DFT Study on the Mechanism of Hydrogen Storage ... - ACS Publications

Article Cite This: Inorg. Chem. 2018, 57, 5903−5914

pubs.acs.org/IC

DFT Study on the Mechanism of Hydrogen Storage Based on the Formate-Bicarbonate Equilibrium Catalyzed by an Ir-NHC Complex: An Elusive Intramolecular C−H Activation Péter Pál Fehér,†,‡ Henrietta Horváth,§ Ferenc Joó,†,§ and Mihály Purgel*,† †

Department of Physical Chemistry, University of Debrecen, Egyetem tér 1, Debrecen H-4032, Hungary Research Centre for Natural Sciences, Hungarian Academy of Sciences, Magyar tudósok körútja 2, H-1117 Budapest, Hungary § MTA-DE Redox and Homogeneous Catalytic Reaction Mechanisms Research Group, Debrecen, P.O. Box 400, H-4002, Hungary ‡

S Supporting Information *

ABSTRACT: A novel iridium based, water-soluble phosphine-NHC (Nheterocyclic carbene) complex, Na2[Ir(emim)(η4-COD)(mtppts)] was previously developed in our research group. It was shown that it is a very effective catalyst for the reversible storage of hydrogen based on the formatebicarbonate equilibrium. In this paper, we present a DFT investigation on the noninnocent behavior of the NHC ligand toward C−H activation of the Nethyl side chain and its possible role in the hydrogen storage mechanism. After preliminary investigations, using both computations and NMR measurements, we conclude that the COD ligand leaves the precatalyst irreversibly and the C−H activation takes place on a monophosphine complex. Two main pathways are considered in which the active Ir(III) complexes are generated differently: One is the cyclometalation path involving the ethyl side chain, the other is the oxidative addition step of a water molecule which has a higher barrier but provide a more stable starting state. We find that though the latter, a catalytic cycle where a hydride is abstracted from formate and gets protonated by solvent molecules gives the lowest calculated energy barrier, +25.8 kcal mol−1. That is, avoiding further redox processes is preferred. There are other pathways involving thermodynamically accessible C−H activated iridacycles but those involve slightly higher overall activation barriers due to the required Ir(I)/Ir(III) transitions. The cycle which involves only iridacycle intermediates offer the lowest energy span (energy difference calculated between only the highest and lowest energy points inside the cycle), however. Together with the experimental results, this implies that C−H activation of the N-ethyl side chain happens off-cycle or the starting solvent addition step of the dominant pathway is blocked kinetically. We also discuss the hydrogen uptake reaction catalyzed by cyclometalated species where the reduction of CO2 is preferred over reversing the first main cycle.

■

INTRODUCTION Hydrogen storage based on the transformation of CO2 has been in focus of research in homogeneous catalysis over the last decades.1−4 Among the possible carrier molecules resulting in the reaction between H2 and CO2, formic acid and methanol are the most prominent, given their high hydrogen content and atomic efficiency. To catalyze these processes many different complexes containing first row (Mn, Co, Ni, Fe) or later (Ru, Rh, Ir) transition metal centers have been developed over the years.5−16 Because of the sheer number of these catalysts, designing one that is clearly superior to the others in every aspect is impossible. A good catalyst would be inexpensive, fast, robust, selective, and would not require organic solvents. Connected with the latter, a central goal in development is to make the reaction processes environmental-friendly if possible. One way to achieve this is to make water-soluble catalysts, for example, by sulfonating one or more phenyl rings of the (phosphine) ligands,17 and utilize the formate−bicarbonate equilibrium18 (1). These are the “greenest” ways of hydrogen © 2018 American Chemical Society

storage because neither CO2 emission nor organic solvents are involved.19,20 Further advantage is that the reaction free enthalpy change is nearly zero meaning that the equilibrium can be shifted by small changes in temperature.21 This is another advantage over the hydrogenation of CO2 to methanol or formic acid because the reduction of CO2 is thermodynamically not favored therefore complicating the realization of actual industrial processes.22,23 HCO−2 + H 2O ⇌ HCO−3 + H 2

(1)

Since in theory any complex that is able to activate the H−H bond in dihydrogen, many catalysts employed in homogeneous reactions are potential candidates for hydrogen storage reactions. Perhaps the latest, most considerable development in the design of such catalysts has been the widespread use of N-heterocyclic carbenes (NHC) as ligands in place of one or Received: February 13, 2018 Published: April 27, 2018 5903

DOI: 10.1021/acs.inorgchem.8b00382 Inorg. Chem. 2018, 57, 5903−5914

Article

Inorganic Chemistry more phosphines.24−26 NHCs were generally considered as stable spectator ligands because of their strong σ-donating affinity. However, it is now known that they are prone to reductive elimination, (Scheme 1) where the release of NHC is

works, however, were done on ruthenium complexes. Only a few examples with other metals exhibiting such behavior were found and they can be described almost exclusively with the general formula Cp*Ir(NHC).30,37,38 Based on these considerations, we aim to investigate the possible role of the C−H activation process in a water-soluble iridium based carbene-phosphine complex, Na2[Ir(emim)(η4COD)(mtppts)] (COD = 1,5-cyclooctadiene, emim = 1-ethyl3-methylimidazol-2-ylidene, mtppts = trisulfonated triphenylphosphine) developed in our research group.39 This catalyst offers exceptionally good performance in the formatebicarbonate system of hydrogen storage.40 As the active intermediates have not been characterized yet, our further goal is to identify possible catalytically active species and develop a rational catalytic cycle.

Scheme 1. Possible Decomposition (Upper) and C−H Activation (Lower) Processes Involving Iridium and emim (1-Ethyl-3-methylimidazol-2-ylidene)

■

COMPUTATIONAL DETAILS All calculations were performed using the Gaussian09 software package.41 The chosen functional was B3LYP together with D3 empirical dispersion corrections.42,43 For iridium, the CRENBL effective core potential (ECP) with the corresponding basis set (5s, 5p, 4d) was used.44 The rest of the atoms were described with the TZVP all electron triple-ζ basis.45 To confirm the stationary nature of the structures frequency calculations were performed. All geometries were optimized in a self-consistent reaction field using the polarizable continuum model (IEFPCM) to account for the solvent (water) unless stated otherwise.46 As the PCM model gives very poor solvation free enthalpies for the solvated anions (formate and bicarbonate), we used experimental values to correct them. The effect of

even more likely when a hydride is cis relative to it.27,28 This presents a major drawback, because catalysts employed in hydrogen storage applications are expected to have at least one hydride species in their catalytic cycle. Another possible deactivation pathway involves the activation of C−H or even C−C bonds in the N-side chain of the NHC.29−31 This process, however, does not always result in decomposition as such metallacycles can be synthesized in good yield and characterized using conventional methods.32−34 In some cases they can even play an active role in catalysis.35,36 Most of the related

Figure 1. Ligand exchange energetics in the precatalyst 1, [Ir(emim)(η4-COD)(PPh3)]+. The values in italics in units of kcal mol−1 correspond to binding energies (Gibbs energy difference between the complex and the sum of its relaxed fragments). Purple color is used for the generation of 2 via the COD hydrogenation pathway, making the following steps more accessible by lowering their energies by 33.1 kcal mol−1; however, these values are not shown for clarity, but discussed in the text. There are also LANL2TZ/SMD optimized results shown in Figure S2. 5904


Article

Inorganic Chemistry

Figure 1. For the reaction of COD + 2H2 = cyclooctane, ΔG is −33.1 kcal mol−1. It is shown in a recent work that COD is hydrogenated even under transfer hydrogenation (iPrOH) conditions where a cis-dihydride is reported as the active species in the transformation.59 Since this process is considered irreversible in aqueous medium, we do not wish to delve into its mechanism but a short discussion is given in Figure S1. However, it might not be trivial how the Ir(I) center can be retained after passing through Ir(III) containing intermediates (other than the reductive elimination of an H2 molecule) or how many phosphines coordinate to the metal center. Following COD dissociation, the two vacant coordination sites can be filled according to Figure 1, which assumes that C− H activation of the ethyl side chain is preferred after the resaturation of the coordination sphere. This kind of a C−H activation can only happen from 6 (Figure S3 shows a direct addition step with a very high, + 46.3 kcal mol−1 barrier) if it precedes the coordination of the second phosphine or there is only one phosphine in the catalyst altogether. In this work, we focus on monophosphine complexes; however, in the presence of excess phosphines, the catalytic activity of Ir(I) complexes is elevated in the formate decomposition reaction,40 but the catalytic activity of monophosphines formed from the various Ir-NHC-phosphines are remarkable, as well.39 Although the bisphosphine complex has better thermodynamic stability, feasibility of a monophosphine species is supported by the observation of an oxalate (ox) stabilized trans-[Ir(III)H2(η2ox)(bmim)(mtppts)] complex.39 Stable complexes containing formate, however, were not found since they are active toward its decomposition. In light of these findings, we have performed NMR measurements on a system containing Na2[Ir(emim)(η4COD)(mtppts)] and NaDCOO in D2O without added phosphine. Hydrogenation of COD in the presence of formate was clearly shown by the disappearance of the caracteristic signals of coordinated COD (29.87, 30.28, 30.73, 31.37, t, CH2,cod; 79.46, 81.17, d, CHcod; 88.62, 90.39, d, CHcod) in the 13 C NMR spectra recorded in DMSO/D2O solvent mixture. The low water solubility of COD (and its hydrogenated derivatives) and the huge excess of formate in the aqueous phase means that the process itself can be considered irreversible. Our NMR measurements without added phosphine confirm some C−H activation process as deuterium exchange is observed on the two methyl end-groups (Figure S5), however, its mechanism needs to be elucidated. It is considered a well-established fact that formate complexes have κ1-O or κ2-O,O resting states and dehydrogenation requires a twisting of the substrate to reach an H-bonded state from which the actual hydride transfer is facile.48,60−62 This step is found to be crucial in previously published mechanisms but the possibility of it providing the TS with the highest energy on a given pathway is rarely discussed.62,63 There are proposed mechanisms where the reorientation happens on a single coordination site (one-site) or via a structure like 3 (κ2-O,O mode) with κ2-O,H bonded formate (two-site).52,63,64 The latter is directly relevant here because in a recent thorough experimental (NMR) investigation done on a water-soluble ruthenium based trisphosphine catalyst, [Ru(II)(H2O)(κ2 HCO2)(mtppts)3] a β-hydride elimination pathway is proposed.65 Since the coordination chemistry of Ru(II) is more like Ir(III) than Ir(I) this process might favor a C−H activated complex which is formed via an oxidative addition pathway instead of a CMD mechanism. For example, 5 is similar to the mentioned trisphosphine complex in a sense that it contains

implicit solvation and the choice of the model is discussed in the Supporting Information. Throughout the work, relative Gibbs free energies are reported as G = E0 + ZPE + H298 − 298.15 × S298, where E0 is the total electronic energy including the PCM correction.

■

RESULTS In a transition metal catalyzed process, eq 1 can be expanded in the following way: HCO−2 + [MLn]m → CO2 + [MHLn]m − 1 −

CO2 + OH ⇌

HCO−3

(2a) (2b)

HCO−3 + [MHLn]m − 1 → HCO−2 + OH− + [MLn]m (2c)

The importance of eqs 2a−2c is in that the catalyst does not necessarily hydrogenate CO2, but rather the bicarbonate, since without using additives, the pH of the solution is 8.3 due to the excess formate. At basic pH, the reverse reaction of eq 2b has a 21.5 (experimental) or 20.7 kcal mol−1 (calculated) activation barrier.47 Another thing to note is that if eq 2a is slower than the forward reaction of eq 2b, which has a barrier of 11.5 (expt) or 13.8 kcal mol−1 (calc), the dehydrogenation of formate produces CO2 that is then hydrated to bicarbonate. Therefore, if a given catalyst is described with eqs 2a−2c, its catalytic cycle should have an energy span between approximately 13 and 21 kcal mol−1. Most of the previous research, however, was done on direct reaction of CO2 and metal hydrides (reverse of eq 2a), either because they modeled acidic conditions (no HCO3−) or because the activation barrier of eq 2c was found higher than the upper limit, 21 kcal mol−1.48,49 In the latter case, the rate-determining step becomes the spontaneous dehydration of bicarbonate (reverse of eq 2b). The H2 uptake/ release process might take place via an acid−base pathway as in eqs 3a−3c.50−52 [M(H2)Ln]m in eqs 3a and 3b may be produced from formate via eq 2a and eq 3a. It can also be deprotonated according to eq 3c, giving [MHLn]m−1, which hydrogenates bicarbonate in eq 2c. H+ + [MHLn]m − 1 → [M(H 2)Ln]m

(3a)

[M(H 2)Ln]m ⇌ [MLn]m + H 2

(3b)

B− + [M(H 2)Ln]m → [MHLn]m − 1 + HB

(3c)

Another possibility is an oxidative addition/reductive elimination equilibrium as in eq 4a, written explicitly for iridium.51 The C−H activation of the N-ethyl side chain might proceed through either eq 4b or via a concerted metalationdeprotonation (CMD)53−56 like in eq 3c with a −CH3 subunit replacing the H2 ligand. [Ir(I)Ln]m + H 2 ⇌ [Ir(III)H 2Ln]m

(4a)

[Ir(I)Ln − 1(κ 1NHC)]m ⇌ [Ir(III)HLn − 1(κ 2 NHC)]m

(4b)

In the first part of the text we consider the formation of a hydride species from [Ir(emim)(η4-COD)(PPh3)]+ (1), where the PPh3 is used instead of mtppts. To find the active form of the catalyst, the displacement of the COD ligand needs to be considered as well, as the catalytic activity of iridium complexes in alkene hydrogenation is well-known.57,58 Also, because formate insertion (3) is endergonic, the hydrogenation of COD to cyclooctene is required to precede this process, as shown in 5905


Article

Inorganic Chemistry

Figure 2. Possible starting structures for formate dehydrogenation. Here and on the following figures, 12 is used as reference for relative Gibbs energies unless indicated otherwise. For the transition structures, ball and stick models are used where the iridium atom is shown in cadet blue, phosphorus in orange, nitrogen in blue, carbon in gray, hydrogen in white. The phenyl rings are represented as wireframes for clarity.

Figure 3. One-site formate dehydrogenation starting from the C−H activated complex, 5. The possible activation barrier between 5 and 13 is discussed in the text.

three strong σ-donor ligands (not counting the hydride), but the lower number of bulky ligands provide less pronounced steric effects. There are also other oxidative processes available, like the addition of H2 or H2O and they should be considered as well. H2 addition, however, is not relevant during the dehydrogenation of formate, but as it is shown on Figure 2, the addition of H2O from 4 is able to compete with the C−H activation, as its TS lies at +26.6 kcal mol−1 and results in a stable intermediate, 12 (Figure 2). Since returning to the starting complex, 1 or accessing bisphosphine complexes is not possible in our case, we chose 12 as the reference for relative energies in this work as it is found to be the lowest energy state before formate dehydrogenation takes place. From this structure, C−H activation may happen via a CMD mechanism where either the coordinated OH− or formate can assist the deprotonation, see SI. The C−H activation between 4 and 5, shown on Figure 2, suggests an equilibrium process (nearly zero free enthalpy

change). It presents an additional possibility for a formate dehydrogenation pathway starting from the structures on the left side of Figure 2 and we discuss this case together with the C−H activation mechanism. This is only possible, however, if the system does not undergo the oxidative addition of water, which reaches a highly stable resting state from which the reverse reaction has a considerably higher barrier. Therefore, we discuss the mechanism involving 12 separately. First, we start with the C−H activation pathway. Although TS9−10 is slightly below the TS of the oxidative addition (TS4−12) in terms of energy, it cannot be said that 12 is kinetically inaccessible without considering further steps of the reaction. The five-membered iridacycle (10 without formate) can adopt two different helicities, as shown in Figure S6, which differ by 2.0 kcal mol−1 in energy. In the two isomers, the N-methyl side group is positioned between two phenyl rings. As the helicity changes the phosphine ligand follows the movement of the methyl group and gets rotated together with the NHC plane, 5906


Article

Inorganic Chemistry Scheme 2. Possible Pathways Connecting 5 and 13a

(a) First step (red): Ir−O bond breaking; Second step (blue): formate rotation to the κ1-H mode. (b) First step: water dissociation; Second step: formate rotation to κ2-O,H state. (c) First step: formate dissociation; Second step: rotation in the bulk of the solvent; Third step: recoordination to the κ1-H bonded state. a

Figure 4. Formation of the κ1-H formate complex and the β-elimination pathway starting from 11. Both directions result in the same cis-dihydride product after the breaking of the formate C−H bond.

resulting in different chemical environment for the ligands cis to the phosphine. In the following discussion, we select the one in which the phenyl rings block the coordination sphere less. The one-site dehydrogenation may take place in 4b or 5, while the β-elimination may proceed from 3 or 11. The difference between 4 and 4b is in that the water and formate ligands are exchanged. The formate binding in 4b is stronger due to the smaller trans influence of the phosphine compared to the NHC. This directing force is what determines the position of the hydride in the product complex after formate dehydrogenation. Starting directly from 4b, the one-site dehydrogenation (Figure S7) is found to be a dead end, which results in an unstable fourcoordinated Ir(I) monohydride product. Note that the reaction of 10 (=5 without the coordinated water molecule) + 2HCOO− − 2CO2 may result in fac or mer trihydrides,66,67 but these species lie above 10 by 3.2 and 5.2 kcal mol−1, respectively. The other dehydrogenation process, starting from the C−H activated complex, 5, is shown in Figure 3. The cis-dihydride product complex, 16, has four possible isomers: two come from

the two possible axial hydride positions and another two where the iridacycle has different helicity. These structures cover an energy range of 2.2 kcal mol−1 (Figure S9). It is clear that from both the κ1- and κ2-NHC complexes the elementary barrier corresponding to the cleavage of the formate C−H bond is very low and most of the activation energy is used to reach the κ1-H intermediate. In the following, focusing on the C−H activated (κ2-NHC) Ir(III) complexes, we investigate how 13 can be reached and whether the barrier is higher than the one corresponding to the bond breaking. Previously, two ways were mentioned for reaching the κ1-H state of the formate (13). The one happening on a single coordination site (Scheme 2a) seems straightforward from Figures 3 and S7, but the two-site process is also viable as the dissociation of the solvent molecule from 5 (Scheme 2b) takes only 2.6 kcal mol−1 energy. There is one additional possibility shown in Scheme 2 that involves the dissociation of the formate from the complex and recoordination with its H atom. We did not calculate activation barriers here as it is assumed to follow a saturation curve with 5907


Article

Inorganic Chemistry

Figure 5. Gibbs energy profile of the catalytic process involving the formation of bicarbonate and the C−H activation resulting in H2 production.

The other mechanism starting from 12 offers the same elementary steps with the only difference being the strong interaction between the coordinated formate and the hydroxide. This results in a change of energy ordering between the κ2-O,H and κ1-H states, where the former transforms to the monodentate mode during geometry optimizations. We did not investigate here the κ1-O to κ1-H transformation, as it is not expected to be fundamentally different from the previously discussed one-site case. The barrier corresponding to the 4 → 12 step is unlikely to be surpassed by this flipping. Since 23 already contains an OH and a CO2 moiety, the further reaction steps involving bicarbonate and hydrogen production are shown in Figure 5 as well. The dissociation of the bicarbonate results in 28, a cis-dihydride complex, where the other two ligands are in trans position relative to each other. This state is essentially an energy sink as the reductive elimination of a H2 molecule (it is easy to see that this would have a barrier that is higher than the energy difference between 28 and 2) can take place from 26 more readily. Compound 28, however, can play a role in a pathway, that includes the C−H activation of the ethyl side chain. The one shown on Figure 5 completes the redox cycle, but the necessary reductive elimination, which is the ringopening step, 31 → 2, comes with a very high barrier. This way, the C−H activation is either a CMD process where the hydride ligand deprotonates the methyl group in the 28 → 29 step or a transient Ir(V) mer-trihydride is formed in the TS which converts to the Ir(III) complex, 29 instantly. The imaginary vibration of the found TS implicates the latter. (Note that all CMD mechanisms are represented in the SI.) The C−H activated pathway, however, cannot be excluded as 31 offers a crossing point to the other main pathway without reverting to Ir(I) intermediates. The resulting 11b is the mirror image of 11, since the hydride ligand is on the other side of the complex. The corresponding rate limiting TS28−29 is at +21.1 kcal mol−1, which is actually a barrier of +24.5 kcal mol−1 from 28. From 11b, however, the catalytic cycle needs to return to 12, which could happen according to Figure 2. This introduces an increase of the overall barrier. We also searched for a different

the end product being the separated formate + the residual complex. This state lies at +10.4 kcal mol−1, which is between 5 and 13; therefore, our model predicts no barrier for the rearrangement process. It also means that 13 is not a real minimum on the PES after correcting for the ΔGsolv of formate. Without the experimental solvation correction, the energy difference between 5 and its separated fragments is +18.2 kcal mol−1, which would result in an activation barrier for this step. This value would exceed the ΔG = +15.7 kcal mol−1 between 5 and TS13−14. The one-site “flipping” in Scheme 2a is a bit similar to this case because it also involves the breaking of the Ir−Oformate bond and results in an estimated (Figure S11) barrier of +18.5 kcal mol−1. In this case, the solvation correction does not matter since formate does not leave the complex. The calculated barrier here can be rationalized by an entropy decrease that is required to keep the loosely coordinated formate together with the complex in the TS. In the case of the two-site process, starting from 11 instead of 5 seems to be a more rational choice (as seen in Scheme 2b, 5 would give a trans-dihydride product). There are two main steps on this pathway: the rearrangement of the κ2-O,O formate, 11, to a κ2-O,H state, 19, followed by a tilting which has the highest energy TS with +24.5 kcal mol−1. The overall energy span of +17.4 kcal mol−1 is found between 11 and TS19−20, see Figure 4. Compared to a recently investigated iron complex,63 the calculated activation values for the one-site flip are remarkably similar while the two-site values are much lower in the case of our iridium complex. Still, this barrier is somewhat higher than the energy of the formate C−H bond breaking TS13−14. The importance of this pathway is therefore may not be in giving 13 but a structure from which βelimination may occur. This is indeed the case as even though 19 is 3.1 kcal mol−1 higher in energy than 13, the calculated βelimination pathway shown in Figure 4 is only slightly less favorable than the one-site dehydrogenation. It is also important to note that in every formate C−H bond breaking process so far the energies of every TS are lower than TS9−10. 5908


Article

Inorganic Chemistry

Figure 6. Catalytic cycle of the water assisted mechanism involving C−H activated complexes. *Gibbs energy corrections place a few TS below the near lying minima in free energy. This is due to the flat potential energy surface, which is most likely the result of the presence of the weakly bonded small molecules. Electronic energies, imaginary frequencies and IRC calculations show that they are indeed true transition structures. Note that the energy difference between the starting 12 (in the upper right) and the ending ones are the reaction Gibbs free energies of eq 1.

is required to complete eq 1. From 16, a reductive elimination of a hydrogen molecule is possible, then the solvent may add oxidatively to the metal center or protonate the ethyl side chain. The main issue with this is that the hydrogen elimination from 16 has a high barrier (TS at +35.8 kcal mol−1, Figure S12). The more likely scenario for hydrogen release from the cis-dihydride is protonation by the solvent, shown in Figure 7. The chemical model was expanded with two explicit water molecules to better describe TS34−35 since it is the highest energy structure in the catalytic cycle. Without the second solvent molecule, the displacement of the iridium-bonded oxygen atom from its coordination site is significant (Figure S13). Although the addition of the second solvent molecule decreases the entropy of the system, it will not overcompensate the removal of the bond strain energy in the TS. This results in a +5.0 kcal mol−1 lowering of the activation barrier. The figure also shows that bicarbonate production is possible and it does not limit the hydrogen evolution as it happens after the rate limiting protonation step. The energy profile of the calculated mechanism involving the C−H activated complex is shown in Figure 8. The CO2 molecule lost in the 22 → 16 step is reintroduced in the 36 → 37 step so the overall stoichiometry of the cycle remains the same as in eq 1. In the reverse reaction, however, it is unlikely

reductive process, namely, the reductive elimination of a hydrogen molecule. Of the two saturated dihydride complexes, 25 gives a very high TS at +54.0 kcal mol−1, but from 26, hydrogen loss is much more facile. This latter case, shown in Figure 6, required explicit solvent molecules (26w) to “fix” the bicarbonate which leaves the complex before the coordinated H2 forms. Without the water molecules the bicarbonate selfdissociates to coordinated OH and CO2 moieties (we get 24 back) and from there, relaxed potential energy scans show a higher barrier. This water assisted mechanism overall has similar energetics until bicarbonate dissociation (26w), compared to Figure 5. The following hydrogen release can proceed via protonation (27w → 29w) or through hydrogen elimination (from 26w via 32), with barriers lying below TS4−12. These two pathways, however, give back the resting state 12 in different ways. The reductive elimination cycle (right-hand side of Figure 5) results in a barrier of +29.4 kcal mol−1, which is the energy of TS4−12 plus the reaction free enthalpy of the main reaction. The water-assisted protonation cycle on the other hand, offers a considerably lower barrier (+22.4 kcal mol−1 or +25.8 kcal mol−1 if we consider the distance from 28) by skipping redox processes. In the iridacycle pathway, which starts with the oxidative C− H addition of the ethyl side chain, an additional water molecule 5909


Article

Inorganic Chemistry

that the bicarbonate encounters an MLn type complex (11 → 38 → 37 → 36 path) for the following reasons: The concentration of these two species follow an inverse proportionality relation according to eqs 2a−2c and that, in practice, a ratio of between 30 and 40% formate, which coordinates more strongly than bicarbonate, remains in the solution after the hydrogen evolution cycle is finished.40 What seems more likely is a hydrogen insertion into 11 to reach directly a structure like 35, where the coordinated hydrogen molecule is deprotonated by a base, as in eq 3c. There are several possible bases: the coordinated OH− left by bicarbonate decomposition (or simply an OH− ion from the bulk of the solvent), the bicarbonate, giving carbonic acid, or less likely the formate. This latter happens in acidic media,68 while the carbonic acid, which is actually formed through bicarbonate according to molecular dynamics simulations,67 pathway gives high barriers.60 Therefore, a coordinated hydroxide ion is assumed to be the best choice in our model. We also considered a protonation pathway, where the CO2 molecule does not leave the catalyst. However, it is likely that, after dissociating 22, the CO2 molecule is hydrated to bicarbonate faster than the protonation of the dihydride 34 occurs, querying the plausibility of the 36 → 37 step. The seemingly more correct model (Figure S14) involves a concerted TS and a +34.4 kcal mol−1 barrier, which is higher than what was found for the dissociation of the bicarbonate. In practice, the charge/discharge cycles are treated as separate processes as well, each having their respective TOF values. In our case the reported TOFs at 80 °C are 15110 h−1 for formate dehydrogenation (using two equivalent phosphine in excess) and 310 h−1 for hydrogen uptake (1 equiv phosphine excess).40 Converting these to ΔG‡ gives 19.8 and 22.5 kcal mol−1, respectively.69 The second value is close to the experimentally determined activation barrier for bicarbonate

Figure 7. Hydrogen release from 16 via the protonation pathway.

Figure 8. Gibbs energy profile of the iridacycle pathway. 5910


Article

Inorganic Chemistry

Figure 9. Concerted reduction of the bicarbonate together with a table that shows the evolution of the indicated bond lengths. The energies are relative to 39, in which the metal complex is the cis dihydride, 16.

dehydration in water at basic pH (21.5 kcal mol−1). So far, the calculated barriers exceed both, as expected, given that we consider only monophosphine complexes in this work. In the following, we investigate whether there is a more favorable pathway for bicarbonate hydrogenation other than reversing the catalytic cycles. This way, for example, the hydrogen release part of the mechanism presented in Figure 8, together with the rate limiting TS34−35 could be skipped. Here we focus on what happens if 16 hydrogenates the bicarbonate. There are two steps involved, namely, the cleavage of the bicarbonate C−OH bond and a new C−H bond formation. However, multiple possibilities result from the order in which these two steps take place and whether they happen simultaneously or separately. The reaction in which the C−H bond is formed first is estimated to have the highest barrier. We did not try to find the corresponding TS as very fine relaxed PES scans showed a very steep uphill process without well-defined maxima. We did find, however, a high energy intermediate at +25.0 kcal mol−1 relative to the starting structure, where the carbon atom of the bicarbonate is tetracoordinated. From this intermediate, the loss of a hydroxide moiety is facile. The concerted process shown in Figure 9 has similarly high activation barrier, but it comes with a well-defined TS. The third pathway, shown in Figure 10, starts with C−OH bond cleavage in 40. Not surprisingly, the calculated barrier of +12.8 kcal mol−1 strengthens the view that the hydride accepting species is a CO2 molecule and not bicarbonate. The calculated barrier is very close to the forward reaction in eq 2b resulting in a dynamic system where CO2 and bicarbonate can rapidly interconvert. This process can be the source of CO2 for the 36

Figure 10. Energy profile of the bicarbonate hydrogenation where the concerted TS40−41 is split into two separate elementary steps. Numbers with * correspond to the minima structures of the PES.

→ 37 step to regenerate 5 in Figure 8 as long as there are dihydride species in the solution.

■

CONCLUSION In this work we have shown that it is possible to construct a working catalytic mechanism using DFT for the reversible 5911


Article

Inorganic Chemistry

NHC type ligands. We suggest the study of stoichiometric reactions of [Ir(I)(NHC)(η4-COD)(PPh3)]+-type monophosphine complexes with no phosphine excess to experimentally detect C−H activation in alkylimidazolylidene complexes.

formate−bicarbonate transformation where the catalytically active species is an iridium-based NHC complex bearing a C− H activated side chain. The water-soluble precatalyst, which had been studied previously in our research group, was modeled as [Ir(I)(emim)(η4-COD)(PPh3)]+, derived from the actual catalyst containing sulfonated triphenylphosphine ligands by omission of the SO3− substituents. We have found four possible pathways for the catalysis of the formate−bicarbonate equilibrium. Three of them start from the product complex, 12, obtained after the oxidative addition of a solvent molecule to the precatalyst, [Ir(I)(emim)(η4-COD)(PPh3)]+, following the loss of the COD ligand. The difference is in the way hydrogen is released: In the most favorable one, hydrogen is released after protonation by solvent molecules through a barrier of +25.8 kcal mol−1. In the other two, the ratedetermining step is the initialization/regeneration (to 12), the oxidative addition of water. One involves the reductive elimination of H2 from a coordinatively saturated cis-dihydride complex; in the other, hydrogen is released via iridacycle formation with +30.0 and +29.5 kcal mol−1 activation barriers, respectively. In the fourth pathway (+31.8 kcal mol−1), the oxidative addition of the NHC N-ethyl group takes place first and contains only iridacycle species. Considering these results, there are two options that lead to the intramolecular C−H activation. The first one is that there are high energy cyclometalated species that are generated off-cycle or that the resting state of the catalyst is blocked kinetically and the iridacycle pathway, only involving κ2-NHC complexes, dominate. This latter case would have a 21.1 kcal mol−1 barrier, which is interesting, because we show at the start of the discussion we show that “optimal” catalysts for the formate−bicarbonate equilibrium have catalytic cycles with an energy span between 11.5 and 21.5 kcal mol−1. The reason, however, why such C−H activated complexes have been elusive in this system so far is most likely due to the applied phosphine excess used in the model hydrogen battery. Calculations suggest that the formation of the [Ir(I)(emim)(κ1 HCO2)(PPh3)2]+ complex is thermodynamically allowed from the diene species even without prehydrogenation. When two phosphines coordinate to the metal center they block the activation of the ethyl side chain by raising its barrier to +46.3 kcal mol−1. It is, however, also known that these catalysts are active toward transfer hydrogenation reactions without excess phosphine. Based on these results we have performed NMR measurements in DMSO/D2O to find out if C−H activation does takes place in the monphosphine complex. The results show that the methyl end groups of the carbene contain active protons, therefore, C−H activation takes place. Further important findings are that our initial assumptions where the COD ligand is hydrogenated and the monophosphine complex is stable also got confirmed. The hydrogenation of bicarbonate by cyclometalated species has also been studied from a different point of view and it was found that the complex prefers the stepwise C−OH cleavage then C−H bond formation instead of steps where the bicarbonate itself is the hydride acceptor. This finding agrees with the ΔG‡ = +22.5 kcal mol−1 calculated from the experimental TOF as this value is near the activation barrier of bicarbonate dehydration at basic pH. We believe that the mechanism developed here is not restricted only to our C−H activated iridium complex but should also work on any Ir(III) catalyst. Some of the calculated energies are also very similar to the ones found for ruthenium or even iron complexes without

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.8b00382. Considerations about the solvent model, figures and relative energies of the bisphosphine complexes, figures about the effect of helicity and additional solvent molecules, and figures about minor pathways like CMD mechanism and reductive elimination of H2, 13C NMR spectra, and Cartesian coordinates of the structures in the main text (PDF).

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Ferenc Joó: 0000-0001-9457-8038 Mihály Purgel: 0000-0002-9982-8590 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was partially supported by the European Union and the European Social Fund through project Supercomputer, the National Virtual Lab, Grant No.: TÁ MOP-4.2.2.C-11/1/ KONV-2012-0010. The research was supported by the EU and co-financed by the European Regional Development Fund under the projects GINOP-2.3.2-15-2016-00008 and GINOP2.3.3-15-2016-00004.

■

REFERENCES

(1) Grasemann, M.; Laurenczy, G. Formic Acid as a Hydrogen Source − Recent Developments and Future Trends. Energy Environ. Sci. 2012, 5, 8171−8181. (2) Singh, A. K.; Singh, S.; Kumar, A. Hydrogen Energy Future with Formic Acid: A Renewable Chemical Hydrogen Storage System. Catal. Sci. Technol. 2016, 6, 12−40. (3) Hull, J. F.; Himeda, Y.; Wang, W.-H.; Hashiguchi, B.; Periana, R.; Szalda, D. J.; Muckerman, J. T.; Fujita, E. Reversible Hydrogen Storage Using CO2 and a Proton-Switchable Iridium Catalyst in Aqueous Media under Mild Temperatures and Pressures. Nat. Chem. 2012, 4, 383−388. (4) Boddien, A.; Gärtner, F.; Federsel, C.; Sponholz, P.; Mellmann, D.; Jackstell, R.; Junge, H.; Beller, M. CO2-“Neutral” Hydrogen Storage Based on Bicarbonates and Formates. Angew. Chem., Int. Ed. 2011, 50, 6411−6414. (5) Burgess, S. A.; Grubel, K.; Appel, A. M.; Wiedner, E. S.; Linehan, J. C. Hydrogenation of CO2 at Room Temperature and Low Pressure with a Cobalt Tetraphosphine Catalyst. Inorg. Chem. 2017, 56, 8580− 8589. (6) Bertini, F.; Glatz, M.; Gorgas, N.; Stöger, B.; Peruzzini, M.; Veiros, L. F.; Kirchner, K.; Gonsalvi, L. Carbon Dioxide Hydrogenation Catalysed by Well-Defined Mn(i) PNP Pincer Hydride Complexes. Chem. Sci. 2017, 8, 5024−5029. (7) Burgess, S. A.; Appel, A. M.; Linehan, J. C.; Wiedner, E. S. Changing the Mechanism for CO2 Hydrogenation Using SolventDependent Thermodynamics. Angew. Chem., Int. Ed. 2017, 56, 15002−15005. 5912


Article

Inorganic Chemistry (8) Burgess, S. A.; Kendall, A. J.; Tyler, D. R.; Linehan, J. C.; Appel, A. M. Hydrogenation of CO2 in Water Using a Bis(diphosphine) Ni− H Complex. ACS Catal. 2017, 7, 3089−3096. (9) Onishi, N.; Laurenczy, G.; Beller, M.; Himeda, Y. Recent Progress for Reversible Homogeneous Catalytic Hydrogen Storage in Formic Acid and in Methanol. Coord. Chem. Rev. 2017, DOI: 10.1016/ j.ccr.2017.11.021. (10) Sordakis, K.; Tang, C.; Vogt, L. K.; Junge, H.; Dyson, P. J.; Beller, M.; Laurenczy, G. Homogeneous Catalysis for Sustainable Hydrogen Storage in Formic Acid and Alcohols. Chem. Rev. 2018, 118, 372−433. (11) Jeletic, M. S.; Mock, M. T.; Appel, A. M.; Linehan, J. C. A Cobalt-Based Catalyst for the Hydrogenation of CO2 under Ambient Conditions. J. Am. Chem. Soc. 2013, 135, 11533−11536. (12) Boddien, A.; Mellmann, D.; Gartner, F.; Jackstell, R.; Junge, H.; Dyson, P. J.; Laurenczy, G.; Ludwig, R.; Beller, M. Efficient Dehydrogenation of Formic Acid Using an Iron Catalyst. Science 2011, 333, 1733−1736. (13) Mellmann, D.; Barsch, E.; Bauer, M.; Grabow, K.; Boddien, A.; Kammer, A.; Sponholz, P.; Bentrup, U.; Jackstell, R.; Junge, H.; Laurenczy, G.; Ludwig, R.; Beller, M. Base-Free Non-Noble-MetalCatalyzed Hydrogen Generation from Formic Acid: Scope and Mechanistic Insights. Chem. - Eur. J. 2014, 20, 13589−13602. (14) Marcos, R.; Bertini, F.; Rinkevicius, Z.; Peruzzini, M.; Gonsalvi, L.; Ahlquist, M. S. G. Mechanistic Studies on NaHCO3 Hydrogenation and HCOOH Dehydrogenation Reactions Catalysed by a FeII Linear Tetraphosphine Complex. Chem. - Eur. J. 2018, 24, 5366. (15) Bertini, F.; Gorgas, N.; Stöger, B.; Peruzzini, M.; Veiros, L. F.; Kirchner, K.; Gonsalvi, L. Efficient and Mild Carbon Dioxide Hydrogenation to Formate Catalyzed by Fe(II) Hydrido Carbonyl Complexes Bearing 2,6-(Diaminopyridyl)diphosphine Pincer Ligands. ACS Catal. 2016, 6, 2889−2893. (16) Bernskoetter, W. H.; Hazari, N. Reversible Hydrogenation of Carbon Dioxide to Formic Acid and Methanol: Lewis Acid Enhancement of Base Metal Catalysts. Acc. Chem. Res. 2017, 50, 1049−1058. (17) Joó, F.; Tóth, Z. Catalysis by Water-Soluble Phosphine Complexes of Transition Metal Ions in Aqueous and Two-Phase Media. J. Mol. Catal. 1980, 8, 369−383. (18) Zaidman, B.; Wiener, H.; Sasson, Y. Formate Salts as Chemical Carriers in Hydrogen Storage and Transportation. Int. J. Hydrogen Energy 1986, 11, 341−347. (19) Jessop, P. G.; Joó, F.; Tai, C.-C. Recent Advances in the Homogeneous Hydrogenation of Carbon Dioxide. Coord. Chem. Rev. 2004, 248, 2425−2442. (20) Papp, G.; Csorba, J.; Laurenczy, G.; Joó, F. A Charge/Discharge Device for Chemical Hydrogen Storage and Generation. Angew. Chem., Int. Ed. 2011, 50, 10433−10435. (21) Engel, D. C.; Versteeg, G. F.; van Swaaij, W. P. M. Chemical equilibrium of hydrogen and aqueous solutions of 1:1 bicarbonate and formate salts with a common cation. Fluid Phase Equilib. 1997, 135, 109−136. (22) Behrens, M. Chemical Hydrogen Storage by Methanol: Challenges for the Catalytic Methanol Synthesis from CO2. Recycl. Catal. 2015, 2, 78−86. (23) Müller, K.; Brooks, K.; Autrey, T. Hydrogen Storage in Formic Acid: A Comparison of Process Options. Energy Fuels 2017, 31, 12603−12611. (24) Levin, E.; Ivry, E.; Diesendruck, C. E.; Lemcoff, N. G. Water in N-Heterocyclic Carbene-Assisted Catalysis. Chem. Rev. 2015, 115, 4607−4692. (25) Hopkinson, M. N.; Richter, C.; Schedler, M.; Glorius, F. An Overview of N-Heterocyclic Carbenes. Nature 2014, 510, 485−496. (26) Sipos, G.; Dorta, R. Iridium complexes with monodentate Nheterocyclic carbene ligands. Coord. Chem. Rev. 2017, DOI: 10.1016/ j.ccr.2017.10.019. (27) Kühl, O. Functionalised N-Heterocyclic Carbene Complexes; Wiley: Chichester, U.K., 2010.

(28) Crabtree, R. H. Deactivation in Homogeneous Transition Metal Catalysis: Causes, Avoidance, and Cure. Chem. Rev. 2015, 115, 127− 150. (29) Diggle, R. A.; Macgregor, S. A.; Whittlesey, M. K. Computational Study of C−C Activation of 1,3-Dimesitylimidazol-2-Ylidene (IMes) at Ruthenium: The Role of Ligand Bulk in Accessing Reactive Intermediates. Organometallics 2008, 27, 617−625. (30) Jazzar, R. F. R.; Macgregor, S. A.; Mahon, M. F.; Richards, S. P.; Whittlesey, M. K. C−C and C−H Bond Activation Reactions in NHeterocyclic Carbene Complexes of Ruthenium. J. Am. Chem. Soc. 2002, 124, 4944−4945. (31) Crudden, C. M.; Allen, D. P. Stability and Reactivity of NHeterocyclic Carbene Complexes. Coord. Chem. Rev. 2004, 248, 2247−2273. (32) Burling, S.; Mas-Marzá, E.; Valpuesta, J. E. V.; Mahon, M. F.; Whittlesey, M. K. Coordination, Agostic Stabilization, and C−H Bond Activation of N-Alkyl Heterocyclic Carbenes by Coordinatively Unsaturated Ruthenium Hydride Chloride Complexes. Organometallics 2009, 28, 6676−6686. (33) Häller, L. J. L.; Page, M. J.; Erhardt, S.; Macgregor, S. A.; Mahon, M. F.; Naser, M. A.; Vélez, A.; Whittlesey, M. K. Experimental and Computational Investigation of C−N Bond Activation in Ruthenium N-Heterocyclic Carbene Complexes. J. Am. Chem. Soc. 2010, 132, 18408−18416. (34) Burling, S.; Mahon, M. F.; Paine, B. M.; Whittlesey, M. K.; Williams, J. M. J. Reversible Intramolecular Alkyl C−H Bond Activation, Alcohol Dehydrogenation, and Trans−Cis Dihydride Isomerization in Ruthenium N-Heterocyclic Carbene Complexes. Organometallics 2004, 23, 4537−4539. (35) Burling, S.; Paine, B. M.; Nama, D.; Brown, V. S.; Mahon, M. F.; Prior, T. J.; Pregosin, P. S.; Whittlesey, M. K.; Williams, J. M. J. CH Activation Reactions of Ruthenium N-Heterocyclic Carbene Complexes: Application in a Catalytic Tandem Reaction Involving CC Bond Formation from Alcohols. J. Am. Chem. Soc. 2007, 129, 1987− 1995. (36) Jongbloed, L. S.; de Bruin, B.; Reek, J. N. H.; Lutz, M.; van der Vlugt, J. I. Reversible Cyclometalation at RhI as a Motif for Metal− ligand Bifunctional Bond Activation and Base-Free Formic Acid Dehydrogenation. Catal. Sci. Technol. 2016, 6, 1320−1327. (37) Tanabe, Y.; Hanasaka, F.; Fujita, K.; Yamaguchi, R. Scope and Mechanistic Studies of Intramolecular Aliphatic C−H Bond Activation of N-Heterocyclic Carbene Iridium Complexes. Organometallics 2007, 26, 4618−4626. (38) Wheatley, J. E.; Ohlin, C. A.; Chaplin, A. B. Solvent Promoted Reversible Cyclometalation in a Tethered NHC Iridium Complex. Chem. Commun. 2014, 50, 685−687. (39) Horváth, H.; Kathó, Á .; Udvardy, A.; Papp, G.; Szikszai, D.; Joó, F. New Water-Soluble Iridium(I)−N-Heterocyclic Carbene−Tertiary Phosphine Mixed-Ligand Complexes as Catalysts of Hydrogenation and Redox Isomerization. Organometallics 2014, 33, 6330−6340. (40) Horváth, H.; Papp, G.; Szabolcsi, R.; Kathó, Á .; Joó, F. WaterSoluble Iridium-NHC-Phosphine Complexes as Catalysts for Chemical Hydrogen Batteries Based on Formate. ChemSusChem 2015, 8 (18), 3036−3038. (41) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Petersson, G. A.; Nakatsuji, H.; Li, X.; Caricato, M.; Marenich, A.; Bloino, J.; Janesko, B. G.; Gomperts, R.; Mennucci, B.; Hratchian, H. P.; Ortiz, J. V.; Izmaylov, A. F.; Sonnenberg, J. L.; Williams-Young, D.; Ding, F.; Lipparini, F.; Egidi, F.; Goings, J.; Peng, B.; Petrone, A.; Henderson, T.; Ranasinghe, D.; Zakrzewski, V. G.; Gao, J.; Rega, N.; Zheng, G.; Liang, W.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Throssell, K.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Keith, T.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Millam, J. M.; Klene, M.; Adamo, C.; Cammi, R.; Ochterski, J. W.; Martin, R. L.; 5913


Article

Inorganic Chemistry Morokuma, K.; Farkas, O.; Foresman, J. B.; Fox, D. J. Gaussian 09, Revision E.01; Gaussian, Inc., Wallingford, CT, 2016. (42) Becke, A. D. Density-functional Thermochemistry. III. The Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648−5652. (43) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. A Consistent and Accurate Ab Initio Parametrization of Density Functional Dispersion Correction (DFT-D) for the 94 Elements H-Pu. J. Chem. Phys. 2010, 132, 154104. (44) Ross, R. B.; Powers, J. M.; Atashroo, T.; Ermler, W. C.; LaJohn, L. A.; Christiansen, P. A. Abinitiorelativistic Effective Potentials with Spin−orbit Operators. IV. Cs through Rn. J. Chem. Phys. 1990, 93, 6654−6670. (45) Schäfer, A.; Huber, C.; Ahlrichs, R. Fully Optimized Contracted Gaussian Basis Sets of Triple Zeta Valence Quality for Atoms Li to Kr. J. Chem. Phys. 1994, 100, 5829−5835. (46) Tomasi, J.; Mennucci, B.; Cammi, R. Quantum Mechanical Continuum Solvation Models. Chem. Rev. 2005, 105, 2999−3094. (47) Stirling, A. HCO3− Formation from CO2 at High pH: Ab Initio Molecular Dynamics Study. J. Phys. Chem. B 2011, 115, 14683−14687. (48) Zhang, D.; Chen, X.; Liu, H.; Huang, X. Mechanistic Studies on the pH-Controllable Interconversion between Hydrogen and Formic Acid in Water: DFT Insights. New J. Chem. 2015, 39, 8060−8072. (49) Jantke, D.; Pardatscher, L.; Drees, M.; Cokoja, M.; Herrmann, W. A.; Kühn, F. E. Hydrogen Production and Storage on a Formic Acid/Bicarbonate Platform Using Water-Soluble N -Heterocyclic Carbene Complexes of Late Transition Metals. ChemSusChem 2016, 9, 2849−2854. (50) Mazzone, G.; Alberto, M. E.; Sicilia, E. Theoretical Mechanistic Study of the Formic Acid Decomposition Assisted by a Ru(II)Phosphine Catalyst. J. Mol. Model. 2014, 20, 2250. (51) Li, J.; Li, J.; Zhang, D.; Liu, C. DFT Study on the Mechanism of Formic Acid Decomposition by a Well-Defined Bifunctional Cyclometalated Iridium(III) Catalyst: Self-Assisted Concerted Dehydrogenation via Long-Range Intermolecular Hydrogen Migration. ACS Catal. 2016, 6, 4746−4754. (52) Zhou, J. Theoretical Investigation on the Ruthenium Catalyzed Dehydrogenation of Formic Acid and Ligand Effect. Appl. Catal., A 2016, 515, 101−107. (53) Boutadla, Y.; Davies, D. L.; Macgregor, S. A.; PobladorBahamondeb, A. I. Mechanisms of C−H bond activation: rich synergy between computation and experiment. Dalton Trans. 2009, 30, 5820− 5831. (54) Pahls, D. R.; Allen, K. E.; Goldberg, K. I.; Cundari, T. R. Understanding the Effect of Ancillary Ligands on Concerted Metalation−Deprotonation by (dmPhebox)Ir(OAc)2(H2O) Complexes: A DFT Study. Organometallics 2014, 33, 6413−6419. (55) Davies, D. L.; Donald, S. M. A.; Al-Duaij, O.; Macgregor, S. A.; Pölleth, M. Electrophilic C-H Activation at {Cp*Ir}: Ancillary-Ligand Control of the Mechanism of C-H Activation. J. Am. Chem. Soc. 2006, 128, 4210−4211. (56) Davies, D. L.; Donald, S. M. A.; Macgregor, S. A. Computational Study of the Mechanism of Cyclometalation by Palladium Acetate. J. Am. Chem. Soc. 2005, 127, 13754−13755. (57) Crabtree, R. H.; Felkin, H.; Morris, G. E. Activation of Molecular Hydrogen by Cationic Iridium Diene Complexes. J. Chem. Soc., Chem. Commun. 1976, 716−717. (58) Deng, J.; Wen, X.; Li, J. Mechanistic Investigation of Rh(I)Catalyzed Alkyne−isatin Decarbonylative Coupling. Org. Chem. Front. 2017, 4, 1304−1312. (59) Rahaman, S. M. W.; Daran, J.-C.; Manoury, E.; Poli, R. The Cyclooctadiene Ligand in [IrCl(COD)]2 Is Hydrogenated under Transfer Hydrogenation Conditions: A Study in the Presence of PPh3 and a Strong Base in Isopropanol. J. Organomet. Chem. 2017, 829, 14− 21. (60) Marcos, R.; Xue, L.; Sánchez-de-Armas, R.; Ahlquist, M. S. G. Bicarbonate Hydrogenation Catalyzed by Iron: How the Choice of Solvent Can Reverse the Reaction. ACS Catal. 2016, 6, 2923−2929. (61) Ohnishi, Y.; Matsunaga, T.; Nakao, Y.; Sato, H.; Sakaki, S. Ruthenium(II)-Catalyzed Hydrogenation of Carbon Dioxide to

Formic Acid. Theoretical Study of Real Catalyst, Ligand Effects, and Solvation Effects. J. Am. Chem. Soc. 2005, 127, 4021−4032. (62) Ohnishi, Y.; Nakao, Y.; Sato, H.; Sakaki, S. Ruthenium(II)Catalyzed Hydrogenation of Carbon Dioxide to Formic Acid. Theoretical Study of Significant Acceleration by Water Molecules. Organometallics 2006, 25, 3352−3363. (63) Sánchez-de-Armas, R.; Xue, L.; Ahlquist, M. S. G. One Site Is Enough: A Theoretical Investigation of Iron-Catalyzed Dehydrogenation of Formic Acid. Chem. - Eur. J. 2013, 19, 11869−11873. (64) Chen, H.-Y. T.; Wang, C.; Wu, X.; Jiang, X.; Catlow, C. R. A.; Xiao, J. Iridicycle-Catalysed Imine Reduction: An Experimental and Computational Study of the Mechanism. Chem. - Eur. J. 2015, 21, 16564−16577. (65) Thevenon, A.; Frost-Pennington, E.; Weijia, G.; Dalebrook, A. F.; Laurenczy, G. Formic Acid Dehydrogenation Catalysed by Tris(TPPTS) Ruthenium Species: Mechanism of the Initial “Fast” Cycle. ChemCatChem 2014, 6, 3146−3152. (66) Ahlquist, M. S. G. Iridium Catalyzed Hydrogenation of CO2 under Basic conditionsMechanistic Insight from Theory. J. Mol. Catal. A: Chem. 2010, 324, 3−8. (67) Praveen, C. S.; Comas-Vives, A.; Copéret, C.; VandeVondele, J. Role of Water, CO2 , and Noninnocent Ligands in the CO 2 Hydrogenation to Formate by an Ir(III) PNP Pincer Catalyst Evaluated by Static-DFT and Ab Initio Molecular Dynamics under Reaction Conditions. Organometallics 2017, 36, 4908−4919. (68) Xia, G.-J.; Liu, J. W.; Liu, Z.-F. A Comparative Study on the CO 2 hydrogenation Catalyzed by Ru Dihydride Complexes: (PMe3)4RuH2 and (Me2PCH2CH2PMe2)2RuH2. Dalton Trans. 2016, 45, 17329−17342. (69) Kozuch, S.; Shaik, S. How to Conceptualize Catalytic Cycles? The Energetic Span Model. Acc. Chem. Res. 2011, 44, 101−110.

5914


DFT Study on the Mechanism of Hydrogen Storage ... - ACS Publications

Recommend Documents