Forum Article pubs.acs.org/IC
Tailoring Electrocatalysts for Selective CO2 or H+ Reduction: Iron Carbonyl Clusters as a Case Study Atefeh Taheri and Louise A. Berben* Department of Chemistry, University of California, Davis, California 95616, United States ABSTRACT: The design of electrocatalysts that will selectively transfer hydride equivalents to either H+ or CO2 to afford H2 or formate is a long-standing goal in molecular electrocatalysis. In this Forum Article, we use experimentally determined thermochemical parameters, hydricity and pKa values, to rationalize our observations that the carbide-containing iron carbonyl cluster [Fe4C(CO)12]2− reduces H+ to H2 in the presence of CO2 in either acetonitrile (MeCN), MeCN with 5% water, or buffered water (pH 5−13), with no traces of formate or other carbon-containing products observed. Our previous work has shown that the closely related nitride-containing clusters [Fe4N(CO)12]− and [Fe4N(CO)11(PPh3)]− will also reduce H+ to H2 in either MeCN with 5% water or buffered water (pH 5−13), but upon the addition of CO2, they selectively generate formate. The thermochemical measurements on [Fe4C(CO)12]2− predict that the free energy for transfer of hydride, in MeCN, from the intermediate [HFe4C(CO)12]2− to CO2 is thermoneutral and to H+ is −32 kcal mol−1. In water, these values are less than −19.2 and −8.6 kcal mol−1, respectively. These results suggest that generation of both H2 and formate should be favorable in aqueous solution and that kinetic effects, such as a fast rate of H2 evolution, must influence the observed selectivity for generation of H2. The hydride-donating ability of [HFe4N(CO)12]− is lower than that of [HFe4C(CO)12]2− by 5 kcal mol−1 in MeCN and by at least 3 kcal mol−1 in water, and we speculate that this more modest reactivity provides the needed selectivity to obtain formate. We also discuss predictions that may guide future catalyst design.
■
INTRODUCTION One promising strategy to mitigate the challenges of a petroleum-based economy is to convert the energy derived from renewable resources, via electrochemical methods, into carbon-neutral fuels.1 The production of H2 from water and the production of formate from CO2 are two of the simplest reactions possible, and both can be achieved by transfer of just one hydride equivalent from a metal hydride intermediate to the substrate: either H+ to give H2 or CO2 to give formate. This hydride intermediate can be generated by the reduction of protons. For this approach to be feasible, reaction of the metal hydride with either H+ or CO2 must occur with complete selectivity if pure H2 or formate is to be obtained. The design of catalysts to achieve this goal is challenging.2,3 On the basis of the results described in this manuscript, we propose hydricity ranges in water that should be targeted for efficient H2 production from water, or alternatively, for selective C−H bond formation with CO2 in water. These measurements also provide a correlation between hydricity measurements in water and those in acetonitrile (MeCN) and a comparison to the large literature of hydricity studies previously performed in MeCN solution. The free energy for release of H− from a metal hydride (M− H) is denoted as ΔG°H−, and the measurement of ΔG°H− (or hydricity) enables some predictive element in understanding hydride transfer and C−H bond-forming processes and, thus, in the design of suitable electrocatalysts for CO2 reduction.4 As an example of a recent work, Linehan and co-workers showed that © XXXX American Chemical Society
by using a cobalt hydride compound (Chart 1A), which is more hydridic than formate, efficient hydrogenation of CO2 to formate was achieved.5 More recently, Peters and co-workers have studied both catalytic and stoichiometric hydrogenation of CO2 using a series of triphosphinoiron compounds (Chart 1B).6 On the basis of hydricity measurements performed in tetrahydrofuran (THF) and converted to the scale in MeCN, formate production was predicted to be unfavorable, but the energy released upon coordination of the formate product to the iron center was thought to provide enough driving force to make the reaction thermally accessible. Catalytic transformations were only possible in methanol. Indeed, recent work from our own laboratory showed that the iron carbonyl cluster [Fe4N(CO)12]− is selective for electrocatalytic formate production in CO2-saturated buffered water7 but produces a mixture of H2 and formate in pure MeCN solution.8 Reactions run in MeCN/ water (95:5) are also selective for formate production, but reaction rates are slower. These solvent dependencies are likely related to changes in the hydricity and illustrate another important parameter in catalyst design. Special Issue: Small Molecule Activation: From Biological Principles to Energy Applications Part 3 Received: October 5, 2015
A
DOI: 10.1021/acs.inorgchem.5b02293 Inorg. Chem. XXXX, XXX, XXX−XXX
Forum Article
Inorganic Chemistry Chart 1. (A) Cobalt Phosphine (Linehan et al.),5 (B) Triphosphinoiron Hydride (Peters et al.),6 (C) Molybdenum/Tungsten Hydride (Bruno et al.),13 (D) Nickel/Platinum/Palladium Phosphine (Dubois et al.)14 and Nickel Phosphine (Yang et al.),23 (E) Ruthenium Pyridyl (Creutz et al.),22 and (F) Chromium and Tungsten Carbonyl Hydride (Darensbourg et al.)29
Method 2 (eqs 6−10) requires an equilibrium measurement of the heterolytic cleavage of hydrogen to form the metal hydride of interest (eq 6).
Another method that determines the relative hydricity measures the equilibrium constant of the reaction between a hydride donor and a hydride acceptor, where the hydride acceptor ability has been estimated based on one of the two methods above. For example, DuBois and co-workers measured the hydricity of [HPd(PNP)2]+ [where PNP is Et2PCH2N(Me)CH2PEt2] based on its reaction with [Pt(PNP)2]2+ (eqs 11 and 12), where the hydricity of [HPt(PNP)2]+ was known.18 To accompany these recent insights and significant progress in in the design of catalysts for selective H2 or formate production,9 still more information on the thermochemistry and elementary reaction steps of metal catalysts is needed. Acidity measurements of metal hydrides are discussed in many literature reports, review articles, and book chapters,10 and these approaches and measurements date back to the 1950s, with Heiber and co-workers.11 The measurement of hydricity values is a more recent subject, only started in recent decades.12 Early measurements of the hydricity for metal hydrides were performed by Sarker and Bruno using IR spectroscopy on a series of molybdenum and tungsten compounds, with the general formula (C5R5)M(CO)2(L)H (Chart 1C).13 The method using thermochemical cycles was further developed, primarily using 31P NMR spectroscopy, by DuBois and coworkers for a series of nickel and platinum phosphine proton reduction electrocatalysts (Chart 1D).14 DuBois and coworkers also reexamined the hydricity values reported by Sarker and Bruno and determined that their results included a significant error due to a failure to establish an equilibrium.15,16 On the basis of these reports, the use of thermochemical cycles to obtain hydricity values has become well-established. The determination of an absolute hydricity value is generally approached using a thermochemical cycle to extract a hydricity value from the experimental measurement of, for example, a pKa value. There are two thermochemical cycles most commonly employed to acquire absolute values.17 In brief, method 1 (eqs 1−5) uses a cycle based on measurements of the pKa value of the hydride and the two-electron redox potential of the conjugate base of the metal hydride.
[HPd(PNP)2 ]+ + [Pt(PNP)2 ]2 + ⇌ [Pd(PNP)2 ]2 + + [HPd(PNP)2 ]+
(11)
ΔG°H−([HPd(PNP)2 ]+ ) = ΔG°H−([HPt(PNP)2 ]+ ) + ΔG°(eq 11)
(12)
The current literature reports on hydricity measurements and calculations are primarily based on experiments performed in MeCN solution.19 However, major efforts by many to develop electrocatalysts for sustainable energy production that operate in water demonstrate a dire need for more thermochemical insight in aqueous solution.20,21 Limited solubility or instability of many metal hydrides in water is likely a key reason that very few reports of hydricity measurements are available in water. Creutz and co-workers made the earliest valuable contributions to our understanding of the hydricity in water. The reaction of [Ru(bpy)2(tpy)H]+ (where bpy is 2,2′-bipyridine and tpy is 2,2′:6′,2″-terpyridine) with CO2 was studied using UV−vis spectroscopy (Chart 1E and Table 1);22 reports on the relative hydricity values of a series of organometallic complexes were present in water. Later on, Creutz and co-workers reported the hydricity of [Ru(η6-C6Me6)(bpy)H]+ in both H2O and MeCN (Table 1).16 Table 1. Thermochemical Properties of Reported Metal Hydrides, H2, and Formate in MeCN and H2O ΔG°H− (kcal mol−1) compound
H2O +
[Ru(bpy)2(tpy)H] [Ru(η6-C6Me6)(bpy)H]+ [HNi(DHMPE)2]+ [HFe4C(CO)12]2− [HFe4N(CO)12]− H2 HCOO−
22
13.8 22.216 30.023 [Fe4C(CO)12]2− and that this decrease correlates with an increase in the hydricity for the reduced hydride intermediates [HFe4N(CO)12]−, [HFe4N(CO)11(PPh3)]−, and [HFe4C(CO)12]2−. These observations are consistent with the conclusion that the ability of the clusters to donate H− increases as the cluster core becomes more electron-rich.
Table 3. IR Absorption Bands for CO Ligands in 0.1 M Bu4NPF6 and MeCN before (Resting State) and after (Reduced State) Application of the Potential compound − 7,8
[Fe4N(CO)12] [Fe4N(CO)11(PPh3)]− 41 [Fe4C(CO)12]2− 34
ground state νCO (cm−1) 2018 1987 1969
1989 1972 1945
reduced state νCO (cm−1) 1940 1897 1874
■
CONCLUSION These results provide an opportunity to compare the reaction selectivity and thermochemical properties of three structurally similar iron electrocatalysts. We conclude that the relatively low hydricity of [HFe4N(CO)12]− [49 kcal mol−1 (MeCN) and 15.5 kcal mol−1 (H2O)] enables selective conversion of CO2 to formate in water and mixtures of water and acetonitrile. For the stronger hydride donor, [HFe4C(CO)12]2− [44 kcal mol−1 (MeCN) and
