethane conversion. When the oxygen coverage further increases to 4/9, the difference in ΔG between decomposition and partial hydrogenation pathways becomes even smaller (only 0.22 eV), indicating that ethylene formation on oxygen modified Mo2C surface is more preferred. Meanwhile, the relative ΔG values of ethylene and C-C cleavage pathways also suggest that the selectivities toward ethylene and methane are mutually exclusive, consistent with the experimental results observed by the pulse experiments (Figure 2). The DFT calculations also show an overall trend that the modification of the Mo2C surface by oxygen would decrease the ΔG values of all reaction routes, which could explain the drop
Supporting Information The supporting information is available free of charge via the Internet at http://pubs.acs.org.
The design of pulse reactor; the BET surface area; the experimental details and quantification of pulse experiments; In-situ XAFS/XRD characterization and STEM images of Mo2C and 1% Fe/Mo2C catalysts; preferential adsorption geometry and energy of specific conversion route collected by DFT calculations.
AUTHOR INFORMATION Corresponding Author *
[email protected]
ACKNOWLEDGMENT The work is supported by the US Department of Energy Office of Science under contract number DE-SC0012704. We also acknowledge support of this work under contract DEAC02-98CH10886 with the U.S. Department of Energy and supported by the Brookhaven National Laboratory Directed Research and Development (LDRD) Project No. 16-045. The in-situ XAFS measurements were performed at the 2-2 beamline at the Stanford Synchrotron Radiation Lightsource (SSRL) at SLAC National Accelerator Laboratory (DE-AC0276SF00515). The in-situ XRD measurements were performed at the 17 BM beamline at the Advanced Photon Source (APS). The DFT calculations were performed at the Center for Functional Nanomaterial, a user facility at Brookhaven National Library which is sponsored by the U.S. DOE Office of Science (DE-AC02-05CH11231).
REFERENCES
ACS Paragon Plus Environment
ACS Catalysis 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1. Cavani, F.; Ballarini, N.; Cericola, A., Oxidative dehydrogenation of ethane and propane: How far from commercial implementation? Catal. Today 2007, 127, 113-131. 2. Gärtner, C. A.; van Veen, A. C.; Lercher, J. A., Oxidative Dehydrogenation of Ethane: Common Principles and Mechanistic Aspects. ChemCatChem 2013, 5, 3196-3217. 3. Ittel, S. D.; Johnson, L. K.; Brookhart, M., Late-Metal Catalysts for Ethylene Homo- and Copolymerization. Chem. Rev. 2000, 100, 1169-1204. 4. Chen, J. Q.; Bozzano, A.; Glover, B.; Fuglerud, T.; Kvisle, S., Recent advancements in ethylene and propylene production using the UOP/Hydro MTO process. Catal. Today 2005, 106, 103-107. 5. Mimura, N.; Takahara, I.; Inaba, M.; Okamoto, M.; Murata, K., High-performance Cr/H-ZSM-5 catalysts for oxidative dehydrogenation of ethane to ethylene with CO2 as an oxidant. Catal. Commun. 2002, 3, 257-262. 6. Bhasin, M. M.; McCain, J. H.; Vora, B. V.; Imai, T.; Pujadó, P. R., Dehydrogenation and oxydehydrogenation of paraffins to olefins. Appl. Catal., A 2001, 221, 397-419. 7. Nakagawa, K.; Okamura, M.; Ikenaga, N.; Suzuki, T.; Nakagawa, K.; Okamura, M.; Suzuki, T.; Kobayashi, T.; Kobayashi, T., Dehydrogenation of ethane over gallium oxide in the presence of carbon dioxide. Chem. Commun. 1998, 9, 1025-1026. 8. Nakagawa, K.; Kajita, C.; Okumura, K.; Ikenaga, N.-o.; Nishitani-Gamo, M.; Ando, T.; Kobayashi, T.; Suzuki, T., Role of Carbon Dioxide in the Dehydrogenation of Ethane over GalliumLoaded Catalysts. J. Catal. 2001, 203, 87-93. 9. Hansen, J.; Sato, M.; Ruedy, R.; Lo, K.; Lea, D. W.; Medina-Elizade, M., Global temperature change. Proc Natl Acad Sci USA 2006, 103, 14288-93. 10. Porosoff, M. D.; Yang, X.; Boscoboinik, J. A.; Chen, J. G., Molybdenum carbide as alternative catalysts to precious metals for highly selective reduction of CO2 to CO. Angew. Chem. Int. Ed. Engl. 2014, 53, 6705-9. 11. Zhang, X.; Zhu, X.; Lin, L.; Yao, S.; Zhang, M.; Liu, X.; Wang, X.; Li, Y.-W.; Shi, C.; Ma, D., Highly Dispersed Copper over β-Mo2C as an Efficient and Stable Catalyst for the Reverse Water Gas Shift (RWGS) Reaction. ACS Catal. 2016, 7, 912-918. 12. Wang, S.; Zhu, Z. H., Catalytic Conversion of Alkanes to Olefins by Carbon Dioxide Oxidative Dehydrogenation A Review. Energy Fuels 2004, 18, 1126-1139. 13. Neylon, M. K.; Choi, S.; Kwon, H.; Curry, K. E.; Thompson, L. T., Catalytic properties of early transition metal nitrides and carbides: n-butane hydrogenolysis, dehydrogenation and isomerization. Appl. Catal., A 1999, 183, 253-263. 14. Solymosi, F.; Németh, R., The oxidative dehydrogenation of ethane with CO2 over Mo2C/SiO2 catalyst. Catal. Lett. 1999, 62, 197-200. 15. Porosoff, M. D.; Myint, M. N.; Kattel, S.; Xie, Z.; Gomez, E.; Liu, P.; Chen, J. G., Identifying Different Types of Catalysts for CO2 Reduction by Ethane through Dry Reforming and Oxidative Dehydrogenation. Angew. Chem. Int. Ed. Engl. 2015, 54, 15501-5. 16. Sullivan, M. M.; Bhan, A., Effects of oxygen coverage on rates and selectivity of propane-CO2 reactions on molybdenum carbide. J. Catal. 2018, 357, 195-205. 17. Lee, J., Molybdenum carbide catalysts I. Synthesis of unsupported powders. J. Catal. 1987, 106, 125-133. 18. Ravel, B.; Newville, M., ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT. J Synchrotron Radiat 2005, 12, 537-41. 19. Kresse, G.; Hafner, J., Ab initio molecular dynamics for liquid metals. Phys. Rev. B: Condens. Matter 1993, 47, 558-561. 20. Kresse, G.; Furthmüller, J., Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 1996, 6, 15-50.
21. Kresse, G.; Joubert, D., From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B: Condens. Matter 1999, 59, 1758-1775. 22. Perdew, J. P.; Burke, K.; Ernzerhof, M., Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865-3868. 23. Wang, T.; Liu, X.; Wang, S.; Huo, C.; Li, Y.-W.; Wang, J.; Jiao, H., Stability of β-Mo2C Facets from ab Initio Atomistic Thermodynamics. J. Phys. Chem. C 2011, 115, 22360-22368. 24. Luo, Q.; Wang, T.; Walther, G.; Beller, M.; Jiao, H., Molybdenum carbide catalysed hydrogen production from formic acid – A density functional theory study. J. Power Sources 2014, 246, 548-555. 25. Fries, R.; Kempter, C. P., 195. Dimolybdenum carbide. Anal. Chem. 1960, 32, 1898-1898. 26. Chen, J. G., Carbide and Nitride Overlayers on Early Transition Metal Surfaces: Preparation, Characterization, and Reactivities. Chem. Rev. 1996, 96, 1477-1498. 27. Hwu, H. H.; Chen, J. G., Surface chemistry of transition metal carbides. Chem. Rev. 2005, 105, 185-212. 28. Ledoux, M. J.; Huu, C. P.; Guille, J.; Dunlop, H., Compared activities of platinum and high specific surface area Mo2C and WC catalysts for reforming reactions I. Catalyst activation and stabilization: Reaction of n-hexane. J. Catal. 1992, 134, 383398. 29. Solymosi, F.; Szoke, A.; Cserenyi, J., Conversion of methane to benzene over Mo2C and Mo2C/ZSM-5 catalysts. Catal. Lett. 1996, 39, 157-161. 30. Darujati, A., Oxidation stability of Mo2C catalysts under fuel reforming conditions. Appl. Catal., A 2003, 253, 397-407. 31. Solymosi, F.; Oszkó, A.; Bánsági, T.; Tolmacsov, P., Adsorption and Reaction of CO2 on Mo2C Catalyst. J. Phys. Chem. B 2002, 106, 9613-9618. 32. Chen, C.-J.; Bhan, A., Mo2C Modification by CO2, H2O, and O2: Effects of Oxygen Content and Oxygen Source on Rates and Selectivity of m-Cresol Hydrodeoxygenation. ACS Catal. 2017, 7, 1113-1122. 33. Lee, W.-S.; Kumar, A.; Wang, Z.; Bhan, A., Chemical Titration and Transient Kinetic Studies of Site Requirements in Mo2C-Catalyzed Vapor Phase Anisole Hydrodeoxygenation. ACS Catal. 2015, 5, 4104-4114. 34. Hensen, E.; Zhu, Q.; Liu, P.; Chao, K.; Vansanten, R., On the role of aluminum in the selective oxidation of benzene to phenol by nitrous oxide over iron-containing MFI zeolites: an in situ Fe XANES study. J. Catal. 2004, 226, 466-470. 35. Oshikawa, K.; Nagai, M.; Omi, S., Characterization of Molybdenum Carbides for Methane Reforming by TPR, XRD, and XPS. J. Phys. Chem. B 2001, 105, 9124-9131. 36. Kresse, G.; Furthmüller, J., Efficient iterative schemes forab initiototal-energy calculations using a plane-wave basis set. Phys. Rev. B: Condens. Matter 1996, 54, 11169-11186. 37. Xing B.; Pang X.; Wang G., C–H bond activation of methane on clean and oxygen pre-covered metals: A systematic theoretical study. J. Catal. 2011, 282, 74-82. 38. Hibbitts D.; Neurock M., Promotional effects of chemisorbed oxygen and hydroxide in the activation of C–H and O–H bonds over transition metal surfaces. Surf. Sci. 2016, 650, 210-220
ACS Paragon Plus Environment
Page 8 of 9
Page 9 of 9 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Catalysis
ACS Paragon Plus Environment
