Bioinspired Design and Computational Prediction of Iron Complexes

Mar 3, 2016 - Olumide Bolarinwa Ayodele , Sara Faiz Hanna Tasfy , Noor Asmawati Mohd Zabidi , Yoshimitsu Uemura. Journal of CO2 Utilization 2017 17, ...
0 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/JPCL

Bioinspired Design and Computational Prediction of Iron Complexes with Pendant Amines for the Production of Methanol from CO2 and H2 Xiangyang Chen†,‡ and Xinzheng Yang*,† †

Beijing National Laboratory for Molecular Sciences, State Key Laboratory for Structural Chemistry of Unstable and Stable Species, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, People’s Repubic of China ‡ University of Chinese Academy of Sciences, Beijing 100049, People’s Repubic of China S Supporting Information *

ABSTRACT: Inspired by the active site structure of [FeFe]-hydrogenase, we built a series of iron dicarbonyl diphosphine complexes with pendant amines and predicted their potentials to catalyze the hydrogenation of CO2 to methanol using density functional theory. Among the proposed iron complexes, [(PtBu2NtBu2H)FeH(CO)2(COOH)]+ (5COOH) is the most active one with a total free energy barrier of 23.7 kcal/mol. Such a low barrier indicates that 5COOH is a very promising low-cost catalyst for high-efficiency conversion of CO2 and H2 to methanol under mild conditions. For comparison, we also examined Bullock’s Cp iron diphosphine complex with pendant amines, [(PtBu2NtBu2H)FeHCpC5F4N]+ (5Cp‑C5F4N), as a catalyst for hydrogenation of CO2 to methanol and obtained a total free energy barrier of 27.6 kcal/mol, which indicates that 5Cp‑C5F4N could also catalyze the conversion of CO2 and H2 to methanol but has a much lower efficiency than our newly designed iron complexes.

A

diphosphine complexes with pendant amines have been developed as electrocatalysts for hydrogen activation.18−30 Among those studies, Bullock and co-workers18−27 reported a series of Mn, Fe, Co, and Ni catalysts and elucidated their mechanisms for oxidation of hydrogen. They found that incorporating pendant amines into the diphosphine ligand can increase the rate of H2 oxidation dramatically through the formation of a metal, M−Hδ−···Hδ+−N, dihydrogen bond.18 For example, an iron catalyst, CpC6F5Fe(PtBu2NBn2)H, achieved turnover frequencies of 0.66−2.0 s−1 with rather low overpotentials of 160−220 mV.19 This is the most active iron catalyst so far for hydrogen oxidation. Later on, they confirmed the existence of the Fe−Hδ−···Hδ+−N dihydrogen bond in [CpC5F4NFeH(PtBu2NtBu2H)]+ using single-crystal neutron diffraction.20 The observed H···H distance of 1.489(10) Å between the protic N−Hδ+ and hydridic Fe−Hδ− part is remarkably short. Appel and co-workers28,29 reported the oxidation of alcohol catalyzed by nickel diphosphine complexes with pendant amines. Murray et al.30 computationally predicted several biomimetic complexes based on the active site structure of [Fe]-hydrogenase and found that a complex with a bidentate diphosphine group and an internal nitrogen base is promising for the H2 activation and hydrogenation reactions. Inspired by the above findings, we computationally examined Bullock’s Cp (cyclopentane) iron diphosphine complex with

s the most basic alcohol, methanol is one of the most versatile and important compounds in the chemical and fuel industries.1−3 It is not only widely used for the production of formaldehyde, acetic acid, olefins, dimethyl ether, and so forth4 but also acts as a high-density hydrogen carrier (12.6 wt %) for different types of fuel cells.5 The direct conversion of carbon dioxide and H2 into methanol is highly attractive because it provides an ideal way for CO2 recycling and hydrogen storage.6 Although significant progress has been made in the hydrogenation of CO2 to methanol, most of the reported catalysts contain precious metals, such as Ir, Rh, Ru, and so forth and have rather low catalytic activities even under rigorous reaction conditions.7−9 The design of high-efficiency and low-cost catalysts with only earth-abundant metals for the production of methanol from CO2 and H2 is highly desirable. In a catalytic hydrogenation reaction, H2 cleavage and hydride transfer are usually rate-determining steps. Recently, some base metal catalysts have been developed for H2 activation through the mimic of the active sites of hydrogenases.10 In our previous study of H2 activation catalyzed by [Fe]-hydrogenase, we found that the pyridone ligand in its active center assists H2 cleavage through the formation of a strong Fe−Hδ−···Hδ+−O dihydrogen bond.11 Inspired by the active site structure of [Fe]-hydrogenase,12−14 we have computationally designed a series of iron complexes with experimentally ready-made acylmethylpyridinol and aliphatic PNP pincer ligands for catalytic hydrogenation of CO2 to formic acid.15 Guided by the pendant SCH2NHCH2S ligand in the active site of [FeFe]-hydrogenase,16,17 a series of base metal © XXXX American Chemical Society

Received: January 25, 2016 Accepted: March 2, 2016

1035

DOI: 10.1021/acs.jpclett.6b00161 J. Phys. Chem. Lett. 2016, 7, 1035−1041

Letter

The Journal of Physical Chemistry Letters

three cascade catalytic reactions, the hydrogenation of CO2 to formic acid, the hydrogenation of formic acid to formaldehyde with the formation of water, and the hydrogenation of formaldehyde to methanol, for the conversion of CO2 and H2 to methanol is predicted and analyzed through density functional theory (DFT) calculations. Further computational details are provided in the Supporting Information. Scheme 1 shows the predicted catalytic cycles for the hydrogenation of CO2 to formic acid. Figure 2 shows the corresponding free energy profile. The optimized structures of key intermediates and transition states are displayed in Figure 3.

pendant amines for catalytic hydrogenation of CO2 to methanol and obtained a total free energy barrier of 27.6 kcal/mol. In order to design iron catalysts with higher activities, we replaced the Cp ligand in Bullock’s iron catalyst using two carbonyls and a functional group and proposed a series of sixcoordinated iron diphosphine complexes (PtBu2NtBu2)FeH(CO)2R (R = H, Cl, NO2, CN, C5H4N, CH3, NH2, OCH3, OH, CHO, COOH, COCH3, COOCH3, CONH2, and CONHCOOH) (Figure 1). All of the functional groups are

Figure 2. Free energy profile for the hydrogenation of CO2 to formic acid (reaction 1).

Figure 1. Cp iron diphosphine complexes with pendant amines (A,B) reported by Liu, Dubois, Bullock, et al. and our newly designed iron complexes (C,D).

1 is a six-coordinated trans-dihydride iron diphosphine complex with two cis carbonyls. The Fe−H bond lengths in 1 are near 1.55 Å. The distances between Fe and two carbonyls are 1.73 and 1.74 Å. The distances between Fe and two P atoms are 2.25 and 2.27 Å, which are close to the observed Fe−P bond lengths of 2.20 Å in the crystal structure of CpC5F4NFeH(PtBu2NtBu2).20 1′ is a 5.7 kcal/mol more stable isomer of 1 with cis-dihydride. At the beginning of the reaction, a CO2 molecule approaches 1 and forms a formate anion by taking a hydride directly from Fe through transition state TS1,2 (Figure 3) with a free energy barrier of 12.2 kcal/mol. The dissociation of formate anion from 2 and the formation of monocation intermediate 3 is 5.2

common ligands and satisfy the 18-electron rule in the complexes.31 Furthermore, such replacements are based on the active site structure of [FeFe]-hydrogenase and could greatly increase the flexibility of adjusting the electron density at the metal center. Among all newly designed iron complexes, the trans-dihydride complex (PtBu2NtBu2)FeH(CO)2H (1) is the simplest one. Therefore, we studied its catalytic mechanism for the formation of methanol from CO2 and H2. The reactions catalyzed by other iron complexes have similar mechanisms but slightly different relative energies. A detailed mechanism with

Scheme 1. Predicted Catalytic Cycles for the Hydrogenation of CO2 to Formic Acid (Reaction 1)

1036

DOI: 10.1021/acs.jpclett.6b00161 J. Phys. Chem. Lett. 2016, 7, 1035−1041

Letter

The Journal of Physical Chemistry Letters

Figure 3. Optimized structures of 1, 3, 5, 7′, TS1,2 (624i cm−1), and TS5,7 (633i cm−1). tert-Butyl groups are omitted for clarity. Bond lengths are in Å.

Scheme 2. Predicted Catalytic Cycle for the Hydrogenation of Formic Acid to Formaldehyde and Water (Reaction 2)

Once 5 is formed, it could also act as a catalyst for the hydrogenation of CO2 to formic acid. When a CO2 molecule approaches 5, it takes a hydride directly from Fe through transition state TS5,7 with an energy barrier of 15.4 kcal/mol (5 → TS5,7). Then, a dication complex 8 is formed with the dissociation of the formate anion from 7. Similar to R1-cycle1, a H2 molecule fills the vacant position in 8 and assists the proton transfer from N1 to HCOO−. With the formation and release of formic acid, 4 is regenerated. The catalytic cycle for the hydrogenation of CO2 to formic acid catalyzed by 5 is named R1-cycle2. Instead of H2, 8 could also attract the dissociated HCOO− and form a much more stable structure 7′ with a strong Fe−O bond of 2.06 Å. 7′ has a relative free energy of −14.2 kcal/mol, which is 4.9 kcal/mol more stable than 5. Therefore, 7′ can be considered as the resting state of the catalytic reaction. According to the energy span model, 7′ and TS5,7 are the rate-determining states in R1-cycle2 with a free energy barrier of 20.3 kcal/mol (7′ → TS5,7). Because of the easy formation of

kcal/mol downhill. The Fe···N1 distance in 3 is 2.23 Å, which indicates a weak interaction between Fe and N1. Then, a H2 molecule inserts into the Fe···N1 bond in 3 and forms a dihydrogen complex 4, which is 0.7 kcal/mol more stable than 3. In addition to H2 insertion, 3 could also attract the dissociated formate anion to form a slightly more stable structure 2′, which is an isomer of 2 with a strong Fe−O bond (2.07 Å). The H2 molecule in 4 could be split easily by Fe and N1 in a fashion of frustrated Lewis pairs (FLPs) for the formation of a much more stable intermediate 5, which is 9.3 kcal/mol more stable than 1. The distance between Hδ− and Hδ+ in the Fe−Hδ−···Hδ+−N dihydrogen bond in 5 is 1.64 Å, which is slightly shorter than the H···H distance range of 1.7− 2.2 Å in most metal dihydrogen bonds reported so far.32 Then, the proton on N1 transfers to an oxygen in the dissociated HCOO− for the formation of formic acid through TS5,6 with an energy barrier of 12.7 kcal/mol. The release of formic acid from 6 and the regeneration of 1 is 2.8 kcal/mol downhill. The catalytic cycle described above is named R1-cycle1. 1037

DOI: 10.1021/acs.jpclett.6b00161 J. Phys. Chem. Lett. 2016, 7, 1035−1041

Letter

The Journal of Physical Chemistry Letters 7′, the total energy barrier of R1-cycle1 is actually 26.4 (7′ → TS1,2). Therefore, R1-cycle1 is much less favorable than R1cycle2 with 5 as the catalyst for the hydrogenation of CO2 to formic acid. Because 5 is much more stable than 1, we examined its catalytic activity for hydrogenation of formic acid (reaction 2). Scheme 2 shows the predicted catalytic cycle for the formation of formaldehyde and water from formic acid and H2 catalyzed by 5. The corresponding free energy profile is shown in Figure 4. The optimized structures of two stable intermediates (11″

released through the cleavage of the C−O bond (TS11′,12) in HOCH2O−, and a hydroxyl group is left on Fe. Then, the proton on N1 transfers quickly to the hydroxyl on Fe and forms a water molecule. The release of H2O from 13 for the regeneration of 3 is 1.7 kcal/mol uphill. Then, a H2 molecule inserts into the Fe−N1 bond in 3 for the regeneration of 4 and 5. However, we may not see the formation of 3 in the reaction because of the quick exchange of H2O and H2. In addition to the formation of 11′, the HOCH2OH molecule could transfer one of the hydroxyl protons back to N1 through TS3,11″ for the formation of 11″ with a strong Fe−O bond (2.06 Å). 11, 11′, and 11″ are isomers with different orientation of HOCH2O− anion groups in them. Among those isomers, 11″ is 15.0 and 12.0 kcal/mol more stable than 11 and 11′, respectively. Once the formaldehyde molecule is formed, it can easily be converted to methanol through a hydrogenation process. Scheme 3 shows the predicted catalytic cycle for the hydrogenation of formaldehyde to methanol catalyzed by 5 (reaction 3). The corresponding free energy profile is shown in Figure 6. The optimized structures of a stable intermediate and the transition state for hydride transfer from Fe to formaldehyde are displayed in Figure 7. When a formaldehyde molecule approaches 5, the hydride on Fe is transferred directly to the carbon atom in formaldehyde for the formation of a methoxy anion through transition state TS5,14, which is the rate-determining step in this catalytic cycle with a rather low barrier of 12.2 kcal/mol (5 → TS5,14). Then, the CH3O− anion can easily catch the proton on N1 through TS14,3 and form CH3OH. The release of CH3OH reproduces intermediate 3, which can attract a H2 molecule and regenerate the catalyst 5 through the H2 cleavage process described in reaction 1. By comparing all relative energies in the above three catalytic reactions, we can conclude that the direct hydride transfer from Fe to formic acid for the formation of an anionic group HOCH2O− is the rate-determining step with a total free energy barrier of 27.4 kcal/mol (7′ → TS5,11) for the hydrogenation of CO2 to methanol catalyzed by 5. The design of new catalysts with higher efficiency rely on a deep understanding of the influence of ligands. In order to find

Figure 4. Free energy profile for the hydrogenation of formic acid to formaldehyde and water (reaction 2).

and 13) and key transition states for hydride transfer (TS5,11), proton transfer (TS3,11″ and TS12,13), and C−O bond cleavage (TS11′,12) are displayed in Figure 5. At the beginning of reaction 2, a formic acid molecule approaches 5, takes the hydride in the Fe−Hδ−···Hδ+−N dihydrogen bond, and forms an anionic group HOCH2O− through transition state TS5,11 (Figure 5) with a free energy barrier of 22.5 kcal/mol (5 → TS5,11). After the formation of an unstable hydrido alkoxo intermediate 11, the transfer of the proton from N1 to HOCH2O− (TS11,3) for the formation and release of HOCH2OH is fast. Then, the HOCH2OH molecule recombines to 3 and forms a slightly more stable intermediate 11′ through TS3,11′, which transfers a proton from oxygen to N1. Once 11′ is formed, a CH2O molecule can easily be

Figure 5. Optimized structures of 11″, 13, TS5,11 (433i cm−1), TS3,11″ (1261i cm−1), TS11′,12 (159i cm−1), and TS12,13 (910i cm−1). tert-Butyl groups are omitted for clarity. Bond lengths are in Å. 1038

DOI: 10.1021/acs.jpclett.6b00161 J. Phys. Chem. Lett. 2016, 7, 1035−1041

Letter

The Journal of Physical Chemistry Letters

as low-cost catalysts for the hydrogenation of CO2 to methanol, the relative energies of 5R, 7′R, and TS5,11‑R are calculated. As listed in Table 1, the largest difference between the calculated free energy barriers of 7′R → TS5,11‑R is 5.9 kcal/mol,

Scheme 3. Predicted Catalytic Cycle for the Hydrogenation of Formaldehyde to Methanol (Reaction 3)

Table 1. Relative Free Energies of 5R → 7′R and 7′R → TS5,11‑R ΔG (kcal/mol) R

expt

complexes

5R → 7′R

7′R → TS5,11‑R

H Cl NO2 CN OH OCH3 NH2 C5H4N CH3 CHO COCH3 CONH2 COOCH3 CONHCOOH COOH 5Cp‑C5F4N 5Cp

−4.9 2.0 1.2 1.9 −2.0 −1.7 −5.5 −2.8 −6.5 −6.0 −3.9 −3.1 −3.2 −1.4 −2.2 −8.4 −7.7

27.4 29.1 29.6 28.0 29.6 29.5 28.1 26.4 26.3 25.6 25.4 24.9 24.8 24.0 23.7 27.6 26.6

which indicates a strong ligand effect. The iron complexes with R = Cl, NO2, and CN groups have much higher barriers of 31.1, 30.8, and 29.9 kcal/mol (5R → TS5,11‑R), respectively. In addition, the ligands OH, NH2, and OCH3 may attack CO in the catalysts and result in catalyst deactivation. Considering their high barriers, we do not recommend them as potential catalysts for hydrogenation of CO2. Different from the step-bystep transfer of hydride and proton for all other ligands, TS5,11‑Cl, TS5,11‑NO2, TS5,11‑CN, TS5,11‑OH, TS5,11‑C5H4N, and TS5,11‑OCH3 simultaneously transfer the hydride on Fe and the proton on N1 to the formic acid molecule for the formation of HOCH2OH. Figure 8 shows the optimized structure of 7′OH

Figure 6. Free energy profile for the hydrogenation of formaldehyde to methanol (reaction 3).

Figure 8. Optimized structures of 7′OH and TS5,11‑OH (1251i cm−1). tert-Butyl groups are omitted for clarity. Bond lengths are in Å.

Figure 7. Optimized structures of TS5,14 (329i cm−1) and 14′. tertButyl groups are omitted for clarity. Bond lengths are in Å.

and TS5,11‑OH. When R = COOH, the energy difference between 7′COOH and TS5,11‑COOH is only 23.7 kcal/mol. Such a low barrier indicates that 5COOH is a highly active catalyst for hydrogenation of carbon dioxide to methanol under mild conditions. It is worth noting that the coordination properties of COOH and other functional groups with carbonyls (CHO, COOH, COCH3, COOCH3, CONH2, and CONHCOOH) are similar to that of the bridging CO in the active center of native [FeFe]-hydrogenase.16,17

more iron complexes with higher catalytic activities, we examined the relative energies of the rate-determining states, 7′ and TS5,11, in the reaction with various ligands. As shown in Figure 1, 14 analogues of 5 are proposed by replacing the hydrogen at the R position. At the same time, the experimental structure, 5Cp‑C5F4N and its simplified structural model 5Cp, in which the C5F4N group is replaced by H, are also examined for comparison. To evaluate the potential of these iron complexes 1039

DOI: 10.1021/acs.jpclett.6b00161 J. Phys. Chem. Lett. 2016, 7, 1035−1041

Letter

The Journal of Physical Chemistry Letters

(5) Ortelli, E. E.; Wambach, J.; Wokaun, A. Methanol Synthesis Reactions over a CuZr Based Catalyst Investigated Using Periodic Variations of Reactant Concentrations. Appl. Catal., A 2001, 216, 227−241. (6) Goeppert, A.; Czaun, M.; Jones, J.-P.; Surya Prakash, G. K.; Olah, G. A. Recycling of Carbon Dioxide to Methanol and Derived Products - Closing the Loop. Chem. Soc. Rev. 2014, 43, 7995−8048. (7) Huff, C. A.; Sanford, M. S. Cascade Catalysis for the Homogeneous Hydrogenation of CO2 to Methanol. J. Am. Chem. Soc. 2011, 133, 18122−18125. (8) Wesselbaum, S.; vom Stein, T.; Klankermayer, J.; Leitner, W. Hydrogenation of Carbon Dioxide to Methanol by Using a Homogeneous Ruthenium−Phosphine Catalyst. Angew. Chem., Int. Ed. 2012, 51, 7499−7502. (9) Wesselbaum, S.; Moha, V.; Meuresch, M.; Brosinski, S.; Thenert, K. M.; Kothe, J.; Stein, T. v.; Englert, U.; Holscher, M.; Klankermayer, J.; et al. Hydrogenation of Carbon Dioxide to Methanol Using a Homogeneous Ruthenium-Triphos Catalyst: From Mechanistic Investigations to Multiphase Catalysis. Chem. Sci. 2015, 6, 693−704. (10) Xu, T.; Chen, D.; Hu, X. Hydrogen-Activating Models of Hydrogenases. Coord. Chem. Rev. 2015, 303, 32−41. (11) Yang, X.; Hall, M. B. Monoiron Hydrogenase Catalysis: Hydrogen Activation with the Formation of a Dihydrogen, Fe−Hδ−··· Hδ+−O, Bond and Methenyl-H4MPT+ Triggered Hydride Transfer. J. Am. Chem. Soc. 2009, 131, 10901−10908. (12) Hu, B.; Chen, D.; Hu, X. Synthesis and Reactivity of Mononuclear Iron Models of [Fe]-Hydrogenase That Contain an Acylmethylpyridinol Ligand. Chem. - Eur. J. 2014, 20, 1677−1682. (13) Song, L.-C.; Hu, F.-Q.; Zhao, G.-Y.; Zhang, J.-W.; Zhang, W.-W. Several New [Fe]Hydrogenase Model Complexes with a Single Fe Center Ligated to an Acylmethyl(Hydroxymethyl)Pyridine or Acylmethyl(Hydroxy)Pyridine Ligand. Organometallics 2014, 33, 6614−6622. (14) Shima, S.; Chen, D.; Xu, T.; Wodrich, M. D.; Fujishiro, T.; Schultz, K. M.; Kahnt, J.; Ataka, K.; Hu, X. Reconstitution of [Fe]Hydrogenase Using Model Complexes. Nat. Chem. 2015, 7, 995− 1002. (15) Yang, X. Bio-Inspired Computational Design of Iron Catalysts for the Hydrogenation of Carbon Dioxide. Chem. Commun. 2015, 51, 13098−13101. (16) Rauchfuss, T. B. Diiron Azadithiolates as Models for the [FeFe]Hydrogenase Active Site and Paradigm for the Role of the Second Coordination Sphere. Acc. Chem. Res. 2015, 48, 2107−2116. (17) Evans, D. J.; Pickett, C. J. Chemistry and the Hydrogenases. Chem. Soc. Rev. 2003, 32, 268−275. (18) DuBois, D. L.; Bullock, R. M. Molecular Electrocatalysts for the Oxidation of Hydrogen and the Production of Hydrogen − the Role of Pendant Amines as Proton Relays. Eur. J. Inorg. Chem. 2011, 2011, 1017−1027. (19) Liu, T.; DuBois, D. L.; Bullock, R. M. An Iron Complex with Pendent Amines as a Molecular Electrocatalyst for Oxidation of Hydrogen. Nat. Chem. 2013, 5, 228−233. (20) Liu, T.; Wang, X.; Hoffmann, C.; DuBois, D. L.; Bullock, R. M. Heterolytic Cleavage of Hydrogen by an Iron Hydrogenase Model: An Fe-H···H-N Dihydrogen Bond Characterized by Neutron Diffraction. Angew. Chem., Int. Ed. 2014, 53, 5300−5304. (21) Bullock, R. M.; Helm, M. L. Molecular Electrocatalysts for Oxidation of Hydrogen Using Earth-Abundant Metals: Shoving Protons around with Proton Relays. Acc. Chem. Res. 2015, 48, 2017−2026. (22) Hulley, E. B.; Kumar, N.; Raugei, S.; Bullock, R. M. ManganeseBased Molecular Electrocatalysts for Oxidation of Hydrogen. ACS Catal. 2015, 5, 6838−6847. (23) Wiedner, E. S.; Appel, A. M.; DuBois, D. L.; Bullock, R. M. Thermochemical and Mechanistic Studies of Electrocatalytic Hydrogen Production by Cobalt Complexes Containing Pendant Amines. Inorg. Chem. 2013, 52, 14391−14403. (24) Wiedner, E. S.; Roberts, J. A. S.; Dougherty, W. G.; Kassel, W. S.; DuBois, D. L.; Bullock, R. M. Synthesis and Electrochemical

In summary, we computationally predicted the potential of Bullock’s Cp iron diphosphine complex with pendant amines, 5Cp‑C5F4N, as a catalyst for the hydrogenation of CO2 to methanol and found a total free energy barrier of 27.6 kcal/mol, which is accessible but slightly too high for a reaction under mild conditions. Inspired by the active site structure of [FeFe]hydrogenase, we replaced the Cp ligand in 5Cp‑C5F4N using two cis carbonyls and a functional group (R). Such replacement gives the flexibility to adjust the electronic structures at the metal center by using various functional groups. A series of iron dicarbonyl diphosphine complexes, [(P tBu 2 N tBu 2 H)FeH(CO)2R]+ (R = H, Cl, NO2, CN, C5H4N, CH3, NH2, OCH3, OH, CHO, COOH, COCH3, COOCH3, CONH2, and CONHCOOH), are proposed and examined computationally. DFT calculations indicate that [(P tBu 2 N tBu 2 H)FeH(CO)2(COOH)]+ (5COOH) is the most active one with a total free energy barrier of 23.7 kcal/mol. Such a low barrier indicates that 5COOH is a very promising low-cost catalyst for high-efficiency production of methanol from CO2 and H2 under mild conditions. In addition, we found that the iron complexes with electron donor groups at R usually have lower barriers. Our findings point the way to developing new catalysts with higher efficiency for hydrogenation reactions as the metal dihydrogen bonds and electron donor groups in metal complexes may be essential for low-energy H transfer. Further computational design of base metal dihphosphine complexes with pendant amines for the hydrogenation of carbonyl compounds and dehydrogenation of alcohols are underway.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.6b00161. Computational details, the mechanism of the hydrogenation of CO2 to methanol catalyzed by 5Cp, and solvent-corrected absolute free energies (PDF) Atomic coordinates of all optimized structures (XYZ)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work is supported by the 100-Talent Program of Chinese Academy of Sciences (CAS) and the National Natural Science Foundation of China (21373228).

■

REFERENCES

(1) Jadhav, S. G.; Vaidya, P. D.; Bhanage, B. M.; Joshi, J. B. Catalytic Carbon Dioxide Hydrogenation to Methanol: A Review of Recent Studies. Chem. Eng. Res. Des. 2014, 92, 2557−2567. (2) Wang, W. H.; Himeda, Y.; Muckerman, J. T.; Manbeck, G. F.; Fujita, E. CO2 Hydrogenation to Formate and Methanol as an Alternative to Photo- and Electrochemical CO2 Reduction. Chem. Rev. 2015, 115, 12936−12973. (3) Nielsen, M.; Alberico, E.; Baumann, W.; Drexler, H.-J.; Junge, H.; Gladiali, S.; Beller, M. Low-Temperature Aqueous-Phase Methanol Dehydrogenation to Hydrogen and Carbon Dioxide. Nature 2013, 495, 85−89. (4) Sun, J.; Yang, G.; Yoneyama, Y.; Tsubaki, N. Catalysis Chemistry of Dimethyl Ether Synthesis. ACS Catal. 2014, 4, 3346−3356. 1040

DOI: 10.1021/acs.jpclett.6b00161 J. Phys. Chem. Lett. 2016, 7, 1035−1041

Letter

The Journal of Physical Chemistry Letters Studies of Cobalt(III) Monohydride Complexes Containing Pendant Amines. Inorg. Chem. 2013, 52, 9975−9988. (25) Fang, M.; Wiedner, E. S.; Dougherty, W. G.; Kassel, W. S.; Liu, T.; DuBois, D. L.; Bullock, R. M. Cobalt Complexes Containing Pendant Amines in the Second Coordination Sphere as Electrocatalysts for H2 Production. Organometallics 2014, 33, 5820−5833. (26) Wiedner, E. S.; Brown, H. J.; Helm, M. L. Kinetic Analysis of Competitive Electrocatalytic Pathways: New Insights into Hydrogen Production with Nickel Electrocatalysts. J. Am. Chem. Soc. 2016, 138, 604−616. (27) Liu, T.; Liao, Q.; O’Hagan, M.; Hulley, E. B.; DuBois, D. L.; Bullock, R. M. Iron Complexes Bearing Diphosphine Ligands with Positioned Pendant Amines as Electrocatalysts for the Oxidation of H2. Organometallics 2015, 34, 2747−2764. (28) Weiss, C. J.; Das, P.; Miller, D. L.; Helm, M. L.; Appel, A. M. Catalytic Oxidation of Alcohol Via Nickel Phosphine Complexes with Pendant Amines. ACS Catal. 2014, 4, 2951−2958. (29) Weiss, C. J.; Wiedner, E. S.; Roberts, J. A.; Appel, A. M. Nickel Phosphine Catalysts with Pendant Amines for Electrocatalytic Oxidation of Alcohols. Chem. Commun. 2015, 51, 6172−6174. (30) Murray, K. A.; Wodrich, M. D.; Hu, X.; Corminboeuf, C. Toward Functional Type III [Fe]-Hydrogenase Biomimics for H2 Activation: Insights from Computation. Chem. - Eur. J. 2015, 21, 3987−3996. (31) Rasmussen, S. C. The 18-Electron Rule and Electron Counting in Transition Metal Compounds: Theory and Application. ChemTexts 2015, DOI: 10.1007/s40828-015-0010-4. (32) Custelcean, R.; Jackson, J. E. Dihydrogen Bonding: Structures, Energetics, and Dynamics. Chem. Rev. 2001, 101, 1963−1980.

1041

DOI: 10.1021/acs.jpclett.6b00161 J. Phys. Chem. Lett. 2016, 7, 1035−1041