Reversing the Rate versus Overpotential Scaling in Electrocatalytic Hydrogen Evolution Induced by Electrostatic Field Engineering at Remote Sites

Abstract

The rate versus overpotential relationship arising from molecular scaling effects is a key challenge in developing efficient electrocatalysts. Circumventing these unyielding scaling relationships is a persistent challenge in the field of electrocatalysis. Escaping this iron law of rate versus overpotential requires judicious modulation, which can balance the electrophilic and nucleophilic/basic characteristics of the electrocatalyst, which is rarely straightforward. Here, we report local electrostatic field-persuaded inverse scaling relationships of the electrocatalytic hydrogen evolution reaction (HER). For this purpose, we report six mononuclear Ni^II complexes 1–6, where 4, 5, and 6 contain positively charged quaternary ammonium moieties at remote sites. Without perturbing the primary coordination sphere, this molecular scaffold allows the installation of a local positive charge, which would increase the electrophilic character, favoring electron transfer processes. This is reflected by the anodic shift in the reductive redox potentials from neutral to charge-appended complexes (1 to 6). Since the catalytically relevant anionic/reduced species are achieved easily via electrostatic stabilization by the local positive charge, this leads to a lower overpotential. Density functional theory (DFT) calculations reveal that the appended local positive charge facilitates the r.d.s. step by ground-state destabilization of the starting intermediate, which leads to a lower activation barrier and facilitates the HER. Thus, the present study demonstrates the optimum balance between the electrophilic character and basicity of the electrocatalyst, assisted by the local positive charge, leading to inverse scaling relationships. To the best of our knowledge, this is the first demonstration of a positive electrostatic field-assisted enhanced HER activity with a gradual decrease in overpotential.

Introduction

The electrocatalytic hydrogen evolution reaction (HER) has attracted rapidly growing attention in the context of carbon-free energy. H₂, being an energy carrier, is considered a promising alternative energy source to traditional fossil fuels. , Owing to their better natural abundance, 3d-transition metal complexes are often tested for the HER. − Apart from the reaction rate, another key factor for the electrocatalytic ability of a catalyst is the overpotential (η), which is the additional potential required in addition to the thermodynamic potential to observe the electrocatalytic process. A higher overpotential would lead to a higher use of electrical energy. Thus, for an effective electrocatalyst, high catalytic activity or maximum turnover frequency (TOF_max) at a lower overpotential (Figure ) would be ideal for practical applications, as it would minimize the use of electrical energy. However, catalysts with a larger TOF_max usually have a higher overpotential not only for electrocatalytic proton reduction − but also for O₂ reduction , and CO₂ reduction. ,

1.

Representative plot of log­(TOF_max) vs overpotential (η) showing normal scaling relationships (red circles). An efficient catalyst (higher TOF_max with lower overpotential) emerges off the normal scaling relationship correlation line, as shown by the blue circle.

The above ubiquitous trend of TOF_max vs overpotential arises due to scaling relationships, which involve a trade-off between the kinetic and thermodynamic performance parameters controlled by the nucleophilic vs electrophilic properties of the electrocatalyst. When structural modification is designed to improve the electrophilicity of an electrocatalyst to favor the electron transfer step for lowering the overpotential, these changes become detrimental to the basicity/nucleophilicity of the electrocatalyst. Consequently, the kinetics of the chemical step, i.e., the protonation or electrophile binding step, is inhibited. This opposite behavior of the electrophilic and nucleophilic character and its consequences on electron transfer and chemical steps result in a lower TOF_max for an electrocatalyst with a lower overpotential. This prevalent iron law of scaling relationships in electrocatalysis presents a difficult challenge for the development of catalysts exhibiting high TOF_max at low overpotentials. Despite significant studies on the electrocatalytic HER, − breaking the molecular scaling relationship trend is very rare. To the best of our knowledge, only one report by Wiedner and co-workers demonstrated the breaking away of the scaling relationship for the electrocatalytic HER by combining factors like electron-withdrawing ligands and controlled structural dynamics using the ligand design and controlling the viscosity of the solvent. Compared to multicomponent control, any strategy with single-component control would be convenient to adopt. Given the scarcity of the reverse trend of the TOF–overpotential for the HER, more studies are required to determine other possible strategies for the inverse scaling relationship.

An underexplored concept for the HER is the utilization of local electrostatic effects to tune the TOF and overpotential. The use of local electrostatic interactions stems from the fact that local electric fields are considered to be crucial factors for enhanced enzymatic activity and selectivity. − Enormous computational studies have supported the significant impact of electrostatic effects on reactivity. −

This has stimulated several important experimental studies on the effect of electrostatic interactions on reactivity across different fields. ,,− Savéant and co-workers demonstrated improved reactivity and lowering of the overpotential for the CO₂ reduction reaction (CO₂RR) by introducing a local positive charge to iron tetraphenylporphyrin. The improved CO₂RR activity was attributed to the stabilization of the negative charge in the Fe⁰–CO₂ adduct by the local positive charge via through-space Coulombic interactions. Mayer and co-workers explored the effect of the electrostatic environment arising due to the position of the positive charge on the electrocatalytic oxygen reduction reaction (ORR) and CO₂RR. This work revealed that a higher cationic charge in the proximity of the metal center can exhibit a significant effect on the kinetics and energetics. These cationic charges facilitate catalysis by favoring substrate binding, which changes the nature of the catalytically active species. McCrory and co-workers reported improved CO₂RR activity with a simultaneous decrease in the overpotential by integrating electrostatic effects, extended conjugation, and electron-withdrawing effects into a single cobalt catalyst. Cao and co-workers demonstrated the effect of the position and number of positively charges decorated on iron porphyrin on the CO₂RR. An increased number of positive charges and their cis-isomers were found to be more active than the trans-isomer via the cooperation between the charged units ensured by ligand rigidity. The same group also reported enhanced CO₂RR activity and selectivity by appending a guanidyl functional group over iron porphyrin. Upon protonation, the appended guanidyl moiety gains a positive charge, which provides electrostatic stabilization to the CO₂-bound intermediate bearing a negative charge. Synergistic porosity and local charge effects were utilized by the Chang group to enhance the photochemical CO₂RR activity. Local electrostatic effects were also found to be effective in controlling the selectivity of the ORR across a vast pH range. The cationic moeity and its location in the second coordination sphere of cobalt corroles were found to be crucial for improved electrocatalytic ORR activity and selectivity. Intermolecular electrostatic interactions can also play a significant role in facilitating the water oxidation reaction (WOR). Yang and co-workers explored local electrostatic effects by installing redox inactive metal ions in the secondary binding pockets for C–H bond oxidation and N–N bond formation reaction. These reports emphasize the impact of local charge on small-molecule activation and catalysis. However, it is always challenging to disentangle the effects of pendant charged moieties on the potential, pK _a, PCET pathways, etc. Specifically, the potential and pK _a effects are opposite, which compromise the TOF against the overpotential. Thus, judicious placing of the charge moiety is required to achieve better TOF without compromising overpotential.

Compared to the CO₂RR and ORR, local electrostatic effects on the electrocatalytic HER are less explored. Only two studies are known to explore the local electrostatic interactions for the electrocatalytic HER, where the inhibition of HER is observed with the installation of local positive electrostatic effects. , Yang and co-workers reported that the installation of Lewis acidic (LA) metal ions (Na⁺ or Ba²⁺) into a prepositioned ancillary pocket around the Pd-complex (Scheme ) inhibits the electrocatalytic HER with an increase in the overpotential up to 260 mV compared to an analogous Pd complex without LA.

1. ChemDraw Representation of Previous Reports Exploring the Effect of Local Positive Charge on the HER.

Cao and co-workers recently demonstrated that the incorporation of positively charged functional moieties into the second coordination sphere of Cu-porphyrin (Scheme ) inhibits the electrocatalytic HER. In both of these cases, the inhibition of HER is attributed to the electrostatic repulsion between protons and the appended cationic moiety(s).

However, we speculate that without affecting the primary coordination sphere, if the positive charge can be installed at a suitable remote place, then it would increase the electrophilicity of the catalyst without compromising the basicity, as the primary coordination sphere is not affected. This increased electrophilicity would favor the reduction processes by stabilizing the LUMO. Thus, the reduced (i.e., more electron-rich) intermediates can be achieved at a more favorable anodic potential. The appended positive charge would also favor the stability of the reduced species via electrostatic stabilization. Since the electron-rich intermediate forms readily, it can more effectively overcome electrostatic repulsion during protonation. As a result, protonation can be facilitated at a more anodic potential, which can lead to a reduced overpotential and enhanced HER catalytic activity. This can drive an inverse molecular scaling relationship between the rate and overpotential by leveraging the positive electric field induced by the appended positive charge.

The above concept motivated us to design a series of square-planar Ni^II complexes 1–6 (Scheme ), where, by keeping the primary coordination sphere unaffected, the local positive charge is installed (4–6) at a remote site. Our aim is to investigate (i) whether the positive electrostatic field, controlled from the remote site, can result in a higher TOF at a lower overpotential for HER, and (ii) the impact of proximity (distance from Ni) of local positive charge on TOF and overpotential. For this purpose, we chose bidentate N,S-coordinating ligands because of the following features (Scheme ): (i) ligands can be functionalized so that a charged moiety can be installed as −NMe ₃ ⁺ or −NMe ₂ Bz ⁺ at a remote site of the nickel complex without perturbing the primary coordination sphere, which provides an ample scope to systematically study the actual impact of the positive electrostatic effect on the HER (Scheme ) controlled from the remote site keeping the similar primary coordination sphere effects; (ii) the present design also provides scope (5 and 6 in Scheme b) to study the effect of the proximity of the positive charge on the HER; (iii) the presence of a Schiff-base moiety and soft-donor S can stabilize the Ni^I state; (iv) the anionic nature after deprotonation of the ligand can provide sufficient stability to Ni^II; (v) the redox noninnocent nature of the Schiff-base moiety (−CN−) can lead to metal–ligand cooperativity, where one electron can be taken by the nickel center and the other electron would be taken by the ligand Schiff-base moiety.

2. (a) (1-4) to Investigate the Electrostatic Effect of Appended Positive Charge on the HER in the Present Work; (b) (5 and 6) to Explore the Distance Effect of Appended Positive Charge on the HER in the Present Work.

Electrochemical studies revealed an anodic shift in the redox potential upon the installation of positive charges. For the positively charged-appended complexes (4–6), electrostatic potential (ESP) surface analysis revealed a dominant positive electrostatic potential, which led to the stabilization of the LUMO, accounting for the anodic shift in the redox potential. The anodic potential shift corroborates our concept that, because of the presence of a positive electrostatic effect, the reduced species, i.e., electron-rich species, can be achieved easily compared to analogous complexes without an appended positive charge. Further electrocatalytic HER studies reveal a rarely observed inverse TOF_max–overpotential trend from 1 to 6. The electrostatic field-assisted inverse scaling relationship for the HER is hitherto not known. Density functional theory (DFT) calculations reveal that the increased reactivity of the local charge-appended complexes is attributed to the ground-state destabilization of the intermediate (reactant complex, RC) involved in the r.d.s. step.

Results and Discussion

The ligand HL¹ was synthesized by the condensation of 3-fluorobenzaldehyde with 4,4-dimethyl-3-thiosemicarbazide (Figure S3). The reaction of HL¹ with Ni­(OAc)₂·4H₂O in a 2:1 ratio resulted in square-planar [Ni^II(L¹)₂] (1), which was structurally characterized by X-ray diffraction (Figure ). The reaction of propyl amine with HL¹ led to the synthesis of HL² (Figure S4). [Ni^II(L²) ₂] (2) was synthesized from HL² following a procedure similar to that of 1. HL³ (Figure S5) was obtained by reacting N,N-dimethyl ethylene diamine with HL¹. [Ni^II(L³)₂] (3) was synthesized using HL³, following a procedure similar to that of 1 and 2. Structural characterization of 3 revealed a dangling −NMe₂ moiety (Figure ) far from the metal center, which is ideal for methylation or benzylation, leading to local positive charge installation (Scheme a,b) without perturbing the primary coordination sphere. Methylation of 3 with methyl iodide resulted in the formation of [Ni^II(L_NMe3 ³)₂]­(I)₂ (4) with two appended positive charges as −NMe₃ ⁺ (Scheme a). Similarly, benzylation of 3 with benzyl bromide led to the formation of [Ni^II(L_NMe2Bz ³)₂]­(Br)₂ (5) with two appended positive charges as −NMe₂Bz⁺ (Scheme b). In order to increase the distance of the appended positive charge from the nickel center, HL⁴ (Figure S6) was synthesized by reacting N,N-dimethyl-1,3-propanediamine with HL¹. The reaction of Ni­(OAc)₂·4H₂O with HL⁴ followed by benzylation with benzyl bromide provided [Ni^II(L_NMe2Bz ⁴)₂]­(Br)₂ (6), where two positive charges were appended as −NMe₂Bz⁺ and the positively charged moieties were one spacer (−CH₂−) apart compared to 5 (Scheme b).

All complexes (1–6) were found to be diamagnetic, as revealed from their diamagnetic ¹H NMR spectra (Figures S7–S12). The ESI–Ms­(+ve) spectra support the 1:2 ratio of nickel and ligand for each complex (Figures S13–S18). Complexes 1–3 were successfully characterized by X-ray diffraction. We could not obtain the X-ray structure of 4–6. However, ¹H NMR (Figures S10–S12) and ESI–Ms­(+ve) analyses (Figures S16–S18) confirm their composition. We attempted to crystallize the above complexes with different counteranions. The crystal quality of [Ni^II(L_NMe2Bz ³)₂]­(BPh₄)₂ (5′) was not sufficiently good for publication; however, the structure (Figure S19) clearly indicated benzylation on the appended amine, resulting in the installation of an appended positive charge. Due to the better solubility of 5 compared to 5′, electrocatalytic studies were conducted using 5. For 4 and 6, changing the counteranion did not help to obtain any structural information.

Molecular Structures

The molecular structure of 1, as revealed by X-ray studies (Figure ), depicts a mononuclear-square-planar Ni^II center with two deprotonated ligands (L¹).

2.

Molecular Structures of 1–3 in (a–c), respectively.

The asymmetric unit of 1 is composed of one Ni^II center, which is coordinated by one imine nitrogen (Ni–N1 = 1.9284(18) Å) and one thiosemicarbazone thiolate (Ni–S1 = 2.1777(6) Å) from the ligand (Table S2). The X-ray structure of 2 (Figure ) also manifests a square-planar Ni^II structure with N2S2 coordination and geometrical parameters similar to those of 1 (Table S2). The molecular structure of 3 also depicts a mononuclear-square-planar geometry around the Ni^II center. Each of the two deprotonated ligands coordinates to the Ni^II center via one imine nitrogen with Ni–N1 = 1.9421(13) Å and one thiosemicarbazone thiolate with Ni–S1 = 2.1810(4) Å, completing the N2S2 coordination to the metal center (Table S2). X-ray structural analysis of 3 confirms the installation of the N,N-dimethyl amine moiety in the second coordination sphere, which would be ideal for local positive charge installation via methylation or benzylation of the dangling N,N-dimethyl amine moiety.

Electrochemical Study

Complexes 1–6 display two reductive responses in dimethylformamide (DMF). The first reductive response (E _1/2 ¹) is observed in the range of −1.51 V to −1.66 V vs Fc⁺/Fc (Figure and Table ), and the second reductive response occurs in the range of −2.34 V to −2.50 V vs Fc⁺/Fc, as determined from DPV (Figure S20 and Table ) for all complexes. The one-electron reduction of these complexes with decamethylcobaltocene (Co­(Cp*)₂) provides EPR signal characteristics of the Ni^I center (Figures and S22). − These Ni^I-based EPR signals confirm the first reductive response to be metal-centered reduction corresponding to the Ni^II/Ni^I redox couple in the range of −1.51 V to −1.66 V for all the complexes.

3.

Cyclic voltammograms of 1–6 in DMF.

1. Electrochemical Parameters for 1–6 .

complex	E 1/2 1 (V vs Fc+/Fc)	E 1/2 2 (V vs Fc+/Fc)	η (V)	diffusion layer TOFmax (s–1)	k 1 (k FOWA) (M–1 s–1)	TON
1	–1.64	–2.49	1.1	0.58 ± 0.04 × 105	0.98 ± 0.03 × 1011	40.8 ± 2.5
2	–1.66	–2.50	1.08	0.75 ± 0.03 × 105	1.03 ± 0.04 × 1011	47.9 ± 5.5
3	–1.62	–2.48	0.77	5.51 ± 0.29 × 105	1.73 ± 0.06 × 1012	154.4 ± 6.2
4	–1.52	–2.36	0.84	4.85 ± 0.52 × 105	1.83 ± 0.07 × 1012	149.3 ± 4.0
5	–1.51	–2.34	0.84	5.28 ± 0.11 × 105	3.99 ± 0.14 × 1012	157.1 ± 4.2
6	–1.56	–2.39	0.86	2.75 ± 0.06 × 105	0.77 ± 0.03 × 1012	111.9 ± 1.9

^aTOF_max was determined from bulk electrolysis without stirring.

^b k ₁ (second-order rate constant for the first protonation step) was evaluated from FOWA.

^cTheoretical TON was determined from bulk electrolysis with stirring.

4.

EPR spectra of [4]⁻. Simulated parameters: g _|| = 2.25 and g _⊥ = 2.06.

DFT calculations on one-electron reduced species of 1–6 manifest a major spin population (∼0.9) on the metal center (Figures and S23), confirming the presence of Ni^I centers in [1]⁻ to [6]⁻. Thus, DFT calculations also support the fact that during the first reduction, the Ni^II center is reduced to Ni^I. The optimization of two-electron reduced species [1]^2– to [6]^2– reveals ligand-centered reduction during the second reduction, resulting in a [Ni^I(L)₂ ^3–.]^2– electronic configuration for [1]^2– to [6]^2–, where the nickel center exists as Ni^I with one unpaired electron and another unpaired electron resides over the ligand (Figure S24). The electronic structure ([Ni^I(L)₂ ^3–.]^2–) of the two-electron reduced forms ([1]^2– to [6]^2–) is similar to that of previously reported square-planar Ni^II complexes with a N₂S₂ coordination and thiosemicarbazone-based Schiff-base containing ligands. ,, [Ni^I(L)₂ ^3–.]^2– can have a high-spin triplet state or a broken-symmetry (BS) singlet state. Both the triplet and broken-symmetry (BS) singlet states were found to be comparable in energy (<1 kcal/mol) and significantly stabilized compared to the closed-shell singlet state (Table S3). The elongated C–N bond lengths (∼1.35 Å) in these two-electron reduced species ([1]^2– to [6]^2–) further confirm that the second reductive response (E _1/2 ²) corresponds to the ligand-centered reduction (imine bond).

5.

Spin density plots for [1]⁻, [3]⁻, and [4]⁻.

Importantly, with the installation of the appended positive charge, there is an anodic shift in the reduction potentials up to 150 mV and 160 mV for the first and second reductive responses, respectively, compared to the analogous complex without the appended positive charge. To understand the effect of the appended local positive charge, the electrostatic potential (ESP) map was explored for 1–6. The ESP demonstrates the repulsive or attractive force experienced by a point positive charge, depending on the electron richness or deficient nature of a structural portion. The electron-rich part of the structure exerts a negative potential (blue color in Figure ), while the electron-deficient segment displays a positive potential (red color in Figure ). The ESP surfaces for 1–6 are displayed in Figure , which reflects the dominant positive electrostatic potential (intense red color in Figure ) for 4, 5, and 6 having appended cationic moieties. The positive electrostatic potential corresponds to an electron-deficient nature, which would favor the reduction process by stabilizing the LUMO, resulting in an anodic shift in the redox potential. , In fact, the anodic shift in the first reductive response (E _1/2 ¹) is linearly correlated with the LUMO energy of 1–6 (Figure S25). Similarly, the redox potential for the second reductive response (E _1/2 ²) is also linearly correlated with the LUMO­(β) energy of [1]⁻–[6]⁻ (Figure S26). Thus, the anodic shift in the reduction potential can be attributed to the local electrostatic potential caused by the appended positive charge, which results in the stabilization of the LUMO energy.

6.

Electrostatic potential surfaces for 1–6. Color bars represent the range of electrostatic potential values (blue for the most negative potential and red for the most positive potential).

Electrocatalytic Hydrogen Evolution

The electrocatalytic proton reduction ability of 1–6 was explored in dry DMF under an argon atmosphere using acetic acid (AcOH) as the proton source. In the presence of AcOH, all the complexes exhibited electrocatalytic waves with enhanced currents (red line in Figure S27) in the CV experiments. With only AcOH and without complexes, no significant current was observed (gray line in Figure S27). The electrocatalytic current was found to be linearly dependent on the complex concentration, indicating a first-order dependence of the complex on the electrocatalytic HER (Figures S28–S29). Bulk electrolysis experiments with the complexes in the presence of excess AcOH generated a noteworthy amount of H₂, as detected by GC (Figure S30). However, bulk electrolysis with the same amount of AcOH in the absence of complexes did not generate any notable H₂ formation (gray line in Figure S30). Thus, the CV and bulk electrolysis experiments confirm electrocatalytic proton reduction by 1–6. The Faraday efficiencies of complexes (1–6) were measured to be in the range of 62% to 78% (Table S4), which is in the range observed for earlier reported thiosemicarbazone-based nickel complexes. ,,,

With increasing amounts of AcOH, the electrocatalytic current increases for all the complexes (Figures and S31). To assess the impact of the local positive charge on electrocatalytic proton reduction, we overlaid the forward CV scans (Figure ) in the presence of a fixed amount of AcOH (72.5 mM) for all complexes. Analysis of these scans reveals that the catalytic peak positions as well as E _cat/2 shifts anodically for complexes with appended positive charges (4, 5, 6) as compared to 1 and 2, which lack the appended positively charged moiety. Thus, with the installation of local positive charges, the overpotential decreases substantially (∼up to 260 mV: η = 1.1 V for 1, η = 1.08 for 2, η = 0.84 V for 4, η = 0.84 V for 5, and η = 0.86 V for 6) without compromising the HER activity (Figure ). The decrease in the overpotential of these complexes follows a linear trend with the anodic shift in the reductive responses (Figure S32) caused by the appended positive charge.

7.

CV scans at different concentrations of AcOH (1.45 mM – 72.5 mM) in the presence of 0.3 mM complex 3 in DMF.

8.

Forward CV scans in the presence of 72.5 mM AcOH for 1–6 (forward scans are shown for clarity).

Next, we carried out a foot-of-the-wave analysis (FOWA) − to assess the rate constant for the first protonation step from the electrocatalytic CV waves. Utilizing FOWA, the I _c/I _p vs 1/(1+ exp­[(F/RT)­(E–E _cat/2)]) plot is linearly correlated near the foot of the electrocatalytic CV wave (see the Supporting Information for details). Using the slope of this linear correlation, the observed rate constant (k _obs) for the first protonation process was evaluated (Figures S33–S38). The k _obs values varied linearly with the AcOH concentration (Figure S39), indicating k _obs = k ₁[H]⁺, where k ₁ is the second-order rate constant for the first protonation step. Thus, the second-order rate constant (k ₁) for the first protonation step was assessed from the slope of the k _obs vs [H⁺] linear plots. The k ₁ values were found to be 0.98 × 10¹¹ M^–1 s^–1 for 1, 1.03 × 10¹¹ M^–1 s^–1 for 2, 1.73 × 10¹² M^–1 s^–1 for 3, 1.83 × 10¹² M^–1 s^–1 for 4, 3.99 × 10¹² M^–1 s^–1 for 5, and 0.77 × 10¹² M^–1 s^–1 for 6. One notable observation from the FOWA was that the I _c/I _p vs 1/(1+exp­[(F/RT)­(E–E _cat/2)]) plots began to deviate from linearity at small x-axis values, leading to a curvature feature (Figures S33–S38) at higher x-axis values. Mechanistically, these curved I _c/I _p vs 1/(1+exp­[(F/RT)­(E–E _cat/2)]) plots together with the linear correlation at the steep rise portion of these plots (Figures S33–S38) are typically consistent with a faster initial protonation step. This implies that the first protonation step cannot be the rate-determining step. ,

Next, the diffusion layer TOF_max values for the overall HER were estimated from the controlled potential electrolysis (CPE) experiment (Figure S40 and Table S4) by following the previously established procedure (see Supporting Information for details). ,− TOF_max values from CPE (Tables and S5) were found to be 0.58 ± 0.04 × 10⁵ M^–1 s^–1 for 1, 0.75 ± 0.03 × 10⁵ M^–1 s^–1 for 2, 5.51 ± 0.29 × 10⁵ M^–1 s^–1 for 3, 4.85 ± 0.52 × 10⁵ M^–1 s^–1 for 4, 5.28 ± 0.11 M^–1 s^–1 for 5, and 2.75 ± 0.06 × 10⁵ M^–1 s^–1 for 6 (extrapolating for 1.0 M acid concentration). These TOF_max values correspond to the overall rate constant k _global. Since k _global is several orders of magnitude less than the k ₁ values (rate constant for the first protonation step obtained from FOWA), this further confirms that the first protonation process is not the rate-limiting process. Thus, utilizing the TOF values from CPE, catalytic Tafel plots (Figure ) were constructed. The catalytic Tafel plots reveal that the installation of a local positive charge in the second coordination sphere (4–6) leads to an increase in reactivity with a substantial decrease in the overpotential compared to analogous complexes without an appended positive charge (1 and 2). Thus, judicious utilization of an internal local positive electrostatic field can be an effective strategy to reverse the general rate–overpotential relationship for the HER. As 3 shows a TOF_max similar to those of the charged analogues (4 and 5), it indicates that the appended −NMe₂ moiety in 3 most likely becomes protonated during the later steps of the HER and gains some charged character (discussed later).

9.

Catalytic Tafel plot for 1–6.

Stability Test

Ultraviolet–visible (UV–vis) measurements of 1–6 were carried out in the presence of excess AcOH to probe the stability of 1–6 under acidic conditions similar to the electrocatalytic HER. Under acidic conditions, no substantial change is observed in the UV–vis spectra for all the complexes (Figure S41), which confirms the stability of 1–6 under excess AcOH conditions. Next, we examined the stability of 1–6 under electrocatalytic CV conditions. For this purpose, we first performed a rinse test after the CV scan to probe the probable complex degradation during the CV scan under acidic conditions. As depicted by the gray line in Figures and S42, the CV experiments after the rinse test did not produce any noteworthy current. This observation suggests that the observed electrocatalytic current is not due to any electrode-deposited heterogeneous species formed during CV scans, because in that case, a significant amount of current would have been observed in the CV after the rinse test. This indicates the homogeneous nature of the observed electrocatalytic proton reduction (i.e., by the dissolved complex in solution). Notably, during multiple sweeps in CV under catalytic conditions, the electrocatalytic wave did not show any change (Figure S43), which also supports the stability of the complexes during CV scans.

10.

Post bulk electrolysis and after the CV rinse test on the working electrode in DMF with 72.5 mM AcOH for 5.

Next, to eliminate the possible formation of any catalytically active heterogeneous species on the working electrode surface during bulk electrolysis, we performed a rinse test after bulk electrolysis. After the bulk electrolysis rinse test, the CV scan did not generate any notable current (blue lines in Figures and S42). Further, after rinsing these electrodes after the bulk electrolysis of complexes with AcOH (72.5 mM), we carried out bulk electrolysis again with 72.5 mM AcOH in a fresh DMF solution without the presence of complexes. These bulk electrolysis experiments with rinsed electrodes did not produce any notable charge (red line in Figure S44) or noteworthy amounts of H₂ (red line in Figure S45). Thus, the above rinse test experiments after bulk electrolysis also confirm that the electrocatalytic HER activity is due to the complex dissolved in solution, i.e., homogeneous in nature and not from any electrode-deposited heterogeneous material. Before and after the bulk electrolysis of the complexes with excess AcOH, there was no substantial change in the UV–vis spectral features of the complexes (Figure S46). This observation further confirms the stability of the complexes during bulk electrolysis over time and discards the possibility of complex decomposition during coulometry conditions.

Impact of Local Positive Charges: Inverse Scaling Relationship

There are several approaches involving the analysis of CV and bulk electrolysis data to obtain qualitative and quantitative information on HER activity. We discuss all these approaches to understand and establish the consequences of appending positively charged moieties on the HER activity. First, we focused on analyzing the reactivity parameters for charged-appended complexes 4–6 as compared to 1 and 2, which lack a local positive charge. The rationalization of the reactivity of 3 is provided toward the end of this section. To obtain a qualitative idea of the effect of the local positive charge on the electrocatalytic HER activity–overpotential correlation, we plotted the I _c/I _p (I _c is the catalytic peak current and I _p is the peak current of the complex in the absence of AcOH) against the overpotential (Figure S47). As depicted in Figure S47, I _c/I _p increases for charged analogues (4–6) with a simultaneous substantial decrease in the overpotential as compared to 1 and 2 lacking remote charge. This is a reverse trend for the commonly observed rate–overpotential relationship.

To further confirm the higher HER activity of complexes with a local positive charge (4–6), we explored bulk electrolysis (with stirring) data at a fixed potential of −2.3 V (Figure S48) for 4–6 and 1 and 2. The turnover number (TON) estimated from the bulk electrolysis (with stirring) experiment (Tables and S6) further supports the higher activity of 4–6 (Figure ) compared to 1 and 2, lacking a local positive electrostatic potential. This result provides compelling evidence that an oriented electric field caused by a proximal positive charge can enhance the HER activity.

11.

Plot of TON for the respective complexes.

Finally, we explored the bulk electrolysis without stirring experiments near the respective catalytic half-wave potentials of all complexes (1–6), as this approach has often been used in the literature to determine the diffusion layer TOF_max values for the overall HER. ,− The analysis of TOF values (Tables and S5) revealed an 8–9 times increase in the TOF with charge installation. When log­(TOF_max) was plotted against the overpotential for the present set of complexes (1–6), an inverse scaling trend (R ² = 0.96) was observed (Figure ). The normal scaling relationship would follow the dotted blue line in Figure , where a decrease in TOF values is expected with a decrease in the overpotential. Our complexes follow the solid gray line in Figure , which shows an increase in TOF values with a simultaneous decrease in the overpotential spanning over a range of 330 mV (0.77 to 1.1 V). The substantial range of overpotential (330 mV) and six data points establish a reasonable inverse relationship, if not a perfectly inverse linear fit. It is worth mentioning that only two reports are known to explore the local electrostatic field for the HER, and both reported the inhibition of HER activity with a concomitant increase in the overpotential with the installation of a local positive charge. , In the present set of complexes, higher reactivity and lower overpotential were observed (Figure ) for complexes with a local positive charge 4–6 (and 3 with an appended basic site, which is discussed later) compared to complexes lacking an appended positive charge (1, and 2).

12.

Inverse scaling relationship (solid gray line, R ² = 0.96) for the electrocatalytic HER by the present set of complexes (1–6). The dotted blue line represents the regular trend of log­(TOF_max) vs overpotential. The dotted blue line is not a fitting line and is only a representative line.

As discussed earlier, the positive electrostatic field caused by the appended quaternary amine stabilizes the LUMO energy. Notably, log­(TOF_max) shows a linear trend with the β-LUMO energies of [1]⁻, [2]⁻, and [4]⁻–[6]⁻ (Figure ), where the HER activity increases with β-LUMO stabilization. This increase in reactivity with the β-LUMO stabilization of one-electron reduced species supports the fact that HER activity is controlled by the ease of generation of two-electron reduced species. This is further supported by the linear plot of log­(TOF_max) vs E _1/2 ² (Figure S49) as well as by the fact that the catalytic half-wave potential for these complexes is similar to their corresponding second reductive response E _1/2 ² (Figure S31). The local appended positive charge in 4–6 creates a positive electrostatic potential (Figure ), which favors the formation of two-electron reduced species, thereby significantly decreasing the overpotential. DFT calculations further reveal that the local electrostatic potential in 4–6 further decreases the transition state free energy barrier of the rate-determining step via ground-state destabilization (see the Mechanistic Aspects of the HER section), and thus higher TOF_max values are observed for 4–6.

13.

Plot of log­(TOF_max) vs β-LUMO energy. The solid gray line represents a linear fit (R ² = 0.98) with the data of [1]⁻, [2]⁻, [4]⁻, [5]⁻, and [6]⁻. The blue circle with a solid circle represents the predicted log­(TOF_max) value (according to the linear fit depicted by the gray line with [1]⁻, [2]⁻, [4]⁻, [5]⁻, and [6]⁻ data points) corresponding to the β-LUMO energy of [Ni^I(L³)­(HL³)], where one appended amine (−NMe₂) moiety is protonated after the first electron transfer. The solid purple circle represents the experimental log­(TOF_max) value for 3.

Typically, regular scaling relationships with molecular catalysts exhibit decreased electrocatalytic activity with a lowering of the overpotential. − Circumventing this relationship to achieve higher activity at lower overpotential has been a difficult task in the molecular electrocatalysis field. In the present work, a linear trend showing more HER activity with a simultaneous decrease in overpotential (inverse scaling) due to the stabilization of LUMO, as well as an anodic shift in the redox potential of complexes caused by a local positive electrostatic field, supports the fact that it is possible to overcome the electrostatic repulsion for protonation by favoring the formation of reduced species via installation of a local positive charge at a remote site without perturbing the primary coordination sphere. An inverse scaling relationship for the electrocatalytic HER assisted by a positive electrostatic field has not yet been reported.

Next, we focused on understanding the increased activity of 3, which has an appended amine moiety (−NMe₂). We noticed that 3 has a similar TOF_max value as that of charged analogues with similar spacers, i.e., 4 and 5. This indicates the possibility of the appended basic amine moiety in 3 being protonated at some stage of the electrocatalytic cycle to form appended −NHMe₂ ⁺. Once the appended amine is protonated, it will bear a local positive charge as −NHMe₂ ⁺ and would favor the reduction process. Initial protonation of 3 (i.e., before the first electron transfer) is not possible, as we did not observe any UV–Vis changes with the addition of excess AcOH to 3 (Figure S41). Also, the first reductive response (E _1/2 ¹) is not affected by AcOH (Figure S50). Thus, this indicates that the protonation of appended amines occurs at a later stage of the HER, most likely after the first electron transfer. Protonation after the first electron transfer is supported by the fact that the catalytic half-wave potential for 3 is substantially shifted anodically (300 mV) to the second reductive response (E _1/2 ²) of 3 in the absence of AcOH. If our hypothesis of 3 behaving like a charged species after the first electron transfer is correct, then its reactivity should be controlled by the linear fit (Figure ) of the β-LUMO energies of [1]⁻, [2]^−, and [4]⁻–[6]⁻ with log­(TOF_max). To prove that we optimized the protonated form of 3 after one-electron reduction, i.e., [Ni^I(L³)­(HL³)], where one appended amine (−NMe₂) moiety is protonated after the first electron transfer. From the β-LUMO energy of [Ni^I(L³)­(HL³)] (Figure S51), the predicted log­(TOF_max) value (blue circle with a solid circle in Figure ) according to the linear fit in Figure is close to the experimental value (purple filled circle in Figure ). We also considered the possible protonation of the dangling amine after two electron transfers. In this case, the predicted log­(TOF_max) deviates significantly from the experimental value (Figure S52). Thus, the high reactivity of 3 is due to the possible protonation of the appended amine after the first electron transfer to develop a local positive electric field, which would favor the second electron transfer, resulting in a decrease in the overpotential and increase in the HER activity, like other charged analogues 4 and 5.

Mechanistic Aspects of the HER

In the presence of excess acid (AcOH), the complexes did not exhibit any change in their UV–vis spectroscopic features (Figure S41). This rules out the possibility of initial protonation prior to electron transfer under acidic conditions. The first reductive response (Ni^II/I redox couple) for all complexes remained unaffected by the addition of AcOH (Figure S50), which indicates discrete electron transfer without the involvement of protons during the first electron transfer process. As discussed previously, the first reductive response corresponds to the Ni^II to Ni^I reduction, as confirmed by EPR (Figures and S22). The optimized structures of one-electron reduced species showed a metal-centered spin population, which further supports nickel-centered reduction (Figures and S23).

After the first electron transfer, the resulting monoanionic Ni^I species can be protonated to form Ni^III-hydride species. However, the formation of the Ni^III-hydride was found to be destabilized by 22.8 to 27.2 kcal/mol as compared to Ni^I species (Figures S53–S58) for all the complexes (energy of Ni^I species + AcOH was compared with Ni^III-hydride + OAc^–). Thus, the significant endergonic nature of this reaction precludes the prospect of Ni^III–hydride formation. Also, the possibility of a bimolecular HER pathway involving the coupling of two Ni^III–hydrides can be ruled out by the observed first-order dependence of the complex concentration (Figure S28) for the electrocatalytic HER.

After the formation of monoanionic Ni^I species, further one-electron reduction leads to the dianionic species. This second electron transfer can correspond to Ni^I to Ni⁰ reduction, or it could be ligand-centered reduction. To understand this second electron transfer process, we analyzed the β-LUMO of the monoanionic Ni^I species, which was found to be located on the imine moiety of the ligand (Figure S59). This supports ligand-centered (imine bond) reduction during the second electron transfer process, leading to dianionic species with a general composition of [Ni^I(L)₂ ^3–.]^2– electronic configuration. The elongated C–N bond lengths (∼1.35 Å) in [Ni^I(L)₂ ^3–.]^2– further corroborate the imine bond reduction. Inspection of the Mulliken spin density for [1]^2––[6]^2– also confirms the [Ni^I(L)₂ ^3–.]^2– electronic configuration, where the nickel center remains as Ni^I with one unpaired electron, and another unpaired electron is located on the ligand (Figure S24). The high-spin triplet state (S = 1) and broken-symmetry (BS) singlet state (S = 0) for [Ni^I(L)₂ ^3–.]^2– are comparable in energy (<1 kcal/mol, Table S3). We first discuss the subsequent reaction steps for the S = 1 state. The involvement of the BS-singlet state is discussed later.

Once the dianionic [Ni^I(L)₂ ^3–.]^2– is formed, there could be protonation at the metal center to form Ni^II-hydride or protonation of ligand nitrogen. We computed the transition state energy barrier for all these possible protonation sites with [4]^2– (S = 1). As shown in Figure S60, the protonation of reduced imine nitrogen involves a significantly lower activation free energy barrier (3.5 kcal/mol) compared to that of metal center protonation (19.6 kcal/mol). Thus, the computational study suggests that the first protonation will lead to ligand center protonation involving transition state TS1, leading to intermediate 4A.

Then, the proton from ligand NH is transferred to the nickel center to form the Ni^II–hydride intermediate (4B) via TS2. The free energy barrier (S = 1) for the conversion of 4A, i.e., [Ni^I{(HL^–.)­(L^–)}]⁻ to 4B, i.e., [Ni^II(H)­(L^–)₂]⁻ was calculated to be 16 kcal/mol. The involvement of Ni–hydride intermediate for the HER was further experimentally supported by inverse KIE using AcOD (Figures and S61), as observed in several earlier literature reports involving nickel/metal-hydride intermediate for HER. ,,− Then, protonation of Ni^II–hydride in 4B leads to the release of H₂ via TS3 (Figure ). In this electrocatalytic HER, Ni^II–hydride formation via the transfer of a proton from the N–H center of the ligand toward the metal center via TS2 is the rate-determining step. Importantly, TS2 is product-like, i.e., it is closer to Ni^II–hydride. This internal proton transfer transition state (TS2) leads to a change in hybridization at the metal center during the partial formation of a relatively shorter and stiffer metal–hydride bond within the transition state than in the reactant. Eventually, this stiffer bond would lead to higher vibrational frequencies and ZPE in the transition state. , Upon isotopic substitution, the absolute ZPE decreases appreciably, providing greater stability to the corresponding transition state. − Consequently, the activation energy for D is less than that of H, resulting in an inverse kinetic isotope effect (k _D > k _H).

14.

Plot of I_c/I_p vs the square root of concentration of AcOH/AcOD for 4.

15.

Reaction coordinate for the HER starting from [4]^2– (S = 1).

The possibility of thiosemicarbazone-N protonation during the first protonation of [4]^2– was not considered as the transition state energy barrier for proton migration from thiosemicarbazone-N to the nickel center was very high (41.37 kcal/mol, Figure S62). The reaction coordinate for the HER from [4]^2– (S = 1) is displayed in Figure . Next, we explored the possible involvement of the BS-singlet state (S = 0) of [4]^2– in the HER. Protonation of the reduced imine nitrogen involves a similar activation free energy barrier for both the triplet and BS-singlet states (Figure S63). However, the subsequent r.d.s. step, i.e., proton migration from ligand N–H to the nickel center, was found to be associated with a significantly higher activation barrier (Figure S64) for the BS-singlet state (25.8 kcal/mol) as compared to the triplet state (16.0 kcal/mol). Given the higher activation barrier of the r.d.s. step, further reaction coordinates were not explored with the BS-singlet state. Following this observation, the HER reaction coordinates from triplets [1]^2–, [2]^2–, [3]^2–, [5]^2–, and [6]^2– are displayed in Figures S65–S69. It should be noted that HER reactivity via the triplet state is theoretically proposed and could not be verified by structural and/or spectroscopic characterization.

Rationalization of the Higher Reactivity of Complexes with a Local Positive Charge

Inspection of the reaction coordinate (Figures and S65–S69) reveals that internal proton migration from the reduced imine NH (1A–6A) to the Ni center to form the Ni^II–hydride intermediate (1B–6B) involving TS2 is the rate-determining step (r.d.s.), as it involves the highest energy barrier compared to other steps involving TS1 or TS3. Thus, we compared the free energy barrier of this r.d.s. step for all the complexes in Figure , where the energy values are w.r.t. the catalysis initiating two-electron reduced form for the respective complexes. For 3, as established earlier, the dangling amine is protonated after the first electron transfer and hence bears a local positive charge. Therefore, 3A in Figure bears one protonated dangling amine.

16.

Comparison of the r.d.s, step free energy barrier for all the complexes. Here, the free energy values are w.r.t. the catalysis, initiating the two-electron reduced form of the respective complexes.

As shown in Figure , the transition state free energy barrier for the charge-appended complexes is lower compared to that of the neutral analogues (1–2), which is consistent with the higher HER activity of 3–6 as compared to 1 and 2. Also, 6 (where the charge is located at a longer distance) exhibits a slightly higher energy barrier than 3–5, which is consistent with the slightly lower TOF_max of 6 compared to 3–5.

Importantly, the r.d.s. step starting intermediate (reactant complex), i.e., the ligand NH protonated intermediate (3A, 4A, 5A, 6A) for charged-appended analogues is more destabilized (w.r.t. the catalysis initiating respective two-electron reduced form) as compared to the respective intermediate (1A and 2A) for neutral complexes. It should be noted that this ground-state destabilization of the reactant complex (RC) ranges within 1.4–4.3 kcal/mol, whereas the transition state destabilization varies within 0.3–1.6 kcal/mol across all the complexes (Figure ). This implies that the change in the activation barrier is mainly controlled by the destabilization of reactant complex/starting intermediate of the r.d.s. step. This destabilization of the intermediate/reactant complex (RC) leads to a smaller activation barrier (Scheme ) for 3–6 as compared to 1 and 2. The ground-state destabilization of the ligand protonated intermediate can be attributed to the local electrostatic repulsion between the appended positive charge and the NH proton.

3. Schematic Presentation of the Local Electrostatic Effect-Assisted Lower Activation Barrier of the R.D.S. Step Via Ground-State Destabilization of the RC (Reactant Complex).

Thus, the appended local positive charge facilitates the r.d.s. step, i.e., internal proton migration from the ligand NH to the nickel center via electrostatic repulsion between the appended positive charge and the NH proton, resulting in ground-state destabilization of the r.d.s. starting intermediate. This leads to a lower activation barrier and facilitates the HER. Lowering of the activation barrier via destabilization of the reactant complex assisted by a local electric field is postulated by Shaik and co-workers for the H atom abstraction reaction. ^a For HER, a similar study is not known yet.

Conclusions

In the present work, we report six mononuclear complexes (1–6) to demonstrate the inverse scaling relationship for the electrocatalytic HER caused by an oriented local positive electric field tuned from a remote site. Out of these six complexes, 4–6 possess a preinstalled positive charge as quaternary nitrogen −NMe₃ ⁺ or −NMe₂Bz⁺ at the remote site, which imparts an oriented positive electric field. The key findings of this work are as follows:

• (i)With the installation of local positive charges, a substantial anodic shift (∼150 mV) in the redox potential was observed, caused by the stabilization of the LUMO.
• (ii)For the charged analogues (4–6), ESP surface analysis revealed a dominant positive electrostatic potential, which leads to the stabilization of the LUMO, accounting for the anodic shift in the redox potential.
• (iii)The electrocatalytic HER study of the six complexes revealed that the HER activity increases with the installation of the local positive charge. An Increase in HER activity with positive electrostatic effects has been unprecedented. The increased reactivity of the local charge-appended complexes is attributed to the ground-state destabilization of the intermediate involved in the r.d.s. step, which lowers the activation barrier.
• (iv)With increasing distance (spacer) of the local charge from the metal center, the reactivity does not drop significantly and still remains higher than that of the complexes without an appended positive charge. Thus, increasing the spacer distance of the local charge does not have a significant impact.
• (v)In contrast to earlier reports, , the complexes with local positive charges not only exhibit higher activity, but also have lower overpotential, which leads to an inverse scaling relationship (Figure ). The increase in HER activity with a simultaneous decrease in the overpotential for the charged analogues is attributed to the favoring of reductive responses and destabilization of the intermediate (RC) involved in the r.d.s. step (Figures , –, Scheme , Figures S32, and S49) by the oriented local positive electric field caused by the remote positive charge. To the best of our knowledge, an inverse scaling relationship with molecular complexes assisted by a local electric field for the electrocatalytic HER has not yet been reported. Our study emphasizes that the strategic deployment of an internal positive electrostatic field via suitable molecular design can drive the inverse trend of rate vs overpotential in the HER by promoting the formation of catalytically active reduced species and facilitating internal proton transfer via ground-state destabilization.

Materials and Methods

Synthesis, experimental details, methods, and measurement details are provided in the Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.6c00424.

• Experimental and computational methods; details of electrochemical data analysis, characterization, control experiments, and experimental data sets in the form of tables and plots (PDF)

§.

Department of Chemistry, National Institute of Technology Agartala, Agartala, Jirania, Tripura 799046, India

S.K.B. conceptualized and supervised this research. A.S. performed the synthesis, experiments, and computational study. S.R. performed the experiments during the revision of the manuscript. S.K. performed the EPR measurements under the supervision of SP. J.D. performed X-ray and NMR data collection. A.S. and S.K.B. analyzed the experimental and computational results. S.K.B. wrote the manuscript with feedback from all authors.

The authors declare no competing financial interest.

This paper is dedicated to Professor Rabindranath Mukherjee