Mechanism of CO2 reduction by H2 on Ru(0 0 0 1) and general selectivity descriptors for late-transition metal catalysts
Graphical abstract
Introduction
More than 85% of the current global energy need is provided by combustion of fossil fuels, which is accountable for continuously increasing atmospheric concentrations of CO2 and accompanying climate change effects [1]. The search for approaches to reduce atmospheric CO2 concentration has become a high priority research area. Recent efforts show the potential promise of directly sequestering CO2 from the atmosphere using amine based sorbent materials, among other methods [2], [3], [4], [5], [6]. If approaches to directly sequester CO2 from the atmosphere prove successful, it will be important to develop efficient, low temperature and pressure processes for converting CO2 to higher value hydrocarbon feedstocks for chemical and fuel production. The coupling of CO2 sorption technologies with solar-based H2 production through catalytic reduction processes would provide an energy efficient, environmentally friendly and carbon neutral approach for chemical and fuel production. This approach relies, in part, on the development of materials that facilitate catalytic conversion of CO2 and H2 into desired products with high selectivity.
Because of high energy requirements, C–C coupling reactions are rare at low temperature and pressure and it is expected that C1 molecules (CO, CH3OH and CH4) will be the dominant products of environmentally friendly CO2 reduction processes. CH3OH synthesis from CO2 and H2 on Cu and “Cu-like” catalysts has received significant attention, due to the extensive use of CH3OH as a precursor for production of chemicals and liquid fuels. However, CH3OH is a minimal side product under low pressure CO2 hydrogenation conditions [7], [8], [9]. On the other hand, highly selective catalytic CH4 and CO production has been demonstrated at low temperature (as low as 150 °C) and atmospheric pressure over a range of supported transition metal catalysts (eg. Ni, Ru, Rh, Pd, Pt) [10], [11], [12], [13], [14], [15]. Supported Ru and Rh catalysts are consistently observed to exhibit high activity and selectivity for CO2 methanation (CO2 + 4H2 → CH4 + 2H2O) at moderate temperatures and atmospheric pressure. Pt and Pd have been shown to facilitate CO production through the reverse Water Gas Shift reaction (rWGS, CO2 + H2 → CO + H2O) and Ni has exhibited a combination of CO and CH4 selectivity [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23]. In spite of extensive theoretical and experimental studies of catalytic reduction of CO2 by H2 on late transition metal catalysts, outstanding mechanistic questions remain, including most importantly the elementary steps and factors that control catalytic performance and selectivity [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33].
Based on the excellent reactivity of Ru for CO2 methanation and extensive experimental evidence, we use this system as a base case to develop mechanistic insights. Three main initial CO2 hydrogenation mechanisms have been proposed in the literature on late transition metal catalysts: (1) direct hydrogenation of CO2 to HCOO∗ or COOH∗ (adsorbed species are denoted by a star), (2) CO2 dissociation followed by hydrogenation of CO∗ [18], [21], [29], [34], [35], and (3) dissociation of CO2, followed by direct dissociation of CO∗ and hydrogenation of surface bound C∗ [14], [36]. Surface formates (HCOO∗) have been observed by in-situ Fourier Transform Infrared (FTIR) spectroscopy; however, results have shown that these species are most likely spectators on the support [37], [38], [39]. Further, the relatively low barrier for CO2 dissociation on Ru indicates that direct CO2 hydrogenation to form COOH∗ is not a primary reaction pathway on Ru [31], [40], [41]. In-situ FTIR studies of CO2 hydrogenation on Ru have repeatedly shown adsorbed CO∗ to be the dominant surface species [11], [12], [13], [17], [25], [35], [36], [37], [42], [43], suggesting that hydrogenation following CO2 dissociation is more feasible than direct hydrogenation of CO2. Therefore it is likely that the main reaction pathway involves hydrogenation of CO∗, which has led to the suggestion that CO2 hydrogenation follows similar mechanism to CO hydrogenation after CO2 is dissociated on the surface. However experiments have shown that CO and CO2 hydrogenation kinetics differ considerably in terms of partial pressure dependences and activation barriers, suggesting important mechanistic differences [17], [22]. As discussed in the Results and Discussion sections, this can be attributed to differences in the surface coverage of CO∗ on Ru under CO and CO2 hydrogenation reaction conditions. CO∗ hydrogenation can evolve through formation of COH∗ or CHO∗ intermediates. Theoretical studies based on density functional theory (DFT) calculations favor CHO∗ route over COH∗ formation [31], [32], [44], [45], [46]. A few studies report traces of elemental carbon (C∗), which can be a result of direct CO∗ dissociation; however, both theoretical and experimental studies have agreed that the H-assisted CO∗ dissociation is the more energetically favorable route [21], [32], [33], [44], [46], [47], [48].
Although there has been significant experimental analysis of the mechanism of CO2 reduction by H2 on late transition metal catalysts, there is no consensus on the initial hydrogenation pathway, or the rate-limiting step (RLS) of the reaction. The most commonly proposed RLSs are direct CO∗ bond breaking, or in more recent studies H-assisted CO∗ dissociation through formation of CHxO∗ [26], [37], [46]. Importantly, there is also no consensus on what properties of late transition metals dictate selective CO production through rWGS (Pt, Pd) versus CH4 production through methanation (Ru and Rh). A recent study showed that the catalytic metal d-band center energy trends with competing selectivity for CO and CH4 production. However, significant variations in the trends observed in this study suggest that d-band center correlation cannot completely describe previously observed variations in selectivity among late transition metal catalysts [19].
In this work we utilize DFT calculations coupled with mean-field microkinetic modeling to develop molecular level insights into the mechanisms that control performance and selectivity of Ru catalysts in CO2 reduction by H2 at atmospheric pressure. Mean-field microkinetic models have been proven successful in describing reaction kinetics semi-quantitatively and have been repeatedly used in computational screening of catalyst materials [7], [49], [50], [51]. To investigate the mechanism of CO2 hydrogenation reaction at realistic operating conditions, without a known rate expression and a priori assumption of RLS, it is essential to conduct a detailed microkinetic analysis using temperature and pressure corrected free energies and entropies. In this study the thermodynamic and kinetic parameters are obtained from DFT calculations and used to build a microkinetic model of competing rWGS and methanation pathways on Ru(0 0 0 1). The microkinetic model captures experimentally observed trends in reaction selectivity and the impact of reactant partial pressures on catalytic rate and surface coverage of reaction intermediates, validating mechanistic insights gained from the model. The initial CO2 hydrogenation pathway on Ru catalysts was shown to occur through CO2 dissociation on the catalyst surface and subsequent hydrogenation of CO∗ to CHO∗, in line with recent FTIR results [11], [12], [13], [17], [25], [35], [36], [37], [42], [43]. Using Campbell’s degree of rate control analysis it was identified that the RLS for CH4 formation on Ru(0 0 0 1) is the dissociation of CHO∗ to form CH∗ and O∗, while the RLS for the competing CO production through rWGS is CO∗ desorption [52], [53]. Insights into competing RLSs for CO and CH4 production on Ru, coupled with linear scaling and Brønsted–Evans–Polanyi (BEP) relationships that extend insights on Ru to Rh, Ni, Pt and Pd catalysts [54], [55], [56], [57], establish that the O∗ adsorption energy is an effective descriptor of competing selectivity between CO and CH4 production on late-transition metal catalysts. That is, more exothermic oxygen adsorption energy is associated with higher selectivity toward CH4 production. This relationship derives from the impact of the O∗ adsorption energy on the exothermicity of CHO∗ dissociation, the resulting influence on the CHO∗ dissociation barrier (the RLS for CH4 formation), and the lack of significant variation in CO∗ binding energy on late transition metal surfaces. The use of oxygen adsorption energy as a selectivity descriptor for CH4 versus CO production in CO2 reduction by H2 accurately explains experimentally observed variations in selectivity on late transition metal catalysts.
Section snippets
DFT calculations
DFT calculations were carried out with the real space grid-based projector-augmented wave (GPAW) code [58]. The Revised Perdew−Burke−Ernzerhof (RPBE) form of the generalized gradient approximation (GGA) was used to approximate exchange and correlation effects [59]. The BEEF-vdW functional was also investigated for geometry optimization of larger intermediate molecules [60]. Comparisons between RPBE and BEEF-vdW calculated results are provided in the Results and Discussion section (see Section
CO2 methanation and rWGS mechanisms on Ru(0 0 0 1)
First we address an un-resolved question of the initial hydrogenation step in the mechanism of CO2 methanation on Ru catalysts. The potential energy diagram shown in Fig. 1 compares three competing free energy pathways: direct hydrogenation of physisorbed CO2∗, hydrogenation of CO∗ after dissociation of adsorbed CO2∗ to CO∗ and O∗ [18], [21], [34], [35], and hydrogenation following complete dissociation of CO2∗ to C∗ and 2O∗ [14], [36]. Our calculated electronic energies are in good agreement
Conclusions
The mechanism of CO2 reduction was investigated over Ru(0 0 0 1) using DFT calculations and mean field microkinetic modeling. A reaction mechanism consisting of 18 elementary steps was proposed. Energetic pathways of competing reactions were explored and direct CO2 dissociation followed by hydrogenation of CO∗ to CHO∗ was identified as the prominent pathway for CH4 production. The microkinetic model successfully predicted experimentally observed trends of surface coverage variations with
Acknowledgments
P.C. acknowledges funding from University of California, Riverside and the National Science Foundation Grant, CHE-1301019. This work used the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation grant number OCI-1053575.
References (85)
- et al.
Catalytic production of methane from CO2 and H2 at low temperature: insight on the reaction mechanism
Catal. Today
(2010) - et al.
Methanation of CO2: further insight into the mechanism over Rh/γ-Al2O3 catalyst
Appl. Catal. B Environ.
(2012) - et al.
Mechanistic study of low temperature CO2 methanation over Rh/TiO2 catalysts
J. Catal.
(2013) - et al.
Hydrogenation of CO2 to CH4 over alumina-supported noble metals
J. Mol. Catal.
(1980) - et al.
Methanation of CO2 on supported rhodium catalysts
J. Catal.
(1981) - et al.
Trends in the catalytic reduction of CO2 by hydrogen over supported monometallic and bimetallic catalysts
J. Catal.
(2013) - et al.
Competitive CO and CO2 methanation over supported noble metal catalysts in high throughput scanning mass spectrometer
Appl. Catal. A Gen.
(2005) - et al.
Hydrogenation of CO2 on group VIII metals, II. Kinetics and mechanism of CO2 hydrogenation on nickel
J. Catal.
(1982) - et al.
Selective methanation of CO over supported noble metal catalysts: effects of the nature of the metallic phase on catalytic performance
Appl. Catal. A Gen.
(2008) - et al.
Mechanistic aspects of the selective methanation of CO over Ru/TiO2 catalyst
Catal. Today
(2012)