Deciphering the Selectivity of the Electrochemical CO₂ Reduction to CO by a Cobalt Porphyrin Catalyst in Neutral Aqueous Solution: Insights from DFT Calculations

Abstract

Density functional theory (DFT) calculations were conducted to investigate the cobalt porphyrin‐catalyzed electro‐reduction of CO₂ to CO in an aqueous solution. The results suggest that Co^II−porphyrin (Co^II−L) undertakes a ligand‐based reduction to generate the active species Co^II−L⋅⁻, where the Co^II center antiferromagnetically interacts with the ligand radical anion. Co^II−L⋅⁻ then performs a nucleophilic attack on CO₂, followed by protonation and a reduction to give Co^II−L−COOH. An intermolecular proton transfer leads to the heterolytic cleavage of the C−O bond, producing intermediate Co^II−L−CO. Subsequently, CO is released from Co^II−L−CO, and Co^II−L is regenerated to catalyze the next cycle. The rate‐determining step of this CO₂RR is the nucleophilic attack on CO₂ by Co^II−L⋅⁻, with a total barrier of 20.7 kcal mol⁻¹. The competing hydrogen evolution reaction is associated with a higher total barrier. A computational investigation regarding the substituent effects of the catalyst indicates that the CoPor−R3 complex is likely to display the highest activity and selectivity as a molecular catalyst.

The cobalt(II) porphyrin‐catalyzed electro‐reduction of CO₂ to CO was investigated by DFT calculations regarding the mechanism of the in an aqueous solution. The rate‐determining step of this CO₂RR is the nucleophilic attack on CO₂ by Co^II−L⋅⁻, with a total barrier lower than that of the competing hydrogen evolution reaction. Substituent effects of the catalyst indicate that the CoPor−R3 complex is likely to display the highest activity and CO selectivity as a molecular catalyst.

Introduction

Electrochemical carbon dioxide reduction ^[1] (CO₂RR) has been widely acknowledged as a promising and environmentally benign strategy to transform CO₂ into value‐added C1 products ^[2] such as carbon monoxide (CO). Due to the electrochemical inertness of CO₂, proton‐coupled reduction approaches could contribute to decreasing the required potentials (Scheme 1). However, an obvious concern comes from the unavoidable competitive hydrogen evolution reaction (HER), ^[3] which is thermodynamically favored over CO₂RR since the redox potential of HER is −0.41 V, more positive than that of CO₂RR to CO (−0.52 V). ^[4] Consequently, considerable efforts have been devoted to developing effective catalysts to improve the efficiency and product selectivity of CO₂RR in the past decades.

Scheme 1.

Experimental reduction potentials (vs SHE) at pH=7 in aqueous solutions.

So far, various types of transition metal catalysts (both molecular and heterogeneous), ^[5] such as Fe,[ ^5h , ⁶ ] Co,[ ^5g , ⁷ ] Ni,[ ^6d , ⁸ ] Zn, ^[9] and Cu‐related ^[10] structures, have been reported to catalyze the electro‐reduction of CO₂ to CO. One advantage of homogeneous molecular catalysts is that it is easy to modify their structures to tune the catalytic activity and selectivity.[ ³ , ¹¹ ] Nevertheless, an inevitable drawback is that organic solvents are required in the CO₂RR to diminish the competitive HER and also the degradation of the catalyst. In contrast, heterogeneous catalysts involving transition metal macrocycle units tend to be applied in an aqueous solution. Moreover, strategies such as changing the substituents of the molecular building blocks[ ^5e , ^7b , ¹² ] and introducing support material[ ^7a , ^7b , ¹³ ] are widely employed in the structural modification of the heterogeneous catalysts to achieve high catalytic performance.

In 2015, Chang and coworkers reported an interesting work that used covalent organic frameworks (COFs) involving modified cobalt porphyrin building units as the electrochemical CO₂RR catalyst. ^[14] As shown in Scheme 2a, COF‐367‐Co, in which the iminophenyl‐substituted cobalt‐porphyrin building units are linked through imine bonds with biphenyl groups, displayed a fantastic performance in the electrochemical CO₂ reduction to CO in neutral aqueous solution. Experiments (pH=7.3) indicate a long‐term Faradaic efficiency for CO (FE_CO for a 24‐hour period) of 83 % with the TON up to 3901. Besides, COF‐366‐Co, where the cobalt‐porphyrin units are linked through imine bonds with phenyl groups, shows slightly higher 24 h period Faradaic efficiency for CO (FE_CO=90 %) yet lower electroactivity (TON=1352). Notably, both COFs illustrate much better catalytic capability in catalyzing CO₂RR in an aqueous solution than the molecular catalyst Co(TAP) (TON=794) that bears four aniline substitutes. Clearly, the precise construction of the active sites in the frameworks through modulation of the building units could contribute to the well‐stabilized catalysts that enable prolonged effective CO₂ conversion and minimize the proton‐induced HER.[ ¹⁴ , ¹⁵ ]

Scheme 2.

a) Schematic structures of the cobalt(II)‐porphyrin building units in COFs. Here, only one cobalt core for each catalyst is presented. b) Simplified cobalt(II) porphyrin bearing different substituents, with a total charge of 0.

Inspired by this experimental work, herein, density functional theory (DFT) calculations have been conducted to shed light on the mechanism of the COFs catalyzed electrochemical CO₂RR. To simplify the calculation, only the cobalt(II) porphyrin core unit was included in the study. Firstly, we concentrate on the mechanism of CO₂RR catalyzed with the model catalyst CoPor (Scheme 2b). Different reaction pathways have been evaluated to reveal the most feasible one. The undesirable hydrogen evolution reaction mechanism was also explored. Furthermore, substituent effects on the catalytic ability have been investigated by incorporating four different groups (R1, R2, R3, and R4, as shown in Scheme 2b) onto the CoPor model. Noteworthy, catalyst CoPor−R2 is the molecular Co(TAP), and catalysts CoPor−R3 and CoPor−R4 originate from the core building unit of COF‐366‐Co and COF‐367‐Co. Overall, we expect that the herein‐presented computational study provides a deeper insight into the mechanism of the electrochemical transformation of CO₂ to CO, and also aids in the design of cobalt porphyrin‐based catalysts for CO₂RR.

Results and Discussion

As shown in Figure 1, the model catalyst CoPor is labeled as 1, which has a ground state of a quartet, and the doublet state is calculated to be 8.6 kcal mol⁻¹ higher in energy. Mulliken spin population analysis of ⁴ 1 (Co^II−L) revealed a spin population of 2.76 on Co. The distances between Co and the four coordinated N atoms are 2.05 Å on average. To determine the lowest energy starting point of the energy profile, the formation of various adducts, including 1‐H₂O, 1‐H₂CO_3, 1‐HCO₃ ⁻ , and 1‐CO₃ ²⁻ , have been considered. The corresponding calculated free energies relative to 1 plus each species (H₂O, H₂CO₃ , HCO₃ ⁻ , and CO₃ ²⁻ , respectively) are displayed in Figure 2. Obviously, the axial coordination to catalyst 1 is all endergonic albeit to a different extent (1‐H₂O: 2.2 kcal mol⁻¹; 1‐H₂CO₃ : 8.6 kcal mol⁻¹ _, 1‐HCO₃ ⁻ : 10.3 kcal mol⁻¹; 1‐CO₃ ²⁻ : 12.0 kcal mol⁻¹). Therefore, catalyst 1 was chosen as the starting point in the energy profiles displayed in Figures 4–5.

Figure 1.

Optimized structure of 1 (total charge of 0). Energies relative to the ground state are given in kcal mol⁻¹. Bond distance is given in Ångström. The spin population on Co is shown in red italics.

Figure 2.

Optimized structures of 1‐H₂O (total charge of 0), 1‐H₂CO₃ (total charge of 0), 1‐HCO₃ ⁻ (total charge of −1), and 1‐CO₃ ²⁻ (total charge of −2). Optimizations were implemented in the gas phase, and the final energies were calculated in the aqueous solution. Energies relative to the ground states and the reaction Gibbs free energy change (ΔG) for the generation of each adduct are given in kcal mol⁻¹. Bond distances are given in Ångström. The spin populations on Co are given in red italics. Unimportant H atoms are omitted.

Reduction of CO₂ to CO. To reach the active state towards attacking CO₂, complex 1 is firstly reduced via a one‐electron‐reduction step to form intermediate 2, with a calculated potential of −1.26 V; hence, this step is endergonic by 3.8 kcal mol⁻¹ given an applied potential of −1.10 V. The ground state of 2 (Co^II−L⋅⁻) is a triplet, whereas the quintet, the broken‐symmetry singlet, and the closed‐shell singlet are 2.5, 5.5, and 11.1 kcal mol⁻¹ higher in energy, respectively. At ³ 2, the Mulliken spin population is 2.68 on Co (S _Co=3/2) and −0.68 on the ligand (S_L=1/2), suggesting an antiferromagnetically interacting manner between the Co center and the ligand radical anion. Besides, at pH=7.3, the direct protonation of the metal coordinated N atom of 1 to form 1 pt, the pK _a of which is calculated to be −3.7 (Figure S1), is found to be endergonic by 15.0 kcal mol⁻¹.

Starting from 2, various pathways have been considered for the CO₂RR to produce CO. As shown in pathway A (Figure 3), 2 (Co^II−L⋅⁻) could perform a nucleophilic attack of CO₂ via TS1 to generate 2‐CO₂ , endergonic by 4.3 kcal mol⁻¹. The ground state of TS1 (Figure 4) is a mixture of the closed‐shell singlet and the triplet since the latter is only 1.0 kcal mol⁻¹ higher in energy than the former. The imaginary frequency of TS1 in the singlet state is 173.1i cm⁻¹, where the vibration mode is related to the Co−C bond formation (the distance of Co−C is 2.80 Å). The calculated total barrier of ¹ TS1 is 20.7 kcal mol⁻¹ relative to 1. Complex 2‐CO₂ has a ground state of the closed‐shell singlet; the triplet is 6.3 kcal mol⁻¹ higher in energy. Therefore, a spin crossing must occur in the transformation of 2 to 2‐CO₂ .

Figure 3.

Gibbs energy diagram for the reduction of CO₂ to CO. The energies are given in kcal mol⁻¹. An applied potential of −1.10 V and pH of 7.3 are used as the reference. Core structures of intermediates are given; the ground spin state and total charge are shown within. Relative energies are shown for low‐, intermediate‐, and high‐spin states (in parentheses).

Figure 4.

Optimized structures of important intermediates and transition states in CO₂RR catalysis. Energies relative to the ground states are given in kcal mol⁻¹, and imaginary frequencies (i cm⁻¹) for transition states are shown. Bond distances are given in Ångström. The spin populations on Co and ligand are given in red italics. Unimportant H atoms are omitted.

Afterward, 2‐CO₂ could abstract a proton from H₂CO₃ to generate complex 4 (Co^III−L−COOH), which lies at −0.9 kcal mol⁻¹ relative to 2‐CO₂ . The corresponding protonation transition state TS6 (Figure 4) has a ground state of a singlet, and the triplet is 24.8 kcal mol⁻¹ higher in energy. TS6 has been identified to be associated with a single imaginary frequency of 469.3i cm⁻¹, the vibration mode of which is related to O1−H1 bond formation (the distances of O1−H1 and O2−H1 are 1.15 Å and 1.27 Å, respectively). The total barrier of TS6 is only 10.3 kcal mol⁻¹, indicating a facile process. Next, a one‐electron reduction on the Co center of 4 (Co^III−L−COOH) leads to the generation of 5 (Co^II−L−COOH). Complex 5 prefers to be a quartet, with a Mulliken spin population of 2.71 on Co, while the doublet is 7.8 kcal mol⁻¹ higher. In this reduction step, the Co−C bond distance elongates from 1.88 Å in 4 to 2.09 Å in 5.

Noteworthy, the nucleophilic attack of CO₂ by 3, which is formed by the reduction of 2, was also calculated, and the results are displayed as pathway B (Figure 3). Calculations revealed that the ground state of 3 is a doublet (with the Mulliken spin population of 0.89 on Co), and the quartet is 6.5 kcal mol⁻¹ higher. The 2/3 transition has a potential of −1.82 V, thereby rendering this reduction process endergonic by 16.6 kcal mol⁻¹. Followingly, the cobalt center of 3 performs a nucleophilic attack on CO₂ to generate 3‐CO₂ via TS4. 3‐CO₂ (Figure S2) has a doublet ground spin state. The total barrier of TS4 (ground state of doublet, see Figure S2), however, is calculated to be as high as 34.6 kcal mol⁻¹, thus, such a reaction manner is ruled out.

Additionally, the protonation of 2 prior to the upcoming reaction to produce complex 5 has also been investigated, ^[16] as seen for pathway C (Figure 3). The protonation of 2 via TS2 results in the generation of 2 pt (Co^II−L⋅⁻−H, Figure 5). TS2 has a triplet ground state, where the spin populations on Co and the ligand are 1.10 and 0.86, respectively. The closed‐shell singlet and the open‐shell quintet are 14.9 kcal mol⁻¹ and 5.4 kcal mol⁻¹ higher than the triplet. Frequency analysis of TS2 indicates a sole imaginary frequency of 1087.5i cm⁻¹¹, the vibration mode of which is related to N−H formation. As for 2 pt, calculations suggest a ground state of a mixture of triplet and quintet since the energy difference between them is only 0.6 kcal mol⁻¹. This protonation step ( ³ 2→ ³ 2 pt) is thus calculated to be endergonic by 4.0 kcal mol⁻¹.

Figure 5.

Optimized structures of important intermediates and transition states in CO₂RR catalysis. Energies relative to the ground states are given in kcal mol⁻¹, and imaginary frequencies (i cm⁻¹) for transition states are shown. Bond distances are given in Ångström. The spin populations on Co and ligand are given in red italics. Unimportant H atoms are omitted.

A further reduction of 2 pt affords intermediate 3 pt (Figure 5), with a redox potential of −1.43 V; the ground state of 3 pt is a doublet (spin population of 1.10 on Co), whereas the quartet lies at +5.0 kcal mol⁻¹ higher. The following attack on CO₂ by 3 pt produces 3 pt‐CO₂ (Co^III−L⋅⁺−CO₂). The associated transition state TS7 (Figure 5) is a quartet (Mulliken spin population is 2.31 on Co), and the doublet is 4.7 kcal mol⁻¹ higher. ⁴ TS7 has a single imaginary frequency of 379.2i cm⁻¹, and its vibration mode is related to the Co−C bond formation. The resulting species 3 pt‐CO₂ prefers to be a doublet, with a spin population of 0.29 on Co and 0.84 on the ligand, indicating that the metal center ferromagnetically interacts with the porphyrin moiety. Calculations suggest a total energy barrier of 26.9 kcal mol⁻¹ for TS7. Subsequently, a nearly barrierless intramolecular proton transfer occurs in 3 pt‐CO₂ via TS8 (Figure 5), yielding intermediate 5 (Co^II−L−COOH). During this process, the proton on the metal‐coordinated N atom transfers to the CO₂ moiety, generating the carboxyl group. Alternatively, the direct nucleophilic attack on CO₂ by 2 pt via TS5 to give complex 2 pt‐CO₂ (pathway D) has a total barrier of 28.7 kcal mol⁻¹, even higher than that of TS7 in pathway C.

Taken together, for the transformation of 2 to 5, the total barrier for pathway C (2→2 pt→3 pt→3 pt‐CO₂→5) is more than 6 kcal mol⁻¹ higher than that of pathway A (2→2‐CO₂→4→5), thus excluding the possibility of CO₂RR taking place at the protonated state of 2, namely, 2 pt. In addition, the critical CO₂ addition step preferably takes place at a formal Co(I) center as seen for complex 2 in pathway A.

Thereafter, protonation of 5 on its COOH group by an H₂CO₃ molecule results in the heterolytic cleavage of the C−O bond to produce intermediate 6 (Co^II−L−CO), accompanied by the release of H₂O and HCO₃ ⁻. The related transition state TS9 has a ground state featuring a mixture of the doublet and the quartet state, where the energy difference is only 0.3 kcal mol⁻¹ between both. TS9 has a single imaginary frequency of 430.0i cm⁻¹, the vibration mode of which is consistent with the C1−O1 bond cleavage as well as the H1−O1 bond formation. The distances of C1−O1, O1−H1, and H1−O2 are 1.76 Å, 1.34 Å, and 1.11 Å, respectively. The formation of complex 6 (doublet ground state) from 5 is calculated to be exergonic by 19.5 kcal mol⁻¹. Eventually, the catalytic cycle closes as CO dissociates from the cobalt center of 6 (Co^II−L−CO) to regenerate 1, and this step is calculated to be exergonic by 3.6 kcal mol⁻¹.

Above all, the rate determining step (RDS) for CO₂RR via pathway A is the nucleophilic attack of 2 on CO₂, with a total energy barrier of 20.7 kcal mol⁻¹ (see TS1 in Figure 4).

Reduction of H⁺ to H₂ . As the main competing reaction to the CO₂RR, possible pathways for the hydrogen evolution reaction have also been calculated. The results are provided in Figure 6 (selected optimized structures are shown in Figure 7). In pathway E (Figure 6), the Co‐hydride species 2 pt’ is formed by the protonation of 2 via TS3, with a barrier of 17.3 kcal mol⁻¹ relative to 2 plus H₂CO₃. TS3 has a singlet ground state (the triplet lies at +13.0 kcal mol⁻¹ relative to the singlet); frequency analysis of ¹ TS3 (Figure 7) suggests a single imaginary frequency of 625.1i cm⁻¹, associated with a vibration mode of the Co−H bond formation. The resulting intermediate 2 pt’ (Co^III−H−L) is also a singlet, lying at +2.1 kcal mol⁻¹ relative to 2. Spin population analysis revealed that the formation of 2 pt’ is accompanied by two electrons, one from the metal center and the other from the porphyrin ligand, transferred from the catalyst to the hydrogen.

Figure 6.

Gibbs energy diagram for the release of H₂. The energies are given in kcal mol⁻¹. An applied potential of −1.10 V and pH of 7.3 are used as the reference. Core structures of intermediates are given; the ground spin state and total charge are shown within. Relative energies are shown for low‐, intermediate‐, and high‐spin states (in parentheses).

Figure 7.

Optimized structures of important intermediates and transition states for HER. Energies relative to the ground states are given in kcal mol⁻¹, and imaginary frequencies (i cm⁻¹) for transition states are shown. Bond distances are given in Ångström. The spin populations on Co and ligand are given in red italics. Unimportant H atoms are omitted.

A further one‐electron reduction of the porphyrin ligand in 2 pt’ gives complex 3 pt’ (Figure 7), with a calculated potential of −1.29 V. 3 pt’ (Co^III−H−L⋅⁻), with a doublet ground state, lies at +4.4 kcal mol⁻¹ above 2 pt’. Next, the Co center of 3 pt’ accepts a proton from H₂CO₃ via TS11 (Figure 7) to produce H₂, and regenerates species 1 to participate in the next cycle. Calculations indicate that TS11 is a mixture of the quartet and the doublet due to the small energy difference (1.0 kcal mol⁻¹) between the two states. The energy barrier of TS11 is 8.8 kcal mol⁻¹ relative to 2 pt’, and the total reaction free energy for the HER is −32.5 kcal mol⁻¹. Besides, the metal‐coordinated N atom in 3 pt’ could be protonated by another H₂CO₃ molecule to yield species 3 dpt via TS10 (pathway F in Figure 6), associated with an energy barrier of 5.0 kcal mol⁻¹ higher than that of TS11. Next, 3 dpt undergoes an H−H coupling step (via TS13) to release H₂ and regenerate species 1. Apparently, the reaction manner in pathway F is less favored compared to pathway E.

In parallel (pathway G in Figure 6), intermediate 2 could first get protonated on its N site via TS2, with a total barrier of 14.5 kcal mol⁻¹, to form complex 2 pt. 2 pt is only 1.9 kcal mol⁻¹ higher in energy than 2 pt’. Nevertheless, the reduction of 2 pt to 3 pt (redox potential of −1.43 V), and the following protonation step (via TS12, total barrier of 30.8 kcal mol⁻¹) are disfavored compared to that in pathway E; thus, pathway G is ruled out in this HER. Additionally, the transformation from 2 pt to 2 pt’ was found to be kinetically disfavored as well (for details, see Figure S4).

Based on the above analysis, the RDS for this HER is the transformation of 2 to 2 pt’, with a total energy barrier of 21.1 kcal mol⁻¹ (see TS3), slightly higher than that of CO₂RR (TS1: 20.7 kcal mol⁻¹).

The substitution effects. To explore the effects of substituents on the reactivity and the selectivity of the CO₂RR, catalysts equipped with different substituents (as shown in Scheme 2b) have been computationally studied. The corresponding energy profiles for the reduction of CO₂RR to CO via pathway A (the preferred reaction pathway based on the calculated results of the model reaction) are displayed in Figure S10. The metal‐hydride formation step was also calculated as it is the RDS for the HER. For convenience, the calculated results of critical steps are summarized in Table 1.

Table 1.

Comparison of the redox potentials, reaction free energies (ΔG), and the energy barriers between CO₂RR and HER. ΔG ^ǂ _CO2RR and ΔG ^ǂ _HER represents the total energy barriers of CO₂RR and HER, ΔΔG ^ǂ =ΔG ^ǂ _CO2RR−ΔG ^ǂ _HER.

Catalyst	1→2 [V]	2→2‐CO2 [kcal mol−1]			2→2 pt’ [kcal mol−1]			ΔΔG ≠ [kcal mol−1] (selectivity)
		ΔG	ΔG ≠	ΔG ≠ CO2RR	ΔG	ΔG ≠	ΔG ≠ HER
CoPor	−1.26	+4.3	+16.9	+20.7	+2.1	+17.3	+21.1	−0.4 (66 %)
CoPor−R1	−1.26	+4.5	+16.9	+20.5	+4.9	+17.1	+20.7	−0.2 (58 %)
CoPor−R2	−1.30	+3.0	+17.0	+21.7	+3.6	+18.3	+23.0	−1.3 (90 %)
CoPor−R3	−1.23	+3.3	+16.3	+19.3	+5.1	+19.0	+22.0	−2.7 (98 %)
CoPor−R4	−1.28	+4.2	+17.7	+21.9	+3.8	+18.8	+23.0	−1.1 (87 %)

As shown in Table 1, the redox potentials for the reduction of 1 to 2 barely change when modifying the catalyst with different substituents. The catalytic performances of different catalysts are also quite similar, favoring the production of CO over H₂. Meanwhile, the energy difference (ΔΔG ^≠) between CO₂RR and HER by using catalysts CoPor−R2 (Co(TAP), see above), CoPor−R3, and CoPor−R4 could lead to more than 85 % of selectivity toward CO over H₂ production; the trend of these findings is in line with the experimental observations that CO is the major product when using Co(TAP) and COFs. Especially, catalyst CoPor−R3 shows the highest selectivity toward CO₂RR over HER, since the total energy barrier of CO₂RR is 2.7 kcal mol⁻¹ lower than that of HER, corresponding to around 98 % of CO selectivity. The total barrier for the CO₂RR to CO (CoPor−R3: 19.3 kcal mol⁻¹) is also the lowest one among them. Therefore, CoPor−R3 is likely to show a high performance both in the activity and the CO selectivity when applied as a molecular cobalt porphyrin catalyst in a homogeneous aqueous solution.

Conclusion

In this work, detailed mechanistic investigations on the CO₂RR to CO by CoPor (1) in neutral aqueous solutions were conducted using DFT calculations. Possible reaction pathways involving critical intermediates and transition state structures were considered. On the basis of the calculated results, a catalytic cycle has been established, as shown in Scheme 3. Initially, the Co^II‐porphyrin catalyst 1 undergoes a ligand‐based one‐electron‐reduction step to form intermediate 2 (Co^II−L⋅⁻), which then performs a nucleophilic attack on CO₂ to generate the critical 2‐CO₂ species. Protonation of 2‐CO₂ by H₂CO₃ affords complex 4 (Co^III−L−COOH), followed by a reduction step to give 5 (Co^II−L−COOH). A proton transfer from H₂CO₃ to the COOH group of 5 results in the heterolytic cleavage of the C−O bond to produce intermediate 6 (Co^II−L−CO), accompanied by the release of an H₂O molecule. Finally, CO dissociates from the cobalt center of 6 (Co^II−L−CO) to regenerate 1 to catalyze the next cycle. Calculations indicate that the RDS of this CO₂RR is the nucleophilic attack on CO₂ by intermediate 2 (Co^II−L⋅⁻), with a total energy barrier of 20.7 kcal mol⁻¹. Meanwhile, the mechanism of the competing HER was also studied and the RDS of which is the formation of the metal hydride, as seen TS3 in Figure 7, with a total barrier of 21.1 kcal mol⁻¹.

Scheme 3.

Mechanism of CO₂RR to CO catalyzed by cobalt‐porphyrin catalyst.

Furthermore, a comparison of the total energy barriers of the CO₂RR and HER catalyzed with a series of modified Co^II‐porphyrin catalysts was conducted (Figure S10 and Table 1). The redox potentials for the reduction of 1 to 2 remain almost unchanged when tuning the substituents. The catalytic performances of different catalysts are in the same trend that favors the production of CO over H₂. Hereto, the introduction of the redox non‐innocent porphyrin ligand could possibly help to circumvent the formation of a low‐valent Co center that leads to the formation of the metal‐hydride, thus contributing to the CO₂RR to produce CO. Importantly, among all the catalysts considered, CoPor−R3 is likely to show a higher catalytic capability, both the activity and the CO selectivity, than the rest ones when applied as a molecular cobalt porphyrin catalyst in homogeneous aqueous solution.

Computational Methods

All DFT calculations presented herein were conducted using the B3LYP‐D3 functional ^[17] (the default Grimme‐D3 dispersion correction) together with the SMD continuum solvation model, ^[18] as implemented in the Gaussian 16 program package. ^[19] As for the geometry optimization, the 6‐31G(d,p) basis sets were used for C, N, O, and H elements, and the SDD pseudopotential was used for Co. ^[20] To verify the features of intermediates and transition states and to obtain the Gibbs free energy corrections at 298.15 K, frequency analysis was implemented at the same level as the geometry optimization. On the basis of optimized geometries, more accurate electronic energies were obtained by single‐point calculations employing the larger basis set 6‐311+G(2df,2p) for C, N, O, and H (Co SDD). Spin population analysis was conducted using Multiwfn. ^[21]

The redox potentials were calculated using the standard hydrogen electrode (SHE, experimental value of 4.28 V) as the reference, ^[22] corresponding to an electron affinity of 98.72 kcal mol⁻¹. The applied potential in the titled reaction is −1.10 V vs. SHE. For a proton, the Gibbs free energy in the gas phase is −6.3 kcal mol⁻¹, and the experimental absolute solvation energy for a proton is −264.0 kcal mol⁻¹ in water. ^[22] Considering the experimental pH of 7.3 in the buffer, the total Gibbs free energy of a proton is −280.3 kcal mol⁻¹.

Additionally, a concentration correction of 1.9 kcal mol⁻¹ at 298.15 K, derived from the Gibbs free energy change from 1 atm (24.5 L mol⁻¹ for an ideal gas) to 1 m (1 mol L⁻¹ in solution), was also added to all species except for water and carbonic acid. For water as solvent (standard concentration of 55.56 mol L⁻¹), the correction is 4.3 kcal mol⁻¹; for carbonic acid, the concentration correction is −3.9 kcal mol⁻¹. ^[23]

Conflict of interest

There are no conflicts to declare.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Cao Y.-C., Shi L.-L., Li M., You B., Liao R.-Z., ChemistryOpen 2023, 12, e202200254.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.