Understanding the Relationship Between Kinetics and

---

Subscriber access provided by GRIFFITH UNIVERSITY

Article

Understanding the Relationship Between Kinetics and Thermodynamics in CO2 Hydrogenation Catalysis Matthew S Jeletic, Elliott Hulley, Monte L Helm, Michael T. Mock, Aaron M. Appel, Eric S. Wiedner, and John Charles Linehan ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.7b01673 • Publication Date (Web): 25 Jul 2017 Downloaded from http://pubs.acs.org on July 26, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Understanding the Relationship Between Kinetics and Thermodynamics in CO2 Hydrogenation Catalysis Authors: Matthew S. Jeletic, Elliott B. Hulley, Monte L. Helm, Michael T. Mock, Aaron M. Appel, Eric S. Wiedner,* and John C. Linehan*

Catalysis Science Group, Pacific Northwest National Laboratory, Richland, Washington, 99352, USA

* E-mail: [email protected], [email protected]

Abstract Catalysts that are able to reduce carbon dioxide under mild conditions are highly sought after for storage of renewable energy in the form of a chemical fuel. This study describes a systematic kinetic and thermodynamic study of a series of cobalt and rhodium bis(diphosphine) complexes that are capable of hydrogenating carbon dioxide to formate under ambient temperature and pressure. Catalytic CO2 hydrogenation was studied under 1.8 and 20 atm of pressure (1:1 CO2:H2) at room temperature in tetrahydrofuran with turnover frequencies (TOF) ranging from 20 to 74,000 h-1. The catalysis was followed by 1H and 31P NMR spectroscopy in real time under all conditions to yield information about the rate determining step. The cobalt catalysts displayed a rate determining step of hydride transfer to CO2, while the hydrogen addition and/or deprotonation steps were rate limiting for the rhodium catalysts. Thermodynamic analysis of the complexes provided information on the driving force for each step of catalysis in terms of the catalyst hydricity (G°H-), acidity (pKa), and free energy for H2 addition (G°H2). Linear free-energy relationships were identified that link the kinetic activity for catalytic hydrogenation of CO2 to formate with the thermodynamic driving force for the rate-limiting steps of catalysis. The 1 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 29

catalyst exhibiting the highest activity, Co(dmpe)2H, was found to have hydride transfer and hydrogen addition steps that were each downhill by approximately 6 to 7 kcal mol–1, and the deprotonation step was thermoneutral. This indicates the fastest catalysts are the ones that most efficiently balance the free energy relationships of every step in the catalytic cycle.

Keywords Carbon dioxide; hydrogenation; hydricity; homogeneous catalysis; cobalt; rhodium

Introduction Photocatalytic and electrocatalytic reduction of CO2 to liquid fuels is widely pursued as a means to store energy generated from geographically disperse renewable sources, such as wind and solar.1-6 As an alternative to direct reduction using protons and electrons, CO2 can also be hydrogenated using clean H2 obtained from renewable sources.7-11 Molecular catalysts for hydrogenation of CO2 are selective for production of formic acid / formate, an attractive chemical hydrogen storage material.12-14 The noble metals, specifically Ir, Rh and Ru, are often the focus for catalytic hydrogenation of CO2 to formic acid.1534

Catalysts based on the more abundant metals Ni, Co, Fe, and Cu have been described,35-51 though these

catalysts often do not rival the activities of noble metal catalysts under optimized conditions. A common link in most of these optimized catalyst systems is the requirement for high temperatures and pressures,5253

which is undesirable for decentralized conversion of CO2 using dispersed renewable energy sources.1,

54-55

For example, Nozaki’s Ir-pincer complex is inactive at 25 °C (>40 atm), and yields a low turnover

frequency (TOF) of 1 h-1 at 100 °C and 1 atm.25-26 Fujita and Himeda et al., demonstrated catalysis under ambient conditions using Ir-bipyridyl catalysts, though the TOF’s were low (400 mM) and different cobalt and rhodium catalysts at room temperature in THF-d8. Figure 1 shows a representative plot of turnovers (mol of [VkdH]+[HCO2]– per mol of catalyst) versus time for Co(BPE5)2H at 20 atm of 1:1 CO2:H2. Pseudo-first order catalytic rate constants were measured from the linear portion of the time-dependent kinetic profiles. These catalytic rate constants are equal to the catalytic turnover frequency (TOF), i.e. moles of [VkdH] +[HCO2]– per mole of catalyst per hour, under the specified conditions and are listed in Tables 1-2. At higher concentrations of the catalysts, the rate of catalysis was mass-transport limited, specifically by the rate of the gas-liquid mixing. Therefore the catalyst concentration was decreased until the TOF remained similar at two different catalyst concentrations (Tables S2-S3), indicating that the reaction was no longer mass-transport limited. Several of the catalysts were too slow at 1.8 atm to measure a significant amount of [VkdH]+[HCO2]– within the first hour, so the TOF of these catalysts were not measured under these conditions.

5 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

Figure 1. A plot of turnovers (mol [VkdH]+[HCO2]– / mol catalyst) versus time for 0.036 mM Co(BPE5)2H at 20 atm of 1:1 H2:CO2 and 570 mM Vkd in THF-d8 at 21 °C. The reaction ceases when all of the Vkd has been consumed. The line is the best linear fit and was used to determine the TOF (R2 = 0.99).

Table 1. Catalytic conversion of CO2 and H2 to formate with Verkade’s base at 1.8 atm.a TOF (h-1) b

Turnovers b

Co(dmpe)2H

Initial Loading (mM) 0.28

6400

1900 c

Co(BPE5)2H

2.6

540

170

Co(dmpbz)2H

2.6

270

200

Co(depe)2H

25

21

Catalyst

+

20 d

Rh(dmpbz)2

0.27

3100

Rh(depe)2+

0.30

1600 d

Rh(dmpe)2

+ +

2.7

90

d

1100 280 55

e

Rh(dmpe)2 2.7 640 88 e Catalytic conditions; 500µL THF-d8, 1.8 atm 1:1 CO2:H2, 21 °C, 400-430 mM 2,8,9-triisopropyl-2,5,8,9-tetraaza-1phosphabicyclo[3,3,3] undecane (Verkade’s base), typical run time 2 h or less. b Uncertainties are 10%. c Previously reported. d Initial TOF. e tert-Butylimino-tris(dimethylamino)phosphorane, pKa ~ 38 for P4tBuH+ in MeCN,62 was used instead of Verkade’s base. a

6 ACS Paragon Plus Environment

Page 7 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Table 2. Catalytic conversion of CO2 and H2 to formate with Verkade’s base at 20 atm. a TOF (h-1) b

Turnovers b

Co(dmpe)2H

Initial Loading (mM) 0.040

74000 c

9400 c

Co(BPE5)2H

0.036

13000

9700

Co(dmpbz)2H

0.037

11000

6700

Co(depe)2H

0.36

610

1300 26

Catalyst

Co(dedpe)2H

22

28

Co(dppe)2H

8.3 d

11 e

+

Rh(dmpbz)2 +

Rh(depe)2

Rh(dmpe)2

+

Rh(dppe)2H +

0.038

17 e f

11000

7100

3000

f

1000

0.38

145

f

100

7.1

130

0.43

d

70

e

Rh(dcpe)2 2.6 33 93 e a Catalytic conditions; 350µL THF-d8, 20 atm 1:1 CO2:H2, 21 °C, 470-600 mM 2,8,9-triisopropyl-2,5,8,9-tetraaza-1phosphabicyclo[3,3,3] undecane (Verkade’s base), typical run time 2 h or less. b Uncertainties are 10%. c Previously reported. d solubility limited catalyst loading. e 40 atm. f initial TOF.

The time-resolved 1H and 31P{1H} NMR spectra reveal the presence of only Co(P2)2H species over the course of the reaction at both low and high pressure, suggesting a rate determining step of hydride transfer from Co(P2)2H to CO2 in all cases. The time-dependent kinetic profiles of the cobalt-catalyzed reactions remain linear over the course of the reaction, though an induction period of ≤ 10 minutes was observed for some of the cobalt catalysts. Time-resolved NMR spectroscopy experiments did not reveal any other species other than Co(P2)2H complex during the induction period, and visual inspection of the reaction did not reveal any noticeable color changes or precipitate formation. Gas-liquid mixing does not appear to be the source of the induction period, as the 1H NMR resonance for dissolved H2 remained constant during the induction period. At low concentrations of the Rh(P2)2+ catalysts, the kinetic plots display a significant decrease in catalyst activity over time, with catalytic activity ceasing after about two hours. We were unable to monitor the catalyst when suspected decomposition is occurring due to the low catalyst loadings. At higher loadings where the Rh(P2)2+ catalyst could be monitored, the maximum number of turnovers is reached 7 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

before decomposition occurs. Mass-transport limitations complicated the determination of an accurate TOF at high loadings of Rh(P2)2+, so the initial TOFs measured at low concentrations of the Rh(P2)2+ catalysts are reported in Tables 1-2. Rh(dmpe)2H2+ is the only species observed in time-resolved NMR spectra at 20 atm of 1:1 CO2:H2, indicating that deprotonation is the rate-determining step under these conditions. At 1.8 atm of 1:1 CO2:H2, time-resolved NMR spectra reveal the presence of Rh(dmpe)2+ and Rh(dmpe)2H2+ in about equal concentrations. The lowering of pressure directly influences the kinetic barrier for hydrogen addition to make it competitive with deprotonation such that two concurrent steady-state intermediates exist. In contrast, only Rh(depe)2+ is observed in time-resolved 1H and 31P NMR spectra at both 1.8 atm and 20 atm, supporting a hydrogen addition rate determining step for this complex. While the focus for this study is more on TOF, the catalysts also afford a modest number of turnovers. Importantly, the number of turnovers is not limited by catalyst decomposition (for the cobalt catalysts), but instead by the amount of Vkd present. Diluting the catalyst concentration infinitely is not a viable solution for increasing the number of turnovers, as residual impurities in the reagents disproportionately start to affect catalyst performance at lower concentrations of catalyst. Since the rhodium catalysts exhibit a different resting state than the cobalt catalysts, we studied the mechanism of Rh(dmpbz)2+ in more detail to verify that the rhodium and cobalt catalysts operate by similar mechanisms. [Rh(dmpbz)2](O3SCF3) displayed low solubility in tetrahydrofuran, so one equivalent of NaBArF4 (ArF = 3,5-(CF3)2C6H3) was added to solubilize the complex. 31P{1H} and 1H NMR spectra indicate that Rh(dmpbz)2+ is in equilibrium with the dihydride complex, cis-Rh(dmpbz)2H2+ (56% conversion), under 1.8 atm of H2. Diagnostic broad peaks attributable to cis-Rh(dmpbz)2H2+ resonate at 35.0 and 33.8 ppm in the 31P{1H} NMR spectrum and –9.28 (hydride, d, 1JHRh ~ 115 Hz) ppm in the 1H NMR spectrum. Only Rh(dmpbz)2+ and Rh(dmpbz)2H (broad resonances at –10.80 (hydride) and 21.6

8 ACS Paragon Plus Environment

Page 9 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

ppm (JPRh = 141 Hz) in the 1H and

31

P{1H} NMR spectra, respectively) were observed when this

experiment was repeated in the presence of ten equivalents of Vkd base. Addition of 1.8 atm of CO2 converts Rh(dmpbz)2H back into Rh(dmpbz)2+ as observed in the 1H and 31P{1H} NMR spectra. In a separate experiment, pressurizing Rh(dmpbz)2+ with 6.7 atm of H2 shifts the equilibrium to favor only cisRh(dmpbz)2H2+. Based on these results, we are reasonably certain that the rhodium complexes follow the same mechanism as the cobalt catalysts, but have a different rate determining step. A similar mechanism was concluded recently in a computational study of Rh(P2)2+ complexes.63 As a last note, unlike the Co(P2)2+ catalysts, the Rh(P2)2+ complexes do not react to form unwanted side products (non-catalytic dimer complex or CO complex).44

Thermodynamics Accurate relative thermodynamic values of the different complexes are needed to properly assess the correlations between driving force and catalytic turnover frequency. Previous studies have demonstrated the challenges associated with experimental measurement of hydricity values for strongly hydridic metal hydrides, such as Co(dmpe)2H.64 Computational methods can often provide accurate thermodynamic values for transition metal hydrides, but complexes possessing ligands with large conformational degrees of freedom (such as depe) still pose a significant challenge.60, 65-66 As a result, we estimated the relative hydricities of the Co(P2)2H and Rh(P2)2H complexes through empirical correlation (described below) instead of by direct experimental measurement or computation. Two of the Co(P2)2H complexes used in this study have known hydricity values: Co(dppe)2H (ΔG°H– = 49.9 kcal mol–1)67-68 and Co(dmpe)2H (ΔG°H– ≈ 36.3 kcal mol–1).65,

69

We relied on two

considerations to estimate hydricities for the remaining Co(P2)2H complexes. First, Co(P2)2H and Ni(P2)2H+ complexes are isoelectronic. Second, the hydricities of Ni(P2)2H+ complexes are linearly

9 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

correlated with the Ni(II/I) couples of the parent Ni(P2)22+ complexes.70 The analogous reduction potential for Co(P2)22+ is the Co(I/0) couple (a d8 to d9 reduction), which has been previously reported for Co(dppe)22+ and Co(dedpe)22+ in acetonitrile.67, 69 Cyclic voltammetry experiments were performed in acetonitrile to measure the potential of the Co(I/0) couple for the remaining Co(P2)22+ complexes, except for Co(BPE5)22+ which was insoluble in acetonitrile and other nitrile solvents. A plot of hydricity versus reduction potential was constructed using Co(dppe)2H and Co(dmpe)2H (both having known hydricities), as well as the the Ni(P2)2H+ analogs of the Co(P2)2H complexes in this study (see the SI for more details). From the linear fit to this plot, hydricity values were estimated for the remaining Co(P2)2H complexes from their Co(I/0) reduction potentials (Table 3). It is important to note that the Co(I/0) couples of Co(dmpbz)2+ and Co(dmpe)2+ were chemically irreversible in acetonitrile. As a result, the hydricity values may have low absolute accuracy, however, the relative trend in hydricities is expected to be accurate.

Table 3. Summary of Estimated Thermochemical Values.a pKa MIII(H)2

Co(dppe)2+

G°H– MI(H) (kcal mol–1) 49.9 b

22.8 b

G°H2 MI (kcal mol–1) -5.0 b

Co(dedpe)2+

44.2

27.4

-5.5

Co(depe)2

40.6

30.3

-5.9

Co(dmpbz)2+

38.0

32.3

Complex

+

36.3

Rh(dcpe)2+

39.9

24.0

4.2 d

Rh(dppe)2+

36.2

26.4

3.8 d

28.3 e

34.3 e

0.9 e

Rh(dmpbz)2

27.4

35.5

0.2 f

Rh(dmpe)2+

26.6 e

36.5 e

-0.4 e

Rh(depe)2+ +

c

-6.1

Co(dmpe)2+

33.7

c

-6.3 c

a

Values are estimated as described in the text unless noted otherwise. b Experimental value, see Ref 67-68. c Computational value, see Ref 60, 65. d Experimental value, see Ref 71. e Experimental value, see Ref 64. f Experimental value, this work.

10 ACS Paragon Plus Environment

Page 11 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Hydricity values have been previously measured for Rh(dmpe)2H (ΔG°H– = 26.6 kcal mol–1) and Rh(depe)2H (ΔG°H– = 28.3 kcal mol–1),64, 68 as well as for the dihydride counterparts Rh(dmpe)2H2+ (ΔG°H– = 49.9 kcal mol–1) and Rh(depe)2H2+ (ΔG°H– = 51.9 kcal mol–1).68, 71 These values correspond to a 24 kcal mol–1 difference between the Rh(P2)2H and Rh(P2)2H2+ complexes of the same diphosphine ligand. A similar difference in hydricities (25 kcal mol–1) was measured for Rh(depx)2H and Rh(depx)2H2+ (depx = α,αʹ-bis(diethylphosphino)xylene), complexes that contain a diphosphine with a large bite angle.72 Therefore we estimated the hydricities of Rh(dcpe)2H and Rh(dppe)2H by subtracting 24 kcal mol–1 from the experimental hydricities of Rh(dcpe)2H2+ and Rh(dppe)2H2+.68, 71 The pKa values of the Rh(P2)2H2+ complexes were determined from a thermochemical cycle using the hydricity of Rh(P2)2H, the free energy for H2 addition to Rh(P2)2+,71 and the constant for heterolytic cleavage of H2 in acetonitrile.68 Examination of the thermochemical data indicates that the hydricity of Rh(P2)2H is inversely linearly correlated with the pKa of Rh(P2)2H2+ and with the free energy for H2 addition to Rh(P2)2+ (see SI). Thus as Rh(P2)2H becomes a stronger hydride donor, Rh(P2)2H2+ becomes harder to deprotonate and H2 addition to Rh(P2)2+ becomes more favorable. These findings were used to estimate additional thermochemical data for the complexes in this study. First, the free energy for H2 addition to Rh(dmpbz)2+ was measured to be 0.2 kcal mol–1 from the reactivity studies described in the previous section. This value was used to estimate the hydricity of Rh(dmpbz)2H and the pKa of Rh(dmpbz)2H2+ using the linear relationships of the other rhodium complexes. Second, pKa values of Co(dmpbz)2H2+, Co(depe)2H2+, and Co(dedpe)2H2+ were estimated by assuming a linear relationship between the pKa and the hydricity of Co(P2)2H, anchored to the pKa values of Co(dppe)2H2+ 67 and Co(dmpe)2H2+.60, 69 The free energy for H2 addition to Co(P2)2+ was then determined from a thermochemical cycle using the hydricity of Co(P2)2H, the pKa of Co(P2)2H2+, and the constant for heterolytic cleavage of H2 in acetonitrile.68 Full details of these estimations are provided in the SI.

11 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

Discussion In this study, we sought to gain a better understanding of how to balance the thermodynamic driving force for each step of catalytic hydrogenation of CO2 in order to maximize the catalytic turnover frequency. This objective relies on being able to correlate the thermodynamic properties of the catalysts with the kinetic data for catalytic CO2 hydrogenation. However, the thermodynamic properties were determined for acetonitrile as a solvent, whereas the kinetic data were collected in tetrahydrofuran. Leito and coworkers have demonstrated that pKa values of organic acids are consistently 5-6 units higher in acetonitrile than in tetrahydrofuran.73-74 Therefore we assume that the thermodynamic parameters of the catalysts have the same relative values between species in both acetonitrile and tetrahydrofuran, thereby allowing a meaningful comparison of driving force and reaction rate. The rate determining step for the Co(P2)2+ catalysts was found to be hydride transfer to CO2 (equation 1). Therefore, the catalytic turnover frequencies are equal to the pseudo-first order rate constant for hydride transfer at a given pressure of CO2. A plot of the catalytic TOF (on a log scale) versus the hydricity of Co(P2)2H is shown in Figure 2 for CO2 partial pressures of 0.9 atm and 10 atm (total pressure of 1.8 atm and 20 atm, respectively). Notably, the hydricity and catalytic rate are linearly correlated at each pressure. This finding is consistent with a recent computational study on Fe and Co complexes containing tripodal ligands, which predicted a linear correlation between hydricity and the activation barrier for hydride transfer to CO2.75 To the best of our knowledge, our present results are the first experimental demonstration of a linear free energy relationship for hydride transfer to CO2 in a series of isostructural transition metal hydride complexes.

Co(P2)2H + CO2 ⇌ Co(P2)2+ + HCO2–

(1)

12 ACS Paragon Plus Environment

Page 13 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Figure 2. Plot of the logarithm of the catalytic turnover frequency versus the hydricity of Co(P2)2H. The data contained in this plot are tabulated in the SI. Analysis of the linear free energy correlation according to the Bell-Evans-Polyani principle76 affords equation 2, where α describes the position of the transition state along the reaction coordinate (0 ≤ α ≤ 1) and C is a constant (see the SI). At 0.9 atm of CO2, the slope of the free energy correlation corresponds to α = 0.78, indicating a late transition state that structurally resembles the product, formate, as opposed to the reactant, CO2. Previous computational studies of Co(dmpe)2H identified two kinetically competent mechanisms for hydride transfer: direct outer-sphere hydride transfer to CO2, and an associative mechanism where CO2 first binds to Co before formation of the C–H bond (Figure 3).77 Both mechanisms require a rehybridization of CO2 from linear to bent in the rate-determining transition state (RDTS), consistent with the experimental determination of a late transition state for hydride transfer.

log(TOF) = –(α/1.364)×ΔG°H– + C

(2)

13 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

Figure 3. Previously calculated mechanisms for hydride transfer from Co(dmpe)2H to CO2.

Few studies examine both the thermodynamics and kinetics for H– transfer to CO2, so it is difficult to compare the activity for hydride transfer from Co(P2)2H to other catalysts. An exception is Ru(terpy)(bpy)H+, for which both thermodynamic and kinetic data is available for hydride transfer in acetonitrile. The hydricity of Ru(terpy)(bpy)H+ is 39.3 kcal mol–1 in acetonitrile,68, 78 corresponding to a favorable driving force of ~5 kcal mol–1 for hydride transfer to 1 atm of CO2. A second-order rate constant of 0.018 M–1 s–1 was reported for hydride transfer from Ru(terpy)(bpy)H+ to CO2 in acetonitrile at 25 °C.79 This second-order rate constant is equivalent to a pseudo first-order rate constant (TOF) of 19 h–1 at 1 atm of CO2, and is much smaller than the TOF of 150 h–1 predicted from the hydricity of Ru(terpy)(bpy)H+ using the free energy correlation of Figure 2. Therefore hydride transfer to CO2 is intrinsically slower for Ru(terpy)(bpy)H+ than for Co(P2)2H, possibly due to a difference in relative stabilities of the metal complexes resulting from hydride transfer. Four-coordinate Co(P2)2+ (16 e–, d8) should be a more stable species than five-coordinate Ru(terpy)(bpy)2+ (16 e–, d6). The BPE5 ligand was considered based on previous evidence by Field that dmpe and BPE5 should have similar chelation strength.80 We were unable to estimate thermodynamic properties for Co(BPE5)2+ 14 ACS Paragon Plus Environment

Page 15 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

due to its insolubility in nitrile solvents. However, the catalytic turnover frequency of Co(BPE5) 2H was found to lie between Co(dmpe)2H and Co(dmpbz)2H, suggesting that the hydricity of Co(BPE5)2H is between that of Co(dmpe)2H (36.3 kcal mol–1) and Co(dmpbz)2H (38.0 kcal mol–1). The Rh(dmpe)2+ catalyst has a low average TOF of 200 h-1 at 20 atm due to a large mismatch between the pKa of Rh(dmpe)2H2+ and Vkd (approximately 3 units).43-44 The mismatch is also apparent when examining the time resolved multinuclear NMR data, as the rate determining step is clearly deprotonation since the only species observed during catalysis is Rh(dmpe)2H2+. Notably, the TOF increases from 90 h-1 to 640 h-1 when a stronger base, tert-Butylimino-tris(dimethylamino)phosphorane, N′-tert-Butyl-N,N,N′,N′,N″,N″-hexamethylphosphorimidic triamide (P4tBu) (pKa ~ 38 in CH3CN)),62 is used at 1.8 atm. In contrast, the other Rh(P2)H2+ species have a lower pKa than Rh(dmpe)2H2+, and the rate determining step changes from deprotonation to hydrogen addition. Therefore, the free energy for H2 addition to Rh(P2)2+ (equation 3) is the appropriate thermodynamic value to compare to the catalytic TOF of the rhodium catalysts (except for Rh(dmpe)2+). A plot of the catalytic TOF (on a log scale) versus the free energy for H2 addition is shown in Figure 4 for catalysis at an H2 partial pressure of 10 atm (total pressure of 20 atm), resulting in a linear correlation between log(TOF) and the driving force.

Rh(P2)2+ + H2 ⇌ Rh(P2)2H2+

(3)

15 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 29

Figure 4. Plot of the logarithm of the catalytic turnover frequency at 20 atm of 1:1 H2:CO2 versus the free energy for H2 addition to Rh(P2)2+. The data contained in this plot is tabulated in the SI.

Using the comprehensive thermodynamic information for the different cobalt and rhodium catalysts, we can analyze the factors that promote catalysis under low pressure, room temperature conditions. A reaction coordinate diagram showing the relative free energy values for each elementary catalytic step of six catalysts that operate under ambient conditions (1.8 atm of total pressure, 25 °C) is shown in Figure 5. The net reaction free energy is defined by the identity of the substrates (H2, CO2, Vkd) and products (HCO2–, VkdH+), while the relative energies of the catalytic intermediates are defined by the thermodynamic properties of each catalyst. An important point for catalyst design is that the sum of the driving forces for each catalytic step must equal the net reaction free energy. As a result, a catalyst that possesses a large driving force for a single step will necessarily have a low driving force for one or more other steps in the catalytic cycle.

16 ACS Paragon Plus Environment

Page 17 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Figure 5. Reaction coordinate diagram showing the relative free energies of six catalysts that operate under ambient conditions (1.8 atm, 25 °C). The lines are not correlations and are only added to guide the eye. Co(dmpe)2+ is the fastest of the catalysts that are active under ambient conditions. As seen in Figure 5, the hydride transfer and H2 addition steps are both downhill in energy, while the deprotonation step is thermoneutral. The other cobalt catalysts, Co(dmpbz)2+ and Co(depe)2+, have a greater driving force for deprotonation than Co(dmpe)2+, but less driving force for hydride transfer to CO2. The extra driving force for deprotonation is “wasted” energy since this is not the rate determining step; as a result, Co(dmpbz)2+ and Co(depe)2+ are slower catalysts than Co(dmpe)2+. In principal, catalysis could be improved over Co(dmpe)2+ by increasing the driving force for hydride transfer, decreasing the driving force for H2 addition, and maintaining the thermoneutrality of deprotonation. As an extreme example, these conditions are fulfilled in Rh(dmpbz)2+ and Rh(depe)2+, which have a very strong driving force for hydride transfer and little to no driving force for H2 addition or deprotonation. However, the thermodynamics for rhodium catalysts have shifted too far, and the catalysts are slower than Co(dmpe)2+ 17 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 29

due to a change in the rate determining step from hydride transfer to H2 addition. Thus, the rhodium catalysts possess too much driving force for the hydride transfer step, and not enough driving force in the H2 addition step. In a computational study of an Ir(PNP) hydride catalyst, Hazari and Bernskoetter similarly concluded that hydride transfer to CO2 was too favorable and H2 addition was not favorable enough.81 These results demonstrate that in a family of structurally related catalysts, the fastest catalyst is the one that most efficiently balances the linear free energy correlations of all the steps in the catalytic cycle (in this case, hydride transfer, H2 addition, and deprotonation). As a result, it is unlikely that a given catalyst will display remarkable activity for CO2 hydrogenation under ambient conditions if any of the reaction steps are energetically unfavorable by more than 1-2 kcal mol–1. This requires avoiding excess driving force for any one step (i.e. greater than the driving force of the net reaction), since the thermodynamic properties of the catalyst (ΔG°H–, ΔG°H2, and pKa) are fundamentally connected in a thermochemical cycle. The optimal balance in driving force between the catalytic steps will likely vary between different catalyst families that have varying correlations between driving force and rate for each step of catalysis. An illustrative example is the complex [Cp*Ir(OH)2(6,6ʹ-dihydroxy-2,2ʹ-bipyridyl)]2+, which is also an active catalyst for hydrogenation of CO2 at room temperature and pressure.23 Computational studies on this Ir complex suggest a similar reaction pathway of H2 addition, deprotonation, and hydride transfer. Of these three steps, H2 addition was found be the rate-limiting step for catalysis despite having a stronger driving force (–2.4 kcal mol–1) than deprotonation (+2.5 kcal mol–1) or hydride transfer (–0.1 kcal mol–1). An initial TOF of 27 h–1 was reported for this Ir catalyst at 1 atm of 1:1 H2:CO2 in aqueous bicarbonate solution, which would correspond to TOF of 49 h–1 at 1.8 atm of 1:1 H2:CO2 (the pressure used in the present study). Rh(dmpbz)2+ displays a larger TOF of 3100 h–1 at 1.8 atm of 1:1 H2:CO2 despite having a

18 ACS Paragon Plus Environment

Page 19 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

lower driving for H2 addition (+0.2 kcal mol–1) than the Ir complex. This difference in TOF is likely due in part to the greater solubility of H2 in organic solvents than in water.82

Summary and Conclusions While many homogeneous catalysts for CO2 hydrogenation to formate have been reported, most require temperatures and pressures in excess of 80 °C and 20 atm. Several groups recognize the need to design catalysts that operate at low temperature and pressure, but very few catalysts have been reported that are active under ambient conditions. We examined the kinetics and thermodynamics for CO2 hydrogenation in a series of Co(P2)2H and Rh(P2)2+ complexes in order to gain a better understanding of the factors that lead to fast catalysis under ambient conditions. Hydride transfer was rate limiting for the cobalt catalysts, while hydrogen addition was rate limiting for all rhodium catalysts except Rh(dmpe)2+. Both the cobalt and rhodium complexes displayed a linear free energy correlation between the catalytic turnover frequency and the thermodynamic driving force for the rate limiting steps. The fastest catalyst in this study was Co(dmpe)2H (TOF = 6400 h–1 at 1.8 atm), which had a similar driving force for both hydride transfer (–7.6 kcal mol–1) and H2 addition (–6.2 kcal mol–1), with the deprotonation step being thermoneutral. This work demonstrates that catalyst design can be greatly aided by consideration of the thermodynamics of every step in the catalytic cycle. Importantly, the catalytic activity is maximized when the linear free energy relationships of the fundamental steps are properly balanced.

Experimental Section General Considerations: All manipulations were performed under an N2 atmosphere unless stated otherwise. Tetrahydrofuran, acetonitrile, ether, and hexane were purified by sparging with N2 and passage through neutral alumina using an Innovative Technology, Inc., PureSolv solvent purification system.

19 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 29

Triethylamine, DBU, the phosphazene base and Verkade’s base were purchased from Sigma-Aldrich and used without further purification. Triethylamine and DBU were dried over sodium and 4Å sieves until a Karl-Fisher reading of less than 20 ppm was obtained. THF-d8 and CD3CN were ordered from Cambridge Isotopes. THF-d8 was dried over a sodium-potassium ball, and CD3CN was dried over P2O5. CoCl2, [Rh(COD)2](CF3SO3), dmpe, depe and dmpm were purchased from Strem Chemical and used without further purification. Syntheses of BPE5,83 dmpbz,84 Co(dmpe)2H,69, 85 Co(dedpe)2H,69 Co(depe)2H,86 [Rh(dmpe)2](CF3SO3)/Rh(dmpe)2H,64,

87

[Rh(depe)2](CF3SO3),71

[Rh(dcpe)2](CF3SO3),71

[Rh(dppe)2](CF3SO3),71 [Co(CH3CN)6](BF4)2,88 and Rh(dppe)2H61 were previously reported elsewhere. Gases were purchased from Oxarc and were delivered to the NMR tube directly from the tank through a gas manifold line at pressures less than 1.8 atm and by an ISCO pump at pressures greater than 6.7 atm. The PEEK tubes were designed and built at Pacific Northwest National Laboratory.89-90 The glass capillary standard was prepared by adding Co(NO3)2 to water until the resonance in the 1H NMR spectrum appeared at a desirable location. All gas mixtures were measured by gas chromatography. GC analyses were carried out using a 100/120 Carbosieve S-II stainless steel column with a length of 10 feet and a diameter of 1/8 in. purchased from Supelco. A thermal conductivity detector was used, and the oven temperature was ramped from 100 to 175 °C. Argon was used as the carrier with a flow rate of 15 mL/min. A CH Instruments 660C potentiostat was used for all voltammetry experiments. Experiments were performed in a nitrogen glovebox (23 ± 2 °C) using a standard three-electrode configuration. A 1 mm diameter PEEK-encased glassy carbon disc (ALS) was used as the working electrode, a glassy carbon rod (Structure Probe, Inc.) was used as the counterelectrode, and a bare silver wire was used as the pseudoreference electrode. Ferrocene was used as an internal standard, and potentials are referenced to the Cp2Fe+/0 couple.

20 ACS Paragon Plus Environment

Page 21 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Representative Synthetic Procedures: Trans-[Co(dmpbz)2(CH3CN)2](BF4)2. To a 50 mL round bottom flask was added [Co(CH3CN)6](BF4)2 (978 mg, 2.04 mmol) and CH3CN (10 mL). To this solution was added dmpbz (829 mg, 4.18 mmol) dropwise. The solution turned from red to brown and was stirred for 10 min at room temperature. The solution was then allowed to sit for 2 h. A brown precipitate formed, which was filtered off. Crystals were grown by diffusing ether into the remaining reaction solution, yielding large dark purple blocks (1.19 g, 77%). Note crystals become orange upon desolvation of one CH3CN molecule. 1H NMR (paramagnetic, CD3CN):  7.67 (br, 2H), 5.26 (br, 2H), 2.10 (br, CH3CN, 6H), -3.76 (br, PCH3, 12H). Anal. Calcd for C24H38B2CoF8N2P4: %C, 40.54; %H, 5.39; %N, 3.94. Found %C, 40.61; %H, 5.20; %N, 3.96. A

similar

procedure

was

used

to

synthesize

[Co(depe)2CH3CN](BF4)2

and

[Co(BPE5)2(CH3CN)x](BF4)2; see the SI for full details. Co(dmpbz)2H. To a 50 mL round bottom flask was added [Co(dmpbz)2(CH3CN)2](BF4)2 (1.18 g, 1.66 mmol) and CH3CN (10 mL). To this solution was slowly added KC8 (475 mg, 3.52 mmol). The solution turned from purple to red and was stirred for 20 min at room temperature. The flask was then sealed with a septum and H2 was added (at a constant flow of 2 atm) over a 15 min period. The reaction was then stirred for 16 h, before filtration. The filtrate was reduced to 5 mL and then placed in a -35 °C to grow microcrystals. The filter cake was washed with pentane providing a red solution, which was removed in vacuo to provide the majority of the product as a red powder (610 mg, 58%). X-ray quality crystals were grown from a pentane:THF (15:1) solution at -30 °C.

1

H NMR (THF-d8):  7.69 (br,

CCH2CH2, 8H), 7.33 (m, CCH2, 8H), 1.54 (CH3, 24H), -15.72 (quintet, CoH, 2JPH = 24 Hz, 1H). NMR (THF-d8): 153.4 (PC), 128.8 (CCH2CH2), 127.2 (CCH2), 26.1 (CH3).

31

13

C{1H}

P{1H} NMR (THF-d8): 

39.4 (br). Anal. Calcd for C20H33CoP4: %C, 52.64; %H, 7.29. Found %C, 52.75; %H, 7.14.

21 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 29

A similar procedure was used to synthesize Co(BPE5)2H and Co(depe)2H; see the SI for full details. [Rh(dmpbz)2](CF3SO3). To a 20 mL vial was added [Rh(COD)2](CF3SO3) (375 mg, 0.801 mmol) and THF (5 mL). To this solution was slowly added dmpbz (325 mg, 1.64 mmol). The solution was stirred at room temperature. During this time the solution changes from red to orange and eventually a yellow precipitate forms. The yellow precipitate is filtered and then washed with 5 mL ether, 5 mL pentane and 5 mL THF to remove free COD. Yield: 489 mg, 94%. CCH2CH2, 8H), 7.69 (m, CCH2, 8H), 1.89 (CH3, 24H). (CCH2CH2), 130.2 (CCH2), 18.1 (CH3).

13

1

H NMR (THF-d8):  8.10 (br,

C{1H} NMR (THF-d8): 145.4 (PC), 132.5

P{1H} NMR (THF-d8):  33.9 (d, JRhP = 121 Hz).

31

19

F{1H}

NMR (THF-d8):  -77. Anal. Calcd for C21H32RhF3O3SP4: %C, 38.90; %H, 4.97. Found %C, 38.77; %H, 4.81. A similar procedure was used to synthesize [Rh(depe)2](CF3SO3) and [Rh(dcpe)2](CF3SO3); see the SI for full details. Rh(dppe)2H.61 To a solution of [Rh(dppe)2](CF3SO3) (180 mg, 0.172 mmol in 5 mL CH3CN) was added KOtBu (27 mg, 0.240 mmol in 5 mL CH3CN). This solution was filtered and then H2 was bubbled through the solution for 10 min. Small orange crystals formed within 2 h that were collected, washed with CH3CN, and then dried. Yield: 96 mg (62%). 1H NMR (THF-d8):  7.32 (br s, PCCH, 16H), 7.06 (t, J = 10 Hz, PCCHCH, 8H), 6.97 (t, J = 10 Hz, PCCHCHCH, 16H), 2.12 (br s, CH2CH2, 8H), 10.55 (m, RhH, 1H).

13

C{1H} NMR (THF-d8): 143.1 (PC), 132.8, 128.0, 127.8, 32.2 (PCH2) .

31

P{1H}

NMR (THF-d8):  56.7 (d, JRhP = 143 Hz). Anal. Calcd for C52H49RhP4: %C, 69.34; %H, 5.48. Found %C, 69.17; %H, 5.36. Catalytic Procedures: Acquisition of NMR Spectra: Operators of high-pressure equipment such as that required for these experiments should take proper precautions to minimize the risk of personal injury. 22 ACS Paragon Plus Environment

Page 23 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

1

H NMR spectra were collected at 500 MHz on a Varian with a 13 degree pulse, 2 sec acquisition

time and a 1 sec recycle time, total acquisition time for 1 transient was 3 sec. Either 8 or 16 transients were collected each cycle for a total NMR acquisition time of 24 or 48 sec. Spectra were referenced to internal residual protonated solvent.

These parameters were found to yield highly reproducible

quantitative results under the reaction conditions listed through testing with known concentrations of formate-containing solutions. At high catalyst concentration the sealed standard containing cobalt(II) and HDO was integrated against the cobalt-hydride resonance before adding gases.

At low catalyst

concentration the hydride resonance was not observable and an external calibration of the sealed HDO standard was used. As an additional internal standard the area of the proton NMR resonance of the acidic proton on the Verkade’s base (which is split into two resonances by scalar coupling from

31

P) was also

used. During kinetic runs no shimming was used before spectral collection and the sample was run without spinning to minimize the time between cycles. The time quoted for each cycle was from the start of the first transient. Turnover frequencies were calculated from the slope of the linear portion of the kinetic curves such as those displayed in the SI. The TOF’s quoted in Table 1 were the average of two to three experiments run under identical conditions.

31

P{1H} NMR spectra were collected at time zero and after

catalysis was complete and for select kinetic runs. Spectra were referenced to external phosphoric acid. Representative preparation in a J Young glass tube: A glass capillary containing the standard was added to the tube. Vkd base was dissolved in 200 µL THF-d8 and then added to the tube, followed by 50 µL of the catalyst at the desired dilution. The volume was then made to a total volume of 500 µL with THF-d8. The tube was sealed and transferred to the gas manifold line. The tube was exposed to a static vacuum to eliminate N2 from the tube, then the CO2/H2 gas mix was delivered and mixed for 30 sec on a vortex mixer before collecting an NMR spectrum. The tube was then connected back to the gas manifold line, fresh gas was added, the sample was mixed for 30 s, and another NMR spectrum was collected. This

23 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 29

cycle was repeated until the reaction was complete. The length of time elapsed between each refill is approximately two minutes. Representative preparation in a PEEK tube: A glass capillary containing the standard was added to the tube. The base was dissolved in 150 µL THF-d8 and then added to the tube, followed by 50 µL of the catalyst at the desired dilution. The volume was then made to a total volume of between 330 and 390 µL with THF-d8. The tube was then sealed and transferred to the ISCO pump apparatus. The tube was exposed to a static vacuum to eliminate N2 from the tube. Then the CO2/H2 gas mix was delivered at a constant pressure by an ISCO pump at the desired pressure. The reaction was placed on a vortex mixer to allow for gas mixing between collection of the NMR spectra. NMR spectra were collected between two and five minutes apart. X-ray

Structural

Determination.

Crystals

of

[Co(depe)2(CH3CN)](BF4)2

and

trans-

[Co(dmpbz)2(CH3CN)2](BF4)2 were grown by ether diffusion into acetonitrile, and crystals of Co(dmpbz)2H were grown from a 15:1 solution of pentane:THF. Crystals selected for diffraction studies were immersed in Paratone-N oil, placed on a Nylon loop, and transferred to a precooled cold stream of N2. A Bruker KAPPA APEX II CCD diffractometer with 0.71073 Å Mo K radiation was used for diffraction studies. The space groups were determined on the basis of systematic absences and intensity statistics. The structures were solved by direct methods and refined by full-matrix least squares on F2. All nonhydrogen atoms were refined anisotropically unless otherwise stated. Hydrogen atoms were placed at idealized positions and refined using the riding model unless otherwise stated. Data collection and cell refinement were performed using Bruker APEX2 software. Data reductions and absorption corrections were performed using Bruker’s SAINT and SADABS programs, respectively.91-92 Structural solutions and refinements were completed using SHELXS-97 and SHELXL-97,93 respectively, using the OLEX2 software package94 as a front-end.

24 ACS Paragon Plus Environment

Page 25 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

Notes on the structure of Co(dmpbz)2H: The complex, which does not possess true two-fold symmetry, is disordered about a two-fold rotation axis. As such, although there are two crystallographically distinct dmpbz ligands, they are forced to almost totally overlap by the two-fold axis. AFIX 66 constraints were placed on the phenyl rings and the SIMU restraint was globally applied. Although all non-H atoms were refined anisotropically, it was necessary to constrain the thermal ellipsoids of two sets of methyl groups to be identical using the EADP command. The two sets of ipso carbons of the dmpbz ligands were also constrained to be identical by the EADP command. The hydride ligand could not be identified in the Fourier difference map; based on the geometry around Co, which itself appears to have pseudo two-fold symmetry (distinct from the crystallographic two-fold), the hydride ligand may be disordered across two positions for each Co environment (thus disordered across four positions in total).

Notes The authors declare no competing financial interests.

Supporting Information (SI) Available: Additional synthetic procedures, multinuclear NMR data, representative kinetic plots, cyclic voltammograms, and thermochemical data (PDF); X-ray crystallographic data (cif)

Acknowledgements The research by M.S.J., A.M.A., E.S.W., and J.C.L. was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences, and Biosciences. The research by E.B.H., M.L.H., and M.T.M. (X-ray crystallography, synthesis) was supported as part of the Center for Molecular Electrocatalysis, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences. The authors thank 25 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 29

Dr. Samantha A. Burgess for assistance in collecting cyclic voltammetry data. Pacific Northwest National Laboratory is operated by Battelle for the U.S. Department of Energy.

References 1. 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. Lim, R. J.; Xie, M.; Sk, M. A.; Lee, J.-M.; Fisher, A.; Wang, X.; Lim, K. H. Catal. Today 2014, 233, 169-180. 3. Kumar, B.; Brian, J. P.; Atla, V.; Kumari, S.; Bertram, K. A.; White, R. T.; Spurgeon, J. M. Catal. Today 2016, 270, 19-30. 4. White, J. L.; Baruch, M. F.; Pander III, J. E.; Hu, Y.; Fortmeyer, I. C.; Park, J. E.; Zhang, T.; Liao, K.; Gu, J.; Yan, Y.; Shaw, T. W.; Abelev, E.; Bocarsly, A. B. Chem. Rev. 2015, 115, 12888-12935. 5. Schneider, J.; Jia, H. F.; Muckerman, J. T.; Fujita, E. Chem. Soc. Rev. 2012, 41, 2036-2051. 6. Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Chem. Soc. Rev. 2009, 38, 89-99. 7. Wang, W.-H.; Himeda, Y.; Muckerman, J. T.; Manbeck, G. F.; Fujita, E. Chem. Rev. 2015, 115, 12936-73. 8. Liu, C.; Colón, B. C.; Ziesack, M.; Silver, P. A.; Nocera, D. G. Science 2016, 352, 1210-1213. 9. Bernskoetter, W. H.; Hazari, N. Acc. Chem. Res. 2017, 50, 1049-1058. 10. Jessop, P. G.; Ikariya, T.; Noyori, R. Chem. Rev. 1995, 95, 259-272. 11. Jessop, P. G.; Joó, F.; Tai, C.-C. Coord. Chem. Rev. 2004, 248, 2425-2442. 12. Dalebrook, A. F.; Gan, W.; Grasemann, M.; Moret, S.; Laurenczy, G. Chem. Commun. 2013, 49, 87358751. 13. Singh, A. K.; Singh, S.; Kumar, A. Catal. Sci. Technol. 2016, 6, 12-40. 14. Mellmann, D.; Sponholz, P.; Junge, H.; Beller, M. Chem. Soc. Rev. 2016, 45, 3954-3988. 15. Hayashi, H.; Ogo, S.; Fukuzumi, S. Chem. Commun. 2004, 2714-2715. 16. Ogo, S.; Kabe, R.; Hayashi, H.; Harada, R.; Fukuzumi, S. Dalton Trans. 2006, 4657-4663. 17. Leitner, W. Angew. Chem., Int. Ed. 1995, 34, 2207-2221. 18. Ezhova, N. N.; Kolesnichenko, N. V.; Bulygin, A. V.; Slivinskii, E. V.; Han, S. Russ. Chem. Bull. 2002, 51, 2165-2169. 19. Graf, E.; Leitner, W. Chem. Commun. 1992, 623-624. 20. Gassner, F.; Leitner, W. Chem. Commun. 1993, 1465-1466. 21. Angermund, K.; Baumann, W.; Dinjus, E.; Fornika, R.; Görls, H.; Kessler, M.; Krüger, C.; Leitner, W.; Lutz, F. Chem. Eur. J. 1997, 3, 755-764. 22. Hull, J. F.; Himeda, Y.; Wang, W. H.; Hashiguchi, B.; Periana, R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E. Nat. Chem. 2012, 4, 383-388. 23. Wang, W. H.; Hull, J. F.; Muckerman, J. T.; Fujita, E.; Himeda, Y. Energy Environ. Sci. 2012, 5, 7923-7926. 24. Wang, W.-H.; Muckerman, J. T.; Fujita, E.; Himeda, Y. ACS Catal. 2013, 3, 856-860. 25. Tanaka, R.; Yamashita, M.; Nozaki, K. J. Am. Chem. Soc. 2009, 131, 14168-14169. 26. Tanaka, R.; Yamashita, M.; Chung, L. W.; Morokuma, K.; Nozaki, K. Organometallics 2011, 30, 6742-6750. 27. Munshi, P.; Main, A. D.; Linehan, J. C.; Tai, C.-C.; Jessop, P. G. J. Am. Chem. Soc. 2002, 124, 7963-7971. 28. Schmeier, T. J.; Dobereiner, G. E.; Crabtree, R. H.; Hazari, N. J. Am. Chem. Soc. 2011, 133, 9274-9277. 29. Himeda, Y.; Onozawa-Komatsuzaki, N.; Sugihara, H.; Kasuga, K. Organometallics 2007, 26, 702-712. 30. Maenaka, Y.; Suenobu, T.; Fukuzumi, S. Energy Environ. Sci. 2012, 5, 7360-7367. 31. Sanz, S.; Benítez, M.; Peris, E. Organometallics 2010, 29, 275-277. 32. Azua, A.; Sanz, S.; Peris, E. Chem. Eur. J. 2011, 17, 3963-3967. 26 ACS Paragon Plus Environment

Page 27 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

33. Drake, J. L.; Manna, C. M.; Byers, J. A. Organometallics 2013, 32, 6891-6894. 34. Huff, C. A.; Sanford, M. S. ACS Catal. 2013, 3, 2412-2416. 35. Inoue, Y.; Izumida, H.; Sasaki, Y.; Hashimoto, H. Chem. Lett. 1976, 5, 863-864. 36. Evans, G. O.; Newell, C. J. Inorg. Chim. Acta 1978, 31, L387-L389. 37. Tai, C.-C.; Chang, T.; Roller, B.; Jessop, P. G. Inorg. Chem. 2003, 42, 7340-7341. 38. Federsel, C.; Boddien, A.; Jackstell, R.; Jennerjahn, R.; Dyson, P. J.; Scopelliti, R.; Laurenczy, G.; Beller, M. Angew. Chem., Int. Ed. 2010, 49, 9777-9780. 39. Langer, R.; Diskin-Posner, Y.; Leitus, G.; Shimon, L. J. W.; Ben-David, Y.; Milstein, D. Angew. Chem., Int. Ed. 2011, 50, 9948-9952. 40. Federsel, C.; Ziebart, C.; Jackstell, R.; Baumann, W.; Beller, M. Chem. Eur. J. 2012, 18, 72-75. 41. Ziebart, C.; Federsel, C.; Anbarasan, P.; Jackstell, R.; Baumann, W.; Spannenberg, A.; Beller, M. J. Am. Chem. Soc. 2012, 134, 20701-20704. 42. Badiei, Y. M.; Wang, W.-H.; Hull, J. F.; Szalda, D. J.; Muckerman, J. T.; Himeda, Y.; Fujita, E. Inorg. Chem. 2013, 52, 12576-12586. 43. Jeletic, M. S.; Mock, M. T.; Appel, A. M.; Linehan, J. C. J. Am. Chem. Soc. 2013, 135, 11533-11536. 44. Jeletic, M. S.; Helm, M. L.; Hulley, E. B.; Mock, M. T.; Appel, A. M.; Linehan, J. C. ACS Catal. 2014, 4, 37553762. 45. Zall, C. M.; Linehan, J. C.; Appel, A. M. ACS Catal. 2015, 5, 5301-5305. 46. Zhang, Y.; MacIntosh, A. D.; Wong, J. L.; Bielinski, E. A.; Williard, P. G.; Mercado, B. Q.; Hazari, N.; Bernskoetter, W. H. Chem. Sci. 2015, 6, 4291-4299. 47. Spentzos, A. Z.; Barnes, C. L.; Bernskoetter, W. H. Inorg. Chem. 2016, 55, 8225-8233. 48. Bertini, F.; Gorgas, N.; Stöger, B.; Peruzzini, M.; Veiros, L. F.; Kirchner, K.; Gonsalvi, L. ACS Catal. 2016, 6, 2889-2893. 49. Bertini, F.; Glatz, M.; Gorgas, N.; Stoeger, B.; Peruzzini, M.; Veiros, L. F. F.; Kirchner, K.; Gonsalvi, L. Chem. Sci. 2017, 8, 5024-5029. 50. Burgess, S. A.; Kendall, A. J.; Tyler, D. R.; Linehan, J. C.; Appel, A. M. ACS Catal. 2017, 7, 3089-3096. 51. Burgess, S. A.; Grubel, K.; Appel, A. M.; Wiedner, E. S.; Linehan, J. C. Inorg. Chem. 2017, 56, 8580-8589. 52. Federsel, C.; Jackstell, R.; Beller, M. Angew. Chem., Int. Ed. 2010, 49, 6254-6257. 53. Cokoja, M.; Bruckmeier, C.; Rieger, B.; Herrmann, W. A.; Kühn, F. E. Angew. Chem., Int. Ed. 2011, 50, 8510-8537. 54. 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.; 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. 55. Cook, T. R.; Dogutan, D. K.; Reece, S. Y.; Surendranath, Y.; Teets, T. S.; Nocera, D. G. Chem. Rev. 2010, 110, 6474-6502. 56. Yang, X. ACS Catal. 2011, 1, 849-854. 57. Urakawa, A.; Jutz, F.; Laurenczy, G.; Baiker, A. Chem. Eur. J. 2007, 13, 3886-3899. 58. Ahlquist, M. S. G. J. Mol. Catal. A: Chem. 2010, 324, 3-8. 59. Kisanga, P. B.; Verkade, J. G.; Schwesinger, R. J. Org. Chem. 2000, 65, 5431-5432. 60. Qi, X.-J.; Liu, L.; Fu, Y.; Guo, Q.-X. Organometallics 2006, 25, 5879-5886. 61. Sacco, A.; Ugo, R. J. Chem. Soc. 1964, 3274-3278. 62. Saame, J.; Rodima, T.; Tshepelevitsh, S.; Kütt, A.; Kaljurand, I.; Haljasorg, T.; Koppel, I. A.; Leito, I. J. Org. Chem. 2016, 81, 7349-7361. 63. Ni, S.-F.; Dang, L. Phys. Chem. Chem. Phys. 2016, 18, 4860-4870. 64. Mock, M. T.; Potter, R. G.; Camaioni, D. M.; Li, J.; Dougherty, W. G.; Kassel, W. S.; Twamley, B.; DuBois, D. L. J. Am. Chem. Soc. 2009, 131, 14454-14465. 27 ACS Paragon Plus Environment

ACS Catalysis

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 29

65. Qi, X.-J.; Fu, Y.; Liu, L.; Guo, Q.-X. Organometallics 2007, 26, 4197-4203. 66. Chen, S.; Rousseau, R.; Raugei, S.; Dupuis, M.; DuBois, D. L.; Bullock, R. M. Organometallics 2011, 30, 6108-6118. 67. Ciancanelli, R.; Noll, B. C.; DuBois, D. L.; Rakowski DuBois, M. J. Am. Chem. Soc. 2002, 124, 2984-2992. 68. Wiedner, E. S.; Chambers, M. B.; Pitman, C. L.; Bullock, R. M.; Miller, A. J. M.; Appel, A. M. Chem. Rev. 2016, 116, 8655-8692. 69. Mock, M. T.; Potter, R. G.; O’Hagan, M. J.; Camaioni, D. M.; Dougherty, W. G.; Kassel, W. S.; DuBois, D. L. Inorg. Chem. 2011, 50, 11914-11928. 70. Berning, D. E.; Miedaner, A.; Curtis, C. J.; Noll, B. C.; Rakowski DuBois, M.; DuBois, D. L. Organometallics 2001, 20, 1832-1839. 71. Wilson, A. D.; Miller, A. J. M.; DuBois, D. L.; Labinger, J. A.; Bercaw, J. E. Inorg. Chem. 2010, 49, 39183926. 72. Raebiger, J. W.; DuBois, D. L. Organometallics 2005, 24, 110-118. 73. Garrido, G.; Koort, E.; Ràfols, C.; Bosch, E.; Rodima, T.; Leito, I.; Rosés, M. J. Org. Chem. 2006, 71, 90629067. 74. Kaljurand, I.; Kutt, A.; Soovali, L.; Rodima, T.; Maemets, V.; Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019-1028. 75. Mondal, B.; Neese, F.; Ye, S. Inorg. Chem. 2016, 55, 5438-5444. 76. Leffler, J. E. Science 1953, 117, 340-341. 77. Kumar, N.; Camaioni, D. M.; Dupuis, M.; Raugei, S.; Appel, A. M. Dalton Trans. 2014, 43, 11803-11806. 78. Matsubara, Y.; Fujita, E.; Doherty, M. D.; Muckerman, J. T.; Creutz, C. J. Am. Chem. Soc. 2012, 134, 15743-15757. 79. Konno, H.; Kobayashi, A.; Sakamoto, K.; Fagalde, F.; Katz, N. E.; Saitoh, H.; Ishitani, O. Inorg. Chim. Acta 2000, 299, 155-163. 80. Field, L. D.; Thomas, I. P.; Hambley, T. W.; Turner, P. Inorg. Chem. 1998, 37, 612-618. 81. Bernskoetter, W. H.; Hazari, N. Eur. J. Inorg. Chem. 2013, 2013, 4032-4041. 82. Purwanto; Deshpande, R. M.; Chaudhari, R. V.; Delmas, H. J. Chem. Eng. Data 1996, 41, 1414-1417. 83. Field, L. D.; Thomas, I. P. Inorg. Chem. 1996, 35, 2546-2548. 84. Steinmetz, G. R.; Larkins, T. H. Organometallics 1983, 2, 1879-1882. 85. Schunn, R. A. Inorg. Chem. 1970, 9, 2567-2572. 86. Basallote, M. G.; Hughes, D. L.; Jimenez-Tenorio, M.; Leigh, G. J.; Vizcaino, M. C. P.; Jimenez, P. V. Dalton Trans. 1993, 1841-1847. 87. James, B. R.; Mahajan, D. Can. J. Chem. 1980, 58, 996-1004. 88. Heintz, R. A.; Smith, J. A.; Szalay, P. S.; Weisgerber, A.; Dunbar, K. R. Inorg. Synth. 2002, 33, 75-83. 89. Yonker, C. R.; Linehan, J. C. J. Organomet. Chem. 2002, 650, 249-257. 90. Yonker, C. R.; Linehan, J. C. Prog. Nucl. Magn. Reson. Spectrosc. 2005, 47, 95-109. 91. Bruker. SAINT Program for Data Reduction, Bruker AXS Inc., Madison, WI, USA, 2007. 92. Sheldrick, G. M. SADABS Program for Absorption Correction of Area Detector Frames, University of Göttingen, Germany, 1996. 93. Sheldrick, G. Acta Crystallogr., Sect. A 2008, 64, 112-122. 94. Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.; Puschmann, H. J. Appl. Crystallogr. 2009, 42, 339-341.

28 ACS Paragon Plus Environment

Page 29 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Catalysis

TOC Graphic

29 ACS Paragon Plus Environment