Article pubs.acs.org/IECR
First-Principles Study of High Temperature CO2 Electrolysis on Transition Metal Electrocatalysts Xiang-Kui Gu, Juliana S. A. Carneiro, and Eranda Nikolla* Department of Chemical Engineering and Materials Science, Wayne State University, Detroit, Michigan 48202, United States ABSTRACT: Electrochemical reduction of CO2 using the electrical energy generated from renewable sources has attracted increasing interest as a potential route for producing high energy molecules from CO2. In this contribution, high temperature electrochemical reduction of CO2 to CO on a series of transition metal electrocatalysts is studied using DFT calculations combined with microkinetic modeling under solid oxide electrolysis cells (SOECs) operating conditions. We show that CO2 dissociation via a two-electron transfer process into adsorbed CO and O2− ions in the electrolyte is favorable on most of the metal electrocatalysts considered, with a dependence of the simulated electrolysis current density on the applied potential consistent with experimental observations. A “volcano”-type relation between the calculated electrolysis rates and the binding energies of atomic O is found, suggesting that the binding energy of O might be a good activity descriptor for high temperature CO2 electrolysis on transition metals. Our structure−activity trends suggest that metallic Ru and Co would exhibit the highest activity for electrochemical reduction of CO2 in SOECs. studied.23−28 It is found that the oxygen vacancies in these oxides play an important role in CO2 adsorption/activation and oxygen diffusion.29 An effect of the nature of the transition metal B-site in these oxides on the CO2 reduction activity has been reported. Ishihara et al.30 showed that the activity of La0.6Sr0.4FeO3‑δ could be improved by doping the Fe site with Mn, as opposed to doping it with Co, Ni, and Cu, which led to a decrease in activity. Although a number of efforts have focused on enhancing the electrochemical activity of MEIC oxides for CO2 electrolysis, it still remains an issue because it is significantly lower than that of Ni-based cathodes.31−33 Xie et al. found that the addition of transition metal nanoparticles (i.e., Ni, Fe, and Ru) to MEIC oxide electrodes significantly lowered the electrode polarization resistance (i.e., overpotential losses) leading to enhanced electrochemical rates.31,32,34−36 Among the transition metal SOEC cathode electrocatalysts explored in the literature, Ni and Co have shown to exhibit higher performance than Cu and Fe, due to the low catalytic activity of Cu for CO2 activation, and the oxidation of Fe during CO2 electrolysis.37 Low electrochemical activities for this process have also been reported for Pt and Ag electrocatalysts.38−41 The performance of the transition metal electrocatalysts can be enhanced by alloying them with another metal. For instance, the alloys of NiFe,42−44 CoFe,45 and NiRu46 have been shown to exhibit higher electrochemical performance and resistance to coking as compared to the parent monometallic electrocatalysts.
1. INTRODUCTION Electrochemical reduction of CO2 using the electrical energy from renewable resources, such as the sun and wind, has attracted wide attention for production of high energy molecules (i.e., CO and hydrocarbons) from CO2.1−4 In addition to recycling CO2, this process also provides an avenue for storing the electrical energy from these intermittent, renewable sources in chemical form. Electrochemical reduction of CO2 to CO using solid oxide electrolysis cells (SOECs) presents a feasible approach due to favorable thermodynamics and kinetics at high operating temperatures, which enhance the process efficiency.5−11 Furthermore, this process produces CO, a valuable intermediate to the production of synthetic fuels using established processes, such as Fischer−Tropsch synthesis.12 In SOECs under an applied potential, CO2 is reduced to CO and oxygen ions at the cathode. The oxygen ions diffuse from the cathode through the electrolyte to the anode where they are evolved as O2 in the gas phase. The commonly used cathode electrocatalyst for this process is Ni due to its good electrical conductivity, thermal compatibility with the other cell components and low cost.5,13 Generally, Ni is mixed with the electrolyte oxide, such as yttriastabilized zirconia (YSZ),14−16 to enhance the triple phase boundary and facilitate oxygen ion transport at the cathode. However, Ni exhibits high overpotential losses for this process.17−19 Mixed ionic and electronic conducting (MIEC) oxides (e.g., perovskite-based materials) have been reported as potential cathodes for SOECs due to their good redox stability and resistance to coking during CO2 electrolysis.20−22 Among these oxides, the doped lanthanum chromates (i.e., La 0.75 Sr 0.25 Cr 0.5 Mn 0.5 O 3‑δ , LSCM), lanthanum ferrites (La0.6Sr0.4FeO3‑δ), and lanthanum titanates have been widely © XXXX American Chemical Society
Received: Revised: Accepted: Published: A
February 27, 2017 May 3, 2017 May 8, 2017 May 8, 2017 DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research
function) is confirmed. For example, our test calculations show that the difference between the binding energies of CO calculated using a higher cutoff energy of 500 eV and that of using a cutoff energy of 400 eV is less than 0.01 eV. During optimization, the bottom two layers of the slab are fixed, whereas the remaining atoms and adsorbates are relaxed until the residual forces are less than 0.02 eV/Å. The binding energies of the intermediates are calculated as
A number of studies have focused on screening various electrocatalysts for reduction of CO2 in SOECs, but limited understanding of the factors that govern their performance exists making the proper optimization of these electrocatalytic systems challenging. In the present work, density functional theory (DFT) calculations are employed to systematically investigate CO2 electrolysis on a series of transition metal electrocatalysts under SOEC conditions. Microkinetic modeling is used to understand the underlying mechanism that governs this process on metal electrocatalysts. Our calculations suggest that on most of the metal electrocatalysts considered (except for Fe, W, and Mo), the most favorable mechanism involves direct electrochemical CO2 dissociation to adsorbed CO on the metal surface and O2− ion in the electrolyte. A “volcano”-type relation between the calculated electrochemical rates and the binding energy of O is obtained, suggesting that the binding energy of O might be a good activity descriptor for CO2 electrolysis on metal electrocatalysts. Our results indicate that Co and Ru electrocatalysts are among the most active transition metals for this process.
BE = Ead/sub − Ead − Esub
(1)
Where Ead/sub, Ead, and Esub are the total energies of the optimized adsorbate/substrate system, the adsorbate in the gas phase, and the clean substrate, respectively. The transition state of the elementary steps is determined by the climbing-image nudged elastic band (CI-NEB) method.52,53 2.2. Proposed Mechanisms for CO2 Electrolysis. Two possible mechanisms for CO2 electrolysis under SOEC conditions are considered as shown in Figure 2. In mechanism
2. METHODOLOGY 2.1. DFT Calculations. Periodic DFT calculations are performed with the Vienna ab initio Simulation Package (VASP).47,48 The exchange-correlation interaction is described by the generalized gradient approximation (GGA) and Perdew− Burke−Ernzerhof (PBE) functional.49 The Kohn−Sham equations are solved in a plane wave basis set with a kinetic energy cutoff of 400 eV. A four-layer slab model with (3 × 3) unit cell is used to model the close packed (111), (110), and (0001) surfaces for metals with FCC (Ni, Cu, Pd, Au, and Ag), BCC (Fe, Mo, and W), and HCP (Co and Ru) crystal structures, respectively (see Figure 1). Spin polarization is considered for Figure 2. Proposed cathode half-cell mechanisms along with the corresponding elementary steps and Gibbs free energies for CO2 electrolysis at solid oxide electrolysis cell conditions.
1, CO2 thermochemically dissociates to CO* and O* on the metal surface, followed by the electrochemical reduction of the adsobed O* to oxygen ions (O2−) in the electrolyte via the transfer of two electrons. On the other hand, in mechanism 2, CO2 is electrochemically reduced via a two-electron transfer to adsorbed CO* on the metal surface and O2− ions in the electrolyte. We note that due to the absence of a complex interfacial model between the transition metal and the electrolyte, the energy for the transition state is assumed to be similar to the thermochemical step on the metal extrapolated to the electrochemical step by accounting for the effect of the potential as descibed below. The consequence of such an assumption would mainly effect the magnitute of the barrier with limited effect on the trends obtained for the different metal electrodes. In both mechanisms, the adsorbed CO* on the metal surface desorbs into the gas phase on the SOEC cathode side, whereas the oxygen ions O2− ions are assumed to migrate through the electrolyte to the anode, where they are evolved as gas-phase O2, without significant ohmic losses. The Gibbs free energies for the thermochemical steps in both mechanisms (i.e., CO2 activation in mechanism 1 and CO desorption) are calculated using DFT and statistical mechanics. The entropic contributions for CO2 adsorption (a loss of 2.03 eV per molecule) and CO desorption (a gain of 1.97 eV per molecule) are calculated based on the method reported in the literature at 1073 K and standard pressure.54 For the electro-
Figure 1. Geometries of (111), (110), and (0001) surfaces for transition metal with FCC, BCC, and HCP crystal structures, respectively.
Ni and Fe. We note that close packed surfaces are used in these studies because they are the most thermodynamically stable and most likely the dominant surface facets under the high operating temperatures of CO2 electrolysis in SOECs. The most stable cubic-ZrO2(111) surface is used to model the YSZ electrolyte with three O−Zr−O layers where one of Zr atoms is replaced by a Y atom. The concentration of Y in this model is ∼8.33%, close to that of 8% in experiments.5 We find that the energetics of surface and subsurface Y doping in ZrO2 are comparable, in line with literature.50,51 In our calculations the surface substitution of Y in ZrO2 is used. In this model, the most stable surface O vacancy is considered as an adsorption site for oxygen ions. We find that the oxygen binding energy on this vacant site is −5.65 eV. A (4 × 4 × 1) k-point mesh is used to sample the surface Brillouin zone, and a 12 Å vacuum is introduced between the repeated slabs along the z-direction. The convergence of the binding energies with respect to electronic parameters (i.e., the kinetic energy cutoff and the partial occupancies of the wave B
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research chemical steps that involve charged O2− ions, which are directly affected by the cathode potential bias, the Gibbs free energies are calculated with respect to a standard oxygen electrode (SOE, 1/ 55,56 2O2 + 2e− → O2− YSZ), as previously discussed in the literature. In this case, ΔV is the applied potential defined as the difference in the potential between the SOE and the cathode (ΔV = VSOE − Vcathode). Because the anode of the SOEC involves the same electrochemical reaction as the SOE, the anode potential can be described as Vanode = VSOE under equilibrium condition. Based on literature reports,55−57 the anode potential under equilibrium is calculated to be 0.98 V at 1073 K and standard pressure. The energy of O2 in the gas phase is calculated relative to gas-phase H2O and H2, due to the reported shortcomings of DFT in describing O2.58 The entropic contribution for O2 desorption at the SOE is calculated to be 1.99 eV per O2 molecule under the same conditions. We note that the effect of the electric field induced by the potential drop across the cathode/electrolyte interface on the Gibbs free energies is not considered because this effect has been reported to be insignificant.55,56 The activation barrier for CO2 dissociation is calculated with respect to CO2 in the gas phase considering the entropic contribution of 2.03 eV for CO2 adsorption at 1073 K and standard pressure. The effect of the potential on the activation barrier associated with the electrochemical step involving electron transfer is described using eq 2.59−62 Ea = Ea0 − nαΔV
3. RESULTS AND DISCUSSION 3.1. Energetics of CO2 Electrolysis on Transition Metal Surfaces. The optimized most stable geometries of the intermediates involved in CO2 electrolysis on transition metal surfaces are shown in Figure 3, and the corresponding calculated
Figure 3. Most stable structures of CO2, CO, and O intermediates adsorbed on the transition metal surfaces. The green, orange, blue, black, and red spheres represent FCC metal, BCC metal, HCP metal, C, and O atoms, respectively.
(2)
where E0a is the thermochemical activation barrier without applied potential in the chemical step,59 n is the number of electron transfer in the elementary step, and α is the transfer coefficient, which ranges from 0.3 to 0.7 for most electrochemical systems.61 Here the reported barriers for the electrochemical steps are based on an value of 0.3 because the simulated plot of current density versus applied potential for CO2 electrolysis on Ni using this value is most consistent with our experimental observations, as shown below. 2.3. Experimental Studies. SOEC button cells are synthesized by initially ball-milling a mixture of NiO (99%, Alpha Aesar), yttria stabilized zirconia (YSZ), and graphite (1:1:1 weight ratio) in ethanol for 48 h. The slurry is dried, grinded, sieved, and pressed (3000 psi) into 13 mm diameter pellets. The pellets are presintered at 1100 °C for 3 h; thereafter, a YSZ electrolyte solution is spin-coated over one side of the porous NiO-YSZ pellet. This is followed by sintering the assembly at 1450 °C for 4 h. This process leads to a NiO/YSZ cathode layer of ∼500 μm and an electrolyte thickness of ∼15 μm. For the anode side, a porous YSZ scaffold layer is sprayed on top of the electrolyte side and sintered at 1450 °C. Lanthanum strontium manganite (LSM) is drop-coated over the YSZ scaffold, and calcined at 400 °C for 3 h. The final geometric area of the anode is ∼0.1 cm2. For both electrodes, gold mesh and wires are used as the current collector and electrical leads, respectively. Gold paste is used to connect the leads and current collectors to the electrodes. A high temperature alumina−based sealant is used to attach the cell to an alumina tubular reactor. The NiO/YSZ electrode is reduced overnight at 700 °C under a 30% H2−70% N2 atmosphere (50 mL min−1). Linear sweep voltammetry tests are performed at 800 °C with the cathode exposed to an atmosphere of 45% CO2−45% CO−10% N2 (100 mL min−1), and anode exposed to air. The polarization curves (I−V curves) are obtained at a scan rate of 30 mV/s from OCV to 1.5 V using a Gamry 300 potentiostat.
Table 1. Calculated Most Stable Binding Energies of Intermediates Involved in CO2 Electrolysis on Transition Metal Electrocatalysts Binding energy (eV) Au(111) Ag(111) Pd(111) Cu(111) Ni(111) Co(0001) Ru(0001) Fe(110) Mo(110) W(110)
CO2
CO
O
−0.01 −0.01 −0.02 −0.01 −0.02 −0.01 −0.19 −0.51 −1.33 −1.28
−0.27 −0.12 −2.02 −0.92 −1.94 −1.65 −1.84 −1.94 −1.86 −1.94
−3.24 −3.67 −4.62 −4.90 −5.53 −5.88 −6.10 −6.68 −7.37 −7.68
binding energies are listed in Table 1. The CO2 molecule adsorbs through a flat laying geometry on the (111) surfaces of Au, Ag, Pd, Cu, and Ni with a C−O bond length of 1.18 Å, which is similar to that of gas-phase CO2. Consequently, the calculated binding energies of CO2 on these surfaces are very weak, with CO2 sitting approximately 3.5 Å from these surfaces. This indicates that CO2 tends to physically adsorb on these (111) surfaces, consistent with the reported literature.63,64 On Co(0001) and Ru(0001) surfaces, CO2 binds to a 3-fold metal surface site, leading to a ∼123° O−C−O angle, consistent with literature findings.65 The calculated binding energy of −0.19 eV on Ru(0001) is slightly stronger than that on Co(0001). The strongest adsorption geometries for CO2 are found on the (110) surfaces of Fe, Mo, and W, in which the C and one of the two O atoms anchor onto two adjacent bridge sites. The calculated CO2 binding energy on Fe(110) is −0.51 eV (similar to the literature C
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research
Table 2. Calculated Barriers and Gibbs Free Energies for the Elementary Steps Involved in CO2 Electrolysis at 1073 K, Standard Pressure, and ΔV = 1.30 V CO2 + 2* → CO* + O* Au(111) Ag(111) Pd(111) Cu(111) Ni(111) Co(0001) Ru(0001) Fe(110) Mo(110) W(110)
CO2 + * + 2e− → CO* + O2−
O* + 2e− → O2− + *
CO* → CO + *
Ea
ΔG
Ea
ΔG
ΔG
ΔG
5.43 4.91 3.52 3.53 2.74 2.47 2.17 1.69 0.95 1.20
5.07 4.78 1.93 2.75 1.10 1.04 0.63 −0.05 −0.66 −1.06
4.65 4.13 2.74 2.75 1.96 1.69 1.39 0.91 0.17 0.42
1.15 1.29 −0.61 0.50 −0.53 −0.24 −0.43 −0.53 −0.44 −0.53
−3.92 −3.48 −2.54 −2.26 −1.63 −1.27 −1.05 −0.47 0.22 0.53
−1.70 −1.85 0.05 −1.05 −0.03 −0.32 −0.13 −0.03 −0.11 −0.03
−9.17 eV, consistent with the reported value.67 Compared to the oxygen binding on the transition metals considered, the strong binding of O on the YSZ model with a subsurface vacancy suggests that oxygen diffusion from the metal surface to the electrolyte would be very facile on all metals. Based on the energetics of the elementary steps involved in CO2 electrolysis discussed below, this would lead to CO2 dissociation becoming the rate-limiting step on all metals, resulting in W, Mo, and Fe exhibiting the highest activities for this process due to the low barriers for CO2 dissociation on these surfaces. This is inconsistent with the experimental observations, where it has been shown that, for example, Fe exhibits low activity due to oxidation, suggesting that oxygen diffusion from Fe to the electrolyte is a challenge.37 Therefore, for the calculations below, we have mainly considered the YSZ model with surface oxygen vacancies because it leads to insights that are consistent with experimental observations. The calculated barriers and Gibbs free energies of the elementary steps involved in CO2 electrolysis at 1073 K and standard pressure are listed in Table 2. An applied potential of 1.3 V is used as an example to evaluate the effect of potential on the energetics associated with the electrochemical steps. We note that the magnitude of the applied potential mainly affects the magnitude of the electrochemical rates and does not affect the overall activity trends derived below. CO2 thermochemical dissociation to CO* and O* on Au(111) and Ag(111) is highly endothermic with high energy barriers of 5.43 and 4.91 eV, respectively. This is due to the weak binding of CO* and O* on these surfaces. Lower CO2 dissociation energy barriers (in the range of 2.17−3.52 eV) are found on Ru, Co, Ni, Cu, and Pd because of the stronger binding strengths of CO* and O* as compared to that on Au and Ag. The much stronger binding strengths of CO* and O* (especially for O*) result in this step being exothermic on Fe, Mo, and W surfaces with calculated energy barriers in the range of 0.95−1.69 eV. In the case of the electrochemical CO2 dissociation to CO* and O2− ions in the electrolyte, the energetics associated with this step highly depend on the applied potential of the SOEC. An increase in the applied potential leads to a lower activation barrier and higher driving force for electrochemical dissociation of CO2. We find that the calculated barriers for both CO2 thermochemical and electrochemical dissociations are linearly correlated to the binding energies of atomic O on the metals as shown in Figure 4, implying that CO2 dissociation is mainly driven by the binding strength of O on these surfaces. This indicates that the binding energy of atomic O can be used to predict the CO2 dissociation barrier. In the case of the CO desorption step on Au(111),
value of −0.56 eV66), and the strongest binding energies of −1.33 and −1.28 eV are found on Mo(110) and W(110), respectively. For CO adsorption on (111) surfaces of Au, Ag, Pd, Cu, and Ni, the most stable configuration is found on a fcc hollow site. Weak CO binding energies of −0.27 and −0.12 eV are found on Au(111) and Ag(111), respectively. On Cu(111), the binding energy of −0.92 eV is modest, whereas the strongest binding energies are found on Pd(111) and Ni(111) with values of −2.02 and −1.94 eV, respectively. On the (0001) surfaces of Co and Ru metals, the CO binding strength on the hcp hollow site is comparable to that on the fcc hollow site, with binding energies of −1.65 eV (Co) and −1.84 eV (Ru). On Fe(110), Mo(110), and W(110) surfaces, CO binding on the top-site is slightly stronger than that on the hollow site with binding energies of −1.94, −1.86, and −1.94 eV, respectively. In the case of atomic O, it favorably binds to a hollow site on all the surfaces considered, and a gradual increase in binding energy is found when the transition metal changes from Au, Ag, Pd, Cu, Ni, Co, Ru, Fe, Mo, to W, consistent with the fact that Au and Ag have the weakest oxygen affinity, whereas Mo and W have the strongest. Moreover, the diffusion of atomic O from one stable site to another on all these surfaces is very facile at high temperatures, and the calculated diffusion barriers are in the range of 0.30−0.65 eV. Oxygen diffusion from the metal surface to the electrolyte is highly dependent on the nature of the metal. For example, the weaker binding of O on the (111) surfaces of Au, Ag, Pd, Cu, and Ni as compared to that on the most stable surface oxygen vacancy of the YSZ electrolyte (−5.65 eV) leads to an exothermic O diffusion step from these surfaces to the electrolyte, suggesting that this diffusion process would be very favorable. However, O diffusion from Co(0001) and Ru(0001) to the electrolyte is slightly endothermic by 0.23 and 0.45 eV, respectively. The more endothermic diffusion energetics (>1.03 eV) are found for oxygen diffusing from the (110) surfaces of Fe, Mo, and W to the electrolyte, implying that the oxygen diffusion from these metals to YSZ would be very limited, whereas the reverse oxygen spillover from YSZ to metal will be very facile. This would lead to the oxidation of these metals under CO2 electrolysis conditions. We have also investigated the effect of the location of oxygen vacancies in the YSZ model on the energetics associated with O diffusion from the metals to the electrolyte. Two YSZ models are considered: one containing only the most stable surface oxygen vacancy, and the other one containing an additional oxygen vacancy per Y2O3 unit in the subsurface layer, similar to that reported in the literature.51,67 We find that the oxygen binding energy on the surface vacancy model is −5.65 eV, whereas on a surface vacancy of the subsurface vacancy model is D
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research
desorption step, in both mechanisms, it is slightly exothermic by 0.03 eV. A microkinetic model is used to evaluate the effect of the applied potential on the electrochemical rates of CO2 electrolysis via both mechanisms on Ni(111). Equation 3 is used to calculate the rate constants for the elementary steps. k=
k bT exp( −Ea /RT ) h
(3)
where kb, h, and Ea are the Boltzmann constant, Planck constant, and activation barrier, respectively. We note that the thermochemical barrier is used for the electrochemical reduction of O*(metal) to O2−(electrolyte), and the absolute value of the CO binding energy is used as the barrier for the CO desorption step. The steady-state approximation is used to determine the coverage of the intermediates and the vacant sites. The electrolysis current density is calculated as i = neσr, where n, σ, and r are the number of electrons transfer in this process, the number of active sites per surface area, and the calculated reaction rate, respectively.60 The simulated plots for the electrolysis current density as a function of applied potential at 1073 K and standard pressure for CO2 electrolysis on Ni(111) via mechanisms 1 and 2 are shown in Figure 6. Figure 6a clearly shows that current density for CO2 electrolysis via mechanism 1 is weakly dependent on the applied potential, and the calculated values are rather low at various applied potentials. This is reasonable given that the rate-limiting step in this mechanism on Ni(111) is the thermochemical CO2 dissociation, which is independent of the applied potential. In this mechanism, the applied potential indirectly affects the rate by providing a driving force for the removal of adsorbed O* from the metal surface via the formation of O2− ions in the electrolyte. However, a strong dependence of the simulated electrolysis current density on the applied potential is found for CO2 electrolysis via mechanism 2 on Ni(111), due to the rate-limiting step (CO2 electrochemical dissociation) being an electrochemical step, which is highly dependent on the applied potential. It is found that the current density increases gradually with an increase in the applied potential (Figure 6b). As a comparison, experimental current−voltage curves are obtained for CO2 electrolysis in a SOEC composed of Ni/YSZ (cathode)| YSZ(electrolyte)|LSM/YSZ (anode) (Figure 6c). A comparison between the simulated and measured I−V curves in Figure 6 suggests that mechanism 2 more appropriately describes the experimental behavior for the electrochemical reduction of CO2 on a Ni cathode. We note that the simulated electrolysis current density using mechanism 2 is overestimated as compared to our
Figure 4. Linear relation between CO2 dissociation barrier and oxygen binding energy on different metal electrocatalysts. The barriers are calculated at 1073 K, standard pressure, and an applied potential (ΔV) of 1.30 V.
Ag(111), and Cu(111), it is facile with Gibbs free energies of −1.70, −1.85, and −1.05 eV, respectively. Whereas on the other surfaces, this step is slightly exothermic, except on Pd(111), where it is slightly endothermic due to the very strong binding of CO (2.02 eV) on this surface. 3.2. Microkinetic Modeling Analysis for CO2 Electrolysis on Ni(111). To understand the effects of the energetics of the elementary steps on the electrochemical rates (the kinetics of the entire electrocatalytic cycle) for CO2 electrolysis on different transition metals, a detailed microkinetic modeling analysis for CO2 electrolysis using the two mechanisms described above is performed. As a first step, we have calculated the electrochemical rates for CO2 electrolysis using the two mechanisms on the commonly used Ni electrocatalyst,13,16 in order to establish the mechanism that best describes this process under experimental conditions. The free energy diagrams for CO2 electrolysis on Ni(111) via mechanisms 1 and 2 at 1073 K and standard pressure are shown in Figure 5, which clearly shows that CO2 dissociation is the rate-limiting step in both mechanisms. In mechanism 1, the barrier for thermochemical CO2 dissociation to CO* and O* is independent of the applied potential. In this mechanism, the applied potential provides a driving force for the removal of the adsorbed O* on the metal surface to form O2− ions in the electrolyte in the presence of two electrons. In mechanism 2, the energy barrier and Gibbs free energy for the electrochemical CO2 dissociation to CO* and O2− are both highly dependent on the applied potential. In the case of the thermochemical CO
Figure 5. Free energy diagrams for CO2 electrolysis on Ni(111) via (a) mechanism 1 and (b) mechanism 2 at 1073 K and standard pressure. E
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research
of CO2 on all metals to consistently compare the activity. In the case of CO2 electrolysis on Ru, Co, Ni, Cu, Pd, Ag, and Au, we find that mechanism 2 is the most favorable with CO 2 dissociation being the rate-limiting step (Table 2). Co and Ru exhibit higher electrochemical activity as compared to Ni, due to facilitating the CO2 activation with lower energy barriers. On the other hand, the higher CO2 activation barriers on Cu, Pd, Au, and Ag result in lower CO2 electrolysis rates on these catalytic systems as compared to Ni. In the case of CO2 electrolysis on Fe, Mo, and W, thermochemical dissociation of CO2 occurs with lower energy barriers as compared to that on the other metals considered. On these systems, the diffusion of oxygen from the metal to the YSZ electrolyte becomes the rate-limiting step. This results in mechanism 1 being more favorable for CO2 electrolysis on these metals. As noted above, the barrier for the oxygen diffusion process is approximated using the binding energy difference between the oxygen adsorbed on these metals and that on the surface oxygen vacancy of YSZ, which leads to an overestimation of the rates, but does not alter the activity trends. The high energy barriers for oxygen diffusion suggest that the removal of oxygen from these metals is challenging resulting in the poisoning of the active sites via oxidation under CO2 electrolysis conditions. On the basis of our microkinetic model, we find that the oxygen coverage on a Fe surface is ∼0.58 monolayer (ML) at an applied potential of 1.3 V, while the surfaces of Mo and W are almost fully covered by oxygen, resulting in limited active site for CO2 activation. To estimate the effect of oxygen coverage on the activity, CO2 dissociation on a Fe surface covered by 2/3 ML oxygen is studied as an example. In this case, the dissociation barrier increases to 3.30 eV at an applied potential of 1.3 V as compared to that of 0.91 eV on a clean Fe surface. These results suggest that the activity of CO2 electrolysis would be even lower on Fe, Mo, and W under reaction conditions than that predicted on the clean surfaces due to oxygen coverage effects. A “volcano”-type relation between the calculated rates for CO2 electrolysis and the binding energies of O on the different transition metal surfaces is found, as shown in Figure 7. This suggests that the binding energy of O could be a good activity descriptor for high-temperature CO2 electrolysis on metal-based electrocatalysts. In Figure 7, the electrochemical activity of the metals on the right of the “volcano” plot is mainly limited by their ability to activate CO2, whereas metals on the left bind oxygen too strongly, making oxygen removal from the metal surface
Figure 6. Simulated plots of calculated current density versus applied potential for CO2 electrolysis on Ni(111) via (a) mechanism 1 and (b) mechanism 2 at 1073 K and standard pressure. (c) Experimentally obtained I−V curve for CO2 electrolysis on a Ni/YSZ cathode SOEC at 1073 K.
experimental data shown in Figure 6c. This is a consequence of assuming that there are no ohmic losses and mass transfer limitations in the simulated conditions, and that the anode acts as a standard oxygen electrode. These assumptions lead to an overestimation of the electrochemical rates, but should not affect the activity trends for CO2 electrolysis on the different metals discussed below because these assumptions are equally applied to all electrocatalytic systems considered. 3.3. Activity Trends for CO2 Electrolysis on Transition Metal Surfaces. Based on the energetics of the elementary steps discussed above, a microkinetic model for CO2 electrolysis on the different transition metals is developed to obtain an understanding of the qualitative activity trend for this process. The electrochemical rates are calculated at 1073 K, standard pressure, applied potential (ΔV) of 1.30 V, and an electron transfer coefficient of 0.3. We have assumed the same conversion
Figure 7. “Volcano”-type plot between the calculated rates for CO2 electrolysis and the binding energies of O on transition metal surfaces at 1073 K, standard pressure and ΔV = 1.30 V. F
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research Biographies
challenging, consequently resulting in the poisoning of the active sites. For these metals, the binding strength of O on a surface oxygen vacancy site of the electrolyte (YSZ considered here) would influence the rates by providing a driving force for O removal from the metal surface to the electrolyte. As such, supporting these metals on a higher oxygen conducting oxide than YSZ would lead to an enhancement in their activity by providing a stronger driving force for the removal of oxygen from their surfaces. Figure 7 suggests that the metals near the top of the “volcano” plot, metallic Co and Ru, would provide the optimal binding strength of oxygen that would lead to the highest activity for CO2 electrolysis on monometallic electrocatalysts.
Dr. Xiang-Kui Gu received his Ph.D. from Dalian Institute of Chemical Physics, Chinese Academy of Sciences. After as a postdoctoral appointment in the School of Chemical Engineering at Purdue University, he joined Prof. Eranda Nikolla’s group as a postdoctoral scholar. His current research interests focus on utilizing state-of-the art computational methods to design efficient and selective heterogeneous catalysts for electrocatalysis and biomass conversion processes.
Therefore, these metals should exhibit higher electrochemical rates for this process when compared to the traditionally used monometallic Ni. This is consistent with experimental report, which has shown that an improvement in the activity of Ni can be achieved by alloying it with Ru.46
4. CONCLUSIONS Periodic DFT calculations are performed to study hightemperature CO2 electrolysis on transition metal electrocatalysts under SOEC operating conditions. Two possible mechanisms for this process are considered. We find that the mechanism involving the electrochemical reduction of CO2 to CO and O2− ions in the electrolyte is favorable on most of the metals studied (except for highly oxophilic surfaces such as Fe, W, and Mo), resulting in a simulated electrolysis current density dependence on the applied potential consistent with experimental behavior. A “volcano”-type relation between the calculated electrochemical
Juliana Carneiro is a Ph.D. candidate in the Department of Chemical Engineering and Material Science at Wayne State University under the supervision of Prof. Eranda Nikolla. The focus of her dissertation is on developing a fundamental understanding of the underlying chemistry that governs high temperature electrolysis of CO2 and H2O using solid oxide electrolysis cells with the aim of improving the energy efficiency of these processes.
rates and the binding energies of O is obtained, suggesting that the binding energy of O can be used as an activity descriptor to screen for optimal SOEC cathode metal-based electrocatalysts for CO2 electrolysis. We find that the metals near the top of the “volcano” plot, which provide the optimal binding of O for this process and thus would lead to the highest electrochemical rates, are Co and Ru. Therefore, SOEC cathodes containing these electrocatalysts would result in enhanced activity toward CO2 electrolysis as compared to the commonly used monometallic Ni electrocatalyst.
■
AUTHOR INFORMATION
Corresponding Author
*E. Nikolla. Email:
[email protected]. ORCID
Eranda Nikolla: 0000-0002-8172-884X
Dr. Eranda Nikolla is an associate professor in the Department of Chemical Engineering and Materials Science at Wayne State University. She received her Ph.D. in Chemical Engineering from the University of
Notes
The authors declare no competing financial interest. G
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research
Temperature Water Electrolysis. J. Electrochem. Soc. 2009, 156, B897− B904. (15) Laguna-Bercero, M. A. Recent advances in high temperature electrolysis using solid oxide fuel cells: A review. J. Power Sources 2012, 203, 4−16. (16) Zhan, Z. L.; Zhao, L. Electrochemical reduction of CO2 in solid oxide electrolysis cells. J. Power Sources 2010, 195, 7250−7254. (17) Duboviks, V.; Maher, R. C.; Kishimoto, M.; Cohen, L. F.; Brandon, N. P.; Offer, G. J. A Raman spectroscopic study of the carbon deposition mechanism on Ni/CGO electrodes during CO/CO2 electrolysis. Phys. Chem. Chem. Phys. 2014, 16, 13063−13068. (18) Duboviks, V.; Lomberg, M.; Maher, R. C.; Cohen, L. F.; Brandon, N. P.; Offer, G. J. Carbon deposition behaviour in metal-infiltrated gadolinia doped ceria electrodes for simulated biogas upgrading in solid oxide electrolysis cells. J. Power Sources 2015, 293, 912−921. (19) Duboviks, V.; Maher, R. C.; Offer, G.; Cohen, L. F.; Brandon, N. P. In-Operando Raman Spectroscopy Study of Passivation Effects on Ni-CGO Electrodes in CO2 Electrolysis Conditions. ECS Trans. 2013, 57, 3111−3117. (20) Shin, T. H.; Myung, J. H.; Verbraeken, M.; Kim, G.; Irvine, J. T. S. Oxygen deficient layered double perovskite as an active cathode for CO2 electrolysis using a solid oxide conductor. Faraday Discuss. 2015, 182, 227−239. (21) Green, R. D.; Liu, C. C.; Adler, S. B. Carbon dioxide reduction on gadolinia-doped ceria cathodes. Solid State Ionics 2008, 179, 647−660. (22) Wang, Y. Q.; Yang, Z. B.; Han, M. F.; Chang, J. S. Optimization of Sm0.5Sr0.5CoO3-δ-infiltrated YSZ electrodes for solid oxide fuel cell/ electrolysis cell. RSC Adv. 2016, 6, 112253−112259. (23) Li, Y. X.; Zhou, J. E.; Dong, D. H.; Wang, Y.; Jiang, J. Z.; Xiang, H. F.; Xie, K. Composite fuel electrode La0.2Sr0.8TiO3‑δ-Ce0.8Sm0.2O2‑δ for electrolysis of CO2 in an oxygen-ion conducting solid oxide electrolyser. Phys. Chem. Chem. Phys. 2012, 14, 15547−15553. (24) Yang, L. M.; Xie, K.; Wu, L.; Qin, Q. Q.; Zhang, J.; Zhang, Y.; Xie, T.; Wu, Y. C. A composite cathode based on scandium doped titanate with enhanced electrocatalytic activity towards direct carbon dioxide electrolysis. Phys. Chem. Chem. Phys. 2014, 16, 21417−21428. (25) Shin, T. H.; Myung, J. H.; Naeem, K. M.; Savaniu, C.; Irvine, J. T. S. Ce(Mn,Fe)O2-(La,Sr) (Fe,Mn)O3 composite as an active cathode for electrochemical reduction of CO2 in proton conducting solid oxide cells. Solid State Ionics 2015, 275, 106−109. (26) Xu, S. S.; Li, S. S.; Yao, W. T.; Dong, D. H.; Xie, K. Direct electrolysis of CO2 using an oxygen-ion conducting solid oxide electrolyzer based on La0.75Sr0.25Cr0.5Mn0.5O3-δ electrode. J. Power Sources 2013, 230, 115−121. (27) Zhang, Y. Q.; Li, J. H.; Sun, Y. F.; Hua, B.; Luo, J. L. Highly Active and Redox-Stable Ce-Doped LaSrCrFeO-Based Cathode Catalyst for CO2 SOECs. ACS Appl. Mater. Interfaces 2016, 8, 6457−6463. (28) Wang, S. J.; Ishihara, T. La0.6Sr0.4Fe0.9Mn0.1O3 Oxide Cathode for the High Temperature CO2 Electrolysis Using LSGM Electrolyte. Solid Oxide Fuel Cells 2013, 57, 3171−3176. (29) Qi, W. T.; Gan, Y.; Yin, D.; Li, Z. Y.; Wu, G. J.; Xie, K.; Wu, Y. C. Remarkable chemical adsorption of manganese-doped titanate for direct carbon dioxide electrolysis. J. Mater. Chem. A 2014, 2, 6904−6915. (30) Ishihara, T.; Wu, K.-T.; Wang, S. High Temperature CO2 Electrolysis on La (Sr) Fe (Mn) O3 Oxide Cathode by Using LaGaO3 Based Electrolyte. ECS Trans. 2015, 66, 197−205. (31) Ruan, C.; Xie, K. A redox-stable chromate cathode decorated with in situ grown nickel nanocatalyst for efficient carbon dioxide electrolysis. Catal. Sci. Technol. 2015, 5, 1929−1940. (32) Li, Y. X.; Xie, K.; Chen, S. G.; Li, H. X.; Zhang, Y.; Wu, Y. C. Efficient Carbon Dioxide Electrolysis Based on Perovskite Cathode Enhanced with Nickel Nanocatalyst. Electrochim. Acta 2015, 153, 325− 333. (33) Yue, X. L.; Irvine, J. T. S. Alternative Cathode Material for CO2 Reduction by High Temperature Solid Oxide Electrolysis Cells. J. Electrochem. Soc. 2012, 159, F442−F448. (34) Gan, L. Z.; Ye, L. T.; Tao, S. W.; Xie, K. Titanate cathodes with enhanced electrical properties achieved via growing surface Ni particles
Michigan in 2009 under the supervision of Prof. Suljo Linic and Prof. Johannes Schwank, followed by a two-year postdoctoral work at California Institute of Technology with Prof. Mark E. Davis. Her research interests lie in the development of active and selective heterogeneous catalysts and electrocatalysts for chemical/electrochemical conversion processes, including electrochemical conversion of CO2 to high energy molecules using a combination of experimental and theoretical techniques. She is the recipient of a number of awards, including the National Science Foundation CAREER Award, the Department of Energy CAREER Award, 2016 Camille Dreyfus TeacherScholar Award and the Young Scientist Award from the International Congress on Catalysis.
■
ACKNOWLEDGMENTS This invited contribution is part of the I&EC Research special issue for the 2017 Class of Influential Researchers. We thank the financial support from the National Science Foundation (CBETCAREER 1350623) and Wayne State University Office of Vice President for Research (OVPR).
■
REFERENCES
(1) Graves, C.; Ebbesen, S. D.; Mogensen, M.; Lackner, K. S. Sustainable hydrocarbon fuels by recycling CO2 and H2O with renewable or nuclear energy. Renewable Sustainable Energy Rev. 2011, 15, 1−23. (2) Porosoff, M. D.; Yang, X. F.; Boscoboinik, J. A.; Chen, J. G. G. Molybdenum Carbide as Alternative Catalysts to Precious Metals for Highly Selective Reduction of CO2 to CO. Angew. Chem., Int. Ed. 2014, 53, 6705−6709. (3) Lu, Q.; Rosen, J.; Zhou, Y.; Hutchings, G. S.; Kimmel, Y. C.; Chen, J. G. G.; Jiao, F. A selective and efficient electrocatalyst for carbon dioxide reduction. Nat. Commun. 2014, 5, 3242. (4) Gu, X. K.; Carneiro, J. S. A.; Nikolla, E. Heterogeneous electrocatalysts for CO2 reduction. Catalysis 2017, 29, 94−121. (5) Ebbesen, S. D.; Jensen, S. H.; Hauch, A.; Mogensen, M. B. High Temperature Electrolysis in Alkaline Cells, Solid Proton Conducting Cells, and Solid Oxide Cells. Chem. Rev. 2014, 114, 10697−10734. (6) Ebbesen, S. D.; Mogensen, M. Electrolysis of carbon dioxide in Solid Oxide Electrolysis Cells. J. Power Sources 2009, 193, 349−358. (7) Jensen, S. H.; Graves, C.; Mogensen, M.; Wendel, C.; Braun, R.; Hughes, G.; Gao, Z.; Barnett, S. A. Large-scale electricity storage utilizing reversible solid oxide cells combined with underground storage of CO2 and CH4. Energy Environ. Sci. 2015, 8, 2471−2479. (8) Hansen, J. B. Solid oxide electrolysis - a key enabling technology for sustainable energy scenarios. Faraday Discuss. 2015, 182, 9−48. (9) Ebbesen, S. D.; Sun, X. F.; Mogensen, M. B. Understanding the processes governing performance and durability of solid oxide electrolysis cells. Faraday Discuss. 2015, 182, 393−422. (10) Singh, V.; Hashigami, S.; Muroyama, H.; Matsui, T.; Inagaki, T.; Eguchi, K. Influence of Fabrication Routes on Microstructure and Electrochemical Performance of Ni-GDC Cathode for High Temperature CO2 Reduction in Solid Oxide Electrolysis Cells. J. Electrochem. Soc. 2016, 163, F3084−F3090. (11) Singh, V.; Muroyama, H.; Matsui, T.; Hashigami, S.; Inagaki, T.; Eguchi, K. Feasibility of alternative electrode materials for high temperature CO2 reduction on solid oxide electrolysis cell. J. Power Sources 2015, 293, 642−648. (12) Dry, M. E. The Fischer−Tropsch process: 1950−2000. Catal. Today 2002, 71, 227−241. (13) Shin, E. C.; Ahn, P. A.; Seo, H. H.; Jo, J. M.; Kim, S. D.; Woo, S. K.; Yu, J. H.; Mizusaki, J.; Lee, J. S. Polarization mechanism of high temperature electrolysis in a Ni-YSZ/YSZ/LSM solid oxide cell by parametric impedance analysis. Solid State Ionics 2013, 232, 80−96. (14) Schefold, J.; Brisse, A.; Zahid, M. Electronic Conduction of YttriaStabilized Zirconia Electrolyte in Solid Oxide Cells Operated in High H
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
Article
Industrial & Engineering Chemistry Research toward efficient carbon dioxide electrolysis. Phys. Chem. Chem. Phys. 2016, 18, 3137−3143. (35) Gan, L. Z.; Ye, L. T.; Wang, S. J.; Liu, M. Z.; Tao, S. W.; Xie, K. Demonstration of direct conversion of CO2/H2O into syngas in a symmetrical proton-conducting solid oxide electrolyzer. Int. J. Hydrogen Energy 2016, 41, 1170−1175. (36) Qi, W. T.; Xie, K.; Qin, Q. Q.; Zhou, Q.; Wang, Y.; Zhang, Y.; Wu, Y. C. Efficient carbon dioxide electrolysis based on ceria cathode loaded with metal catalysts. J. Solid State Electrochem. 2014, 18, 3415−3425. (37) Ahn, J. Y.; Kim, B. K.; Park, J. S. Effects of Metal Catalysts on CoElectrolysis of Steam and Carbon Dioxide. J. Electrochem. Soc. 2016, 163, F1288−F1293. (38) Tao, G.; Sridhar, K. R.; Chan, C. L. Study of carbon dioxide electrolysis at electrode/electrolyte interface: Part II. Pt-YSZ cermet/ YSZ interface. Solid State Ionics 2004, 175, 621−624. (39) Tao, G.; Sridhar, K. R.; Chan, C. L. Study of carbon dioxide electrolysis at electrode/electrolyte interface: Part I. Pt/YSZ interface. Solid State Ionics 2004, 175, 615−619. (40) Xie, Y. M.; Xiao, J.; Liu, D. D.; Liu, J.; Yang, C. H. Electrolysis of Carbon Dioxide in a Solid Oxide Electrolyzer with Silver-GadoliniumDoped Ceria Cathode. J. Electrochem. Soc. 2015, 162, F397−F402. (41) Alioshin, Y.; Kohn, M.; Rothschild, A.; Karni, J. High Temperature Electrolysis of CO2 for Fuel Production. J. Electrochem. Soc. 2016, 163, F79−F87. (42) Wang, S. J.; Inoishi, A.; Hong, J.; Ju, Y.; Hagiwara, H.; Ida, S.; Ishihara, T. Ni-Fe bimetallic cathodes for intermediate temperature CO2 electrolyzers using a La0.9Sr0.1Ga0.8Mg0.2O3 electrolyte. J. Mater. Chem. A 2013, 1, 12455−12461. (43) Wang, S. J.; Tsuruta, H.; Asanuma, M.; Ishihara, T. Ni-FeLa(Sr)Fe(Mn)O3 as a New Active Cermet Cathode for IntermediateTemperature CO2 Electrolysis Using a LaGaO3-Based Electrolyte. Adv. Energy Mater. 2015, 5, 1401003. (44) Liu, S. B.; Liu, Q. X.; Luo, J. L. Highly Stable and Efficient Catalyst with In Situ Exsolved Fe-Ni Alloy Nanospheres Socketed on an Oxygen Deficient Perovskite for Direct CO2 Electrolysis. ACS Catal. 2016, 6, 6219−6228. (45) Liu, S. B.; Liu, Q. X.; Luo, J. L. CO2-to-CO conversion on layered perovskite with in situ exsolved Co-Fe alloy nanoparticles: an active and stable cathode for solid oxide electrolysis cells. J. Mater. Chem. A 2016, 4, 17521−17528. (46) Kim-Lohsoontorn, P.; Bae, J. Electrochemical performance of solid oxide electrolysis cell electrodes under high-temperature coelectrolysis of steam and carbon dioxide. J. Power Sources 2011, 196, 7161−7168. (47) Kresse, G.; Furthmuller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B: Condens. Matter Mater. Phys. 1996, 54, 11169−11186. (48) Kresse, G.; Furthmuller, 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. (49) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (50) Cadi-Essadek, A.; Roldan, A.; de Leeuw, N. H. Ni Deposition on Yttria-Stabilized ZrO2(111) Surfaces: A Density Functional Theory Study. J. Phys. Chem. C 2015, 119, 6581−6591. (51) Ammal, S. C.; Heyden, A. Combined DFT and Microkinetic Modeling Study of Hydrogen Oxidation at the Ni/YSZ Anode of Solid Oxide Fuel Cells. J. Phys. Chem. Lett. 2012, 3, 2767−2772. (52) Henkelman, G.; Uberuaga, B. P.; Jonsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J. Chem. Phys. 2000, 113, 9901−9904. (53) Henkelman, G.; Jonsson, H. Improved tangent estimate in the nudged elastic band method for finding minimum energy paths and saddle points. J. Chem. Phys. 2000, 113, 9978−9985. (54) Gu, X. K.; Liu, B.; Greeley, J. First-Principles Study of Structure Sensitivity of Ethylene Glycol Conversion on Platinum. ACS Catal. 2015, 5, 2623−2631.
(55) Mukherjee, J.; Linic, S. First-principles investigations of electrochemical oxidation of hydrogen at solid oxide fuel cell operating conditions. J. Electrochem. Soc. 2007, 154, B919−B924. (56) Ingram, D. B.; Linic, S. First-Principles Analysis of the Activity of Transition and Noble Metals in the Direct Utilization of Hydrocarbon Fuels at Solid Oxide Fuel Cell Operating Conditions. J. Electrochem. Soc. 2009, 156, B1457−B1465. (57) Tsiplakides, D.; Vayenas, C. G. Electrode work function and absolute potential scale in solid-state electrochemistry. J. Electrochem. Soc. 2001, 148, E189−E202. (58) Norskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J. R.; Bligaard, T.; Jonsson, H. Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J. Phys. Chem. B 2004, 108, 17886− 17892. (59) Rostamikia, G.; Mendoza, A. J.; Hickner, M. A.; Janik, M. J. Firstprinciples based microkinetic modeling of borohydride oxidation on a Au(111) electrode. J. Power Sources 2011, 196, 9228−9237. (60) Suthirakun, S.; Ammal, S. C.; Munoz-Garcia, A. B.; Xiao, G. L.; Chen, F. L.; zur Loye, H. C.; Carter, E. A.; Heyden, A. Theoretical Investigation of H2 Oxidation on the Sr2Fe1.5Mo0.5O6 (001) Perovskite Surface under Anodic Solid Oxide Fuel Cell Conditions. J. Am. Chem. Soc. 2014, 136, 8374−8386. (61) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamentals and Applications; John Wiley & Sons, Inc., 2001. (62) Gu, X. K.; Nikolla, E. Fundamental Insights into HighTemperature Water Electrolysis Using Ni-Based Electrocatalysts. J. Phys. Chem. C 2015, 119, 26980−26988. (63) Peng, G. W.; Sibener, S. J.; Schatz, G. C.; Ceyer, S. T.; Mavrikakis, M. CO2 Hydrogenation to Formic Acid on Ni(111). J. Phys. Chem. C 2012, 116, 3001−3006. (64) Gu, X. K.; Li, W. X. First-Principles Study on the Origin of the Different Selectivities for Methanol Steam Reforming on Cu(111) and Pd(111). J. Phys. Chem. C 2010, 114, 21539−21547. (65) Feng, X.; Cerda, J. I.; Salmeron, M. Orientation-Dependent Interaction between CO2 Molecules Adsorbed on Ru(0001). J. Phys. Chem. Lett. 2015, 6, 1780−1784. (66) Wang, S. G.; Liao, X. Y.; Cao, D. B.; Huo, C. F.; Li, Y. W.; Wang, J. G.; Jiao, H. J. Factors controlling the interaction of CO2 with transition metal surfaces. J. Phys. Chem. C 2007, 111, 16934−16940. (67) Shishkin, M.; Ziegler, T. Hydrogen Oxidation at the Ni/YttriaStabilized Zirconia Interface: A Study Based on Density Functional Theory. J. Phys. Chem. C 2010, 114, 11209−11214.
I
DOI: 10.1021/acs.iecr.7b00854 Ind. Eng. Chem. Res. XXXX, XXX, XXX−XXX
