Toward Rational Design of 3d Transition Metal Catalysts for CO2

May 10, 2016 - Synopsis. Pushing electron density to the metal center dramatically improves the CO2 hydrogenation reactivity for complexes with relati...
0 downloads 18 Views 2MB Size
Article pubs.acs.org/IC

Toward Rational Design of 3d Transition Metal Catalysts for CO2 Hydrogenation Based on Insights into Hydricity-Controlled RateDetermining Steps Bhaskar Mondal, Frank Neese, and Shengfa Ye* Max-Planck Institut für Chemische Energiekonversion, Stiftstraße 34-36, D-45470 Mülheim an der Ruhr, Germany S Supporting Information *

ABSTRACT: Carbon dioxide functionalization attracts much interest due to the current environmental and energy challenges. Our earlier work (Mondal, B.; Neese, F.; Ye, S. Inorg. Chem. 2015, 54, 7192−7198) demonstrated that CO2 hydrogenation mediated by base metal catalysts [M(H)(η2-H2)(PP3Ph)]n+ (M = Co(III) and Fe(II), n = 1, 2; PP3Ph = tris(2-(diphenylphosphino)phenyl)phosphine) features discrete ratedetermining steps (RDSs). Specifically, the reaction with [CoIII(H)(η2-H2)(PP3Ph)]2+ passes through a hydride-transfer RDS, whereas the conversion with [FeII(H)(η2-H2)(PP3Ph)]+ traverses a H2-splitting RDS. More importantly, we found that the nature and barrier of the RDS likely correlate with the hydride affinity or hydricity of the dihydride intermediate [M(H)2(PP3Ph)](n−1)+ generated by H2-splitting. In the present contribution, following this notion we design a series of potential Fe(II) and Co(III) catalysts, for which the respective dihydride species possess differential hydricities, and computationally investigated their reactivity toward CO2 hydrogenation. Our results reveal that lowering the hydrictiy of [CoIII(H)2(PP3Ph)]+ by introducing anionic anchors in PP3Ph dramatically decreases the hydride-transfer RDS barrier, as shown for the enhanced reactivity of [Co(H)(η2-H2)(CP3Ph)]+ and [Co(H)(η2-H2)(SiP3Ph)]+ (CP3Ph = tris(2-(diphenylphosphino)phenyl)methyl, SiP3Ph = tris(2-(diphenylphosphino)phenyl)silyl), while the same ligand modification increases the H2-splitting RDS barriers for [Fe(H)(η2-H2)(CP3Ph)] and [Fe(H)(η2-H2)(SiP3Ph)] relative to that for [Fe(H)(η2-H2)(PP3Ph)]+. Conversely, upon increasing the hydricity of [FeII(H)2(PP3Ph)] by adding an electron-withdrawing group to PP3Ph, the transformation with [Fe(H)(η2-H2)(PP3PhNO2)]+ (PP3PhNO2 = tris(2-(diphenylphosphino)-4-nitrophenyl)phosphine) is predicted to encounter a lower barrier for H2-splitting and a higher barrier for hydride transfer than those for [Fe(H)(η2-H2)(PP3Ph)]+. Thus, we have shown that hydricity can be used as a guide to direct the rational design and development of more efficient catalysts.



INTRODUCTION Research related to the transformation of carbon dioxide into liquid fuels and useful chemicals has been growing considerably over the past decade in order to develop carbon-neutral energy sources.1 However, the thermodynamic stability and the kinetic inertness of CO2 pose significant challenges to chemists.2 Homogeneous CO2 hydrogenation catalyzed by transition metals, for which formate, formic acid, or their derivatives are the primary products, is one of the most efficient pathways to transform CO2 into useful chemicals.3 To date, noble metals such as rhodium,4 ruthenium,5 and iridium6 have been extensively employed, and impressively high catalytic activity has been reported. For instance, CO2 reduction mediated by a pincer-supported iridium catalyst Ir(III)-PNP (PNP = 2,6bis(dialkylphosphinomethyl)pyridine) shows a turnover frequency (TOF) of as high as 150 000 h−1 and a turnover number (TON) of 3 500 000 for CO2-to-formate (HCOO−) conversion.6a Unfortunately, the extortionate price of precious metals significantly limits large-scale applications of such catalysts. The base metals such as iron, cobalt, and nickel are much cheaper and biorelevant, and therefore deserve more © XXXX American Chemical Society

attention. However, 3d transition metal catalysts are often less efficient in CO2 activation. Recently, several Fe- and Co-based catalysts have been reported to show comparable reactivity to noble metals. For example, [Fe(H)(η2-H2)(PP3Ph)]+ (PP3Ph = tris(2-(diphenylphosphino)phenyl)phosphine) displays a maximum TON of 5100 with a TOF of 255 h−1 toward formation of dimethylformamide (the reaction takes place in the presence of HNMe2).7 The TOF of Co(dmpe)2H (dmpe = 1,2bis(dimethylphosphino)ethane) converting CO2 to formate was found to reach 3400 h−1 at ambient conditions.8 Yet, in order to accomplish optimal performances, harsh reaction conditions, such as high temperature and pressure for the Fe case (100 °C and 100 atm, respectively)7 and the use of a strong base (Verkade’s base, pKa 33.7 in acetonitrile)8 for the Co case, have to be applied. Thus, the design and development of base metal catalysts that can operate in mild conditions is highly desirable. Undoubtedly, this requires a deep chemical understanding of the reaction mechanism, which may shed light Received: February 25, 2016

A

DOI: 10.1021/acs.inorgchem.6b00471 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Scheme 1. CO2 Hydrogenation Catalytic Cycle Proposed for Complexes 1Fe and 1Co

Scheme 2. Fe(II) and Co(III) Complexes Generated from 1Fe and 1Co with Modified Ligands

energy change of the hydride dissociation reaction, MH → M+ + H−, which measures the strength of the M−H− interaction in a hydride acceptor.11 By definition, a more positive ΔG°H−(MH) value, higher hydricity, corresponds to greater hydride-accepting ability. Earlier studies have established an elegant correlation between the hydricity of metal-hydride catalysts and their hydride transfer reactivity.8,11a,12 In the present case, our work even suggested that not only the barrier of hydride transfer but also that of H2-splitting are likely to correlate with the hydricity of the dihydride intermediate 3.10 Specifically, for complex 3Fe, possessing a relatively low hydricity (65.0 kcal/ mol), H2-splitting was identified as the RDS, whereas for complex 3Co, having a much higher hydricity (99.3 kcal/mol), the reaction passes through a hydride-transfer RDS. On the basis of this observation, hydricity could be used as a practical guide to tailor the catalytic activity. Thus, we proposed that pulling electron density from the metal center in complex 1Fe likely increases the catalytic activity, whereas for complex 1Co catalytic activity could be improved by pushing more electron

on pivotal factors that control the nature of the RDS and modulate its barrier. Eventually, the insights may trigger new ideas for designing more efficient catalysts. To the best of our knowledge, no systematic studies following this route have been published yet, since most mechanistic and theoretical investigations deal with individual reaction mechanisms only.9 In our earlier work, we explored the reaction mechanism of CO2 hydrogenation catalyzed by [M(H)(η2-H2)(PP3Ph)]n+ (1) (M = Fe(II), Ru(II), and Co(III); n = 1, 2) types of complexes.10 The reactions with the three different metal complexes (1Fe, 1Ru, and 1Co) follow a common mechanism as shown in Scheme 1. The overall catalytic cycle exhibits three distinct reaction phases, namely, base-promoted heterolytic H2splitting, (2 → 3·Et3NH+), hydride transfer from the metal center to CO2 (4 → 5), and product dissociation (5 → 1). We found that the hydride affinity or hydricity of the dihydride intermediate 3 dictates the nature of the RDS of the overall catalytic cycle, which can be either H2-splitting or hydride transfer. Hydricity, ΔG°H−(MH), is defined as the Gibbs free B

DOI: 10.1021/acs.inorgchem.6b00471 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. DLPNO-CCSD(T)-computed free energy profiles for the key steps of CO2 hydrogenation by the Co(III) complexes. Schematic drawings of the structures for the species involved are provided on top for clarity.

in order to make a comparison between the results from two different levels of theory.

density to the metal center, provided that the nature of the RDS is the same. Herein, we apply the strategy proposed above to designing more efficient 3d transition metal based catalysts for CO2 hydrogenation. To this end, a series of modified ligands with differential electron-donating ability are employed, and the corresponding complexes are presented in Scheme 2. PP3PhNO2 (PP 3 PhNO2 = tris(2-(diphenylphosphino)-4-nitrophenyl)phosphine), generated by adding NO2 groups to the second coordination sphere of PP3Ph, is expected to have an attenuated electron-donating ability, whereas it is enhanced by replacing the neutral phosphine anchor in PP3Ph with anionic C− and Si− atoms termed CP3Ph (tris(2-(diphenylphosphino)phenyl)methyl) and SiP3Ph (tris(2-(diphenylphosphino)phenyl)silyl), respectively. Note that SiP3Ph and the closely related ligands CP3iPr and SiP3iPr have been used by Peters and co-workers to prepare a range of Fe−N2 complexes.12b,13 We have undertaken a detailed mechanistic investigation on the new systems, [Fe(H)(η 2-H2 )(CP3 Ph)] (1 Fe/C ), [Fe(H)(η 2-H 2)(SiP 3Ph)] (1Fe/Si), [Fe(H)(η2-H2)(PP3PhNO2)]+ (1Fe/NO2), [Co(H)(η2H2)(CP3Ph)]+ (1Co/C), and [Co(H)(η2-H2)(SiP3Ph)]+ (1Co/Si), and compared the reactivity with their parent complexes 1Fe and 1Co. Therefore, this work focuses on the barriers of the possible RDSs, H2-splitting, and hydride transfer, and a complete mechanistic investigation for complexes 1Fe and 1Co has been published.10 The recent developments in highly correlated wave function based ab initio methods allow us to produce reliable reaction energies and barriers within the chemical accuracy with a reasonable computational cost. Specifically, the domain-based local pair natural orbital coupled-cluster approach with single and double excitations and perturbative triple corrections (DLPNO-CCSD(T))14 has been used to compute the reaction energetics. Density functional theory (DFT) calculations have also been performed



COMPUTATIONAL SETUP

All calculations were performed by using the ORCA program package.15 The meta-GGA density functional M06-L16 was employed for geometry optimizations and frequency calculations. The relativistic effects were accounted for with the scalar relativistic zeroth-order regular approximation (ZORA).17 To accelerate calculations with meta-GGA functionals, resolution of the identity (RI) approximation18 was used in conjunction with the auxiliary basis set def2-TZV/J.18b All geometry optimizations were carried out by employing Ahlrichs tripleζ quality basis set without f-polarization functions def2-TZVP(−f)19 for the first coordination sphere (for Fe, P, C, Si, N, H, H2, and CO2), whereas the rest of the atoms were described with double-ζ-quality split-valence basis set def2-SVP.20 To mimic solvation effects, the conductor-like screening solvation model (COSMO)21 was applied during geometry optimizations and energy calculations, for which methanol (ε = 32.63) was chosen as the solvent, consistent with the experiment. We also tested an explicit solvation model in the hydridetransfer step by placing two CH3OH molecules within H-bonding distances to CO2, but in the course of the geometry optimization the solvent molecules moved apart and no H-bonding effect was found eventually. For the H2-splitting processes, we also investigated the competitive solvent binding to the metal center. The H2 binding to the Fe(II) and Co(III) centers is computed to be ∼10 and ∼3 kcal/mol more favorable than CH3OH, respectively. Both observations suggest that the implicit solvation model COSMO may account for most solvation effects in the reactions. Harmonic vibrational frequencies were computed on the optimized geometries to ensure that all local minima display real frequencies only, whereas the transition states (TSs) were characterized by a single imaginary frequency. The zeropoint vibrational energies, thermal corrections to electronic energies, and entropy terms for the optimized geometries were obtained from the frequency calculations at the M06-L level. Final single-point energies were calculated by using conventional DFT as well as highly correlated ab initio approaches. For DFT, we C

DOI: 10.1021/acs.inorgchem.6b00471 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 2. DLPNO-CCSD(T)-computed free energy profiles for the key steps of CO2 hydrogenation by the Fe(II) complexes. Schematic drawing of the structures of the species involved are provided on top for clarity.



employed the M06-L functional and hybrid-GGA B3LYP22 functional in conjunction with the more flexible triple-ζ def2-TZVPP23 (including high angular momentum polarization functions for all elements) basis set for all atoms. For DFT-B3LYP calculations, an atom-pairwise dispersion correction involving Becke−Johnson damping (D3BJ) proposed by Grimme24 was used to account for noncovalent interactions. All the DFT-B3LYP single-point energy calculations were performed with the density fitting (RI-J) and “chain of spheres” (COSX)25 approximation RIJCOSX, which treats the Coulomb term via RI and the exchange term via seminumerical integration, for which auxiliary Coulomb fitting basis set def2-TZVPP/ J18 was used. Accurate electronic energies were computed with the domain-based local pair natural orbital coupled cluster method with single and double excitations and perturbative triple corrections (DLPNO-CCSD(T)) in conjunction with the def2-TZVPP basis set. The correlation fitting auxiliary basis set def2-TZVPP/C was used to accelerate the DLPNO-CCSD(T) calculations. The RIJCOSX approximation was also applied during the DLPNO-CCSD(T) calculations with the auxiliary Coulomb fitting basis set def2TZVPP/J. The solvent corrections to the electronic energy were taken from the M06-L single-point energy calculations. To estimate the basis set incompleteness error in our DLPNO-CCSD(T)/def2TZVPP results, we have calculated the reaction barriers of 1Co/Si at the basis set limit using a direct basis set extrapolation technique26 with the cc-pVDZ27 and cc-pVTZ27 basis sets. A detailed description of this extrapolation scheme for transition metal complexes is described elsewhere.28 In comparison with the basis-set-extrapolated values (Table S1), the calculations with def2-TZVPP show a maximum deviation of approximately 2 kcal/mol, falling within the uncertainty limit of the DLPNO-CCSD(T) method itself.14b Thus, the DLPNOCCSD(T)/def2-TZVPP values are reported in the following. The Gibbs free energy surfaces are constructed by intrinsic barriers, which are referenced with respect to the corresponding reactant complexes, because errors in the estimated entropic contributions for association reactions are lower than those computed by choosing the infinitely separated reactants as the zero point.29

RESULTS AND DISCUSSION Figures 1 and 2 present the DLPNO-CCSD(T) free energy profiles for the key reaction steps of CO2 hydrogenation by the Co(III) and Fe(II) complexes, respectively. The M06-L calculations produce similar reaction energetics with a maximum deviation of ∼5 kcal/mol compared to the DLPNO-CCSD(T) results (Table S2). It is evident that introducing the charge-neutralizing anionic donors C− and Si− in complex 1Co dramatically reduces the barrier for the hydridetransfer RDS. Specifically, the difficult hydride-transfer process (4 → 5)10 for complex 1Co becomes accessible for complexes 1Co/C and 1Co/Si with a barrier of 8.3 and 11.8 kcal/mol, respectively. At the same time the H2-splitting barriers increase moderately by 2.2 and 7.0 kcal/mol for complexes 1Co/C and 1Co/Si, respectively. The RDS for complex 1Co/Si appears to be ambiguous since the barriers for H2-splitting and hydride transfer are nearly identical (11.4 and 11.8 kcal/mol, respectively). For complexes 1Fe, 1Fe/C, and 1Fe/Si, a similar situation is observed. The same ligand modification leads to increases of 2.1 and 5.2 kcal/mol in the barrier for the H2splitting RDS, respectively, whereas the corresponding barrier for hydride transfer drops by ∼8 kcal/mol for both the complexes. Conversely, upon adding the electron-withdrawing group NO2 in complex 1Fe, the H2-splitting barrier of complex 1Fe/NO2 declines to 8.5 kcal/mol and the hydride-transfer barrier enlarges to 11.8 kcal/mol, the latter value close to that computed for the H2-splitting RDS of complex 1Fe within the uncertainty limit14b of our calculations. Therefore, the RDS for complex 1Fe/NO2 switches from H2-splitting to hydride transfer, and the overall reactivity remains similar to complex 1Fe. Taken together, the theoretical results verify the viability of our proposed ligand modification strategy. In our previous work10 we established the role of the conjugate acid Et3NH+ in enlarging the corresponding driving force and hence lowering the hydride-transfer barrier. Similar effects have also been D

DOI: 10.1021/acs.inorgchem.6b00471 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry observed here; the barrier decreases by ∼3−7 kcal/mol and the driving force increases up to ∼14 kcal/mol depending on the systems. The H2-splitting process for [Fe(H)(η2-H2)(SiP3iPr)] (SiP3iPr = tris(2-(dipisopropylphosphino)phenyl)silyl) was suggested experimentally to involve a prohibitively high barrier, in accordance with our calculations on complex 1Fe/Si, and the entire reaction of CO2 hydrogenation hence follows a distinct pathway.12b The theoretical study on this new mechanism is definitely beyond the scope of the current work and will be reported in a separate publication. For all complexes under consideration, a good linear correlation (R2 = 0.91, Figure S3) between the H2-splitting barrier (ΔG⧧) and the reaction free Gibbs energy change (ΔG) is identified, as predicted by the Bell−Evans−Polanyi (BEP) principle.30 In comparison with the parent complexes 1Co and 1Fe, introducing the anionic anchors is found to appreciably lower the driving force for H2-splitting as observed for complexes 1Co/C, 1Co/Si, 1Fe/C, and 1Fe/Si (Figures 1 and 2). On the contrary, the driving force for complex 1Fe/NO2 is increased by 7.8 kcal/mol compared to complex 1Fe. The reaction energy of H2-splitting measures the relative stability of metal-dihydride complex 3 relative to metal-dihydrogen complex 1. In our earlier study we show that it is the differential M−H− interaction in 3 that dictates the large variation of the reaction energies (∼50 kcal/mol), because the M−H2 binding enthalpies are similar for Co(III) and Fe(II) complexes with a difference of ∼5 kcal/mol only.10 The strength of the M−H− interaction in complex 3 can be quantified by its hydricity (ΔG°H−).11 The calculated ΔG°H− values (in acetonitrile) for all the species are presented in Table 1. The details of the employed calculation protocol are

The standard deviation of the computed hydricity is approximately 8 kcal/mol.10

present case a good correlation (R2 = 0.92) (Figure 3b) is established. This correlation can be rationalized by the BEP principle as well. The reaction energy is largely controlled by the differential energy penalty required for breaking the respective M−H− bond, because formate (HCOO−) or its salt (HCOOH·NEt3) is the common product for all the hydride-transfer processes. The correlation between the hydride-transfer barrier and the driving force is, in fact, found to be of the same quality (R2 = 0.92, Figure S4) as that for the H2-splitting step. The overwhelming hydricity of complex 3Co (99.3 kcal/mol) hinders subsequent hydride transfer to CO2 (3 → 5). In contrast, the more electron rich metal centers in complexes 3Co/C and 3Co/Si result in the substantially weakened M−H− interaction, as evidenced by their remarkably reduced hydricities, ∼30 kcal/mol lower than that for 3Co. As a result, the hydride transfers for complexes 3Co/C and 3Co/Si occur with moderate barriers (Figure 1). Similar effects have also been observed for complexes 3Fe/C and 3Fe/Si. On the contrary, the electron-withdrawing effect of NO2 in complex 3 Fe/NO2 strengthens the M−H− interaction, and consequently the hydricity increases from 65.0 kcal/mol (3Fe) to 72.3 kcal/mol (3Fe/NO2). This in turn raises the hydride-transfer barrier by 2.1 kcal/mol and the associated RDS step switches from H2splitting to hydride transfer. Taken together, for H2-splitting and hydride transfer, complexes 1Co/C and 1Co/Si have accessibly low barriers and hence appear to be very promising catalysts toward CO2 hydrogenation besides 1Fe. As shown in Figure 3, the barriers for both key steps elegantly correlate with the hydricity of the hydride-transfer agent. If a dihydride species has a rather high hydricity, hydride transfer may encounter a exceedingly high barrier as found for complex 3Co. Conversely, if a facile hydride-transfer process takes place, because the hydride-transfer reactant possesses an extremely low hydricity, then the H2-splitting process to generate this species is expected to traverse a tremendous barrier as observed for complexes 3Fe/Si and 3Fe/C. For a useful catalyst, its hydricity should be close to that of HCOOH·NEt3, the final product of the reaction. Of all complexes under investigation, only complexes 3Fe, 3Co/C, and 3Co/Si have similar hydricity to that calculated for HCOOH·NEt3 (57.7 kcal/ mol),10 and in Figure 3 they are closest to the horizontal line representing the hydricity of HCOOH·NEt3 in both correlation plots. This reflects that a delicate balance between H2-splitting and hydride transfer has been achieved for complexes 3Fe, 3Co/C, and 3Co/Si. In line with the above reasoning, the reaction barriers of the two steps are comparable with the largest difference of 4.4 kcal/mol computed for complex 1Fe.

documented in the Supporting Information. Our computed hydricity of HCOO− matches the experimental value well, lending credence to our employed methodology. In fact, a reasonable correlation (R2 = 0.81, Figure S6) between ΔG°H− and ΔG is found. This corroborates that hydricity is a determining factor that modulates the H2-splitting driving forces. As such, ΔG⧧ should correlate with ΔG°H−. In fact, a correlation with a similar quality (R2 = 0.82, Figure 3a) between ΔG⧧ and ΔG°H− can be identified, indicating that the H2splitting process to generate a low-hydricity species encounters a higher barrier, whereas that to yield a high-hydricity product features a diminished barrier. Hydricity, by definition, measures the strength of a given M− H− interaction. Thus, one may expect a correlation between ΔG°H− and the hydride-transfer barrier.12,32 Indeed, in the

CONCLUSIONS In conclusion, our current calculations using the accurate ab initio DLPNO-CCSD(T) method along with DFT verify our proposed catalyst-design strategy10 for CO2 hydrogenation. The systematic investigations on a range of Co(III) and Fe(II) complexes show a good correlation between the key reaction barrier and the hydricity of the metal-dihydride complex. Dihydride species with relatively high hydricity typically feature a hydride-transfer RDS (e.g., complex 1Co). Enhancing the charge-donating power of the ligand (e.g., complexes 1Co/C and 1Co/Si), which weakens the metal-hydride interaction, is found to speed up the hydride-transfer process. In contrast, dihydride species possessing lower hydricity often encounter a high barrier for H2-splitting (e.g., complex 1Fe). Their reactivity can be fine-tuned by increasing the electron-withdrawing power of

Table 1. Hydricity ΔG°H− Computed by M06-La species −

HCOO HCOOH HCOOH·NEt3 3Fe 3Co 3Fe/NO2 3Fe/C 3Fe/Si 3Co/C 3Co/Si

ΔG°H−, kcal/mol

experiment31

45.2 134.8 57.7 65.0 99.3 72.3 38.5 41.2 68.4 65.5

43.0

a



E

DOI: 10.1021/acs.inorgchem.6b00471 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 3. Correlation plots between H2-splitting (a) and hydride-transfer (b) barriers with calculated hydricity. For comparison, a horizontal dashed line marks the calculated hydricity of HCOOH·NEt3. Fujita, E.; Gibson, D. H.; Goddard, W. A.; Goodman, D. W.; Keller, J.; Kubas, G. J.; Kung, H. H.; Lyons, J. E.; Manzer, L. E.; Marks, T. J.; Morokuma, K.; Nicholas, K. M.; Periana, R.; Que, L.; Rostrup-Nielson, J.; Sachtler, W. M. H.; Schmidt, L. D.; Sen, A.; Somorjai, G. A.; Stair, P. C.; Stults, B. R.; Tumas, W. Chem. Rev. 2001, 101, 953−996. (i) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois, D. L.; Dupuis, M.; Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.; Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.; Portis, A. R.; Ragsdale, S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.; Waldrop, G. L. Chem. Rev. 2013, 113, 6621−6658. (2) (a) Mondal, B.; Song, J.; Neese, F.; Ye, S. Curr. Opin. Chem. Biol. 2015, 25, 103−109. (b) Rakowski DuBois, M.; DuBois, D. L. Acc. Chem. Res. 2009, 42, 1974−1982. (c) Schneider, J.; Jia, H.; Muckerman, J. T.; Fujita, E. Chem. Soc. Rev. 2012, 41, 2036. (d) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Chem. Soc. Rev. 2009, 38, 89−99. (3) (a) Wang, W.-H.; Himeda, Y.; Muckerman, J. T.; Manbeck, G. F.; Fujita, E. Chem. Rev. 2015, 115, 12936−12973. (b) Saeidi, S.; Amin, N. A. S.; Rahimpour, M. R. J. CO2 Util. 2014, 5, 66−81. (c) Himeda, Y. Eur. J. Inorg. Chem. 2007, 2007, 3927−3941. (d) Leitner, W. Angew. Chem., Int. Ed. Engl. 1995, 34, 2207−2221. (e) Jessop, P. G.; Joó, F.; Tai, C.-C. Coord. Chem. Rev. 2004, 248, 2425−2442. (f) Jessop, P. G.; Ikariya, T.; Noyori, R. Chem. Rev. 1995, 95, 259−272. (g) Federsel, C.; Jackstell, R.; Beller, M. Angew. Chem., Int. Ed. 2010, 49, 6254−6257. (h) Wang, W.; Wang, S.; Ma, X.; Gong, J. Chem. Soc. Rev. 2011, 40, 3703. (4) (a) Leitner, W.; Dinjus, E.; Gassner, F. J. Organomet. Chem. 1994, 475, 257−266. (b) Tsai, J. C.; Nicholas, K. M. J. Am. Chem. Soc. 1992, 114, 5117−5124. (c) Hutschka, F.; Dedieu, A.; Eichberger, M.; Fornika, R.; Leitner, W. J. Am. Chem. Soc. 1997, 119, 4432−4443. (d) Gassner, F.; Leitner, W. J. Chem. Soc., Chem. Commun. 1993, 1465. (e) Gassner, F.; Dinjus, E.; Görls, H.; Leitner, W. Organometallics 1996, 15, 2078−2082. (f) Burgemeister, T.; Kastner, F.; Leitner, W. Angew. Chem., Int. Ed. Engl. 1993, 32, 739−741. (5) Ohnishi, Y.-Y.; Matsunaga, T.; Nakao, Y.; Sato, H.; Sakaki, S. J. Am. Chem. Soc. 2005, 127, 4021−4032. (b) Yin, C.; Xu, Z.; Yang, S.-Y.; Ng, S. M.; Wong, K. Y.; Lin, Z.; Lau, C. P. Organometallics 2001, 20, 1216−1222. (c) Tominaga, K.-I.; Sasaki, Y.; Kawai, M.; Watanabe, T.; Saito, M. J. Chem. Soc., Chem. Commun. 1993, 629. (d) Wesselbaum, S.; vom Stein, T.; Klankermayer, J.; Leitner, W. Angew. Chem. 2012, 124, 7617−7620. (e) Munshi, P.; Main, A. D.; Linehan, J. C.; Tai, C.C.; Jessop, P. G. J. Am. Chem. Soc. 2002, 124, 7963−7971. (f) Tai, C.C.; Pitts, J.; Linehan, J. C.; Main, A. D.; Munshi, P.; Jessop, P. G. Inorg. Chem. 2002, 41, 1606−1614. (g) Getty, A. D.; Tai, C.-C.; Linehan, J. C.; Jessop, P. G.; Olmstead, M. M.; Rheingold, A. L. Organometallics 2009, 28, 5466−5477. (6) (a) Tanaka, R.; Yamashita, M.; Nozaki, K. J. Am. Chem. Soc. 2009, 131, 14168−14169. (b) Hull, J. F.; Himeda, Y.; Wang, W.-H.; Hashiguchi, B.; Periana, R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E.

the supporting ligand (e.g., complex 1Fe/NO2). It is important to realize that the electronic-structure requirements for H2splitting and hydride transfer contradict each other. Hence, a successful catalyst must strike a delicate balance in order to accomplish both steps efficiently. To this end, in silico studies prior to catalyst synthesis can clearly be of major utility in guiding catalyst design.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b00471. Supporting Tables S1−S4, Figures S1−S6, hydricity calculation details, and Cartesian coordinates of all the species involved (PDF)



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We gratefully acknowledge the financial support from the MaxPlanck Society. Thanks are due to Dr. D. G. Liakos for helpful discussions about DLPNO-CCSD(T) calculations. We are also grateful to Prof. J. C. Peters (California Institute of Technology) for fruitful discussions and the learned reviewers for their constructive comments and suggestions to improve the quality of the manuscript.



REFERENCES

(1) (a) Sakakura, T.; Choi, J.-C.; Yasuda, H. Chem. Rev. 2007, 107, 2365−2387. (b) Mikkelsen, M.; Jørgensen, M.; Krebs, F. C. Energy Environ. Sci. 2010, 3, 43. (c) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Chem. Soc. Rev. 2009, 38, 89−99. (d) Centi, G.; Perathoner, S. Catal. Today 2009, 148, 191−205. (e) Yu, K. M. K.; Curcic, I.; Gabriel, J.; Tsang, S. C. E. ChemSusChem 2008, 1, 893−899. (f) Langanke, J.; Wolf, A.; Hofmann, J.; Böhm, K.; Subhani, M. A.; Müller, T. E.; Leitner, W.; Gürtler, C. Green Chem. 2014, 16, 1865. (g) Ola, O.; Mercedes Maroto-Valer, M.; Mackintosh, S. Energy Procedia 2013, 37, 6704−6709. (h) Arakawa, H.; Aresta, M.; Armor, J. N.; Barteau, M. A.; Beckman, E. J.; Bell, A. T.; Bercaw, J. E.; Creutz, C.; Dinjus, E.; Dixon, D. A.; Domen, K.; DuBois, D. L.; Eckert, J.; F

DOI: 10.1021/acs.inorgchem.6b00471 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Nat. Chem. 2012, 4, 383−388. (c) Schmeier, T. J.; Dobereiner, G. E.; Crabtree, R. H.; Hazari, N. J. Am. Chem. Soc. 2011, 133, 9274−9277. (7) Ziebart, C.; Federsel, C.; Anbarasan, P.; Jackstell, R.; Baumann, W.; Spannenberg, A.; Beller, M. J. Am. Chem. Soc. 2012, 134, 20701− 20704. (8) Jeletic, M. S.; Mock, M. T.; Appel, A. M.; Linehan, J. C. J. Am. Chem. Soc. 2013, 135, 11533−11536. (9) (a) Ahlquist, M. R. S. G. J. Mol. Catal. A: Chem. 2010, 324, 3−8. (b) Yang, X. ACS Catal. 2011, 1, 849−854. (c) Matsubara, T. Organometallics 2001, 20, 19−24. (d) Tanaka, R.; Yamashita, M.; Chung, L. W.; Morokuma, K.; Nozaki, K. Organometallics 2011, 30, 6742−6750. (e) Kumar, N.; Camaioni, D. M.; Dupuis, M.; Raugei, S.; Appel, A. M. Dalton Trans. 2014, 43, 11803. (f) Musashi, Y.; Sakaki, S. J. Am. Chem. Soc. 2002, 124, 7588−7603. (g) Ohnishi, Y.-Y.; Nakao, Y.; Sato, H.; Sakaki, S. Organometallics 2006, 25, 3352−3363. (h) Cao, L.; Sun, C.; Sun, N.; Meng, L.; Chen, D. Dalton Trans. 2013, 42, 5755. (i) Huang, F.; Zhang, C.; Jiang, J.; Wang, Z.-X.; Guan, H. Inorg. Chem. 2011, 50, 3816−3825. (j) Osadchuk, I.; Tamm, T.; Ahlquist, M. S. G. Organometallics 2015, 34, 4932−4940. (10) Mondal, B.; Neese, F.; Ye, S. Inorg. Chem. 2015, 54, 7192−7198. (11) (a) DuBois, D. L.; Berning, D. E. Appl. Organomet. Chem. 2000, 14, 860−862. (b) Qi, X.-J.; Fu, Y.; Liu, L.; Guo, Q.-X. Organometallics 2007, 26, 4197−4203. (12) (a) Muckerman, J. T.; Achord, P.; Creutz, C.; Polyansky, D. E.; Fujita, E. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 15657−15662. (b) Fong, H.; Peters, J. C. Inorg. Chem. 2015, 54, 5124−5135. (c) Ni, S.-F.; Dang, L. Phys. Chem. Chem. Phys. 2016, 18, 4860−4870. (d) Pitman, C. L.; Brereton, K. R.; Miller, A. J. M. J. Am. Chem. Soc. 2016, 138, 2252−2260. (13) (a) Whited, M. T.; Mankad, N. P.; Lee, Y.; Oblad, P. F.; Peters, J. C. Inorg. Chem. 2009, 48, 2507−2517. (b) Creutz, S. E.; Peters, J. C. J. Am. Chem. Soc. 2014, 136, 1105−1115. (c) Moret, M.-E.; Peters, J. C. Angew. Chem., Int. Ed. 2011, 50, 2063−2067. (d) Moret, M.-E.; Peters, J. C. J. Am. Chem. Soc. 2011, 133, 18118−18121. (e) Lee, Y.; Mankad, N. P.; Peters, J. C. Nat. Chem. 2010, 2, 558−565. (f) Ye, S.; Bill, E.; Neese, F. Inorg. Chem. 2016, 55, 3468−3474. (14) (a) Riplinger, C.; Neese, F. J. Chem. Phys. 2013, 138, 034106. (b) Riplinger, C.; Sandhoefer, B.; Hansen, A.; Neese, F. J. Chem. Phys. 2013, 139, 134101. (15) Neese, F. WIREs Comput. Mol. Sci. 2012, 2, 73−78. (16) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101. (17) van Wüllen, C. J. Chem. Phys. 1998, 109, 392. (18) (a) Eichkorn, K.; Weigend, F.; Treutler, O.; Ahlrichs, R. Theor. Chem. Acc. 1997, 97, 119−124. (b) Eichkorn, K.; Treutler, O.; Ö hm, H.; Häser, M.; Ahlrichs, R. Chem. Phys. Lett. 1995, 240, 283−290. (19) Schäfer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829. (20) Schäfer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571. (21) Klamt, A.; Schüümann, G. J. Chem. Soc., Perkin Trans. 2 1993, 799−805. (22) (a) Becke, A. D. Phys. Rev. A: At., Mol., Opt. Phys. 1988, 38, 3098−3100. (b) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (c) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785−789. (23) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297. (24) Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput. Chem. 2011, 32, 1456−1465. (25) Neese, F.; Wennmohs, F.; Hansen, A.; Becker, U. Chem. Phys. 2009, 356, 98−109. (26) (a) Klopper, W.; Kutzelnigg, W. J. Mol. Struct.: THEOCHEM 1986, 135, 339−356. (b) Halkier, A.; Helgaker, T.; Jørgensen, P.; Klopper, W.; Koch, H.; Olsen, J.; Wilson, A. K. Chem. Phys. Lett. 1998, 286, 243−252. (c) Helgaker, T.; Klopper, W.; Koch, H.; Noga, J. J. Chem. Phys. 1997, 106, 9639. (27) (a) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007−1023. (b) Woon, D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1993, 98, 1358− 1371. (c) Balabanov, N. B.; Peterson, K. A. J. Chem. Phys. 2005, 123, 064107. (d) Balabanov, N. B.; Peterson, K. A. J. Chem. Phys. 2006, 125, 074110.

(28) Liakos, D. G.; Sparta, M.; Kesharwani, M. K.; Martin, J. M. L.; Neese, F. J. Chem. Theory Comput. 2015, 11, 1525−1539. (29) (a) Xue, G.; Geng, C.-Y.; Ye, S.; Fiedler, A. T.; Neese, F.; Que, L., Jr. Inorg. Chem. 2013, 52, 3976−3984. (b) Ye, S.; Riplinger, C.; Hansen, A.; Krebs, C.; Bollinger, M., Jr.; Neese, F. Chem. - Eur. J. 2012, 18, 6555−6567. (c) Song, J.; Klein, E. L.; Neese, F.; Ye, S. Inorg. Chem. 2014, 53, 7500−7507. (d) Fang, H.; Jing, H.; Ge, H.; Brothers, P. J.; Fu, X.; Ye, S. J. Am. Chem. Soc. 2015, 137, 7122−7127. (30) Evans, M. G.; Polanyi, M. Trans. Faraday Soc. 1938, 34, 11−23. (31) Ellis, W. W.; Miedaner, A.; Curtis, C. J.; Gibson, D. H.; DuBois, D. L. J. Am. Chem. Soc. 2002, 124, 1926−1932. (32) Berning, D. E.; Miedaner, A.; Curtis, C. J.; Noll, B. C.; Rakowski DuBois, M. C.; DuBois, D. L. Organometallics 2001, 20, 1832−1839.

G

DOI: 10.1021/acs.inorgchem.6b00471 Inorg. Chem. XXXX, XXX, XXX−XXX