Next Article in Journal
Upgrading Mixed Agricultural Plastic and Lignocellulosic Waste to Liquid Fuels by Catalytic Pyrolysis
Next Article in Special Issue
A Review of Biomass-Derived Heterogeneous Catalysts for Biodiesel Production
Previous Article in Journal
Site-Specific Covalent Immobilization of Methylobacterium extorquens Non-Blue Laccse Melac13220 on Fe3O4 Nanoparticles by Aldehyde Tag
Previous Article in Special Issue
A Review of CeO2 Supported Catalysts for CO2 Reduction to CO through the Reverse Water Gas Shift Reaction
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Catalysts
Volume 12
Issue 11
10.3390/catal12111380
catalysts-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Effects of Site Geometry and Local Composition on Hydrogenation of Surface Carbon to Methane on Ni, Co, and NiCo Catalysts

by
Sebastian Godoy
1,
Prashant Deshlahra
2,*,
Francisco Villagra-Soza
1,
Alejandro Karelovic
1 and
Romel Jimenez
1,*
1
Carbon and Catalysis Laboratory (CarboCat), Department of Chemical Engineering, Faculty of Engineering, University of Concepción, Concepción 4070409, Chile
2
Department of Chemical and Biological Engineering, Tufts University, Medford, MA 02155, USA
*
Authors to whom correspondence should be addressed.
Catalysts 2022, 12(11), 1380; https://doi.org/10.3390/catal12111380
Submission received: 9 October 2022 / Revised: 30 October 2022 / Accepted: 31 October 2022 / Published: 7 November 2022
(This article belongs to the Special Issue Applications of Heterogeneous Catalysts in Green Chemistry)
Browse Figures
Versions Notes

Abstract

:
Surface carbon deposits deactivate Ni and Co catalysts in reactions involving hydrocarbons and COx. Electronic properties, adsorption energies of H, C, and CHx species, and the energetics of the hydrogenation of surface C atom to methane are studied for (100) and (111) surfaces of monometallic Ni and Co, and bimetallic NiCo. The bimetallic catalyst exhibits a Co→Ni electron donation and a concomitant increase in the magnetization of Co atoms. The CHx species resulting from sequential hydrogenation are more stable on Co than on Ni atoms of the NiCo surfaces due to more favorable (C-H)–Co agostic interactions. These interactions and differences between Co and Ni sites are more significant for (111) than for (100) bimetallic surfaces. On (111) surfaces, CH is the most stable species, and the first hydrogenation of C atom exhibits the highest barrier, followed by the CH3 hydrogenation steps. In contrast, on (100) surfaces, surface C atom is the most stable species and CH2 or *CH3 hydrogenations exhibit the highest barriers. The Gibbs free energy profiles suggest that C removal on (111) surfaces is thermodynamically favorable and exhibits a lower barrier than on the (100) surfaces. Thus, the (100) surfaces, especially Ni(100), are more prone to C poisoning. The NiCo(100) surfaces exhibit weaker binding of C and CHx species than Ni(100) and Co(100), which improves C poisoning resistance and lowers hydrogenation barriers. These results show that the electronic effects of alloying Ni and Co strongly depend on the local site composition and geometry.

Graphical Abstract

1. Introduction

H2 and small carbon-containing species such as CO, CO2, and CH4 are produced and utilized as feedstocks in large industrial processes, making their interconversions important for the economy and environment. Prominent examples of such processes include CO and CO2 methanation, Fischer–Tropsch synthesis, methane reforming using H2O (steam reforming) or CO2 (dry reforming of methane, DRM), decomposition of larger hydrocarbons, and H2 production. The rational selection and design of catalysts that are active at milder reaction conditions, selective for the desired products, and can be used uninterruptedly for longer periods lead to more efficient and greener processes with lower energy requirements. Noble metals (e.g., Ru, Rh, and Pt) show high activity for methanation and DRM reactions [1,2] while being resistant to deactivation, but their scarcity and high cost limit their industrial applications. More abundant and cheaper Ni-based catalysts exhibit high selectivity for CH4 in methanation processes, and high activity for both methanation and DRM, but suffer from rapid deactivation due to carbon deposition. Therefore, understanding molecular details and site requirements for C deposition in Ni-based catalysts is important.
Cobalt is also more abundant than noble metals and shows high activity for methanation and DRM. More importantly, Co shows a higher resistance to carbon poisoning. However, it forms carbon deposits (sheets and filaments) in other reactions, such as ethanol dehydrogenation [3], and has been found to deactivate by oxidation at high temperatures used in DRM processes [4,5]. Co forms homogeneous alloys with nickel over wide composition and temperature ranges [6], making it a good candidate for a bimetallic CoNi catalyst. Alloying or doping Ni-based catalysts with transition metals has been shown to affect the resistance to carbon deposition and, in turn, the activity and selectivity for DRM and methanation [2,4,7,8,9]. This fine-tuning of catalytic (and electronic) properties via alloying has received significant attention in recent theoretical and experimental studies [2,8,10,11,12] and is a promising area for the development of more active, selective, and stable catalysts for a wide range of reactions.
For example, activity for the decomposition of ethanol on Co-mixed oxides was improved by the formation of active NiCo phases, but began to decrease after 14 h, showing significant deposition of carbonaceous species (carbon nano-filaments and multi-walled nanotubes according to the particle size) [13]. Takanabe et al. [4] experimentally studied Co-Ni/TiO2 catalysts in DRM conditions, showing that while >80%Ni alloys underwent C formation and >90%Co alloys deactivate by metal oxidation, the intermediate NiCo alloys performed stably with high activity. Adding oxygen to the feed has been reported to considerably decrease coke formation during DRM on Ni-rich NiCo catalysts [14]. The higher oxophilicity (affinity to bind O and OH) of Co surfaces has been proposed to propitiate oxidation and removal of surface carbon deposits, making it more resistant to coke poisoning [8,15]. Tu et al. [8] studied (111) surfaces by DFT and proposed that the higher surface O coverage on Co(111) and CoNi(111) allows an O-assisted dissociative adsorption of CH4 (not favored on pure Ni surfaces) that enhances methane activation in the DRM conditions. Chen and Yang recently studied NiCo step B5 sites by DFT, suggesting that higher *C coverage on Ni steps may relate to carbon deposition and higher *O coverage on Co steps may relate to deactivation by oxidation. Ou et al. [2] used DFT to study CH4 dehydrogenation to surface *C on clean and Co-doped Ni(111) surfaces, finding a negative effect of the Co doping for direct dissociative adsorption of CH4, which was suggested to be the rate-determining step. They also reported that the incorporation of Co weakened *C adsorption and lowered the barriers for the desired *C + *H → *CH step, which suggested the thermodynamically and kinetically enhanced removal of surface carbon on Co-doped (111) surfaces. The *CH dehydrogenation has been found to be a kinetically relevant barrier, sometimes competing with the dissociative adsorption of CH4 as the rate-determining step for surface *C production in various DRM-DFT studies of the close-packed Co [1,16], Ni(111) [17,18], and NiCo(111) surfaces [7,9,19], with one barrier or the other being higher depending on the specific system and conditions considered. For example, Liu et al. [19] agree with the results of Ou et al. [2] for higher CH4 dissociative adsorption barriers on NiCo(111) systems, but the formation of surface *C from *CH has been proposed as the highest overall barrier on Ni(111) [19] and Co(111) [7], also in agreement with Li et al.’s results [9] for Ni alloying with 9-group metals, and closely competing with CH4 dissociation on NiCo(111) [7].
Notably, these and most other studies use the close-packed (111) (and similar (0001) for Co) surfaces in density functional theory (DFT) modeling, with a few recent exceptions [20]. Although the (111) surfaces show clear differences between Ni and Co, other surfaces that are also present on monometallic Ni [21,22] and Co [23] clusters in smaller but still significant proportions may exhibit very different binding properties that make them worth studying in comparison to the (111) surface. For example, on Ni-based single-atom alloys, the CO dissociation has been shown to be endothermic on (111) surfaces but exothermic on (100) surfaces [11]. The dissociation routes (direct or H-assisted) are favored for Ni surfaces on (100) and step sites [24], which are also more reactive than close-packed facets for both CH4 steam reforming and the formation of C deposits over certain crystal sizes [25]. The hydrogenation of *CH3 (or the methane dissociative adsorption) has been suggested as the rate-limiting step for CH4 formation (or DRM) on various fcc and hcp Co surfaces [26,27] and activation barriers related to the corresponding surface d-band centers [26], but with different dominating paths and deactivation methods for Co(111) and Co(211) [26]. Stronger *C binding with lower mobility and considerably lower *CH dehydrogenation barriers are reported for Co [16] and Ni [17] (100) surfaces compared to the close-packed (111) surfaces [16], and the strong square planar adsorption of *C on Ni and Co (100) sites also promotes significant surface restructuration [28].
Thus, the surface geometry significantly affects CHx reactions on both monometallic Ni and Co surfaces. The most studied close-packed (111) facets tend to be less active for CHx conversions, but also more resistant to the formation of surface C formation, than the (100) facets. Overall activity, selectivity, and deactivation resistance may come from a combination of these (and other relevant) surfaces. These strong differences between monometallic facets and reported properties of NiCo(111) alloys led us to comparatively study paths for the elimination of surface *C on the monometallic and NiCo alloys of both (111) and (100) surfaces. The results presented here analyze in detail the differences between the reactivity of these surfaces and their links to the surface electronic properties resulting from the from the Ni–Co alloying.

2. Results and Discussion

2.1. Structure and Electronic Properties of Ni, Co, and NiCo

Nickel nanoparticles exhibit an fcc crystal structure. Cobalt can form both fcc and hcp (hexagonal close-packed) phases, depending on the preparation temperature and nanoparticle size [29], which also affects its catalytic properties [23]. For homogeneous NiCo alloys, the fcc phase is experimentally observed when the Ni content is greater than 35 wt.% [6] and forms naturally from NiCoO2 mixed oxide after reduction [13]. Only the fcc phase is considered in this work to avoid crystallographic effects and focus on alloying effects. The lattice constants fitting the RPBE-derived bulk energies to the Murnaghan equation of state are 3.55 Å for Ni, Co, and NiCo, within the accepted DFT error [20] but slightly larger than experimental values (near 3.52 Å) [6,30,31,32].
From the bulk model, (111) and (100) surfaces were constructed. Figure 1 shows the unique adsorption sites of the monometallic (a and b, Co as an example) and bimetallic (b and c) slab models. The Ni2Co or NiCo2 local surface composition of three-fold (111) sites distinguishes Ni-rich or Co-rich sites. The four-fold (100) sites are analogously distinguished by a subsurface Ni or Co atom (hN and hC in Figure 1d, respectively).
The atomic magnetic moments derived from the DDEC6 method for bulk Co and Ni are 1.66 and 0.65 μB, respectively (Table 1), in agreement with Bader partitions (SI.2, Table S2) and previously published values [5,33,34]. Surface atoms in monometallic (111) and (100) surfaces remain essentially neutral (DDEC6 electronic charges ~±0.01 e, Table 1) but exhibit higher magnetic moments than the bulk atoms (1.72 and 1.78 μB for Co, 0.67 and 0.70 μB for Ni, Table 1). DDEC6 and Bader charges show electron density donation from Co atoms to the more electronegative Ni atoms in bulk and surface models. The magnetic moment of Co@NiCo is higher than monometallic bulk Co and monometallic surface Co, but the magnetic moment of Ni@NiCo for the bulk and surfaces remains essentially the same as monometallic bulk Ni and slightly lower than monometallic surface Ni. The extent of charge donation and magnetic moment change in Co is higher for the less coordinated (100) surfaces than the (111) surface. This increase in magnetic moment of Co upon alloying but slight decrease for Ni is not explained by simple electronic donation and requires consideration of their corresponding density of states, as discussed next.
The projected density of states for the d-band (d-PDOS) of Co and Ni atoms in mono- and inter-metallic bulk, (111), and (100) surfaces are shown in Figure 2.
The spin-up components of Co atoms (Figure 2a,b) appear at significantly lower energies than the spin-down component. Spin-up bands are completely filled, while the spin-down bands span across the Fermi level, leading to a considerable fraction of empty states above the Fermi level. The Ni atoms (Figure 2c,d) contain relatively more symmetric spin-up and spin-down components at the low-energy end with the spin-up also completely filled but fewer unoccupied states above the Fermi level for the spin-down component. These differences in d-bands are consistent with the higher magnetization of Co than Ni. This asymmetry in band structure suggests that a simple Co→Ni electron donation will make Co band occupancies more asymmetric by withdrawing spin-down electrons near the Fermi level, consistent with the observed increase in Co magnetic moment. A corresponding increase in Ni spin-down electrons will decrease the Ni magnetic moment, qualitatively explaining the trends in Table 1. However, these magnetization changes are also affected by changes in the band structure caused by the interactions of Ni and Co bands in the alloy, as discussed next.
For bulk metals and surfaces, the alloying and the Co→Ni electron donation increase the amount of Co empty states just above the Fermi level, with only a slight reduction of occupied states around −0.5 eV. A similar reduction around that energy is also observed for the Ni atoms in the alloy, but in this case the density of unoccupied states is also reduced around the Fermi level and spread to higher energies, while the occupied states concentrate at lower energies compared to the monometallic Ni (~−1.2 eV).
These changes in band and filled states’ distributions can be represented by the Fermi-corrected d-PDOS first moment (i.e., the d-band center ed, Table 2 and vertical lines in Figure 2). For monometallic surfaces, the reported ed values are higher for Ni(100) than Ni(111) [17], but lower for Co(100) than Co(111) [1], which is consistent with the trends in Table 2. Such changes are attributed to the polarization of Co atoms. For Ni and Ni@NiCo atoms, there is little difference in ed between bulk and both surfaces, consistent with the smaller effect on magnetization in Table 1 compared to Co atoms. More importantly, the overall effect of the Ni–Co interaction is a reduction of the d-band centers for both Co and Ni in all bulk, (111), and (100) surfaces. Ni atoms also show a larger downward shift from mono- to bi-metallic even when only occupied levels are considered (see SI.3).
The atomic charges, band structures, and band center energies suggest that alloying significantly modifies the electronic properties of both Ni and Co atoms. Next, we discuss the effect of such modification on binding energies of surface species on monometallic and bimetallic surfaces.

2.2. Geometries and Adsorption Energies of C, H, and CHx Species

Previous DFT studies have shown that C, H, and CHx (x = 1–3) species preferentially bind to three- and four-fold sites of (111) and (100) Ni and Co surfaces, respectively, with the exception of CH3 that prefers a bidentate bridge site on (100) surfaces [35,36]. Our results are consistent with those reports, and we further probe in detail the differences between hcp and fcc sites, and differences from the local site composition on the (111) and (100) intermetallic surfaces. Figure 3 shows adsorption energies for the most stable geometries of adsorbed H, C, CH, CH2, and CH3 species on surfaces, and the symbol shapes for the different site geometries are ▲ fcc,▼ hcp, ▬ bridge, and ■ hollow (same as Figure 1, open and filled symbols represent Ni-rich and Co-rich sites, respectively) (images, energies, and geometric parameters can be found in SI.7).

2.2.1. H and Cn Adsorptions

Figure 3 shows that C and H atoms exhibit similar binding strengths on Ni(111) and Co(111) surfaces (Figure 3a,b), but both bind significantly more strongly on Ni(100) than Co(100) (Figure 3f,g). Moreover, while H binds more strongly on the (111) fcc sites than on the four-fold sites of (100) surfaces, C binds over 100 kJ/mol more strongly on the (100) surface than on the (111) hcp sites. For C on (111) surfaces, Co and the Co-rich intermetallic sites show similar binding energies, close to Ni and significantly stronger than the Ni-rich intermetallic site. For the intermetallic sites, weaker adsorptions are found for H on both surfaces and for C on (100), which is qualitatively consistent with the lower band centers according to the d-band model [37]. This anti-synergistic effect is more clear and significantly greater on the (100) surface than on the previously studied (111) surface [9,19], especially for C on NiCo(100), being 9 kJ/mol less stable than on Co(100) and 30 kJ/mol than on Ni(100). Weaker C and H binding also suggests a higher surface mobility that can promote hydrogenation, lessen carbon deposits, and render intermetallic sites less prone to poisoning.
The strong binding of C is consistent with the well-reported deactivation due to the formation of surface C deposits, especially for Ni-based catalysts. Such deactivation is also observed for Co in DRM conditions [27] but not usually reported, for example, under milder methanation conditions. The build-up of a small amount of C was studied through the formation of C-chains from atomic *C (Figure 4; images, energies, and geometric parameters are presented in SI.8), that are reported to be thermodynamically stable and have higher mobility than adatoms and serve as a precursor of star-like formations, rings, and later graphitic overlayers [38].
On (111) surfaces, the C-chains become more stable with increasing chain length (Figure 4a), compared to surface C species. In contrast, on (100) surfaces, the longer chains are less stable, which suggests that the high stability of C adsorbed on the four-fold M4 sites hinders the formation of C-C bonds. Other reconstruction or deactivation processes [28,39] may be more favorable on (100) surfaces. For the (111) intermetallic surface, two chain directions, one along and the other across the Co–Ni interface lines (Figure 4a, insets l- and t-), exhibit some differences in the energies of C2 and C3 chains but give similar energies for longer chains. For (100), C5 chains are similar or more stable than C4 chains, and C3 are more stable than C2. On both C2 and C4 chains, one terminal C ends on a less stable bridge site. Thus, the position of terminal C atoms instead of the direction of chains appears to dominate the formation energy of short linear Cn.

2.2.2. Adsorption of CHx Species

CH: From Figure 3c, on (111) surfaces, the hcp sites are preferred on monometallic Co and the NiCo alloy, but fcc sites are preferred on Ni(111). The CH species binding is strongest on Ni, followed by Co-rich NiCo sites, Co(111), and lastly, Ni-rich NiCo, for their respective preferred sites. These energy differences span a range of about 10 kJ/mol from the strongest binding Ni(111) and weakest binding Ni-rich NiCo(111). From Figure 3h, all (100) sites bind CH species at least 30 kJ/mol more strongly than the (111) sites of similar composition. Among the (100) sites, the CH binding is the strongest on Ni(100), followed by Co(100) and the Ni-rich NiCo(100) sites (17 kJ/mol weaker). In this case, the energy difference between the strongest binding Ni and weakest binding Co-rich NiCo sites is over 30 kJ/mol. Compared to pure Ni and Co surfaces, both NiCo sites appear less effective to stabilize C-containing surface species.
CH2: CH2 species adsorbs with an H atom above the C atom and the second H (with longer C-H bond) in a top or bridge position (Figure 3d,i insets, and SI.7, M-H distances between 1.7 and 1.8 Å). The fcc sites are preferred on Co(111), Ni(111), and the Co-rich NiCo(111) sites. The Co-rich NiCo(111) site adsorbs CH2 most strongly, followed by Ni(111), Co(111), and 10 kJ/mol more strongly than the Ni-rich NiCo(111) sites. On NiCo(111), *CH2 orientations with the second H on top of a Co atom are preferred over that on top of Ni atoms by at least 5 kJ/mol, suggesting that (C-H)–Co agostic interactions (i.e., interactions of C-H bonds with metal atoms [18,40]) are stronger than (C-H)–Ni interactions. Adsorption of CH2 on Ni(100) is about 12 kJ/mol stronger than on Co(100) and the Ni-rich NiCo(100) sites, but 27 kJ/mol stronger than the Co-rich sites on NiCo(100). On the NiCo(100) surfaces, the preference for Ni-rich sites suggests a predominant effect of the more stable Ni–Carbon interaction, as observed for C and CH (Figure 3g,h), rather than the (C-H)–Co agostic interactions that exist in (111) surfaces.
CH3: On (111) surfaces, CH3 sits in a three-fold site with its three H atoms on top of surface metal atoms (inset in Figure 3e, and SI.7, M-H distance between 2.0 and 2.1 Å). The fcc sites bind CH3 more strongly than the hcp sites by about 3 kJ/mol on both monometallic sites and by 5.6 kJ/mol on the NiCo(111) Co-rich sites. The Co and Co-rich (111) fcc sites exhibit the strongest binding, 6.7 kJ/mol stronger than Ni(111) and over 11 kJ/mol stronger than the hcp Ni-rich site on NiCo(111). On (100) surfaces, the CH3 species binds preferentially through C on a bridge site and one of the three H atoms over a surface metal (inset Figure 3j). The binding on Ni(100) is slightly stronger than on Co(100) (2 kJ/mol) but both monometallic sites bind CH3 more strongly than the NiCo(100) surface. In contrast to other CHx species, the Co-rich NiCo(100) sites bind CH3 more strongly than the Ni-rich NiCo(100), and all (100) surfaces bind CH3 weaker than the (111) surfaces and the Co-rich NiCo(100) site is preferred to the Ni-rich NiCo(100) site.
In summary, adsorption energies of H, C, and CHx species exhibit complex trends resulting from electronic and geometric factors. The adsorption energy of CHx species on (100) and (111) surfaces becomes less negative (weaker binding) for more hydrogenated species. On the (100) surface, all species exhibit similar trends for composition effects on adsorption energy, with bimetallic sites binding species less strongly than monometallic sites, which potentially makes *C adsorption energy a descriptor for CHx species. Similar trends have been described for CHx species on other (111) surfaces [41], but no significant trends exist between CH3 and C adsorption energies for Ni, Co, and NiCo (111) surfaces (see CHx vs. C relations in SI.9). For the three-fold sites on NiCo, adsorption of C and CHx species is stronger on the Co-rich sites than on the Ni-rich sites regardless of the distinction between fcc and hcp geometries. In contrast, on the four-fold NiCo(100) sites C, CH, and CH2 show a preferred adsorption on Ni-rich NiCo sites that is consistent with the stronger adsorption on Ni(100) compared to Co(100). For CH2 on (111) and CH3 on both surfaces, the preferred adsorption configuration allows favorable alignment of H atoms with surface metal atoms at short distances (1.7–2.1 Å), suggesting a role of agostic interactions between electron-deficient metal and C-H σ-bonds. The (C-H)–Co agostic interactions are stronger than (C-H)–Ni interactions, which accounts for stronger adsorption on Co(111) than on Ni(111), contrary to other CHx species. For the (100) surfaces, the Ni(100) is preferred to Co(100) and the Ni-rich sites are preferred on NiCo(100), which suggests that the strong interaction with Ni atoms in four-fold sites is the predominant effect. Weaker adsorption on Ni-rich sites is observed for all species and surfaces compared to the monometallic Ni and it is also weaker than monometallic Co in most cases, which suggest a strong anti-synergy from alloying Ni and Co in the less coordinated surface. From the Co → Ni electronic donation, the slightly more negative Ni atoms (also with fewer states near the Fermi level) form weaker binding sites for CHx. In turn, the more electronically unsaturated Co atoms offer stronger binding sites and are able to further stabilize C-H σ-bonds through agostic interactions. These Co atoms in the alloy provide more stable M-(C-H) centers that may also function as transition states for hydrogenation/dehydrogenation reactive steps, as discussed later.

2.3. Elementary Steps and Energetics for Hydrogenation of Surface Carbon

The sequential hydrogenation of an adsorbed C atom to methane was studied through the elementary steps in Scheme 1. Figure 5 shows the free energy reaction profile on the (111) and (100) surfaces (265 °C, 25 kPa H2, and 1 kPa CH4). Free energies, electronic energies, and the electronic energy profile are shown in the Supplementary Information (SI.10 and SI.11, respectively). The free energy for dissociative H2 co-adsorption is positive, suggesting absence of a pool of bound H atoms. Therefore, H atoms are added sequentially for each hydrogenation and referenced to gas-phase H2-free energy. Regardless, H2 dissociation and migration steps exhibit low barriers [42,43,44,45] (~10 kJ/mol) and are considered kinetically irrelevant compared to the hydrogenation steps. The transition states for the H2 dissociation are therefore not shown in Figure 5 for brevity. Figure 6 and Figure 7 show the geometries of steps R1–R4 on the (111) and (100) surfaces of the alloy (other surfaces are shown in SI.10, geometric parameters in Tables S12 and S13), and the exploration of different reaction paths for the kinetically relevant steps converged to the paths presented here.

2.3.1. Hydrogenation Steps on (111) Surfaces

The sequential hydrogenation of *C species requires H2 dissociation at an empty site and the migration of a *H species close to a CHx species. The C-H bond formation transition state for the first H addition (step R1, Scheme 1) can involve an H atom atop a surface metal atom or in the bridge position between two metal atoms (Figure 6a–c). The transition states with H atoms on top of metal atoms are more stable on Co(111) and NiCo(111) surfaces, consistent with the importance of (C-H)–metal agostic interactions observed in Section 2.2.2. In the case of NiCo(111), transitions started with H atop a Co atom are more stable than those atop Ni atoms. Subsequent H additions (steps R2–R4) also involve similar interactions involving H atoms in top positions and favoring Co atoms over Ni in the intermetallic surfaces.
On all (111) surfaces, the first hydrogenation step R1 (Scheme 1) is exoergic (dotted lines; ΔGr < 0; Figure 5), favoring the formation of *CH from *C. The *CH is the lowest free energy CHx species on the (111) surface for Co(111), NiCo(111), and Ni(111), making it the most abundant surface species. The transition states for the second (R2; TS.2) and third hydrogenations (R3; TS.3) on all (111) surfaces exhibit lower free energy than the first hydrogenation (TS.1). Thus, interconversion between *CH, *CH2, and *CH3 can occur more frequently than between *C and *CH. The transition state for the hydrogenation of *CH3 to form methane (TS.4) has higher free energy than the hydrogenation of *CH (TS.2) and *CH2 (TS.3) on all (111) surfaces, close to the free energy barrier for the first hydrogenation (TS.1) on Co(111) and NiCo(111), but 16 kJ/mol lower on Ni(111). This suggests that the first (R1) and last (R4) hydrogenations have similar kinetic relevance on Co(111) and NiCo(111) surfaces, but the first hydrogenation limits the rate on Ni(111) at the simulated conditions. Conditions that inhibit the H2 dissociative adsorption (higher ΔGr, e.g., for lower H2 pressure or higher temperature) will push the later hydrogenation barriers (TS.2, TS.3, and TS.4) to higher energies, thus making the last hydrogenation more relevant (see SI.11). At the reaction conditions, the complete hydrogenation of *C towards CH4(g) is thermodynamically favored on all (111) surfaces (ΔGr < −100 kJ/mol), similar on Co(111) and NiCo(111), and slightly more negative on Ni(111). The free energies of *CH, CH4(g), and the TS.4 transition state on NiCo(111) are similar to Co(111) (e.g., TS.4 ΔG 65, 61, and 47 kJ/mol in Co, NiCo, and Ni; Figure 5), suggesting that the equimolar intermetallic behaves kinetically more similar to the monometallic Co(111) than Ni(111). Ni catalysts are experimentally known to deactivate by C poisoning more than Co catalysts, but the TS.1 barriers are slightly lower on Ni(111) than on Co(111) and the TS.4 barriers for Ni(111) are 14 kJ/mol lower than for Co(111) or NiCo(111), suggesting a more facile elimination of *C through hydrogenation on Ni(111) surfaces. Therefore, even though the combination of *C in *Cn chains is favored on (111) surfaces (Section 2.2.1), the flat (111) surfaces are not the primary reason for deactivation processes by C poisoning.

2.3.2. Hydrogenation Steps on (100) Surfaces

The transition states for sequential hydrogenation of *C involve a bridging position of the first and second added H atoms (R1, R2; Figure 7). The third and the fourth H atoms are near the top of a metal atom in their respective hydrogenation transition states (R3, R4). The free energy of the first hydrogenation transition states (TS.1; Figure 5) on (100) surfaces is within 5 kJ/mol of values on corresponding (111) surfaces. However, on (100) surfaces, the first hydrogenation step R1 is endoergic (ΔGr > 17 kJ/mol) and all TS.2 free energies are higher than TS.1. This suggests a favored equilibrium towards *C rather than *CH. Moreover, for all (100) surfaces, the transition state energies are in the order TS.1 < TS.2 < TS.3~TS.4. Thus, all *CHx species favor further dehydrogenation and *C is the most abundant surface species. In particular, the small TS.2 (reverse) barriers compared to *CH2 suggest an equilibrium entirely displaced towards the dehydrogenation and any *CH2 formed quickly dissociates to *CH, especially on the Ni(100) surface. TS.4 presents the highest barrier for methane formation on all Ni(100) and Co(100) surfaces, but within 5 kJ/mol of the *CH2 hydrogenation barrier (TS.3). The significant dehydrogenation barriers for *CH3 (>57 kJ/mol, with TS.3 as the reverse barrier) suggest that *CH3 species may also be relevant for other reaction conditions. For instance, previous studies of the DMR process on (100) monometallic surfaces [16,17] coincide with the dissociative adsorption of CH4(g) as the highest barrier towards *C.
The TS.1 free energies are similar on all (100) surfaces (~74 kJ/mol), but the values for TS.2, TS.3, and TS.4 are lower on NiCo(100) than on monometallic (100) surfaces, suggesting a more facile hydrogenation of *C on NiCo(100) surfaces consistent with the experimentally observed C poisoning resistance of NiCo catalysts [46]. In particular, the higher TS.4 barrier on Ni(100) compared to Co(100) and NiCo(100) suggest a more difficult elimination of *C from the four-fold sites by hydrogenation on Ni(100). These *C do not recombine favorably to Cn chains (Section 2.2.1) but are hard to eliminate by hydrogenation (specially on the Ni(100) surface), and this suggests that *C species may distribute on (100) surfaces, blocking four-fold sites. The accumulation of *C on four-fold sites is related to deactivation near step edges [39] and propitiates surface restructuring of flat surfaces [28] (though such reconstructions are easier on Co than Ni surfaces). These results suggest that (100) surfaces are more relevant for explaining deactivation phenomena and the greater resistance of NiCo bimetallic sites to C poisoning.
The formation of CH4(g) from *C is thermodynamically more favored on NiCo(100) than on Co(100) and significantly less favored on Ni(100). This trend is consistent with the trends in adsorption energy discussed in Section 2.2, with all species (especially *C) binding more strongly on Ni(100) and significantly weaker on the NiCo(100) surface.

2.3.3. Effect of Density of States on Transition State Stabilization

As discussed in Section 2.3.1 and 2.3.2, all transition states on (111) and (100) surfaces involving H atoms on top of a metal atom are stabilized more when that metal atom is Co rather than Ni (Figure 5, Figure 6 and Figure 7). Similarly, as shown in Section 2.2, CHx species involving agostic interactions with Co are more stable than that with Ni. To analyze the difference between top-Co and top-Ni TS.4 transition states on the NiCo(111) and NiCo(100) surfaces in more detail, the atom-projected density of states (PDOS) for metal (Ni or Co), C, and H atoms was computed as shown in Figure 8. The interaction of the forming bond with the metal, i.e., M–(C-H(1)), is observed between −7.3 and −6.3 eV, while the non-interacting H(3) and H(4) only have a peak in the same position as the C atoms near −5.6 eV. On the (100) surfaces, the positions of the reacting peaks (−6.78 eV for Co, −6.83 eV for Ni; Figure 8b,d) are closer to the Fermi level than on (111) surfaces (−7.16 eV for Co, −7.22 eV for Ni; Figure 8a,c). On both (111) and (100) surfaces, these peaks are slightly closer to the Fermi level for the top-Co transition states, in agreement with reports for (111) bimetallic surfaces [2]. The H(2) atom in TS.4 is near the metal surface in the opposite position to the forming C-H(1) bond. This H(2) atom shows populations overlapping with the M-H(1) populations, and the non-reactive H(3) and H(4) populations, suggesting some interaction of H(2) with the metal that is consistent with the M–(C-H) interactions proposed in Section 2.2.2 and the high stability of the Co over Ni transition states. Therefore, the electronic and binding effects of alloying Ni and Co translate to kinetic effects that differentiate NiCo surfaces from the expected average of monometallic Ni and Co surfaces. Agostic interactions of the Co and Ni atoms (and to a smaller extent on Ni atoms) with the forming C-H(1) bond and for the opposite C-H(2) bond are also observed in the spin density isosurfaces in Figure S6 of the Supplementary Information (SI.13).
The electronic adsorption energies (Section 2.2.2), the reaction energies to form key intermediates, and transition states’ barriers suggest that the intermetallic surfaces behave more similarly to monometallic Co than Ni. Although, it should be noted that the effects of the Ni–Co interaction discussed here are specific to the local coordination degree, a specific reaction path, and are strongly related to the site composition, with Ni atoms in the alloy providing generally less stable adsorption and reaction sites compared to Co atoms in the alloy and monometallic Ni surfaces. The (111) surfaces are usually the only surfaces considered when studying the effects of alloying metals, but for a better description of the differences between bimetallic and monometallic catalysts, the key insights on the Ni–Co interactions derived from this work should be extended to study sites on other kinetically relevant surfaces (e.g., steps and kinks).

3. Methodology

The direct hydrogenation path from surface C to methane on Ni, Co, and NiCo surfaces was studied using density functional theory (DFT), as implemented in the VASP 5.4 code [47,48,49,50]. The revised PBE functional (RPBE [51]) was employed in spin-polarized calculations to appropriately deal with nickel and cobalt magnetism. Core electrons were described by the projector-augmented wave method [52,53]. A set of plane waves up to 460 eV was employed and a 0.05 Methfessel-Paxton [54] smearing of orbital populations was used to speed up calculations. All electronic energies were then extrapolated to 0 K. Energy differences under 10−6 eV between iterations are required for electronic convergence.
Face-centered cubic (fcc) crystal models were used to fit lattice parameters according to the Murnaghan equation of state [55] (see Section 2.1) and estimate atomic charges and magnetization of bulk monometallic Ni, Co, and equimolar NiCo intermetallic sites (see Supplementary Information, SI.1). One-sided slab models of (111) and (100) surfaces were constructed in 20 Å periodic cells using 4 layers of 16 atoms with o(4 × 4) arrangement and about a 15 Å vacuum between a slab and its periodic image. The resulting supercell sizes were 10.0 Å × 8.7 Å × 22.3 Å for (111) surfaces and 10.0 Å × 10.0 Å × 22.2 Å for (100) surfaces. Then, 12 × 12 × 12 and 4 × 4 × 1 Monkhorst-Pack [56] k-point meshes were employed for bulk and surface models, respectively. The top two layers and any adsorbed species were relaxed until all forces were below 0.02 eV/Å. Atomic charges and magnetism were estimated through Bader partitions [57] and the density-derived electrostatic and chemical (DDEC6) method [58] in the chargemol code [59] (see Supplementary Information, SI.2). Projected densities of states (PDOS) and their d-band centers were computed (see SI.3).
The electronic adsorption energy for A is defined as ΔEAd = ES·A − ES − EA, where ES·A, ES, and EA are, respectively, the electronic energy of the surface with the adsorbed species, clean surface, and free relaxed A species in a 15 Å empty box. Lower EAd values correspond to a stronger binding to the surface. Vibrational frequencies were computed through diagonalization of the hessian matrix constructed from finite displacements of atomic coordinates assuming parabolic potentials. Only the degrees of vibrational freedom of the free gases (H2, CH4) and adsorbed species were considered. Small frequencies were truncated to 100 cm−1 (see SI.4). Real frequencies allow checking for configurational stability and estimating zero-point vibrational energies (ZPVE), entropy, and enthalpy contributions to compute Gibbs free energies (see SI.5). Transition states (TS) were approximated by climbing image-nudged elastic band (CI-NEB) reaction paths [60] and then refined through a dimer relaxation [61] until all forces were below 0.02 eV/Å (see SI.6). A single imaginary frequency and its visualization confirmed that the TS connected initial and final structures. The reaction-free energy (ΔGr,s) of a step is the difference between its final and initial free energies, and activation-free energy barriers (ΔGa) are the difference between the free energy of the activated geometry and the free energy of the reference initial state (*C + 4H2(g) on the different surfaces at conditions described before).

4. Conclusions

For the NiCo alloy, DDEC6 and Bader charge partitions show that Ni atoms draw electronic charge from Co atoms. These charge donations are greater for the less coordinated (100) surfaces and disproportionately affect the spin magnetic moment of surface atoms, with a slight decrease for Ni atoms in the alloy regardless of the coordination degree but a significant increase for Co atoms in the alloy bulk and surfaces. The PDOS for the intermetallic surfaces shows a significant reorganization with more occupied and unoccupied states around the Fermi level for Co atoms in the alloy compared to monometallic Co. In contrast, Ni atoms show a lower population near the Fermi level, with grouping of occupied states below and unoccupied states above the Fermi level. The electronic adsorption energies of gaseous CHx fragments become less negative (weaker binding) with the hydrogenation degree. C, CH, and CH2 species bind more strongly on (100) surfaces, while H and CH3 species bind slightly more strongly on (111) surfaces. On NiCo(111), adsorption of CHx species is more stable on Co-rich sites and favors orientations that allow (C-H)–Co interactions. C and CHx species bind more strongly on Ni(100) than on Co(100). On NiCo(100), the Ni-rich sites are more stable than Co-rich sites, though both are significantly less stable than the average of the stability on monometallic Ni(100) and Co(100), suggesting a strong anti-synergy effect of alloying on binding strengths. Unlike the (111) surfaces, the formation of short Cn chains from *C is unfavorable on (100) surfaces due to the particularly strong binding of C on four-fold sites.
The hydrogenation of *C to CH4(g) is similarly favored on all (111) surfaces. The first hydrogenation is the highest barrier on (111) surfaces, followed by the last hydrogenation barrier, and *CH is the most stable surface species. Ni(111) is found to offer lower hydrogenation barriers than Co(111), with the first hydrogenation barrier being close to NiCo(111). Thus, the elimination of *C is more favored on Ni(111) and NiCo(111) than on Co(111). On (100) surfaces, the full hydrogenation towards CH4 is less favorable on Ni(100) than Co(100) and most favored on NiCo(100), which is attributed to the weakly bonded CHx species on NiCo(100). The hydrogenation of *CH2 and *CH3 are the main reaction barriers. *C instead of *CH is the most stable surface species. These computed free energies suggest that the greater deactivation by C poisoning of Ni catalysts compared to Co catalysts can be explained based on the energetics on the (100) surfaces, rather than (111). These energetics on (100) surfaces also show enhanced resistance to *C formation for the NiCo catalysts.
For the kinetically relevant steps on the NiCo surfaces, lower barriers are observed for transition states involving H on top of surface Co rather than surface Ni atoms, which is consistent with electron-deficient Co atoms on the alloy surfaces better-stabilizing hydrogenation transition states. This also explains the binding strength trends for NiCo(111), though the interaction of C with Ni is still more stable on the four-fold (100) sites, showing that the effects of the bimetallic interaction are strongly affected by surface coordination. Less coordinated sites and surface restructuring may also be involved in deactivation, but (111) and (100) are likely more abundant sites in fresh catalysts. DFT results in this work provide insight at the electronic level of key aspects that differentiate the kinetic behavior of alloyed NiCo from monometallic Ni and Co surfaces of different coordination degrees.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/catal12111380/s1, Table S1: Optimized lattice parameters. Table S2: Bader and DDEC6 atomic electronic charges. Table S3: Bader and DDEC6 atomic spin magnetic moments. Table S4: Average occupied d-band centers (up to the Fermi level) for bulk and surfaces. Figure S1: d-band projected density of states (PDOS) in bulk, (100), and (111) surfaces. Table S5: Frequencies from hessian matrices (different steps). Table S6: Translational and rotational thermodynamic corrections. Table S7: Vibrational frequencies, physisorbed CH4. Figure S2: *C adsorption energies as descriptors of *CHx adsorption energies. Tables S8 and S9: Gibbs free energy differences for initial geometries (ΔGi) with *C+4H2(g) as a reference, for reaction steps (ΔGr,s), activation step barriers (ΔGa,s), and reverse step barriers (in parenthesis) on (111) and (100) surfaces. Tables S10 and S11: Electronic energy differences for initial geometries (ΔEi) with *C+4H2(g) as a reference, for reaction steps (ΔEr,s), activation step barriers (ΔEa,s), and reverse step barriers (in parenthesis) on (111) and (100) surfaces. Scheme S1: Reaction steps for the construction of energy profiles. Figures S3–S5: Alternative electronic and free energy profiles. Tables S12 and S13: Relevant geometric parameters of the hydrogenation steps (with H*) on (111) and (100) surfaces. Figure S6: Spin densities for the activated state of the *CH3 hydrogenation [6,11,26,30,31,32,55,62].

Author Contributions

Conceptualization, S.G., P.D. and R.J.; methodology, S.G. and P.D.; software, S.G.; formal analysis, S.G. and P.D.; data curation, S.G. and F.V.-S.; investigation, S.G., F.V.-S., A.K. and R.J.; writing—original draft preparation, S.G.; writing—review and editing, P.D.; supervision, P.D., A.K. and R.J.; funding acquisition, R.J. All authors have read and agreed to the published version of the manuscript.

Funding

Authors acknowledge financial support of the FONDECYT regular project 1170610. S.G. acknowledges financial support of the ANID national scholarship, ANID BECAS/DOCTORADO NACIONAL 2018/21180468. F.V.-S. acknowledges financial support of the ANID national scholarship, ANID BECAS/DOCTORADO NACIONAL 2021/21210586. Powered@NLHPC: This research was partially supported by the supercomputing infrastructure of the NLHPC (ECM-02). Some computations were performed with resources provided by the Kultrun Astronomy Hybrid Cluster via the projects Conicyt Programa de Astronomia Fondo Quimal 2017 QUIMAL170001, Conicyt PIA ACT172033, and Fondecyt Iniciacion 11170268. P.D. acknowledges support from the National Science Foundation (award number 2034911) and computing resources from eXtreme Science and Engineering Discovery environment (project number TG-CTS150005).

Data Availability Statement

The data presented in this study are detailed in the supplementary information, models and further details are available on request from the corresponding author.

Acknowledgments

S.G acknowledges the technical support and recommendations from the supercomputing infrastructure of the NLHPC team and the Kultrun Astronomy Hybrid Cluster team.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Huang, H.; Yu, Y.; Zhang, M. Mechanistic insight into methane dry reforming over cobalt: A density functional theory study. Phys. Chem. Chem. Phys. 2020, 22, 27320–27331. [Google Scholar] [CrossRef] [PubMed]
  2. Ou, Z.; Ran, J.; Niu, J.; Zhang, Z.; Deng, T.; He, Z.; Qin, C. Effect of active site and charge transfer on methane dehydrogenation over different Co doped Ni surfaces by density functional theory. Int. J. Hydrogen Energy 2020, 45, 31849–31862. [Google Scholar] [CrossRef]
  3. Ashok, A.; Kumar, A.; Bhosale, R.; Saad, M.A.S.; AlMomani, F.; Tarlochan, F. Study of ethanol dehydrogenation reaction mechanism for hydrogen production on combustion synthesized cobalt catalyst. Int. J. Hydrogen Energy 2017, 42, 23464–23473. [Google Scholar] [CrossRef]
  4. Takanabe, K.; Nagaoka, K.; Nariai, K.; Aika, K. Titania-supported cobalt and nickel bimetallic catalysts for carbon dioxide reforming of methane. J. Catal. 2005, 232, 268–275. [Google Scholar] [CrossRef]
  5. Kocić, S.; Corral Valero, M.; Schweitzer, J.-M.; Raybaud, P. Surface speciation of Co based Fischer-Tropsch catalyst under reaction conditions: Deactivation by coke or by oxidation? Appl. Catal. A Gen. 2020, 590, 117332. [Google Scholar]
  6. Liu, Y.; Yang, H.; Liu, Y.; Jiang, B.; Ding, J.; Woodward, R. Thermally induced fcc to hcp martensitic transformation in Co–Ni. Acta Mater. 2005, 53, 3625–3634. [Google Scholar] [CrossRef]
  7. Liu, H.; Wang, B.; Fan, M.; Henson, N.; Zhang, Y.; Towler, B.F.; Harris, H.G. Study on carbon deposition associated with catalytic CH4 reforming by using density functional theory. Fuel 2013, 113, 712–718. [Google Scholar] [CrossRef]
  8. Tu, W.; Ghoussoub, M.; Singh, C.V.; Chin, Y.-H.C. Consequences of Surface Oxophilicity of Ni, Ni-Co, and Co Clusters on Methane Activation. J. Am. Chem. Soc. 2017, 139, 6928–6945. [Google Scholar] [CrossRef]
  9. Li, K.; Jiao, M.; Wang, Y.; Wu, Z. CH4 dissociation on NiM(111) (M = Co, Rh, Ir) surface: A first-principles study. Surf. Sci. 2013, 617, 149–155. [Google Scholar] [CrossRef]
  10. Ray, K.; Sandupatla, A.S.; Deo, G. Activity and stability descriptors of Ni based alloy catalysts for dry reforming of methane: A density functional theory study. Int. J. Quantum Chem. 2020, 121, e26580. [Google Scholar] [CrossRef]
  11. Sprowl, L.H.; Adam, B.M.; Tucker, J.D.; Árnadóttir, L. First-principles study of the products of CO2 dissociation on nickel-based alloys: Trends in energetics with alloying element. Surf. Sci. 2018, 677, 219–231. [Google Scholar] [CrossRef]
  12. Turap, Y.; Wang, I.; Fu, T.; Wu, Y.; Wang, Y. Co–Ni alloy supported on CeO2 as a bimetallic catalyst for dry reforming of methane. Int. J. Hydrogen Energy 2020, 45, 6538–6548. [Google Scholar] [CrossRef]
  13. Ashok, A.; Kumar, A.; Ponraj, J.; Mansour, S.A.; Tarlochan, F. Effect of Ni incorporation in cobalt oxide lattice on carbon formation during ethanol decomposition reaction. Appl. Catal. B Environ. 2019, 254, 300–311. [Google Scholar] [CrossRef]
  14. Leba, A.; Yıldırım, R. Determining most effective structural form of nickel-cobalt catalysts for dry reforming of methane. Int. J. Hydrogen Energy 2020, 45, 4268–4283. [Google Scholar] [CrossRef]
  15. Guo, X.; Liu, H.; Wang, B.; Wang, Q.; Zhang, R. Insight into C + O(OH) reaction for carbon elimination on different types of CoNi(111) surfaces: A DFT study. RSC Adv. 2015, 5, 19970–19982. [Google Scholar] [CrossRef]
  16. Hao, X.; Wang, Q.; Li, D.; Zhang, R.; Wang, B. The adsorption and dissociation of methane on cobalt surfaces: Thermochemistry and reaction barriers. RSC Adv. 2014, 4, 43004–43011. [Google Scholar] [CrossRef]
  17. Li, J.; Croiset, E.; Ricardez-Sandoval, L. Methane dissociation on Ni (100), Ni (111), and Ni (553): A comparative density functional theory study. J. Mol. Catal. A Chem. 2012, 365, 103–114. [Google Scholar] [CrossRef]
  18. Mueller, J.E.; van Duin, A.C.T.; Goddard, W.A. Structures, Energetics, and Reaction Barriers for CHx Bound to the Nickel (111) Surface. J. Phys. Chem. C 2009, 113, 20290–20306. [Google Scholar] [CrossRef] [Green Version]
  19. Liu, H.; Zhang, R.; Yan, R.; Wang, B.; Xie, K. CH4 dissociation on NiCo (111) surface: A first-principles study. Appl. Surf. Sci. 2011, 257, 8955–8964. [Google Scholar] [CrossRef]
  20. Chen, S.; Yang, B. Activity and stability of alloyed NiCo catalyst for the dry reforming of methane: A combined DFT and microkinetic modeling study. Catal. Today 2021, 400–401, 59–65. [Google Scholar] [CrossRef]
  21. Meltzman, H.; Chatain, D.; Avizemer, D.; Besmann, T.M.; Kaplan, W.D. The equilibrium crystal shape of nickel. Acta Mater. 2011, 59, 3473–3483. [Google Scholar] [CrossRef]
  22. Zhang, W.-B.; Chen, C.; Zhang, S.-Y. Equilibrium Crystal Shape of Ni from First Principles. J. Phys. Chem. C 2013, 117, 21274–21280. [Google Scholar] [CrossRef]
  23. Liu, J.-X.; Su, H.-Y.; Sun, D.-P.; Zhang, B.-Y.; Li, W.-X. Crystallographic Dependence of CO Activation on Cobalt Catalysts: HCP versus FCC. J. Am. Chem. Soc. 2013, 135, 16284–16287. [Google Scholar] [CrossRef] [PubMed]
  24. Andersson, M.; Abild-Pedersen, F.; Remediakis, I.; Bligaard, T.; Jones, G.; Engbæk, J.; Lytken, O.; Horch, S.; Nielsen, J.H.; Sehested, J. Structure sensitivity of the methanation reaction: H2-induced CO dissociation on nickel surfaces. J. Catal. 2008, 255, 6–19. [Google Scholar] [CrossRef]
  25. Bengaard, H.S.; Nørskov, J.K.; Sehested, J.; Clausen, B.S.; Nielsen, L.P.; Molenbroek, A.M.; Rostrup-Nielsen, J.R. Steam Reforming and Graphite Formation on Ni Catalysts. J. Catal. 2002, 209, 365–384. [Google Scholar] [CrossRef]
  26. Huang, H.; Yu, Y.; Zhang, M. Structure sensitivity of CH4 formation from successive hydrogenation of C on cobalt: Insights from density functional theory. Chem. Phys. Lett. 2019, 737, 136824. [Google Scholar] [CrossRef]
  27. Chen, S.; Zaffran, J.; Yang, B. Dry reforming of methane over the cobalt catalyst: Theoretical insights into the reaction kinetics and mechanism for catalyst deactivation. Appl. Catal. B Environ. 2020, 270, 118859. [Google Scholar] [CrossRef]
  28. Nandula, A.; Trinh, Q.T.; Saeys, M.; Alexandrova, A.N. Origin of Extraordinary Stability of Square-Planar Carbon Atoms in Surface Carbides of Cobalt and Nickel. Angew. Chem. Int. Ed. 2015, 54, 5312–5316. [Google Scholar] [CrossRef] [Green Version]
  29. Kitakami, O.; Sato, H.; Shimada, Y.; Sato, F.; Tanaka, M. Size effect on the crystal phase of cobalt fine particles. Phys. Rev. B 1997, 56, 13849–13854. [Google Scholar] [CrossRef]
  30. Pearson, W.B.; Thompson, L.T. The lattice spacings of nickel solid solutions. Can. J. Phys. 1957, 35, 349–357. [Google Scholar] [CrossRef]
  31. Wang, H.; Miller, J.T.; Shakouri, M.; Xi, C.; Wu, T.; Zhao, H.; Akatay, M.C. XANES and EXAFS studies on metal nanoparticle growth and bimetallic interaction of Ni-based catalysts for CO2 reforming of CH4. Catal. Today 2012, 207, 3–12. [Google Scholar] [CrossRef]
  32. Zhang, H.; Yao, T.; Sun, Z.; Li, Y.; Liu, Q.; Hu, F.; Pan, Z.; He, B.; Xie, Z.; Wei, S. Structural Study on Co−Ni Bimetallic Nanoparticles by X-ray Spectroscopy. J. Phys. Chem. C 2010, 114, 13596–13600. [Google Scholar] [CrossRef]
  33. Liu, X.; Sun, L.; Deng, W.-Q. Theoretical Investigation of CO2 Adsorption and Dissociation on Low Index Surfaces of Transition Metals. J. Phys. Chem. C 2018, 122, 8306–8314. [Google Scholar] [CrossRef]
  34. Peng, G.; Sibener, S.; Schatz, G.C.; Mavrikakis, M. CO2 Hydrogenation to Formic Acid on Ni(111). J. Phys. Chem. C 2012, 116, 3001–3006. [Google Scholar] [CrossRef]
  35. Watwe, R.; Bengaard, H.; Rostrup-Nielsen, J.; Dumesic, J.; Nørskov, J. Theoretical Studies of Stability and Reactivity of CHx Species on Ni(111). J. Catal. 2000, 189, 16–30. [Google Scholar] [CrossRef]
  36. Wang, S.-G.; Cao, D.-B.; Li, Y.-W.; Wang, J.; Jiao, H. CH4 dissociation on Ni surfaces: Density functional theory study. Surf. Sci. 2006, 600, 3226–3234. [Google Scholar] [CrossRef]
  37. Hammer, B.; Nørskov, J. Electronic factors determining the reactivity of metal surfaces. Surf. Sci. 1995, 343, 211–220. [Google Scholar] [CrossRef]
  38. Cheng, D.; Barcaro, G.; Charlier, J.-C.; Hou, M.; Fortunelli, A. Homogeneous Nucleation of Graphitic Nanostructures from Carbon Chains on Ni(111). J. Phys. Chem. C 2011, 115, 10537–10543. [Google Scholar] [CrossRef]
  39. Tan, K.F.; Xu, J.; Chang, J.; Borgna, A.; Saeys, M. Carbon deposition on Co catalysts during Fischer–Tropsch synthesis: A computational and experimental study. J. Catal. 2010, 274, 121–129. [Google Scholar] [CrossRef]
  40. Robinson, J.; Woodruff, D.P. The local adsorption geometry of CH3 and NH3 on Cu(111): A density functional theory study. Surf. Sci. 2002, 498, 203–211. [Google Scholar] [CrossRef]
  41. Abild-Pedersen, F.; Greeley, J.; Studt, F.; Rossmeisl, J.; Munter, T.R.; Moses, P.G.; Skúlason, E.; Bligaard, T.; Nørskov, J.K. Scaling Properties of Adsorption Energies for Hydrogen-Containing Molecules on Transition-Metal Surfaces. Phys. Rev. Lett. 2007, 99, 016105. [Google Scholar] [CrossRef] [PubMed]
  42. Greeley, J.; Mavrikakis, M. Surface and Subsurface Hydrogen: Adsorption Properties on Transition Metals and Near-Surface Alloys. J. Phys. Chem. B 2005, 109, 3460–3471. [Google Scholar] [CrossRef] [PubMed]
  43. Catapan, R.C.; Oliveira, A.A.M.; Chen, Y.; Vlachos, D.G. DFT Study of the Water–Gas Shift Reaction and Coke Formation on Ni(111) and Ni(211) Surfaces. J. Phys. Chem. C 2012, 116, 20281–20291. [Google Scholar] [CrossRef]
  44. Zhu, Y.-A.; Chen, D.; Zhou, X.-G.; Yuan, W.-K. DFT studies of dry reforming of methane on Ni catalyst. Catal. Today 2009, 148, 260–267. [Google Scholar] [CrossRef]
  45. Ren, J.; Guo, H.; Yang, J.; Qin, Z.; Lin, J.; Li, Z. Insights into the mechanisms of CO2 methanation on Ni(111) surfaces by density functional theory. Appl. Surf. Sci. 2015, 351, 504–516. [Google Scholar] [CrossRef] [Green Version]
  46. Bian, Z.; Kawi, S. Highly carbon-resistant Ni–Co/SiO2 catalysts derived from phyllosilicates for dry reforming of methane. J. CO2 Util. 2017, 18, 345–352. [Google Scholar] [CrossRef]
  47. Kresse, G.; Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 1993, 47, 558–561. [Google Scholar] [CrossRef]
  48. Kresse, G.; Hafner, J. Ab initio molecular-dynamics simulation of the liquid-metal--amorphous-semiconductor transition in germanium. Phys. Rev. B 1994, 49, 14251–14269. [Google Scholar] [CrossRef]
  49. Kresse, G.; Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 1996, 6, 15–50. [Google Scholar] [CrossRef]
  50. Kresse, G.; Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996, 54, 11169–11186. [Google Scholar] [CrossRef]
  51. Hammer, B.; Hansen, L.B.; Nørskov, J.K. Improved adsorption energetics within density-functional theory using revised Perdew-Burke-Ernzerhof functionals. Phys. Rev. B 1999, 59, 7413–7421. [Google Scholar] [CrossRef]
  52. Blöchl, P.E. Projector augmented-wave method. Phys. Rev. B 1994, 50, 17953–17979. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  53. Kresse, G.; Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 1999, 59, 1758–1775. [Google Scholar] [CrossRef]
  54. Methfessel, M.; Paxton, A.T. High-precision sampling for Brillouin-zone integration in metals. Phys. Rev. B 1989, 40, 3616. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  55. Tyuterev, V.G.; Vast, N. Murnaghan’s equation of state for the electronic ground state energy. Comput. Mater. Sci. 2006, 38, 350–353. [Google Scholar] [CrossRef]
  56. Monkhorst, H.J.; Pack, J.D. Special points for Brillouin-zone integrations. Phys. Rev. B 1976, 13, 5188. [Google Scholar] [CrossRef]
  57. Tang, W.; Sanville, E.; Henkelman, G. A grid-based Bader analysis algorithm without lattice bias. J. Phys. Condens. Matter 2009, 21, 084204. [Google Scholar] [CrossRef]
  58. Manz, T.A.; Limas, N.G. Introducing DDEC6 atomic population analysis: Part 1. Charge partitioning theory and methodology. RSC Adv. 2016, 6, 47771–47801. [Google Scholar] [CrossRef] [Green Version]
  59. Manz, T.A.; Sholl, D.S. Improved Atoms-in-Molecule Charge Partitioning Functional for Simultaneously Reproducing the Electrostatic Potential and Chemical States in Periodic and Nonperiodic Materials. J. Chem. Theory Comput. 2012, 8, 2844–2867. [Google Scholar] [CrossRef]
  60. Henkelman, G.; Uberuaga, B.P.; Jónsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 2000, 113, 9901–9904. [Google Scholar] [CrossRef] [Green Version]
  61. Henkelman, G.; Jónsson, H. A dimer method for finding saddle points on high dimensional potential surfaces using only first derivatives. J. Chem. Phys. 1999, 111, 7010–7022. [Google Scholar] [CrossRef]
  62. Liu, B.; Yuan, F.; Jin, K.; Zhang, Y.; Weber, W.J. Ab initio molecular dynamics investigations of low-energy recoil events in Ni and NiCo. J. Phys. Condens. Matter 2015, 27, 435006. [Google Scholar] [CrossRef] [PubMed]
Catalysts 12 01380 g001 550
Figure 1. Unique adsorption sites on surface models: (a) Co(111), (b) Co(100), (c) NiCo(111), and (d) NiCo(100). Symbols according to site geometry: ● t for top, ▲ fcc, ▼ hcp, ▬ b for bridge, and ■ h for four-fold hollow sites, filled symbols for Co or Co-rich sites, open symbols on Ni or Ni-rich sites.
Figure 1. Unique adsorption sites on surface models: (a) Co(111), (b) Co(100), (c) NiCo(111), and (d) NiCo(100). Symbols according to site geometry: ● t for top, ▲ fcc, ▼ hcp, ▬ b for bridge, and ■ h for four-fold hollow sites, filled symbols for Co or Co-rich sites, open symbols on Ni or Ni-rich sites.
Catalysts 12 01380 g001
Catalysts 12 01380 g002 550
Figure 2. d-band projected density of states (PDOS) in bulk, (100), and (111) surfaces for (a) Co, (b) Co in NiCo, (c) Ni in NiCo, and (d) Ni. Vertical lines show the corresponding d-band centers.
Figure 2. d-band projected density of states (PDOS) in bulk, (100), and (111) surfaces for (a) Co, (b) Co in NiCo, (c) Ni in NiCo, and (d) Ni. Vertical lines show the corresponding d-band centers.
Catalysts 12 01380 g002
Catalysts 12 01380 g003 550
Figure 3. Adsorption energies (ΔEad) on Co, NiCo, and Ni for H (a,f), C (b,g), CH (c,h), CH2 (d,i), and CH3 (e,j) on (111) and (100) surfaces, respectively. Sites geometries: ▲ fcc, ▼ hcp, ▬ bridge, and ■ hollow, using filled symbols on Co and Co-rich sites and open symbols on Ni and Ni-rich sites. Top views of geometries are shown as insets.
Figure 3. Adsorption energies (ΔEad) on Co, NiCo, and Ni for H (a,f), C (b,g), CH (c,h), CH2 (d,i), and CH3 (e,j) on (111) and (100) surfaces, respectively. Sites geometries: ▲ fcc, ▼ hcp, ▬ bridge, and ■ hollow, using filled symbols on Co and Co-rich sites and open symbols on Ni and Ni-rich sites. Top views of geometries are shown as insets.
Catalysts 12 01380 g003
Catalysts 12 01380 g004 550
Figure 4. Surface C addition, reaction of electronic energies on (a) (111) and (b) (100) surfaces.
Figure 4. Surface C addition, reaction of electronic energies on (a) (111) and (b) (100) surfaces.
Catalysts 12 01380 g004
Catalysts 12 01380 sch001 550
Scheme 1. Steps involved in the hydrogenation of *C to CH4. Surface sites are represented by *.
Scheme 1. Steps involved in the hydrogenation of *C to CH4. Surface sites are represented by *.
Catalysts 12 01380 sch001
Catalysts 12 01380 g005 550
Figure 5. Gibbs free energy profiles for the hydrogenation of *C to CH4(g) on (a) Co, (b) NiCo and (c) Ni, (111) (dotted lines) and (100) (solid lines) surfaces (265 °C, 25 kPa H2, 1 kPa CH4).
Figure 5. Gibbs free energy profiles for the hydrogenation of *C to CH4(g) on (a) Co, (b) NiCo and (c) Ni, (111) (dotted lines) and (100) (solid lines) surfaces (265 °C, 25 kPa H2, 1 kPa CH4).
Catalysts 12 01380 g005
Catalysts 12 01380 g006 550
Figure 6. Hydrogenation sequence on the NiCo(111) surface to form (a–c) *CH, (d) *CH2, (e) *CH3 and (f,g) *CH4(phys). Atom colors: Cyan—Co, purple—Ni, gray—C, white—H.
Figure 6. Hydrogenation sequence on the NiCo(111) surface to form (a–c) *CH, (d) *CH2, (e) *CH3 and (f,g) *CH4(phys). Atom colors: Cyan—Co, purple—Ni, gray—C, white—H.
Catalysts 12 01380 g006
Catalysts 12 01380 g007 550
Figure 7. Hydrogenation steps on the NiCo(100) surface to form (a) *CH, (b) CH2, (c,d) *CH3 and (e,f) *CH4(phys). Atom colors: Cyan—Co, purple—Ni, gray—C, white—H.
Figure 7. Hydrogenation steps on the NiCo(100) surface to form (a) *CH, (b) CH2, (c,d) *CH3 and (e,f) *CH4(phys). Atom colors: Cyan—Co, purple—Ni, gray—C, white—H.
Catalysts 12 01380 g007
Catalysts 12 01380 g008 550
Figure 8. PDOS for the transition state *CH3 + *H → CH4(phys.) + 2*: (a) top-C path on NiCo(111), (b) top-Co path on NiCo(100), (c) top-Ni on NiCo(111), and (d) top-Ni on NiCo(100).
Figure 8. PDOS for the transition state *CH3 + *H → CH4(phys.) + 2*: (a) top-C path on NiCo(111), (b) top-Co path on NiCo(100), (c) top-Ni on NiCo(111), and (d) top-Ni on NiCo(100).
Catalysts 12 01380 g008
Table
Table 1. DDEC6 atomic (average) electronic charges and spin magnetic moments.
Table 1. DDEC6 atomic (average) electronic charges and spin magnetic moments.
Electronic Charges (e)Spin Magnetic Moment (μB)
Bulk(111)surf.(100)surf.Bulk(111)surf.(100)surf.
Co0−0.01+0.011.661.721.82
Co(@ NiCo)+0.04+0.04+0.101.721.811.89
Ni(@ NiCo)−0.04−0.05−0.090.650.640.65
Ni0−0.01+0.010.650.670.70
Table
Table 2. Average d-band centers (ed in eV) for bulk, (111), and (100) surfaces.
Table 2. Average d-band centers (ed in eV) for bulk, (111), and (100) surfaces.
Bulk(111)surf.(100)surf.
Co−1.01−1.23−1.32
Co@ NiCo−1.08−1.32−1.38
NiCo (average)−1.24−1.38−1.40
Ni @ NiCo−1.39−1.43−1.41
Ni−1.21−1.26−1.18
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Godoy, S.; Deshlahra, P.; Villagra-Soza, F.; Karelovic, A.; Jimenez, R. Effects of Site Geometry and Local Composition on Hydrogenation of Surface Carbon to Methane on Ni, Co, and NiCo Catalysts. Catalysts 2022, 12, 1380. https://doi.org/10.3390/catal12111380

AMA Style

Godoy S, Deshlahra P, Villagra-Soza F, Karelovic A, Jimenez R. Effects of Site Geometry and Local Composition on Hydrogenation of Surface Carbon to Methane on Ni, Co, and NiCo Catalysts. Catalysts. 2022; 12(11):1380. https://doi.org/10.3390/catal12111380

Chicago/Turabian Style

Godoy, Sebastian, Prashant Deshlahra, Francisco Villagra-Soza, Alejandro Karelovic, and Romel Jimenez. 2022. "Effects of Site Geometry and Local Composition on Hydrogenation of Surface Carbon to Methane on Ni, Co, and NiCo Catalysts" Catalysts 12, no. 11: 1380. https://doi.org/10.3390/catal12111380

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop