Al2O3 Interfaces in WaterâGas Shift and

Subscriber access provided by READING UNIV

Article 2

3

Confronting the Role of Ni/AlO Interfaces in Water-Gas Shift and Dry Reforming of Methane Lucas Foppa, Tigran Margossian, Sung Min Kim, Christoph Mueller, Christophe Copéret, Kim Larmier, and Aleix Comas-Vives J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08984 • Publication Date (Web): 27 Oct 2017 Downloaded from http://pubs.acs.org on October 27, 2017

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 free 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 accessible to all readers and 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.

Journal of the American Chemical Society 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 15

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

Journal of the American Chemical Society

Confronting the Role of Ni/Al2O3 Interfaces in Water-Gas Shift and Dry Reforming of Methane a

a

b

b

a

a,

Lucas Foppa, Tigran Margossian, Sung Min Kim, Christoph Müller, Christophe Copéret, Kim Larmier * and a, Aleix Comas-Vives * a

Department of Chemistry and Applied Biosciences, ETH Zurich, Vladimir Prelog Weg 1-5, CH-8093 Zurich, Switzerland. b Laboratory of Energy Science and Engineering, Department of Mechanical and Process Engineering, ETH Zurich, Leonhardstrasse 21, 8092 Zurich, Switzerland. KEYWORDS Nickel, Alumina, Metal-Support Interface, Water-Gas Shift Reaction, Dry Reforming of Methane, Density Functional Theory, Microkinetic Modeling. ABSTRACT: Transition metal nanoparticles (NPs) are typically supported on oxides to ensure their stability, which may result in modification of the original NP catalyst reactivity. In a number of cases, this is related to the formation of NP/support interface sites that play a role in catalysis. The metal/support interface effect verified experimentally is commonly ascribed to stronger reactants adsorption or their facile activation on such sites compared to bare NPs, as indicated by DFT-derived Potential Energy Surfaces (PESs). However, the relevance of specific reaction elementary steps to the overall reaction rate depends on the preferred reaction pathways at reaction conditions, which usually cannot be inferred based solely on PES. Hereby, we use a multi-scale (DFT/microkinetic) modeling approach and experiments to investigate the reactivity of the Ni/Al2O3 interface towards Water-Gas Shift (WGS) and Dry Reforming of Methane (DRM), two key industrial reactions with common elementary steps and intermediates, but held at significantly different temperatures: 300 vs. 650 °C, respectively. Our model shows that despite the more energetically favorable reaction pathways provided by the Ni/Al2O3 interface, such sites may or may not impact the overall reaction rate depending on reaction conditions: the metal/support interface provides the active site for WGS reaction, acting as a reservoir for oxygenated species, while all Ni surface atoms are active for DRM. This is in contrast to what PESs alone indicate. The different active site requirement for WGS and DRM is confirmed by the experimental evaluation of the activity of a series of Al2O3-supported Ni NP catalysts with different NP sizes (2-16 nm) towards both reactions.

1. Introduction Metal nanoparticles (NPs) supported on oxides are among the most common class of industrial catalysts. The reactivity of these systems is affected by the choice of support (e.g. SiO2, ZrO2, TiO2, MgO, CeO2, Al2O3) for a broad range of reactions such as Water-Gas Shift (WGS),1-5 Dry (CO2) Reforming of Methane (DRM),6-9 CO oxidation10-12 and methanol synthesis,13 to name but a few. In most of the cases, support effects on the activity of the catalyst cannot be explained by simply summing the contribution of each component (NP + support). The formation of different types of active sites at the metal/support interface is one of the ways through which the support is believed to influence the reactivity. For instance, the amount of sites in the NP/support boundary correlates with the reaction rates for CO oxidation or WGS2,12 reactions carried out on series of CeO2, Al2O3 and TiO2supported metal NPs with different sizes, indicating that the metal/oxide interface is the active site for these reac-

tions. However, how such interfaces influence the reaction rates at a molecular level remains ill understood. A common approach to obtain an atomistic description of the metal/support interfaces consists in using Density Functional Theory (DFT) calculations to evaluate reaction Potential Energy Surfaces (PESs) on models of oxidesupported metal NPs (Figure 1a).3,4,10,14-16 The interface can in principle modify the energy profiles of catalytic processes in two ways (Figure 1b): (i) providing adsorption sites where relevant intermediates are more or less stabilized compared to the bare NP, modifying the overall reaction rate by a thermodynamic effect17 and (ii) providing less energy-demanding transition states (TSs) for relevant bond cleavage/forming elementary reaction steps (said rate-determining, RDSs) compared to the NP, owing rate enhancement by a kinetic effect.18 Based on PESs, metal/support interfaces were shown to strongly adsorb CO19,20 or activate H2, CH421 and CO215 molecules, presumably making the overall process easier than on bare metal sites.

ACS Paragon Plus Environment


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

Nevertheless, it is not straightforward to understand whether the effect of the interface sites on the PESs actually translates into an effect on the overall activity of the catalyst, leading to incomplete, if not misleading interpretations for the role of NP/support interface sites. There are two main reasons for that: (i) The relevant intermediates and TSs during catalysis, i.e. those whose stability affect the overall reaction rate, depend on the relative population of surface species and on the preferred reaction routes, which usually cannot be inferred based

Figure 1. (a) Representation of a supported NP catalyst.15 (b) PES for an arbitrary reaction on the metal NP (blue) and possible effects of NP/support interface in the reactivity (red). solely on DFT-derived PESs because they are sensitive to reaction conditions (temperature, pressure and gas-phase chemical potential). In particular, the availability of the sites for key reaction steps is difficult to assess when they may suffer from poisoning or inhibition during catalysis. (ii) Even if PESs are usually drawn as a linear succession of elementary steps, in most cases the various intermediates are part of several competing intricate pathways within very complex reaction networks. This makes the determination of preferred reaction routes during catalysis very difficult from the PESs alone. The use of DFT-driven microkinetic modeling applied to surface reactions has recently emerged as an approach to alleviate this caveat by taking into account the population of surface species and the complex reaction networks under reaction conditions.22-27 However, only few studies explicitly address the effect of metal/oxide interface sites in the reactivity of supported metal catalysts so far,1,3,4 especially considering interface and metal sites simultaneously. Herein, we question the role of metal/support interfaces versus metal sites at the NP surface by examining at the same time two different and yet closely related prototypical reactions, namely WGS (Eq. 1) and DRM (Eq. 2). The

Page 2 of 15

WGS is fundamental for H2 production28 and DRM is an alternative route to Steam Reforming of Methane (SRM) converting CO2 into value-added products, also used to tune the CO/H2 ratio of reforming products.29 CO(g)+H2O(g)CO2(g)+H2(g)

∆ °=-41.1kJ.mol-1

CH4(g)+CO2(g)2H2(g)+2CO(g) ∆ ° = 247.4 kJ.mol-1

(1) (2)

Both WGS and DRM are efficiently catalyzed by Ni NPs supported on Al2O3 and involve several common elementary steps and intermediates (Scheme 1). Furthermore, both reactions have been proposed to benefit from the NP/support interface,1-4,6-9

Scheme 1. Proposed mechanisms for WGS and DRM reactions. The common intermediates CO2*, CO* and O* are highlighted in yellow. which was ascribed to the strong CO binding19,20 or easy CO2 activation15 on the Ni/Al2O3 interface sites, as pointed out by DFT-derived PESs. However, because of very diverse thermodynamics, these reactions are performed at different temperatures: 300 °C and ≥ 600 °C for WGS and DRM, respectively. Under such different experimental conditions, which also include different feed compositions, it is not clear how the stronger binding to CO or the facile CO2 activation will affect the overall reactivity, since the relevant intermediates and RDS may not be the same. In order to address the Ni/Al2O3 reactivity on WGS and DRM, we use a multi-scale modeling approach combining DFT-calculations and microkinetic modeling, and use experimental kinetic evaluation to probe the predictions of the model. The DFT calculations are performed on a model for a nanometric Ni NP supported on γ-Al2O3, including both interface and bare NP (metal) types of sites. The DFT-calculated kinetic parameters (adsorption and rate constants) are then used to construct a microkinetic model, which takes into account the effect of reaction conditions (temperature and chemical potential of the 2


Page 3 of 15

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


reactants) on the reactivity and enables the identification of relevant intermediates and reaction pathways during catalysis through a Degree of Rate Control (DRC) analysis.17,18,30 We show that the role of the Ni/Al2O3 interface is not only related to the easy bond forming/breaking steps as expected from the analysis of DFT-based PESs, but also to the population of intermediates taking part in the RDSs. In particular, the availability of oxygen-containing species (O*, OH*) in the interface is found as a key factor controlling the reactivity, especially when the Ni NP is completely covered with CO*, which is the case for WGS but not for DRM. For this reason, the Ni/Al2O3 interface plays a key role in WGS, but not in DRM. The conclusions derived from our microkinetic model are fully supported by the experimental evaluation of WGS and DRM rate dependence on the particle size on a series of Al2O3supported Ni NPs. We show here how the combination of the molecular-level understanding provided by PESs with a surface species population analysis obtained from experiment-validated microkinetic modeling provides a proper understanding on the reactivity of metal/oxide interfaces.

2. Computational Methods Periodic DFT calculations were carried out with Vienna Ab Initio Simulation Package (VASP) code,31-33 using a planewave basis set with pseudopotentials (PAW method) and the Perdew-Wang (PW91) exchange-correlation functional.34 The Climbing Image Nudge Elastic Band (CI-NEB) method35 was used to locate TSs. The entropy of adsorbed species at finite temperatures was evaluated within the harmonic approximation, all degrees of freedom being treated as vibrational. The supported Ni NP model is described in detail in section 4.1. Kinetic modeling simulations were performed using the Chemkin® software.36 A plug-flow reactor model was used to calculate reaction rates, population of surface species during catalysis and concentration profiles along the reactor bed. We considered two kinetically different types of elementary reactions: non-activated adsorption/desorption reactions (treated within the Hertz-Knudsen model)37 and activated surface reactions, whose rate constants were calculated from Transition State Theory through the Eyring equation.38 We did not take into account lateral interactions between adsorbates. The DRC analysis17,18,39 was used to identify RDSs. Full computational details are available in the ESI.

3. Experimental Procedures A series of four Al2O3-supported Ni NP catalysts has been prepared using procedures described elsewhere, using either nickel nitrate or [Ni(μ2OCHO)(OCHO)(TMEDA)(μ2-H2O)] (TMEDA = tetramethyleethylenediamine) as a nickel precursor.40 The samples were characterized by elemental analysis (Inductively Coupled Plasma - Optical Emission Spectroscopy, Pascher Laboratory), Scannning Transmission Electronic

Microscopy (STEM) (High Angle Annular Dark Field (HAADF STEM) on a Hitachi HD2700CS microscope or bright field TEM imaging on a Philips CM12 microscope), H2 chemisorption at 25 °C (BelMax apparatus, BelJapan). The catalytic activity of the samples towards WGS and DRM was measured in a fixed-bed quartz flow reactor at ambient pressure (10 mm internal diameter, catalyst bed length of about 7 mm for 200 mg of catalyst, see details in ESI).We verified that negligible mass transfer limitations occur under our conditions (see calculations in ESI). Hence, we assume that the observed rates are limited by the chemical kinetics only. WGS was carried out at 300 °C, while DRM was carried out at 650 °C. The conversion was kept in the same range for all catalysts by varying the space velocity (12-16 % CO conversion range for WGS and 29-32 % CH4 conversion range for DRM, see Table S2). The rates are expressed with respect to the consumption of carbon monoxide (WGS) or methane (DRM) in mol molNi-1 s-1, respectively. In the case of WGS, steady-state conditions are reached while DRM undergoes deactivation with time on stream (see ESI).40 In the latter case, we report the measured activity after 0.5 h on stream (initial activity), before significant deactivation occurs.

4. Results 4.1. Evaluation of PES by DFT Calculations We used a model composed of a Ni55 NP (ca. 1 nm diameter) supported on a γ-alumina (110) fully dehydroxylated surface, shown in Figure 2. The shape of the NP was found after an extensive analysis of different Ni55 NPs supported on γ-Al2O3, described in details elsewhere.15 It should be noted that the size of the Ni NP model used here is much closer to technical Ni catalysts (containing at least a few nanometers diameter) than commonly used Al2O3-supported Nin cluster models (n0) for a given i elementary step indicates that increasing its rate results in overall reaction

rate enhancement (the elementary step is then said RDS), whereas a zero value (XRC,i>0) indicates non-RDS steps, i.e. steps whose energetics is not relevant for the overall reaction rate. Negative DRC values (XRC,i>0) are also possible for elementary steps that may inhibit the overall reaction.17 Table 2. Elementary reaction steps included in kinetic model for Al2O3-supported Ni NPs. Label

Reaction

A1 A2 A3 A4

CO + *NP ⇌ CO*NP CO2 + *NP ⇌ CO2*NP H2O + *NP ⇌ H2O*NP CO + *IN ⇌ CO*IN

A5 A6

CO2 + *IN ⇌ CO2*IN H2O + *IN ⇌ H2O*IN

R1 R2 R3 R4

OH*NP + H*NP ⇌ H2O*NP + *NP O*NP + H*NP ⇌ OH*NP + *NP CO2*NP + *NP ⇌ CO*NP + O*NP H2 + 2 *NP ⇌ 2 H*NP

R5 R6 R7

CO*NP + *NP ⇌ C*NP + O*NP HCO*NP + *NP ⇌ CO*NP + H*NP CH*NP + O*NP ⇌ HCO*NP + *NP

R8 R9 R10

CH*NP + *NP ⇌ C*NP + H*NP CH2*NP + *NP ⇌ CH*NP + H*NP CH3*NP + *NP ⇌ CH2*NP + H*NP

R11 R12

CH4 + 2*NP ⇌ CH3*NP + H*NP CH4+ *NP + O*NP ⇌ CH3*NP + OH*NP

R13 R14 R15 R16

OH*IN + H*NP ⇌ H2O*IN + *NP O*IN + H*NP ⇌ OH*IN + *NP H2O + O*IN + *NP ⇌ OH*NP + OH*IN CO2*IN + *NP ⇌ CO*NP + O*IN

R17 R18 R19

CO2*IN + H*NP ⇌ CO*NP + OH*IN CO*IN + *NP ⇌ O*IN + C*NP HCO*IN + *NP ⇌ CO*IN + H*NP

R20 R21 R22

CH*NP + O*IN ⇌ HCO*IN + *NP CH4(g) + *IN + *NP ⇌ CH3*NP + H*IN CH4(g) + *NP + O*IN ⇌ CH3*NP + COOH* 2*IN + H*NP ⇌ HCOO*IN + *NP

R23 R24 R25

HCOO*IN + H*NP ⇌ H2COO*IN + * + *NP ⇌ *IN + O*NP O*IN

R26 R27

C*IN + *NP ⇌ *IN + C*NP H*IN + *NP ⇌ *IN + H*NP

In order to probe the role of interface sites in Al2O3supported Ni NP catalysts reactivity towards WGS and DRM, we performed two different simulations for each reaction, using always the same set of elementary reaction steps -only temperature, feed composition and proportion 7



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

of NP/IN sites were changed from one simulation to another. For the first simulation, only NP sites were included. For the second one, 90% and 10% of NP and IN sites were consider, respectively, while keeping the same total number of sites constant. The reaction network for each simulation with net reaction rates for the most relevant elementary steps is represented in Scheme 2. The complete set of rates is available in ESI Table S7. WGS on Ni NP sites. For the simulation with NP sites only (Scheme 2a.1), the reaction proceeds via water adsorption (A3) and dissociation (R1) to form adsorbed OH*NP and H*NP followed by OH*NP dissociation (R2), generating O*NP and H*NP. The O*NP then reacts with CO*NP to form CO2*NP (R3). The WGS overall reaction rate, taken as the rate of CO2 production at 10% CO conversion, is equal to 1.95×10-15 mol.cm-2.s-1. A negligible amount of CO methanation products (CH4) is observed under these conditions, in agreement with experiments. During the WGS reaction on Ni NP, the surface is covered with CO*NP with a coverage value of ca. 9.99 ×10-1 monolayer (ML). Other minor surface species are OH*NP (2.09×10-4 ML) and C*NP (1.74×10-4 ML). According to the DRC analysis, OH*NP dissociation (R2, XRC,R2 = 0.15) is the main rate-determining surface reaction. The overall reaction rate is also sensible to water dissociation (R1, XRC,R1 = 0.04) to a minor extent. WGS on NP and IN sites. The analysis of the individual elementary steps rates for the simulation including both NP and IN sites (Scheme 2a.2) shows that in this case the main reaction channel (94% of the total rate of CO2 production) is associated to IN sites. Even if water activation occurs at an equivalent extent on NP or IN sites (R1 and R13, respectively), ca. 82% of the OH* is dissociated by a disproportionation reaction involving OH*NP and OH*IN species on IN sites (R15), generating O*IN and H2O(g) in an quasi-equilibrated reaction step (net rate equals to zero in Scheme 2). The remaining OH*IN is converted to O*IN by its direct cleavage on the IN sites (R14). The CO2 formation occurs almost exclusively (94%) by the reaction of CO*NP with O*IN species on IN sites (R16). The carboxyl pathway (R17, not shown in Scheme 2) at the IN site has a negligible contribution to the total rate (2.00×10-25 mol.cm-2.s-1), probably due to the high activation enthalpy for that transformation (282 kJ.mol-1). The rest of the conversion (6%) comes from the pathway which takes place exclusively on NP sites, which is the same one identified in the previous simulation with NP sites only. The Turnover Frequency (TOF), i.e. the rate per number of sites, is 130 times higher for IN compared to NP, showing that the formers are much more active than the latters. Regarding the total CO2 production via WGS, it is equal

Page 8 of 15

to 6.25×10-15 mol.cm-2.s-1, i.e. more than 3 times higher compared to the simulation on the bare Ni NP (1.95×10-15 mol.cm-2.s-1). For this simulation, the Ni NP surface is covered with CO*NP (9.99×10-1 ML), and the interface contains almost exclusively OH*IN species (9.97×10-1 ML). From the DRC analysis, the reaction channel via IN sites has the OH*IN dissociation reaction as the most rate-determining step (R14, XRC,R14 = 0.04). For the reaction path via NP sites, the most RDS is the water dissociation (R1, XRC,R1 = 0.19). Therefore, the inclusion of IN sites to the model in addition to NP sites shifts the WGS reaction bottleneck (OH* dissociation) from pathways associated to the Ni NP to those occurring on the Ni/Al2O3 interface, boosting the reaction rate. This effect can be related to the more facile water dissociation on the Ni/Al2O3 interface compared to the NP (activation enthalpy of 49 vs. 24 kJ.mol-1 for NP and IN sites, respectively), but most importantly to the high population of OH*IN species at the interface. This is because the availability of OH* species is a crucial factor affecting the rate of the most RDS of the reaction network (O-H bond breaking in adsorbed hydroxyls). The OH*IN coverage (9.97×10-1 ML) is at least four orders of magnitude higher than that of OH*NP on the Ni NP, which is highly covered with CO*NP at WGS reaction conditions. Additionally, OH*IN species are transformed to O* by the OH* disproportionation reaction on the interface (R15). This is a bimolecular reaction involving the participation of two surface species in adjacent sites. Indeed, R15 displays the most important contribution to the formation of O* (Scheme 2a.2). It should be noted that our model does not consider lateral interactions among co-adsorbed species and thereby assumes adsorption energy values independent of coverage. This approximation likely results in over-estimation of CO*NP coverage during WGS. However, the repulsive interactions among co-adsorbates lead to a reduction of the adsorption strengths which scales similarly for the main intermediates: CO, O and OH* (see ESI). Therefore, we do not expect changes in the reaction mechanism when co-adsorbed interactions are taken into account.58 DRM on NP sites. The DRM reaction on the bare Ni NP (Scheme 2b.1) occurs almost exclusively (>99.9 % of the total rate of CH4 conversion) following the direct activation of both CH4 (R11) and CO2 (R3) on NP sites, followed by CH*NP oxidation to form the HCO*NP intermediate (R7), that decomposes yielding the products CO*NP and H*NP (R6) (HC-O oxidation route). The C*NP oxidation by O*NP (R8 and R7, C-O oxidation route) and the oxygenassisted pathway for CH4 activation (R12, not shown in Scheme 2b) are therefore not relevant for CO production.

8


Page 9 of 15

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


Scheme 2. Reaction networks and net rates for individual elementary steps for (a) WGS and (b) DRM reactions calculated with our kinetic model for Al2O3-supported Ni NPs. Panels (1): rates for the simulation with NP sites only. Panels (2): rates for the simulation with both NP and IN sites. Thick arrows indicate the preferred reaction channel. WGS reaction conditions: 300 °C, 1 bar total pressure, feed composition (molar fraction): 0.10 H2O, 0.10 CO, 0.80 N2, CO conversion: 10%. DRM reaction conditions: 650 °C, 1 bar total pressure, feed composition (molar fraction): 0.45 CH4, 0.45 CO2, 0.1 N2, CH4 conversion: 25%. The same major DRM reaction pathway was identified on HCO*NP intermediate also on NP sites after a CH-O oxidation step. 94% of the total CH4 conversion rate is due to the Ni (111) surface from microkinetic modeling.55 The DRM rate, taken as the methane consumption rate (R12) this pathway. Only 5% of the rate is related to IN sites, at ca. 25% CH4 conversion, is equal to 8.42×10-10 mol.cmwhich correspond to the reaction pathway via C*NP oxida2 -1 .s . tion by O*IN (R18) to form CO*IN. The reaction pathway Concerning the surface species, the most abundant ones via the HCO*IN intermediate on IN sites (not shown in are C*Np (5.64×10-1 ML), free sites (*NP, 1.70×10-1 ML), O*NP Scheme 2) is not significant, since it accounts for only 1% (1.27×10-1 ML) and CO*NP (1.19×10-1 ML). The high amount of the DRM rate. The TOF of IN sites for this simulation is of surface carbon is in line with the coking process, which 40% lower than that of NP sites. The total DRM rate (CH4 takes place at such temperatures on Al2O3-supported Ni consumption) for this simulation is equal to 7.25×10-10 40 NP catalysts. From the DRC analysis, the most RDS mol.cm-2.s-1, i.e. ca. 14% lower than in the simulation of surface reaction of the mechanism is the oxidation of DRM on the bare Ni NP. CH*NP by O*NP (R7, XRC,R7 = 0.57). The step in which C-O The major surface species are C*NP (5.53×10-1 ML), *NP bond is formed (oxidation reaction) was also identified as (1.75×10-1 ML), O*NP (1.30×10-1 ML), CO*NP (1.23×10-1 ML), 59 the RDS for DRM on Ni by DFT studies on slab models. OH*IN (7.32×10-1 ML) and O*IN (2.68×10-1 ML). Concerning the DRC analysis, CH* oxidation reaction on the NP site DRM on NP and IN sites. When both NP and IN sites (R7) was identified as the most RDS (XRC,R7 = 0.57), simiare considered (Scheme 2b.2), the main reaction channel larly to the previous simulation. Therefore, the Ni/Al2O3 is the same that was identified for the simulation with NP interface sites do not contribute to an increase in the total sites only, i.e. via direct CH4 and CO2 activation reactions reaction rate, but rather to a slight decrease, since the on NP sites (R3 and R11, respectively) and forming the 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

TOF of IN sites is not as high as the one of NP sites. This is because the O* species, key intermediates for the CH* or C* oxidation reactions (surface RDS of the prevalent

Page 10 of 15

mechanisms) are available at the same extent in both NP (1.30×10-1 ML) or IN (2.68×10-1 ML) sites.

Figure 5. Influence of the proportion of IN sites considered in the kinetic model for Al2O3-supported Ni NPs on (1) preferred reaction pathways and (2) overall reaction rates for (a) WGS and (b) DRM. WGS reaction conditions: 300 °C, 1 bar total pressure, feed composition (molar fraction): 0.10 H2O, 0.10 CO, 0.80 N2, CO conversion: 10%. DRM reaction conditions: 650 °C, 1 bar total pressure, feed composition (molar fraction): 0.45 CH4, 0.45 CO2, 0.1 N2, CH4 conversion: 25%. The yellow area shows the range expected for the particles prepared in this study (between 2 and 16 nm diameter, vide infra).12 WGS and DRM reaction rate dependence on the proportion of IN sites. So far, all the simulations considered a proportion of 10% IN sites. We hereby evaluate the influence of Ni/Al2O3 interface site percentage on the preferred pathways and overall rates of WGS and DRM (Figure 5). The predicted12 range of interface sites proportion for the synthesized Ni catalysts (2-16 nm NP size, vide infra) is shown in yellow in Figure 5. For the WGS reaction, the contribution of the reaction pathways via IN sites to the total rate sharply increases with the proportion of such sites (Figure 5a.1). Already 54% of the CO is converted via reaction pathways associated occurring on the Ni/Al2O3 interface when the proportion of IN sites is as low as 1%. Regarding the WGS overall reaction rate (Figure 5a.2), it increases with the percentage of interface sites, even at low IN proportion. For instance, when taking a proportion of 1% of IN sites,

the rate is significantly increased by a factor of 1.8 compared to the bare NP without considering IN sites. For the DRM reaction (Figure 5b), the relative contribution of the different reaction pathways via NP and IN sites to the total rate is only marginally modified with the proportion of IN sites. At the IN proportion value of 1%, NP sites account for 99% of the reaction rate (via HC-O pathway) and only 1% of the rate comes from reactions on IN sites. The DRM reaction rate slightly decreases (

Al2O3 Interfaces in WaterâGas Shift and

Recommend Documents

Al2O3 Interfaces in WaterâGas Shift and