Low-Temperature Restructuring of CeO2 ... - ACS Publications

Subscriber access provided by Kaohsiung Medical University

Article

Low-temperature restructuring of CeO2-supported Ru nanoparticles determines selectivity in CO2 catalytic reduction Aisulu Aitbekova, Liheng Wu, Cody J. Wrasman, Alexey Boubnov, Adam S. Hoffman, Emmett D. Goodman, Simon R. Bare, and Matteo Cargnello J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b07615 • Publication Date (Web): 25 Sep 2018 Downloaded from http://pubs.acs.org on September 25, 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 34 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

Low-Temperature Restructuring of CeO2-Supported Ru Nanoparticles Determines Selectivity in CO2 Catalytic Reduction

Aisulu Aitbekovaa,§, Liheng Wua,b,§, Cody J. Wrasmana, Alexey Boubnovb, Adam S. Hoffmanb, Emmett D. Goodmana, Simon R. Bareb,*, Matteo Cargnelloa,*

a

Department of Chemical Engineering and SUNCAT Center for Interface Science and Catalysis,

Stanford University, Stanford, CA 94305, USA b

Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, Menlo

Park, CA 94025, USA

ACS Paragon Plus Environment

Journal of the American Chemical Society 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

ABSTRACT

CO2 reduction to higher value products is a promising way to produce fuels and key chemical building blocks while reducing CO2 emissions. The reaction at atmospheric pressure mainly yields CH4 via methanation and CO via the reverse water gas shift (RWGS) reaction. Describing catalyst features that control the selectivity of these two pathways is important to determine the formation of specific products. At the same time, identification of morphological changes occurring to catalysts under reaction conditions can be crucial to tune their catalytic performance. In this contribution we investigate the dependency of selectivity for CO2 reduction on size of Ru nanoparticles (NPs) and on support. We find that even at rather low temperatures (210 °C), oxidative pre-treatment induces redispersion of Ru NPs supported on CeO2 and leads to a complete switch in the performance of this material from a well-known selective methanation catalyst to an active and selective RWGS catalyst. By utilizing in-situ x-ray absorption spectroscopy we demonstrate that the low-temperature redispersion process occurs via decomposition of the metal oxide phase with size-dependent kinetics, producing stable single-site RuOx/CeO2 species strongly bound to the CeO2 support that are remarkably selective for CO production. These results show that reaction selectivity can be heavily dependent on catalyst structure and that structural changes of the catalyst can occur even at low temperatures and can go unseen in materials with less defined structures.


Page 2 of 34

Page 3 of 34 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


INTRODUCTION

Catalytic CO2 reduction is a promising way to make higher value products, such as fuels and key building blocks in the chemical industry, while reducing net CO2 emissions.1 The reaction at atmospheric pressure mainly yields CH4 via CO2 methanation, and/or CO via the reverse water gas shift (RWGS) reaction. While CO2 methanation is a way to produce synthetic natural gas, the RWGS produces CO that can be used as a feedstock in other processes such as Fischer-Tropsch synthesis.1 Depending on the application, catalysts that are active and selective for either one of the two pathways are needed. Furthermore, an optimal balance between the two reactivity patterns is important for the production of hydrocarbons at higher pressure, where CO hydrogenation and CO insertion in C-H bonds are required to build molecules of increasingly higher complexity.2 For the catalytic reduction of CO2, the interfacial region is important in supported metal catalysts, where the support activates CO2 while the metal activates H2.1, 3 An interplay between these two functionalities determines which pathway is dominant, and it is generally agreed that desorption of CO determines the selectivity in this process: if the desorption is favorable, the dominant pathway is RWGS; otherwise, hydrogenation is more likely.4-6 For this reason, materials are usually classified as either methanation or RWGS catalysts. Active metal-based RWGS catalysts are mainly represented by supported Cu, Pd and Pt, while methanation catalysts are mainly presented by supported Ni, Rh, and Ru.1, 3 In many cases, however, the activity of the supported metal phase is heavily affected by the support and, more importantly, by the oxidation state, particle size, and structure of the supported metal.7-9 The ability to control the selectivity of CO2 reduction catalytically provides an opportunity for



controlled production of more complex molecules in order to harvest the potential of CO2 as a C1 building block. The literature teaches that supported ruthenium catalysts are among the most active and stable methanation catalysts due to their ability to dissociate hydrogen and bind CO.10-14 It was found that reducible supports (e.g. CeO2, TiO2) are better at activating CO2 and since the metalsupport interface is directly involved in the reaction, CeO2- and TiO2-supported Ru catalysts are usually considered more active compared to Ru/Al2O3.15-17 The effect of Ru particle size is, however, very important because small clusters and atomic species supported on Al2O3 are found to be very selective for the RWGS pathway, whereas larger particles are selective towards CH4 formation.9 Nevertheless, this size dependence is debated in the literature since many reports show that even single Ru atoms and small Ru NPs (1-3 nm) supported on CeO2 exhibit 99% selectivity to CH4, and that Ru-doped CeO2 is also an active methanation catalyst.15, 18-22 These contrasting results highlight the difficulty in correlating catalyst structure and activity in materials where particle size, structure and loading can be highly inhomogeneous, leading to catalytic properties being an average of the multiple types of active sites present in the materials. Additionally, materials may be heavily affected by structural changes occurring under reaction conditions that are known to be able to cause drastic changes in catalytic behavior, including changes in rates and selectivity.23 The ability to track and explain changes in the supported metal particles as a function of reaction conditions is therefore essential to fully understand and therefore lead to controlling the reaction selectivity towards specific desired products. Among the morphological changes that can occur to supported systems under reaction conditions, redispersion of particles into clusters and atomic species has attracted scientific interest due, in particular, to the opportunity to maximize dispersion of active sites and to


Page 4 of 34

Page 5 of 34 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


regenerate catalysts following sintering processes.24 Several strategies are employed to induce redispersion in metal-based systems. The most common approach is oxidation-reduction cycles, where active phase is redispersed during oxidative treatment and the dispersed species precipitate into smaller NPs during reduction.25 Redispersion of Fe, Pt, Pd, Rh, Ag, and Au has been studied extensively.24 In most cases, high temperatures (>400 °C) are required during the oxidation step to induce the redispersion.24 The mechanism of redispersion via oxidation can be explained by a “strain” model.24, 26-27 In particular, redispersion of Pt NPs supported on γ-Al2O3 thin films after heating in air at 400 °C was explained by interfacial strain between the support and an oxidized Pt phase, which is relaxed by the fracture of the metal crystallites.26 Clearly, the support plays an important role in these processes: silica, which is usually considered inert, can lead to Cu particle redispersion at room temperature to form atomically stable copper oxide species in a sizedependent process.28 In the case of Ru, several studies show that active sites in Ru/CeO2 catalysts, synthesized via conventional methods such as precipitation-deposition, are presented by two distinct types of species: metallic Ru NPs and a small fraction of highly dispersed species that have Ru=O bond.29-31 Although these studies report distinct catalytic properties of the two types of active sites in various reactions (i.e. coupling of alkynes with acrylates30, N2O decomposition in the presence of C3H629, synthesis of indole via dehydrogenative Nheterocyclization31), they lack quantitative characterization and explanations as to what governs the formation of dispersed species possibly due to the inhomogeneous nature of synthesized catalysts. In this work we investigate the effect of size and support in Ru-catalyzed CO2 reduction systems by utilizing pre-formed uniform Ru NPs of different sizes (1.4, 2.6, and 4.4 nm). We show that low-temperature oxidative pretreatment (210 °C) induces structural changes in



supported Ru catalysts leading to redispersion of the Ru NPs and to drastic changes in CO2 reduction selectivity. By utilizing in-situ X-ray absorption spectroscopy (XAS), we demonstrate that the low temperature redispersion produces stable atomically dispersed RuOx/CeO2 species, which are responsible for the switch in the catalytic selectivity from methanation to RWGS character. The redispersion of Ru is complete on CeO2 but only partial on TiO2 and Al2O3 due to the strong stabilization of RuOx species on the CeO2 surface. These results show that the CO2reduction selectivity can be effectively tuned by restructuring processes at low temperatures that can go unseen in materials with less defined structure.

EXPERIMENTAL SECTION

Materials Triruthenium dodecarbonyl (Ru3(CO)12, 99%), trioctylamine (TOA, 98%), 1-oleylamine (OLAM, 70%) and benzyl ether (98%) were purchased from Sigma Aldrich. Cerium oxide (98.5%) was purchased from Treibacher Industrie AG and calcined at 500 °C for 5 h before use. Al2O3 Pluralox TH100/150 was obtained from Sasol and calcined at 900 °C for 24 h. The calcined material was predominantly γ-Al2O3. Titanium (IV) oxide Aeroxide P25 was purchased from Acros Organics and calcined at 500 °C for 5 h. All solvents were of reagent grade and all reagents were used as-received. All calcined supports and samples were ground and sieved below 180 µm grain size. Synthesis of Ru NPs


Page 6 of 34

Page 7 of 34 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


Ru NPs were prepared by thermal decomposition of Ru3(CO)12 via colloidal synthesis using standard Schlenk techniques using a literature procedure with slight modifications.32 Ru NPs with three different sizes were synthesized. For the synthesis of 1.4 nm Ru NPs, 8 mL of TOA were added to 50 mg of Ru3(CO)12 in a three-neck flask. The reaction content was degassed (95% in all cases. The overall reaction rates based on CO2 conversion for the LTR catalysts were in line with literature reports, with the TiO2- and CeO2-supported samples being more active than the Al2O3-supported ones (Figure 2ac).16, 37 The Ru phase in the LTR catalysts was metallic according to previous work and our x-ray absorption spectroscopy results (see below and Figure S2).35 The calculated values for the apparent activation energy of the 2.6 nm Ru/TiO2, Ru/Al2O3, and Ru/CeO2 (48, 70, and 76 kJ mol-1, respectively) were in good agreement with previous studies.9, 15, 20, 38 In all cases, CO2 was converted to methane with selectivity >90% (and quantitatively in the case of Ru/TiO2). This result is not surprising given the well-known ability of metallic Ru to dissociate H2, making it a strong methanation catalyst. Reducible supports are known for their strong ability to activate


Page 12 of 34

Page 13 of 34 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


CO2 and metals supported on reducible supports such as TiO2 and CeO2 are known for their high methanation activity, where the metal-support interface serves as active sites for the reaction.39-40

Table 1. Apparent activation energies (in kJ mol-1) based on CO2 conversion rate for 2.6 nm Ru NPs supported on Al2O3, TiO2 and CeO2.

LTR

OX-LTR

OX-HTR

2.6 nm Ru/Al2O3

70

80

79

2.6 nm Ru/TiO2

48

59

48

2.6 nm Ru/CeO2

76

63

67



Figure 2. Arrhenius plots of: (a) 2.6 nm Ru/Al2O3; (b) 2.6 nm Ru/TiO2; (c) 2.6 nm Ru/CeO2. (d) CO selectivity at 240 °C. CO2 conversion was less than 5% in all cases. Experimental errors for the rate and selectivity measurements were consistently less than ±8%.

Although the catalytic results for the LTR pretreated catalysts were in line with expectations, an unexpected change in catalytic selectivity for the CeO2-supported sample occurred after OX-LTR treatment. The overall CO2 conversion rate slightly decreased, but the CO selectivity showed a dramatic change from 90% for the OXLTR one (Figure 2d). The oxidative pretreatment at 230 °C switched the catalyst activity from


Page 14 of 34

Page 15 of 34 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


almost pure methanation to RWGS character. Meanwhile, the CO selectivity after OX-LTR treatment increased from 10% to roughly 60% for the Al2O3-supported sample but did not change for the TiO2-supported catalyst. Given the surprising changes after OX-LTR, we also tested the OX-HTR pretreatment in which reduction at 530 °C was used to study the effect of reduction temperature on catalytic performance. Reduction at temperatures higher than 500 °C is known to promote strong-metal support interactions (SMSI) between metals and reducible oxides.41-43 We found that after the OX-HTR pretreatment, the TiO2 sample showed much decreased CO2 conversion rates compared to the OX-LTR sample, and much enhanced CO selectivity from 1 % to 82 % (Figure 2b,d). The suppressed activity and change in selectivity are known traits of a catalyst in SMSI state.41 In systems exhibiting SMSI such as group 8-10 noble metals supported on TiO2, reduction at temperatures above 500 °C leads to partial reduction of surface TiO2 to TiOx that migrate on the metal surface, leading to partial coverage of the metal particle.36 This phenomenon leads to a decrease in H2 adsorption and, thus, catalytic activity decreases as well as selectivity switches to producing CO rather than methane.41 CeO2 is also known as a support that promotes SMSI.44 For the CeO2 support, previous work showed that in the case of Rh/CeO2, reduction at 500 °C led to comparable catalytic activity for CO2 hydrogenation under steady-state conditions compared to that after reduction at 227 °C.44 Likewise, we observed similar catalytic activity under steady-state conditions for both the OXHTR and OX-LTR Ru/CeO2 samples with slight enhancement in the CO selectivity from 92% to 97%, respectively (Figure 2c,d). For the Al2O3 support, the Ru/Al2O3 sample did not show obvious changes in activity and selectivity after the OX-HTR, which is in line with the fact that the Al2O3 support is inert under these conditions.



To understand the reasons for the unexpected changes in catalytic activity and selectivity after different pretreatments, electron microscopy characterization of these samples was performed. Initially, in the LTR state, all catalysts showed the presence of uniform and welldispersed Ru NPs on the support surface (Figure S3). However, after OX-LTR pretreatment and catalysis the Ru NPs on the Ru/TiO2 and Ru/Al2O3 were smaller (1.3 ± 0.5 nm and 1.7 ± 0.4 nm, respectively) than those in the fresh catalysts (Figure S4). Even more surprisingly, the Ru NPs were not visible by TEM in the OX-LTR Ru/CeO2 sample. To determine whether the mild oxidative pretreatment was responsible for the morphological changes of the catalysts, the Ru/Al2O3 and Ru/CeO2 samples were oxidized only and then investigated by TEM. The results for Ru/Al2O3 again showed particles smaller (1.2 ± 0.6 nm) than those in the fresh catalysts (Figure S5a), while still no particles could be found in Ru/CeO2 (Figure S5b). The fact that no Ru NPs on CeO2 could be discerned after oxidation could be due to either the Ru species becoming incorporated into the support structure or that the Ru NPs dispersed into very small NPs and/or single Ru species that are below the detection limit of conventional TEM. TEM techniques are not sensitive enough to distinguish highly dispersed Ru species on CeO2 given the very close electron density between these two elements. With respect to the first possibility, it is known that Ru and CeO2 can form RuOx-CeO2 solid solutions where Ru atoms substitute Ce atoms in CeO2.21-22 Such a catalyst has been described as active and selective for CO2 methanation.21-22 We note that these materials are synthesized by combustion at high temperatures (>550 °C). Therefore, it is not likely that the solid solution forms at such low oxidation treatment temperature of 230 °C.25 Redispersion of Ru in the form of small clusters or atomic species is a more plausible option. However, small Ru NPs (90% (and almost quantitative CO selectivity for the 1.4 nm Ru/CeO2 sample after OX-HTR), the selectivity over the 4.4 nm Ru/CeO2 sample was only around 40%. The fact that the activity and selectivity of all catalysts did not change when the reductive pretreatment was performed either at 230 °C or 550 °C suggest that the switch in the catalyst selectivity is induced by the mild oxidation pretreatment performed.



Figure 3. (a) Arrhenius plots for 1.4, 2.6, and 4.4 nm Ru/CeO2 for CO2 hydrogenation reaction. (b) CO selectivity at 240 °C collected at similar conversion of CO2 (less than 5%).


Page 18 of 34

Page 19 of 34 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 4. Steady-state rates for CO and CH4 formation, and CO selectivity at 240 °C (CO2 conversion less than 5%) after oxidative pretreatments at increasing temperatures for: (a) 1.4 nm Ru/CeO2. (b) 2.6 nm Ru/CeO2. (c) 4.4 nm Ru/CeO2.



To further explore the role of the oxidative pretreatment on Ru/CeO2, a series of catalytic experiments were conducted where 1.4, 2.6 and 4.4 nm Ru/CeO2 were oxidized at increasing temperatures for 30 min with a ramp rate of 20 °C min-1, and steady-state activity for CO2 reduction after each pretreatment was measured at 240 °C (Figure 4). As expected, all reduced catalysts at the beginning of the experiments exhibited strong methanation character. 1.4 and 2.6 nm Ru/CeO2 samples show overall rates that were higher than the 4.4 nm Ru/CeO2 sample, in agreement with the literature that small Ru NPs (1-3 nm) on CeO2 are very active for CO2 methanation.15, 19 However, oxidation at increasing temperatures led to a gradual decrease in the CH4 formation rate and a concomitant increase in the CO formation rate. It is important to notice that as the starting Ru NP size increased from 1.4 to 4.4 nm, higher oxidative pretreatment temperatures were required to achieve CO selectivity of >90% (Figure 4). In particular, for 1.4 nm Ru/CeO2, oxidation at a temperature as low as 210 °C was sufficient to obtain CO selectivity >90%, whereas the same selectivity over 2.6 and 4.4 nm was achieved after oxidation at around 230 °C and 280 °C, respectively. Additionally, once the CO selectivity reached >90%, all catalysts had very similar CH4 and CO rates, suggesting that the final state of the Ru was the same regardless of the initial Ru NP size. The same trend across all catalysts suggests a similar underlying mechanism for the change in reaction selectivity. The previous work on Ru/Al2O3 showed that CO selectivity increases as Ru metal cluster size decreases.9 The authors claimed that at low Ru loadings ( 90% (Figure S6b). Therefore, we conclude that the redispersion is not dependent on the predominant CeO2 facet. Additionally, to rule out the effect of the diluent, we performed experiments with non-diluted catalysts and observed the same phenomenon.

Figure 5. Ex-situ EXAFS results in the R space: (a) 2.6 nm Ru/CeO2; (b) 1.4 nm Ru/CeO2. Black thin lines indicate the fits.


Page 22 of 34

Page 23 of 34 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


Table 2. EXAFS fitting results for 2.6 and 1.4 nm Ru/CeO2

Path

2.6 nm Ru/CeO2 Fresh

Ru-O

Ru-Ru

OX-LTR

1.4 nm Ru/CeO2 Post-cat.

Fresh

OX-LTR

Post-cat.

R, Å

1.99±0.01

1.99±0.01

1.99±0.01

2.00±0.01

2.00±0.01

2.00±0.01

CN

4.5±0.4

5.5±0.6

5.2±0.5

4.7±0.5

5.4±0.6

5.2±0.6

σ2

0.005±0.001

0.005±0.001

0.005±0.001

R, Å

2.69±0.01

-

-

2.69±0.03

-

-

CN

1.6±0.3

-

-

0.7±0.3

-

-

σ2

0.003*

-

-

0.003**

-

-

R, Å

3.15±0.06

-

-

3.18±0.03

-

3.18±0.03

CN

0.8±0.5

-

-

1.0±0.5

-

0.9±0.8

σ2

0.006*

-

-

0.003**

-

0.003*

0.004±0.001 0.004±0.001 0.004±0.001

(metal)

Ru-Ru (oxide)

* σ2 was set to 0.003 and 0.006 for Ru-Ru (metal) and Ru-Ru (oxide) paths, respectively; ** σ2 was set to 0.003 for Ru-Ru (metal) and Ru-Ru (oxide) paths. R-factors were 2.08% and 1.87% for 2.6 nm Ru/CeO2 and 1.4 nm Ru/CeO2, respectively. ∆E was 0±1 eV and 1±1 eV for 2.6 nm Ru/CeO2 and 1.4 nm Ru/CeO2, respectively.

The redispersion process and the nature of the catalysts were further investigated by ex-situ and in-situ x-ray absorption spectroscopy (XAS). The ex-situ EXAFS results for 1.4 and 2.6 nm Ru/CeO2 after different treatments are shown in Figure 5 and Figure S7, whereas coordination



numbers and EXAFS parameters for each sample are presented in Table 2 and Table S2. The best fit model to the EXAFS data confirms that the fresh catalysts comprise both metallic Ru and Ru oxide, and the coordination numbers are consistent with the Ru being present as nanoparticles. However, after OX treatment both the metallic Ru-Ru scattering path at 2.69 Å for metallic Ru and the Ru-Ru scattering path at 3.15 Å for Ru oxide disappeared (Figure S7). Instead, the best-fit model to the EXAFS data is consistent with a single contribution due to RuO scattering, as ascertained by previous studies,53 and no contribution from Ru-Ru scattering, which is not entirely true in other systems reporting Ru single sites on ceria for CO2 methanation.18 This result confirmed that low-temperature oxidation induces redispersion of Ru NPs into atomic species. The EXAFS pattern of the OX-LTR treated, and post-catalysis samples show very similar features and coordination numbers (Figure 5, Table 2). These results suggest that once redispersed during the oxidative treatment, the RuOx species remain in the dispersed state during the reductive treatment and even under the reaction conditions, likely due to a strong bonding with the CeO2 support. This hypothesis is consistent with DFT work showing that single Ru atoms are strongly stabilized by the ceria support, as demonstrated by the strong binding energies and high diffusion barriers of a single Ru atom.54 The strong Ru-O-Ce bonds are the reason why these materials form solid solutions when appropriately prepared.21-22 However, unlike our atomically redispersed Ru on CeO2 being active RWGS catalyst, Ru-doped ceria has been shown as active methanation catalyst.21 The Ru atoms in the solid solution are located within the CeO2 lattices, which is expected to provide a quite different reactivity from the CeO2supported Ru atomically dispersed species described in this work. In our case the oxidative pretreatment temperature is too low to form the solid solution. The different chemical environment of Ru atoms may be responsible for their different catalytic performance.


Page 24 of 34

Page 25 of 34 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


To further demonstrate the stability of the redispersed Ru species, the catalytic performance of the OX pretreated 2.6 nm Ru/CeO2 catalyst was compared with the OX-LTR pretreated catalyst (Figure S8). Similar CO selectivity for the two samples further suggests the high stability of atomically dispersed RuOx species after oxidative pretreatment. The EXAFS results strongly suggest that the switch in CO selectivity is related to the change in the catalyst’s structure. Additionally, the similar EXAFS patterns and coordination numbers of 1.4 and 2.6 nm Ru/CeO2 after redispersion explain very similar catalytic activity of the samples, confirming that Ru redisperses to almost the atomic level independently of the initial NP size. To gain more insights into the catalyst structure during reaction conditions, the samples were further analyzed by in-situ XAS. Figure 6a shows the in-situ XANES spectra for 4.4 nm Ru/CeO2 after different treatments and Figure S9 shows the results for 2.6 nm Ru/CeO2. The XANES region of the OX pretreated samples showed a pre-edge feature centered at 22120 eV. The appreciable intensity of this pre-edge feature, assigned to the forbidden 1s to 4d transition, is indicative of non-centrosymmetric symmetry of the atomically dispersed Ru species, likely either in four-fold (tetrahedral) or five-fold coordination.29-31 The Ru in RuO2 is in octahedral coordination and thus the XANES spectrum of bulk RuO2 does not display this feature (Figure 6a). During the reductive pretreatment, the pre-edge feature disappeared indicating that Ru=O groups reacted with hydrogen and were converted from a four- or fivecoordinated species to a six-coordinated configuration.29-31 Further proof is provided by ex-situ DRIFTS performed on the 4.4 nm Ru/CeO2 sample (Figure 6b). The DRIFTS spectra of the sample after oxidative pretreatment at 280 oC showed a strong feature centered around 975 cm-1 that can be assigned to the atomically dispersed Ru=O moiety.29-31 However, this feature was absent in the fresh and the post-catalysis samples. These results further highlight the single-site



nature of the Ru/CeO2 material after low-temperature oxidative pretreatment which are consistent with the XAS measurement. Additionally, in-situ XANES (Figure S10) showed that the pre-edge feature, initially absent in the fresh samples, developed as the oxidative pretreatment temperature increased. Once fully formed, it stopped changing with a further increase in the temperature. The integrated intensity of the pre-edge peak in the samples increased as a function of the oxidative pretreatment temperature due to the contribution of atomic Ru=O species (Figure S11).55 The formation temperature of the pre-edge feature is lower for the 2.6 nm Ru/CeO2 sample than for 4.4 nm Ru/CeO2 (Figure S11), which is consistent with the fact that lower temperatures are needed to oxidize and redisperse smaller NPs into atomic species.


Page 26 of 34

Page 27 of 34 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 6. (a) In-situ XANES spectra and (b) ex-situ DRIFTS spectra of the 4.4 nm Ru/CeO2 catalyst under different environments. Here the OX temperature was 280 °C as the redispersion of the 4.4 nm Ru/CeO2 catalyst occurred after oxidation at 280 °C with a CO selectivity >90%.

To further corroborate the formation of atomically-dispersed species, CO chemisorption was performed. The oxidative pretreatment led to expected significant increases in dispersion compared to the reduced catalysts. About 100% Ru dispersion values were obtained after the OX-LTR pretreatment for all sizes of the Ru/CeO2 catalysts indicating the dispersed nature of these materials (Table S3). Despite the CO selectivity over the OX-LTR treated 4.4 nm sample was only 42% (Figure 3b), we hypothesize that the presence of small clusters formed after the OX-LTR treatment may explain some residual methanation activity while still providing large exposed metal surface to achieve 100% Ru dispersion during chemisorption. To the best of our knowledge, redispersion of Ru NPs into single sites by oxidative treatment at low temperatures have not been reported before. We believe that the reason this phenomenon has been not observed in conventional catalysts is the difference in thermal treatment conditions used for preparing catalysts. During conventional impregnation the support is infiltrated with Ru salts and calcination of the impregnated sample is performed to remove the inorganic byproducts and convert the salts into RuO2 particles. These RuO2 particles are then reduced to achieve a desired metallic Ru phase active for CO2 hydrogenation. Such preparation usually results in relatively large particle size distributions (PSD). In our case, the use of colloidal synthesis to make preformed well-defined metallic Ru NPs provides us with a different starting material compared to impregnation catalysts. Since it is oxidative pretreatment of



metallic NPs that induces their redispersion, we believe this process has not been observed with conventional impregnation catalysts because this step is not performed. Additionally, due to the relatively large PSD, it is possible that only part of the population of particles redisperses, and that microscopy techniques are not sensitive to this small fraction of particles and the redispersion process could have gone unnoticed. Given the catalytic results and characterization of the Ru/CeO2 system showing redispersion of the oxide phase, we can hypothesize that a similar phenomenon likely happened in the Ru/Al2O3 and Ru/TiO2 systems. Since we observed particles shrinking but not disappearing after oxidative pretreatment and after catalysis on these supports (Figure S4-S5), we suggest that oxidation of these systems led to only partial redispersion of the Ru NPs. This hypothesis is consistent with the DFT work showing that unlike CeO2, Al2O3 and TiO2 cannot stabilize single Ru resulting in their relatively high mobility and formation of Ru clusters.54 To elucidate the effect of the oxidative pretreatments on these catalysts, we treated Ru/TiO2 and Ru/Al2O3 in diluted oxygen at increasing temperatures and measured their steadystate activity at 240 °C after each treatment (Figure S12-S13). For 2.6 nm Ru/Al2O3, oxidation at 230 °C led to a decrease in the CH4 rate and an increase in the CO rate compared to the reduced catalyst. The TEM results showed a decrease in the average Ru NP size from 2.6 nm in the LTR sample (Figure S3a) to 1.7 nm in the OX-LTR catalyst (Figure S4a). Since it has been demonstrated that in Ru/Al2O3 systems CO selectivity increases as NP size decreases, the increased CO selectivity over the OX-LTR sample can be explained by the smaller NP size, in line with the previous studies.9 After oxidation at 280 °C the CO selectivity increased further, although a decrease in both the CH4 and CO rates was observed with a more pronounced change in the CH4 formation. However, after oxidation at 330 °C, and subsequently at 430 °C, we


Page 28 of 34

Page 29 of 34 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


observed an opposite trend: the CH4 rate increased and the CO rate decreased. These results suggest that while oxidation at 230 °C led to partial redispersion of the metal NPs, oxidation at higher temperatures (>330 °C) caused sintering of the Ru NPs, as verified by TEM analysis (Figure S14). Unlike Ru/Al2O3, 2.6 nm Ru/TiO2 retained its strong methanation character even as the oxidation temperature increased from 230 °C to 480 °C. Although the TEM results also showed the decrease in an average NP size from 2.6 nm in the LTR sample (Figure S3b) to 1.3 nm in the OX-LTR catalyst (Figure S4c), no significant changes in the CO selectivity were observed. Thus, although oxidation induced partial redispersion of Ru, the small NPs still serve as active methanation catalyst, which are consistent with the literatures showing that small Ru NPs supported on TiO2 are strong methanation catalyst.14, 20, 56 Finally, we studied the stability of the redispersed Ru/CeO2 catalysts (Figure S15). The catalysts were relatively stable for CO production over 12-14 h, while CH4 production decreased rather quickly with time-on-stream, resulting in further increased CO selectivity from >90% to ~99% with time on stream. It is possible that the active sites for methane formation became poisoned with carbonaceous deposits, whereas single-sites that weakly bind carbon remained active and stable for much longer period, thus providing stable catalysts to produce CO with high selectivity and noble metal utilization efficiency.

CONCLUSIONS Using size-controlled uniform Ru NPs, we have shown that the structure of the oxide-supported Ru catalysts is strongly dependent on the pretreatment conditions. In particular, we have demonstrated that after reductive pretreatment in H2, the Ru NPs on CeO2 support remain



metallic, resulting in selective methanation catalysts for the CO2 reduction reaction. In contrast, after oxidative pretreatment in O2 even at rather low temperatures (210 °C), the catalytic performance switched dramatically from methanation to preferential CO formation with selectivity above 90 % via RWGS. In-situ and ex-situ XAS analysis revealed that this oxidative treatment induced the redispersion of initial Ru NPs into atomically dispersed RuOx species, which substantially changed the reaction pathway from selective CO2 methanation into dominant RWGS. Our results demonstrate the importance of pretreatment conditions on tuning the structures of supported Ru NPs toward selective CO2 reduction. This pretreatment-induced structural change may be extended to other well-defined supported NPs to design selective catalysts for important chemical reactions. ASSOCIATED CONTENT Supporting Information Supporting Information Available: particle size distributions, TEM images, XAS results, catalytic results (Figures S1-S15 and Tables S1-S3). This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Authors *[email protected] (S.R.B.); [email protected] (M.C.) Author Contributions §

A.A. and L.W. contributed equally to the work.

Notes The authors declare no competing financial interest.


Page 30 of 34

Page 31 of 34 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


ACKNOWLEDGMENTS We gratefully acknowledge support from the Stanford Precourt Institute for Energy. M.C. acknowledges support from the School of Engineering at Stanford University and from a Terman Faculty Fellowship. A.A. acknowledges support from a Stanford Graduate Fellowship (SGF) and an EDGE fellowship. Part of this work was performed at the Stanford Nano Shared Facilities (SNSF), supported by the National Science Foundation under award ECCS-1542152. Part of this work was performed at Stanford Synchrotron Radiation Lightsource (SSRL) of SLAC National Accelerator Laboratory and use of the SSRL is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Contract No. DE-AC02- 76SF00515. We would like to thank Arun Asundi and Prof. Stacey Bent (Stanford University) for performing the DRIFTS measurements.

REFERENCES 1. Porosoff, M. D.; Yan, B.; Chen, J. G. Energy Environ. Sci. 2016, 9, 62-73. 2. Luk, H. T.; Mondelli, C.; Ferré, D. C.; Stewart, J. A.; Pérez-Ramírez, J. Chem. Soc. Rev. 2017, 46, 1358-1426. 3. Kattel, S.; Liu, P.; Chen, J. G. J. Am. Chem. Soc. 2017, 139, 9739-9754. 4. Graciani, J.; Mudiyanselage, K.; Xu, F.; Baber, A. E.; Evans, J.; Senanayake, S. D.; Stacchiola, D. J.; Liu, P.; Hrbek, J.; Sanz, J. F.; Rodriguez, J. A. Science 2014, 345, 546-550. 5. Iizuka, T.; Kojima, M.; Tanabe, K. J. Chem. Soc., Chem. Commun. 1983, 638-639. 6. C., M. I.; Hai-Yan, S.; Federico, C.-V.; A., H. H.; I., M. J.; G., I. N.; John, K.; F., J. T.; K., N. J.; Jan, R. ChemCatChem 2011, 3, 1159-1165. 7. Siwei, L.; Yao, X.; Yifu, C.; Weizhen, L.; Lili, L.; Mengzhu, L.; Yuchen, D.; Xiaoping, W.; Binghui, G.; Ce, Y.; Siyu, Y.; Jinglin, X.; Yongwang, L.; Xi, L.; Ding, M. Angew. Chem. Int. Ed. 2017, 56, 10761-10765. 8. Matsubu, J. C.; Yang, V. N.; Christopher, P. J. Am. Chem. Soc. 2015, 137, 3076-3084. 9. Kwak, J. H.; Kovarik, L.; Szanyi, J. ACS Catal. 2013, 3, 2449-2455. 10. Leitenburg, C. d.; Trovarelli, A.; Kašpar, J. J. Catal. 1997, 166, 98-107.



11. Scirè, S.; Crisafulli, C.; Maggiore, R.; Minicò, S.; Galvagno, S. Catal. Lett. 1998, 51, 4145. 12. Gao, J.; Liu, Q.; Gu, F.; Liu, B.; Zhong, Z.; Su, F. RSC Adv. 2015, 5, 22759-22776. 13. Wang, W.; Wang, S.; Ma, X.; Gong, J. Chem. Soc. Rev. 2011, 40, 3703-3727. 14. Abdel-Mageed, A. M.; Widmann, D.; Olesen, S. E.; Chorkendorff, I.; Behm, R. J. ACS Catal. 2018, 8, 5399-5414. 15. Wang, F.; Li, C.; Zhang, X.; Wei, M.; Evans, D. G.; Duan, X. J. Catal. 2015, 329, 177186. 16. Dreyer, J. A. H.; Li, P.; Zhang, L.; Beh, G. K.; Zhang, R.; Sit, P. H. L.; Teoh, W. Y. Appl. Catal. B 2017, 219, 715-726. 17. Tada, S.; Ochieng, O. J.; Kikuchi, R.; Haneda, T.; Kameyama, H. Int. J. Hydrogen Energy 2014, 39, 10090-10100. 18. Guo, Y.; Mei, S.; Yuan, K.; Wang, D.-J.; Liu, H.-C.; Yan, C.-H.; Zhang, Y.-W. ACS Catal. 2018, 8, 6203-6215. 19. Wang, F.; He, S.; Chen, H.; Wang, B.; Zheng, L.; Wei, M.; Evans, D. G.; Duan, X. J. Am. Chem. Soc. 2016, 138, 6298-6305. 20. Abe, T.; Tanizawa, M.; Watanabe, K.; Taguchi, A. Energy Environmen. Sci. 2009, 2, 315-321. 21. Sharma, S.; Hu, Z.; Zhang, P.; McFarland, E. W.; Metiu, H. J. Catal. 2011, 278, 297309. 22. Upham, D. C.; Derk, A. R.; Sharma, S.; Metiu, H.; McFarland, E. W. Catal. Sci. Technol. 2015, 5, 1783-1791. 23. Matsubu, J. C.; Zhang, S.; DeRita, L.; Marinkovic, N. S.; Chen, J. G.; Graham, G. W.; Pan, X.; Christopher, P. Nat Chem. 2017, 9, 120-127. 24. Morgan, K.; Goguet, A.; Hardacre, C. ACS Catal. 2015, 5, 3430-3445. 25. Metal-support interactions in catalysis, sintering, and redispersion. Stevenson, S. A., Ed. Van Nostrand Reinhold Co.: New York :, 1987. 26. Ruckenstein, E.; Malhotra, M. L. J. Catal. 1976, 41, 303-311. 27. Okal, J.; Kubicka, H.; K epiński, L.; Krajczyk, L. App. Catal. A 1997, 162, 161-169. 28. Kim, D.; Becknell, N.; Yu, Y.; Yang, P. Nano Lett. 2017, 17, 2732-2737. 29. Hosokawa, S.; Nogawa, S.; Taniguchi, M.; Utani, K.; Kanai, H.; Imamura, S. Appl. Catal. A 2005, 288, 67-73. 30. Hiroki, M.; Shun, S.; Saburo, H.; Seiji, Y.; Kenji, W.; Masashi, I. Adv. Synth. Catal. 2011, 353, 2837-2843. 31. Shimura, S.; Miura, H.; Wada, K.; Hosokawa, S.; Yamazoe, S.; Inoue, M. Catal. Sci. Technol. 2011, 1, 1340-1346. 32. Vanden Brink, M.; Peck, M. A.; More, K. L.; Hoefelmeyer, J. D. J. Phys. Chem. C 2008, 112, 12122-12126. 33. Cargnello, M.; Chen, C.; Diroll, B. T.; Doan-Nguyen, V. V. T.; Gorte, R. J.; Murray, C. B. J. Am. Chem. Soc. 2015, 137, 6906-6911. 34. Ravel, B.; Newville, M. J. Synchrotron Radiat. 2005, 12, 537-541. 35. Lanza, R.; Järås, S. G.; Canu, P. Appl. Catal. A 2007, 325, 57-67. 36. Zhang, S.; Plessow, P. N.; Willis, J. J.; Dai, S.; Xu, M.; Graham, G. W.; Cargnello, M.; Abild-Pedersen, F.; Pan, X. Nano Lett. 2016, 16, 4528-4534. 37. Kuśmierz, M. Catal. Today 2008, 137, 429-432. 38. Solymosi, F.; Erdöhelyi, A. J. Mol. Catal. 1980, 8, 471-474.


Page 32 of 34

Page 33 of 34 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


39. Solymosi, F.; Erdöhelyi, A.; Bánsági, T. J. Catal. 1981, 68, 371-382. 40. Burch, R.; Flambard, A. R. J. Catal. 1982, 78, 389-405. 41. Tauster, S. J. Acc. Chem. Res. 1987, 20, 389-394. 42. Ou, D. R.; Mori, T.; Togasaki, H.; Takahashi, M.; Ye, F.; Drennan, J. Langmuir 2011, 27, 3859-3866. 43. Bernal, S.; Calvino, J. J.; Cauqui, M. A.; Gatica, J. M.; Larese, C.; Pérez Omil, J. A.; Pintado, J. M. Catal. Today 1999, 50, 175-206. 44. Trovarelli, A.; Dolcetti, G.; de Leitenburg, C.; Kaspar, J.; Finetti, P.; Santoni, A. J. Chem. Soc., Faraday Trans. 1992, 88, 1311-1319. 45. Jeong, H.; Bae, J.; Han, J. W.; Lee, H. ACS Catal. 2017, 7, 7097-7105. 46. Johnson, M. F. L.; Keith, C. D. J. Phys. Chem. 1963, 67, 200-201. 47. Wanke, S. E., Sintering and Redispersion of Conventional Supported Metal Catalysts in Hydrogen and Oxygen Atmospheres. In Materials Science Research: Volume 16 Sintering and Heterogeneous Catalysis, Kuczynski, G. C.; Miller, A. E.; Sargent, G. A., Eds. Springer US: Boston, MA, 1984; pp 223-242. 48. Yao, H. C.; Shelef, M. J. Catal. 1976, 44, 392-403. 49. Jones, J.; Xiong, H.; DeLaRiva, A. T.; Peterson, E. J.; Pham, H.; Challa, S. R.; Qi, G.; Oh, S.; Wiebenga, M. H.; Pereira Hernández, X. I.; Wang, Y.; Datye, A. K. Science 2016, 353, 150-154. 50. M., G. A.; Maria, C.; Philippe, V.; Stéphane, L.; J., C. S. A. F.; Thierry, E.; Benjamin, B.; Rüdiger, H.; Jan-Dierk, G. Angew. Chem. Int. Ed. 2017, 56, 13078-13082. 51. Chin, Y.-H.; García-Diéguez, M.; Iglesia, E. J. Phys. Chem. C 2016, 120, 1446-1460. 52. Wang, Z. L.; Feng, X. J. Phys. Chem. B 2003, 107, 13563-13566. 53. Wang, X.; Fu, X.-P.; Yu, W.-Z.; Ma, C.; Jia, C.-J.; Si, R. Inorg. Chem. Front. 2017, 4, 2059-2067. 54. Li, F.; Li, L.; Liu, X.; Zeng, X. C.; Chen, Z. ChemPhysChem 2016, 17, 3170-3175. 55. Boubnov, A.; Roppertz, A.; Kundrat, M. D.; Mangold, S.; Reznik, B.; Jacob, C. R.; Kureti, S.; Grunwaldt, J.-D. Appl. Surf. Sci. 2016, 386, 234-246. 56. Kim, A.; Sanchez, C.; Patriarche, G.; Ersen, O.; Moldovan, S.; Wisnet, A.; Sassoye, C.; Debecker, D. P. Catal. Sci. Technol. 2016, 6, 8117-8128.



TOC Graphic

Supported Ru/CeO2 for CO2 reduction 100

Before O2

o

230 C

After

CO selectivity (%)

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

80 60 40 20 0

Before


After

Page 34 of 34

Low-Temperature Restructuring of CeO2 ... - ACS Publications

Recommend Documents