Design and Testing Model Cobalt Catalysts for Reactions Involving

Subscriber access provided by University of South Dakota

Article

Design and Testing Model Cobalt Catalysts for Reactions Involving CO and HO 2

2

Carlos Andre C. Pérez, Neuman S de Resende, Vera Maria M. Salim, and Martin Schmal J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.8b07416 • Publication Date (Web): 12 Nov 2018 Downloaded from http://pubs.acs.org on November 13, 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 36 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

The Journal of Physical Chemistry

Design and Testing Model Cobalt Catalysts for Reactions Involving CO2 and H2O Carlos A. C. Perez a,b, Neuman S. de Resende b; Vera M. M. Salim b; Martin Schmal b* a

Instituto Federal do Rio de Janeiro – Campus Nilópolis.

b

Universidade Federal do Rio de Janeiro, Chemical Engineering Program (PEQ), NUCAT, Centro de Tecnologia, Ilha do Fundão, Rio de Janeiro, 21941-914, RJ, Brazil.

*Corresponding author. Tel.: +55 21 39388348 E-mail adress:[email protected]

Abstract In this work, we developed experimental model of cobalt heterogeneous catalyst, inserting the active phase into a nanoporous material ordered with the purpose of structurally elucidating the processes that occur during the interaction of the model with molecules. We synthesized the models by inserting cobalt species into a Faujasite zeolite and performed detailed studies of the complexes formed in the interaction of these models with water and carbon dioxide emphasizing electron densities. Regarding the interaction of Faujasite with water, X-ray diffraction showed the physical adsorption of a dense monomolecular film in the nanoporous space. There was no evidence of water entering the sodalite cage. The cell parameter increased with the degree of hydration for Faujasite containing cobalt (CoY). In the case of dehydrated material, cobalt occupied the sites within the hexagonal prism (I) and the sodalite cage (I'). The carbon dioxide adsorbs bridges between adjacent compensation cations located in II. We tested the catalyst models for the oxidation of carbon monoxide and the combustion of hydrogen at low temperatures. The CoY and CoYt models showed different performances in these reactions. The electron density map obtained during the exposure of the CoY model to the oxidation reaction conditions of the CO showed the cobalt within the smallest cavities of the zeolite. In the combustion of hydrogen in CoY, there was consumption of reagents in the temperature range of 30o to 100o C.

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 Natural gas is an important source for different processes and in particular for production of hydrogen for fuel cells. Reforming is used to produce synthesis gas, CO+H2 with H2/CO ratio about 2. Fuel cells need pure hydrogen and PROX is the most used process. In applied catalysis, such as PROX, we used different metals and supports1. The stream contains about 60% H2 and CO+O2. One observes CO oxidation at low temperatures and H2 oxidation forming respectively H2O and CO2. On the other hand, carbon dioxide and water are the lowest free energy substances in the cycles of generation and use of energy, which are also produced by converting chemicals from molecules, by burning of fuels and by breathing of living beings. Nowadays, the activation and transformation of CO2 and H2O molecules can also be a way to control the emission of carbon dioxide. Although high activation energy is required, the reorganization of chemical bonds allows transforming them into higher value-added compounds. Indeed, according to Izumi 2 solar energy could be a favorable economic option in the photocatalytic conversion of CO2 into fuel. However, the CO2 molecule is thermodynamically stable and is very difficult to activate because it requires modification of its electronic structure for a chemical reaction. The interactions between CO2 molecules and the metals are too weak, as shown by Freund 3 in adsorption studies on single crystals. Although there are exceptions, such as iron, nickel, and alkali-promoted metals, there are still contradictions about the spontaneous activation of CO2 by other metals. De la Peña O’Shea et al. 4 showed through density functional theory (DFT) that CO2 can interact with cobalt surfaces. Their DFT calculations predicted the structure


Page 2 of 36

Page 3 of 36 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


sensibility and suggested that cobalt nanoparticles terminated by (110) and (100) surfaces were active for reactions involving CO2 and facilitate its molecular dissociation. In contrast to the activation of CO2, there are only a few reports related to the activation of water on metals. Gong 5 calculated the formation of water from OH and H on cobalt surfaces, registering that the energy barrier was of the order of 155KJ/mol on Co (111) surface. Moreover, Tang 6 and Gohake 7, in water formation on Cu (110) and Cu (111) surfaces, found energy barriers of 60 KJ/mol and 120 KJ/mol, respectively. They also observed lower energy values when a monolayer of oxygen atoms covered these surfaces. This study demonstrated the influence of oxygen at the surface on activation of water molecules. Van Grootel 8 studied the activation of water on rhodium surfaces, considering the effect of the surface topology and the presence of oxygen pre-adsorbed on the activation process. For this system, authors claim that the activation of water does not depend either on the coordination atoms of rhodium surface and adsorbed oxygen species. Liu et al. 9 found by ambient pressure angle resolved X-ray Photoelectron Spectroscopy that a polycrystalline Co surface adsorbed CO2 as carbonates and co-adsorbed CO2 and H2O molecules forming methoxy and formate species. Liao et al. 10 studied photocatalytic water splitting using sunlight over cobalt nanoparticles. The process generated stoichiometric values of H2 and O2 without external electric polarization or sacrificial reagents, which makes the reaction simple and less expensive. Their results showing a quantum efficiency of 5% under visible radiation without co-catalysts were significant since the conventional catalysts for water-splitting have quantum efficiencies of 0.1-0.3%. Cobalt oxide nanoparticles-based materials are promising



photocatalysts for reactions involving the formation of oxygen and hydrogen 11 and artificial photosynthesis12. On the other hand, Bonenfant et al. 13 showed that the structure influences the adsorption of CO2 on zeolite materials. The presence and type of exchanged cations, besides the topology and ration Si/Al in the structure induced higher or lower basicity and differentiated electrical fields in the space cavities, provoking greater interaction with CO2 molecules, since they have quadrupole moments. The capacity of zeolite interaction with molecules depend on their topology, like geometry and channel sizes, or shape selectivity which allows to introduce nanoparticles and influence the molecules and their transition states during the reaction. Faujasites have approximately half of volume of the cell unit empty. The presence of species with charge in the zeolite structure produces electrical fields, which are enough to polarize confined molecules in the nanopores for activation of the molecules. In our previous study 14 we investigated the interaction of water with sodium Faujasite and determined the sodium and water sorption sites as a function of water content. We found that water forms a physisorbed shell-like layer inside the supercages, located at a distance of 2.8 Å from the pore wall oxygens. For water saturated Faujasite, this shell had a surface density comparable to liquid water. The ion exchange sites and ordered water found inside the super cage are within this layer. These results, although limited by the use of a conventional lab source Xray diffraction, were confirmed by a more detailed synchrotron study that found water forming tetrahedral clusters and ice-like hexamers inside the Faujasite cages15 and also investigated the thermal behavior of these structures. These results lead us to propose ordered nanoporous solids as model supports for heterogeneous catalysts. We recently showed 14 by using in situ X-ray diffraction (XRD) that


Page 4 of 36

Page 5 of 36 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


the crystal symmetry of a zeolite allows us to investigate the structure of adsorbatenanoporous solid complex in atomic detail. Inspecting the experimental electron density within the pores allows investigating the structural features of physisorbed matter inside the Faujasite nanochannels and nanocages. Our proposal is not intended to be an alternative to the well-established surface science methods to model heterogeneous catalysts, but we think of it as a complementary approach. For instance, by hosting active sites in the nanopores of a model-support, we lack the chance to ‘see’ them by the current analytical surface science techniques. Therefore, we can include porosity and surface curvature effects in the model. For a nanoporous solid its representative ‘surface’ is also part of this bulk structure. Heterogeneous model catalysts are traditionally made on single crystal surfaces in ultra-high vacuum (UHV), an ideal condition to investigate them through current surface science techniques. Although this approach provided an atomic-scale understanding of many catalyst surfaces 16, it can be performed in an environment quite different from that of industrial processes, where pressures are at least twelve orders of magnitude higher. In the language of model catalysis, this is called the pressure gap 17. The role of porosity is also overlooked in these model systems, since they are made on flat surfaces. The topological difference between model and industrial catalysts is called the material gap.

17

For the ordered

nanoporous model system, there is no need to concern about the pressure gap, since structural studies by x-rays and/or neutrons can be performed under the typical pressure, temperature and flow conditions of industrial processes. We choose a Faujasite (Y) zeolite as a model support, because its structure has a three dimensional system of interconnected nanosized channels and cavities that represent 50% (by volume) of the material. The sizes of these channels and cavities are big enough for the



transit of small molecules. Faujasite is an aluminosilicate framework made by (Si,Al)O4 tetrahedra linked by their corners. This structure is built by sodalite cages (6846), each one tetrahedrally linked to another four by the hexagonal prisms (46). Each silicon ion substitution by aluminum causes a unit negative charge in the framework. In order to turn the structure electrically neutral, positive ions sit at specific sites, these are called extra framework cation sites. Faujasite topology and extra framework cation site nomenclature 18 are shown in Figure 1. Site I is at the center of the hexagonal prism at x=0, y=0, z=0; site I’ is inside the sodalite cage, near the aperture to the prism; site II at the center of the hexagonal window between sodalite and supercage and sites III and III’ are in the supercage, close to the sodalite square window.

Figure 1. Faujasite topology and extraframework sites 16.

The objective of this work was to design and test heterogeneous model catalysts made by dispersing cobalt into a nanoporous zeolite Faujasite. We focus on in situ methods that can reveal the structure of the adsorbate-nanoporous complex during water and carbon dioxide interaction with isolated cobalt ions and cobalt oxide supported nanoparticles.


Page 6 of 36

Page 7 of 36 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


Materials and methods The material used was a commercial sodium saturated Faujasite in powder form (code CBV100) from Zeolyst International. The Faujasite, in powder form, was initially saturated with sodium, is the model support called NaY. Cobalt incorporation was made by ion exchange in an aqueous solution of cobalt nitrate hexahydrate at a concentration of 0.1 molar, producing a model named CoY. Another cobalt oxide-nanoparticle model was produced by treatment of the CoY with 0.1 molar sodium hydroxide solution, following the procedure suggested by Tang et al. 19 . The latter was named CoYt.

Chemical analysis Chemical analyses were performed by X-ray fluorescence (XRF), using a Rigaku RIX3100 sequential X-ray spectrometer with Rh tube (4kW). The zeolite powder was pressed into a pellet and calcined at 500oC in a muffle furnace during 4h before XRF analysis. The pellets were maintained at ~200oC and in vacuum (13 mPa) before and during XRF spectroscopy, in order to minimize adsorption of species from the atmosphere. In situ X-ray diffraction (conventional and synchrotron sources) We use two experimental setups, a conventional Cu k (=1.54056 Å) source X-ray diffractometer and another installed in the X-ray powder diffraction (XPD) beamline at Brazilian Synchrotron Light Laboratory (LNLS). Both have reaction chambers installed on the goniometers and allow in situ data collection in the temperature range from ambient to 900oC under gas flow. The conventional setup has been described in detail elsewhere. 14 For the synchrotron data collection, we tuned the beam energy to 7.5 KeV, (=1.6501 Å) which



is slightly less than the CoK absorption edge (7.712 KeV). This precaution avoided fluorescence by cobalt atoms. We use a scanning linear Mythen 1024 channel detector and vacuum path between both monochromator and chamber and chamber and detector. Exiting gases were analyzed by a quadrupole mass spectrometer Pfeiffer model Prisma 100. Simulations and Rietveld refinements of the powder diffraction data were performed by using the Fullprof Suite 20, Prima program 21 for maximum entropy calculations and Vesta 22,23 for crystal structure drawing.

Ultraviolet-Visible absorption spectroscopy (UV-VIS) UV-visible diffuse reflectance spectroscopy (DRS) analyses were performed using Varian Cary 5000 UV-vis spectrometer with a reduction chamber coupled to a Harrick diffuse reflectance accessory. Prior to analysis, the samples were dried at 300 °C under He flow (30 mL/min) for 30 min and then reduced at 300 ºC, 400 ºC or 500 °C using 50 mL/min flow rate of a reducing mixture containing 10% H2/He (v/v) for 1 h. The reflectance measurements were in the UV-Visible region (200-800 nm) scanning the whole range. The contribution of the support was removed by dividing the catalyst reflectance by the support reflectance measured in the same reduction temperature, before calculating the "Kubelka-Munk" function, F(R).


Page 8 of 36

Page 9 of 36 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


Results and Discussion Chemical analysis The unit cell composition in Table 1 was calculated from the X-ray fluorescence data. The total number of electrons in the unit cell (e/cell) was obtained by summing the atomic numbers from the formula components. Table 1. Chemical composition and formulas

% Sample

Formula

e/cell

Na2O

CoO

Al2O3

SiO2

NaY

13.6(2)

-

22.4(2)

63.8(3)

Na56Al56Si136O384

6320

CoY

4.23(3)

10.8(2)

22.6(2)

62.4(4)

Co19Na18Al56Si136O384

6415

CoYt

9.93(2)

9.89(2)

21.3(2)

58.9(4)

(Co3O4)4Co6Na43Al56Si130O384 6579

The zeolite NaY is fully saturated by Na+ ions, as the atomic ratio Na/Al determined of 1:1, resulting in electrostatic equilibrium structure. The CoY structure shows partial substitution of sodium by cobalt. The structural charge neutrality is achieved if we suppose that the cobalt ions are divalent. There are 19 divalent (Co2+) and 18 monovalent ions (Na+) inside the Faujasite unit cell, which means 56 positive charges to compensate for the 56 Al negative framework charges. However, after treatment with NaOH, it was observed the incorporation of an amount of Na and no significant loss of cobalt ions (CoYt). In fact, the NaOH treatment caused the precipitation of cobalt hydroxide in the zeolite cages, considering the cobalt moved from the position of charge compensation to form another structure. About six atoms of cobalt per unit cell remained in the compensation sites, while twelve Co atoms migrate to form cobalt oxide inside the zeolite channels. After calcination, we identified the presence of a nanocrystalline Co3O4 phase by X-ray diffraction as we will show forward in this text.



UV-visible spectroscopy Figure 2 shows UV-visible spectrum of CoY sample at different temperatures (30oC, 100oC, 200oC, 300ºC, and 400oC). As the temperature increases, besides the band at 520 nm attributed to cobalt in octahedral coordination (Co2+ (H2O)6), the spectrum exhibited bands at 534, 584 and 653 nm, assigned to Co2+ species in tetrahedral coordination 23. Moreover, the band at 300 nm corresponds to the Co2+ species in trigonal coordination. Figure 3 shows the spectra of CoYt sample also at 30oC, 100oC, 200oC, 300ºC, and 400oC. The absorption band at 300 nm observed at temperatures lower than 300º C is assigned to the trigonal coordination of cobalt

23.

Above this temperature, the absorption is significantly

higher due to the transformation of hydroxide (Co(OH)2) to the oxide (Co3O4).

Figure 2. UV-VIS absorption spectra of CoY after heating in He.


Page 10 of 36

Page 11 of 36 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 3. UV-VIS absorption spectra of CoYt after heating in He.

Electron density map by in-situ XRD The model catalyst must have free nanopores for the transit of molecules, so the inner surfaces were initially cleaned by calcination. In sequence, we carried out the adsorption of probe molecules (CO2 and H2O) and after, we performed the in situ reaction of CO oxidation and water formation. Experimental electron densities were obtained by in situ XRD for all these steps.

Calcination of the models The diffraction patterns of NaY, CoY and CoYt after calcination are displayed in Figure 4. The diffraction pattern of the model support NaY (Figure 4a) is characteristic of a Faujasite structure with space group 𝐹𝑑3𝑚 and matches the corresponding reference pattern. 24 As calcination temperature increases, the cell parameter increases. No additional peaks were observed in the diffraction patterns during calcination, apart from those that belong to the 𝐹𝑑



3𝑚 space group assigned to the NaY structure, evidencing no symmetry changes upon heating. Besides the 𝐹𝑑3𝑚 symmetry, the diffractogram of the CoY model evidenced significant alterations of relative peak intensities when compared to that of NaY. These are probably due to the different electron densities between cobalt and sodium and/or different site locations. The exchange of cobalt ions preserves Faujasite lattice symmetry, and no other phase appeared. Further, the diffraction peak widths of the NaY and CoY are similar, meaning that the cobalt ionic exchange does not alter the crystal domain sizes. In short, NaY and CoY diffraction patterns exhibited a single crystalline phase with Faujasite topology. Besides that, phase, CoYt showed segregated nanoparticles (Co3O4) with mean domain size around 30 Å. During calcination, cobalt hydroxide precipitated in the larger cavities and as the temperature increased, Co3O4 nanoparticles were formed.


Page 12 of 36

Page 13 of 36 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. Rietveld Refinement of the models (dark: experiments, red: calculated, and blue: difference of observed and calculated intensities). a) NaY calcined at 400oC; b) CoY calcined at 400oC; c) CoYt calcined at 400oC; highlighting the presence of Co3O4 nanoparticles (~30 Å).



Noteworthy is that in the calcination step the intensities of the peaks of all patterns increased significantly in the initial Bragg angular range (2

Design and Testing Model Cobalt Catalysts for Reactions Involving

Recommend Documents