
==== Front
Proc Natl Acad Sci U S A
Proc Natl Acad Sci U S A
PNAS
Proceedings of the National Academy of Sciences of the United States of America
0027-8424
1091-6490
National Academy of Sciences

38261614
202315362
10.1073/pnas.2315362121
datasetDatasetresearch-articleResearch ArticlechemChemistry410
Physical Sciences
Chemistry
Boosted hydrogen evolution kinetics of heteroatom-doped carbons with isolated Zn as an accelerant
Li Yang a https://orcid.org/0000-0002-5380-7460

Zuo Shouwei a
Wei Fen b https://orcid.org/0000-0002-9292-8708

Chen Cailing c https://orcid.org/0000-0003-2598-1354

Zhang Guikai d
Zhao Xiaojuan d
Wu Zhipeng a
Wang Sibo b https://orcid.org/0000-0003-2656-9169

Zhou Wei e
Rueping Magnus a
Han Yu c
Zhang Huabin huabin.zhang@kaust.edu.sa
a 1 https://orcid.org/0000-0003-1601-2471

aKing Abdullah University of Science and Technology Catalysis Center, Physical Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal 23955, Saudi Arabia
bState Key Laboratory of Photocatalysis on Energy and Environment, College of Chemistry, Fuzhou University, Fuzhou 350116, People’s Republic of China
cAdvanced Membranes and Porous Materials Center, Physical Sciences and Engineering Division, King Abdullah University of Science and Technology, Thuwal 23955, Saudi Arabia
dBeijing Synchrotron Radiation Facility, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049, People’s Republic of China
eDepartment of Applied Physics, Tianjin Key Laboratory of Low Dimensional Materials Physics and Preparing Technology, Faculty of Science, Tianjin University, Tianjin 300072, People’s Republic of China
1To whom correspondence may be addressed. Email: huabin.zhang@kaust.edu.sa.
Edited by Di-Jia Liu, Argonne National Laboratory, Lemont, IL; received September 6, 2023; accepted December 9, 2023 by Editorial Board Member Peter J. Rossky

23 1 2024
30 1 2024
23 7 2024
121 5 e231536212106 9 2023
09 12 2023
Copyright © 2024 the Author(s). Published by PNAS.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).

Significance

Carbon-based single-atom materials are important electrocatalysts for hydrogen evolution reaction (HER). While considerable research has been dedicated to investigating the HER activity and kinetics on the metal center, little attention has been given to the contribution of the peripheral alien atom–carbon bonding environment, yet the relationship between the peripheral environment of metal center and the HER activity, kinetics and stability has not been clearly recognized. Here, we developed a carbon-based single Zn-atom catalyst with asymmetric ZnN4S1 configuration. The Zn's electron injection effect leads to fast proton-feeding kinetics on neighboring atoms, and the self-adaption structure evolution of ZnN4S1 moiety leads to the catalyst being highly stable and active. This work provides insight into the coordination design of single-atom materials.

Carbon-based single-atom catalysts, a promising candidate in electrocatalysis, offer insights into electron-donating effects of metal center on adjacent atoms. Herein, we present a practical strategy to rationally design a model catalyst with a single zinc (Zn) atom coordinated with nitrogen and sulfur atoms in a multilevel carbon matrix. The Zn site exhibits an atomic interface configuration of ZnN4S1, where Zn's electron injection effect enables thermal-neutral hydrogen adsorption on neighboring atoms, pushing the activity boundaries of carbon electrocatalysts toward electrochemical hydrogen evolution to an unprecedented level. Experimental and theoretical analyses confirm the low-barrier Volmer–Tafel mechanism of proton reduction, while the multishell hollow structures facilitate the hydrogen evolution even at high current intensities. This work provides insights for understanding the actual active species during hydrogen evolution reaction and paves the way for designing high-performance electrocatalysts.

single-atom catalysis
asymmetric-coordination design
electron injection effect
hydrogen evolution
activity origin
King Abdullah University of Science and Technology (KAUST) 501100004052 140144 Yang LiShouwei ZuoFen WeiCailing ChenGuikai ZhangXiaojuan ZhaoZhipeng WuWei ZhouMagnus RuepingYu HanHuabin Zhang
==== Body
pmcAmidst the urgent global imperative to steer away from conventional fossil fuels, hydrogen (H2) has emerged as a remarkably promising alternative distinguished by its unparalleled energy density and remarkable superiority to produce zero emissions (1–3). Electrochemical water electrolysis, a key technology for green H2 fuel supplies, encounters a formidable obstacle in the form of a scarcity of catalysts capable of efficaciously catalyzing the intricate hydrogen evolution reaction (HER) (4, 5). While platinum-group-based materials have shown unrivaled HER performance, their large-scale application is hindered by their high cost and limited abundance (6, 7). Consequently, there is a pressing demand of searching for cost-efficient and highly active alternatives to propel the hydrogen economy forward.

One such alternative that has garnered significant attention is carbon-based single-atom catalysts (CSACs). These catalysts employ elements such as nitrogen (N), oxygen (O), sulfur (S), phosphorus (P), and others to anchor dispersed metal atom, offering unique electronic properties and high atom-utilization efficiency (8–11). While considerable research has been dedicated to investigating the HER activity and reaction kinetics on the metal center of CSACs, little attention has been given to the contribution of the peripheral alien atom–carbon (C) bonding environment (12–14). The electronic coupling between the isolated metal center and coordinated atoms results in the electron redistribution and hence substantial modifications in electronic orbitals (15–17). However, in the case of standard planar symmetrical metal-N4 structure, the simplicity of its structural design results in significantly different Gibbs free energy change of the hydrogen intermediate (ΔGH∗) of continuously adsorbed H (18–23). This disparity poses challenges for the HER process, which involves two proton–electron transfer reactions occurring at a single site or adjacent two sites (24, 25). As a consequence, these limitations increase the potential barriers in the reaction processes and consequently hamper the catalytic activity of HER. Therefore, breaking the symmetric electron distribution of the metal-N4 moiety via electronic perturbation is highly desirable to achieve a fast proton-feeding kinetics.

Here, we investigate the influence of near-range coordination of a single zinc (Zn) atom on the activity of HER by combining experimental observations and theoretical calculations. The synthesized carbon-based Zn single-atom catalyst comprises an optimized ZnN4S1 moiety incorporated into a N and S co-doped multilevel carbon matrix (Zn1/NS-MCM). Our findings demonstrate that this catalyst exhibits exceptional HER activity, characterized by a significantly reduced Tafel slope and a rapidly increasing current densities at more negative potentials, surpassing those of conventional carbon electrocatalysts. Theoretical computations further reveal that the electron injection effect of Zn leads to a more thermally neutral hydrogen adsorption-free energy on adjacent atoms and a low-barrier Volmer–Tafel mechanism for the proton reduction process. This work sheds light on the role of asymmetric structures in facilitating fast proton-feeding kinetics and provides valuable insights into the coordination design of CSACs for efficient hydrogen production and beyond.

Results

Materials Synthesis and Characterization.

The schematic illustration describes the synthetic process of the Zn1/NS-MCM (Fig. 1). Zn-based zeolitic imidazolate framework (ZIF-8) polyhedrons synthesized by a co-precipitation method are directly used as the precursor (26). The successful preparation of the ZIF-8 precursor is confirmed by the X-ray diffraction (XRD) technique (SI Appendix, Fig. S1). The subsequent examination of the materials through scanning electron microscopy (SEM) and transmission electron microscopy (TEM) reveals their uniform polyhedral morphology, characterized by a smooth surface and solid structure (SI Appendix, Fig. S1). The ZIF-8 polyhedrons, once collected, are uniformly dispersed in a solution consisting of tannic acid (TA) mixed with ethanol and water in a 1:1 volume ratio. During this process, a gradual formation of an amorphous network shell takes place around the ZIF-8 scaffold, resulting from the coordination reaction between TA and ZIF-8. The subsequent chemical etching of this complex structure leads to the formation of yolk–shell ZIF-8@TA-Zn polyhedral nanostructures.

Fig. 1. Schematic illustration of the formation of Zn1/NS-MCM. Step I: Etching of ZIF-8 by TA to form ZIF-8@TA-Zn. Step II: Calcination treatment with sulfur to form ZnS@NS-MCM. Step III: Etching of ZnS@NS-MCM to obtain Zn1/NS-MCM.

Further characterization of ZIF-8@TA-Zn by XRD technique reveals the residue of ZIF-8 cores (SI Appendix, Fig. S2). SEM examinations further confirm that the overall polyhedral morphology is maintained in the integrated structure, but the surface becomes rough (SI Appendix, Fig. S2 B and C). The formation of yolk–shelled structure can be unambiguously examined from the cracked particle (SI Appendix, the Inset of Fig. S2B). A comparison between the original ZIF-8 polyhedron and the ZIF-8 core within the TA-Zn shell reveals that the core's surface becomes rough due to continuous dissolution during the etching process (27). TEM confirms a significant void space between the ZIF-8 core and the TA-Zn shell (SI Appendix, Fig. S2D). Additionally, a high-resolution TEM (HRTEM) image confirms the amorphous nature of the TA-Zn shell (SI Appendix, Fig. S3). The yolk–shell structure of ZIF-8@TA-Zn is further supported by high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) along with elemental mapping images (SI Appendix, Fig. S4). The presence of N in the TA-Zn shell can be attributed to the inclusion of organic ligands during the coordination reaction between TA and ZIF-8.

The sulfuration-thermochemical treatment enables the conversion of Zn components in the as-synthesized ZIF-8@TA-Zn yolk–shell structure to lattice confined Zn and ZnS decorated onto N, S-co-doped MCM (ZnS@NS-MCM). An acid etching method is employed to remove all the ZnS components, while the lattice-confined Zn species can be effectively retained within the carbon matrix, resulting in the formation of Zn1/NS-MCM catalyst. The XRD pattern confirms that Zn1/NS-MCM exhibits typical diffraction peaks of graphitic carbon (SI Appendix, Fig. S5A). SEM images reveal that the Zn1/NS-MCM retains the polyhedral shape inherited from ZIF-8@TA-Zn (Fig. 2 A–C and SI Appendix, Fig. S5B). The inner carbon cores, which feature a hollow structure, can be observed from the cracked particle and TEM image (Fig. 2 C and D). The polyhedral carbon shell has a wall thickness of approximately 30 nm (Fig. 2 D and E). The measured interlayer spacing of carbon is about 0.335 nm, which is consistent with graphite (Fig. 2 F and G) (28). Through the aberration-corrected HAADF-STEM image, individual Zn atoms are distinctly observed throughout the carbon matrix (Fig. 2 H and I), without the formation of clusters and particles. Even distribution of N, S, and Zn is evident in elemental mapping images of the multilevel carbon skeleton (Fig. 2 J–L). The multilevel carbon structure exposes more active sites and ensures a higher instantaneous electrolyte concentration during HER (29). For comparison, Zn1/N-MCM (without S) and Zn1/NS-HCM (single Zn atom decorated on N and S co-doped hollow carbon matrix) are also prepared (SI Appendix, Figs. S6–S10).

Fig. 2. (A–C) SEM images of Zn1/NS-MCM. (D and E) TEM images of Zn1/NS-MCM. (F and G) HRTEM image (F) and the inverse fast Fourier transformation (IFFT) image with corresponding line scan (G) of Zn1/NS-MCM. (H and I) Atomic resolution ADF-STEM image of Zn1/NS-MCM (H) and pseudocolour surface plot of a randomly selected single Zn site. (J–L) HAADF-STEM image of Zn1/NS-MCM, and related elemental mappings.

Chemical State and Atomic Structure Analysis.

The coordination environment of Zn atoms is examined using Zn K-edge X-ray absorption spectroscopy. The analysis of extended X-ray absorption fine structure (EXAFS) reveals a dominant peak at 1.75 Å in Zn1/NS-MCM. This value is larger than the Zn–N signal at 1.58 Å observed in Zn1/N-MCM and shorter than the Zn–S path (1.91 Å) observed in ZnS (Fig. 3A). Thus, it can be concluded that this dominant peak represents the hybridization of Zn–N and Zn–S coordination. No Zn–Zn peak is observed at 2.3 Å, corroborating the atomic dispersion of single Zn sites in Zn1/NS-MCM. Quantitative analysis indicates that the Zn–N coordination number is approximately 4.1 and the Zn–S coordination number is around 1.0, with average bond lengths of 2.1 Å and 2.3 Å, respectively (Fig. 3 B and C and SI Appendix, Fig. S11 and Table S1). Furthermore, wavelet transform (WT) analysis is used to study the atomic configuration of Zn in Zn1/NS-MCM (Fig. 3E). Considering the contributions of Zn–N and Zn–S, the WT contour plots of Zn1/NS-MCM exhibit the maximum peak at 5.7 Å−1. Notably, no intensity maximum corresponding to Zn–Zn is observed, further confirming the isolated nature of Zn species in Zn1/NS-MCM. X-ray absorption near-edge structure (XANES) spectra show the position of the absorption threshold for Zn1/NS-MCM is slightly higher than that of Zn1/N-MCM, confirming the additional S coordination increases the oxidation state of Zn in Zn1/NS-MCM.

Fig. 3. (A) FT k3-weighted Zn K-edge EXAFS spectra of Zn1/NS-MCM and the references. (B and C) Fits of the EXAFS spectrum of Zn1/NS-MCM in R space (B) and K space (C). (D and E) Zn K-edge EXAFS spectra (D) and WT-EXAFS (E) of Zn1/NS-MCM and the references. (F) C K-edge XANES spectra of Zn1/NS-MCM and Zn1/N-MCM. (G and H) The S L-edge (G) and K-edge (H) XANES spectra of Zn1/NS-MCM.

The electronic structure evolution following S coordination is further investigated using synchrotron-radiation-based soft XANES. The carbon K-edge spectra of Zn1/NS-MCM and Zn1/N-MCM display three distinct peaks at 285.7 eV, 288.7 eV, and 292.8 eV, respectively. These peaks can be attributed to the dipole transition of the C 1s core electron to the π*C=C, π*C-N/S-C, and σ*C-C orbitals (Fig. 3F) (30). The intensity of the π*C=C resonance in Zn1/NS-MCM is observed to be weaker compared to Zn1/N-MCM, indicating a decrease in the 1s–π* transition after S coordination. This reduction may arise from the involvement of partial carbon π electrons in C–S bonding (31). Moreover, Zn1/NS-MCM exhibits an increased intensity ratio of π*C–N/C–S and σ*C–C of 1.32 compared to that of Zn1/N-MCM (1.2), demonstrating that the incorporation of additional S atoms induces extensive electronic modulation of the carbon matrix (SI Appendix, Table S2) (32). The presence of S in Zn1/NS-MCM is confirmed by the S L-edge XANES spectrum (Fig. 3G), where small peaks within the 163 to 167 eV range correspond to the C–S–C coordination species (33). The S atomic state in Zn1/NS-MCM is further probed through the S K-edge XANES spectrum. Two primary composite peaks at about 2,473 and 2,482 eV can be assigned to reduced and oxidized S species, respectively (Fig. 3H) (24, 25, 34). Within the reduced S species, the dominant peak, along with a narrow shoulder peak, corresponds to thiophene (th-S) and th-S-Zn species, respectively.

The chemical bonding state of the samples is investigated by X-ray photoelectron spectroscopy (XPS). The survey spectra clearly demonstrate the presence of C, N, S, and Zn elements in Zn1/NS-MCM and Zn1/N-MCM, whereas the S peak is absent in Zn1/N-MCM (SI Appendix, Figs. S12 and S13). The N 1s spectrum of Zn1/NS-MCM exhibits three distinct peaks corresponding to pyridinic-N/Zn-Nx (398.4 eV), pyrrolic-N (399.9 eV), and graphitic-N (401.2 eV) (SI Appendix, Fig. S12C) (35, 36). In the high-resolution S 2p spectrum of Zn1/NS-MCM, two major peaks at 163.6 and 164.9 eV are observed, corresponding to S 2p3/2 and S 2p1/2, respectively (SI Appendix, Fig. S12D). The emergence of a new peak at a low binding energy of 161.5 eV indicates the presence of Zn-S coordinated bonds (37). In the Zn 2p spectrum, two peaks at 1,022.3 and 1,045.3 eV correspond to Zn2p3/2 and Zn2p1/2, respectively (SI Appendix, Fig. S13D). Notably, the peak positions of Zn2p3/2 and Zn2p1/2 in Zn1/NS-MCM are slightly shifted to higher binding energies compared to those in Zn1/N-MCM, suggesting the electronic reconfiguration of Zn due to the formation of Zn–S coordination. Quantitative XPS measurements combined with inductively coupled plasma optical emission spectrometry reveal that the Zn content is 2.4 wt% for Zn1/NS-MCM. The graphitization degree of the carbon-based samples is assessed through Raman spectra (SI Appendix, Fig. S14). Two characteristic peaks are observed at 1,344 and 1,574.8 cm−1, corresponding to sp3-type disordered carbon (D-band) and sp2-type graphitized carbon (G-band) (38). Zn1/NS-MCM exhibits a higher ID/IG ratio of 1.09 compared to Zn1/N-MCM (1.02), indicating the presence of more structural defects resulting from S doping. N2 sorption experiment of Zn1/NS-MCM shows a high specific surface area of 814.187 m−2 g−1 and the existence of both micropore and mesopores, which allow the high exposure of active sites and are conducive to electrolyte penetration during HER (SI Appendix, Fig. S15).

Electrocatalytic Performance.

The electrochemical performances of the samples are evaluated using a three-electrode setup. As shown by the linear sweep voltammetry (LSV) curves (Fig. 4A and SI Appendix, Figs. S16 and S17), the onset potential (ηonset) of Zn1/NS-MCM (236.2 mV) is more positive than those of Zn1/NS-HCM (249.6) and Zn1/N-MCM (270.6 mV). The overpotential required to drive a 10 mA cm−2 (η10) HER current density is 270.8 mV for Zn1/NS-MCM, lower than those of Zn1/NS-HCM (295.1 mV) and Zn1/N-MCM (343.8 mV) (Fig. 4A and SI Appendix, Fig. S17). The increment of the current density for Zn1/NS-MCM is much faster than those of Zn1/NS-HCM and Zn1/N-MCM. To achieve a current density of 100 mA cm−2, Zn1/NS-MCM only requires an overpotential (η100) of 331.4 mV, whereas Zn1/NS-HCM and Zn1/N-MCM require overpotentials of 346.6 and 448 mV, respectively. These results demonstrate the synergistic effect of the advantageous multilevel carbon structure and active ZnN4S1 moiety in enhancing HER activity.

Fig. 4. (A and B) Multiple-measured polarization curves (A) and corresponding Tafel plots (B) of as-prepared samples. (C) Comparison of the Tafel slope for the as-prepared samples with Pt/C, MoS2, B-G, N-G, O-G, S-G and P-G and comparison of the j0 for the Zn1/N-MCM and Zn1/NS-MCM. (D) The increment of the current density with the rising scan rate at 0.5 V for Zn1/N-MCM and Zn1/NS-MCM. (E) Calculated Cdl value of the as-prepared samples based on multiple-measured CV curves at different scan rates. (F) Time dependence of cathodic current density for Zn1/NS-MCM under a fixed overpotential of −300 mV. (G) Comparison of the HER activity over Zn1/NS-MCM with the reported metal-free electrocatalysts.

The Tafel plots extracted from the LSV curves illustrate that the Zn1/NS-MCM catalyst exhibits the smallest Tafel slope value of 65.9 mV dec−1 compared to Zn1/NS-HCM (70.3 mV dec−1) and Zn1/N-MCM (123.2 mV dec−1), indicating a rapid increase in the hydrogen evolution rate with the rising overpotential (Fig. 4 B and C). Moreover, the Tafel slope value of Zn1/NS-MCM is much smaller than those of B, N, O, S, and P-doped graphene (B–G, N–G, O–G, S–G, and P–G) materials (120 mV dec−1) and Zn1/N-MCM (123.2 mV dec−1), suggesting greatly enhanced HER kinetics (Fig. 4C and SI Appendix, Table S3) (25). The exchange current density (j0) is determined by extrapolating the Tafel plots at low current density (SI Appendix, Fig. S18). As anticipated, Zn1/NS-MCM shows a higher j0 value of 25.37 μA cm−2, demonstrating its high intrinsic electrocatalytic activity (Fig. 4C). The HER electrocatalytic activity of Zn1/NS-MCM is comparable to those of the most widely reported metallic and metal-free HER electrocatalysts (Fig. 4G and SI Appendix, Table S3), demonstrating the positive effect of Zn’s electron injection effect on the boosted HER activities of carbon material.

To assess the electrochemically active surface area of the catalysts, cyclic voltammetry at non-faradaic region is performed to measure the double-layer capacitance (Cdl). The current response of Zn1/NS-MCM exhibits a faster increase with the rising scan rate compared to Zn1/N-MCM (Fig. 4D and SI Appendix, Fig. S19). The calculated Cdl value of Zn1/NS-MCM is 52.41 mF cm−2, outperforming that of Zn1/N-MCM (17.19 mF cm−2) (Fig. 4E and SI Appendix, Fig. S20). The unique multilevel porous structure of Zn1/NS-MCM enables exceptional exposure of active sites, leading to superior HER activity. The remarkably low charge transfer resistance (Rct) implies rapid electron transfer at the electrode/electrolyte interface during HER for Zn1/NS-MCM (SI Appendix, Fig. S21 and Table S4). To probe the stability of the Zn1/NS-MCM catalyst, continuous HER is performed at a constant overpotential of −300 mV (Fig. 4F). Remarkably, only negligible current degradation is observed for Zn1/NS-MCM, showcasing its excellent electrochemical durability. Furthermore, the morphology of the catalyst remains unchanged, without the formation of aggregated particles after HER testing, further confirming its high stability (SI Appendix, Figs. S22 and S23). The full-scan XPS spectrum of Zn1/NS-MCM after HER testing reveals the coexistence of Zn, N, S, and C (SI Appendix, Fig. S24). High-resolution N 1s and S 2p spectra indicate the presence of Zn–N and Zn–S coordination, showing the stability of isolated Zn species.

Density Functional Theory (DFT) Calculations.

The mechanisms of isolated Zn atom on boosted HER activity are investigated using DFT calculations. Two proposed models, including planar symmetrical ZnN4 moiety with a subsurface S atom (ZnN4-Ssub) and planar unsymmetrical ZnN3S moiety with a subsurface N atom (ZnN3S-Nsub), are constructed based on the experimental results (SI Appendix, Fig. S25). Analysis of the charge density difference on ZnN4-Ssub and ZnN3S-Nsub reveals a tendency for Zn atoms to lose electrons, with the delocalized electrons accumulating around neighboring C–N/C–S bonds, demonstrating the electron injection effect of Zn (Fig. 5 A and B).

Fig. 5. Calculated charge density differences of (A) ZnN4-Ssub and (B) ZnN3S-Nsub model. Free energy diagram of the HER following the Volmer–Heyrovsky pathway and Volmer–Tafel pathway at (C) dsite-ZnN4-Ssub site and (D) esite-ZnN3S-Nsub site. (E) The structure evolution at dsite-ZnN4-Ssub site according to Volmer–Heyrovsky pathway. (F) The structure evolution at esite-ZnN3S-Nsub site according to Volmer–Tafel pathway. (G) The DOS of C site at the second Volmer step of Tafel step for ZnN4-Ssub (Top) and ZnN3S-Nsub (bellow) model. (H) Energy level diagram showing orbital hybridization of active sites and hydrogen adsorbate. Ef is the Fermi level of the substrate; σ and σ∗ indicate bonding and anti-bonding states, respectively. (I and J) Charge density difference plot of ZnN4-Ssub (I) and ZnN3S-Nsub (J) models before and after H adsorption at the second Volmer step of Volmer–Tafel pathway.

The theoretical assessment of the HER activity on ZnN4-Ssub and ZnN3S-Nsub models relies on the calculations of the Gibbs free energy of hydrogen adsorption (ΔGH∗) (SI Appendix, Fig. S26) (39). The ZnN4-Ssub model showcases an optimal ΔGH∗ value at the N atom (d site in SI Appendix, Fig. S26A, represented by dsite-ZnN4-Ssub), while the ZnN3S-Nsub model demonstrates a “just right” ΔGH∗ value at the C atom (e site in SI Appendix, Fig. S26C, represented by esite-ZnN3S-Nsub). Remarkably, these models exhibit comparable theoretical activities toward the HER. To reveal the reaction mechanism of ZnN4-Ssub and ZnN3S-Nsub model, the HER free energy diagram through either Volmer–Heyrovsky or Volmer–Tafel pathway is further constructed.

In the quest to uncover the HER reaction mechanism, Volmer–Heyrovsky and Volmer–Tafel pathways are further explored (Fig. 5 C and D and SI Appendix, Table S5). The dsite-ZnN4-Ssub site within the ZnN4-Ssub model favors the Volmer–Heyrovsky pathway, with the highest free energy change of 0.42 eV at the Volmer step of the overall reaction pathway, serving as the rate-determining step (RDS). On the other hand, the Volmer–Tafel pathway shows a higher free energy change of 0.67 eV (Fig. 5C and SI Appendix, Fig. S27 and Table S5). These findings highlight the dominance of the self-adaptive Volmer–Heyrovsky pathway on the ZnN4-Ssub model (Fig. 5E). In contrast, the Volmer–Tafel pathway dominants on the ZnN3S-Nsub model, with a RDS at the Tafel step (Fig. 5 D and F and SI Appendix, Fig. S28 and Table S5).

The hydrogen adsorption at the second Volmer step through the Volmer–Tafel pathway distinct variations between the ZnN4-Ssub and ZnN3S-Nsub models, specifically at the peripheral C site (Fig. 5F and SI Appendix, Figs. S27 and S29 and Table S5). This finding motivated us to study the carbon support’s intrinsic activity. A strong correlation is found between the hydrogen adsorption strength and the highest peak of the active center DOS (Ep). A higher Ep position indicates a stronger adsorption strength for H* due to the interaction between the carbon site and the adsorbed hydrogen (Fig. 5G) (25). This interaction results in the splitting of their hybridized energy levels into two groups, particularly the anti-bonding states (σ∗), which typically cross the Fermi Level (Fig. 5H). The location of σ∗ influences the strength of hydrogen adsorption. With the active carbon Ep is positioned higher, σ∗ moves up with lower occupancy, resulting in enhanced interaction between H and the active carbon site. In comparison to the ZnN4-Ssub model, the Ep position for the ZnN3S-Nsub model crosses the Fermi Level, suggesting more favorable H adsorption (Fig. 5G). Moreover, compared with ZnN4-Ssub model, an enhanced charge depletion in the charge density difference plot at the second Volmer step of ZnN3S-Nsub for H adsorption is observed (represented by the green areas), which is consistent with the band structure analysis (Fig. 5 I and J).

Discussion

This work delves into the intricate interplay between catalytic activity and the peripheral atom of a single Zn site, unveiling the HER mechanism of the ZnN4S1 moiety. Through a remarkable process of self-adaption structure evolution, the ZnN4S1 moiety within the Zn1/NS-MCM catalyst exhibits exceptional stability and activity, surpassing carbon electrocatalysts with a significantly reduced Tafel slope of 65.93 mV dev−1 (compared to 120 mV dev−1). DFT calculations elucidate the connection between the low Tafel slope and the presence of the peripheral carbon atom, attributing the accelerated kinetics of the HER to a more constrained electronic structure. These findings not only shed light on the coordination effects of SACs but also pave the way for the development of future catalysts tailored to specific application requirements.

Materials and Methods

Synthesis of ZIF-8.

In a typical synthesis, a solution of 0.3 g of Zn(OAc)2·2H2O in 5 mL of deionized (DI) water was added into a solution of 1.12 g of 2-methylimidazole in 5 mL of DI water, and the resulting mixture was stirred for a few seconds. Then, the mixture was left at room temperature for 24 h to form polyhedral ZIF-8 nanocrystals. The resulting white powder was collected by centrifuging, washed with water and methanol in sequence for several times, and finally dried in vacuum at 50 °C overnight.

Synthesis of ZIF-8@TA-Zn, ZnS@NS-MCM, Zn1/NS-MCM, Zn1/N-MCM, and Zn1/NS-HCM.

The as-prepared ZIF-8 was first dispersed into 20 mL of ethanol, then poured into 300 mL of ethanol, and DI water mixture solution (VH2O:Vethanol = 1:1) containing 2 mg mL−1 of TA and stirred at room temperature for 5 min. The product (ZIF-8@TA-Zn) was collected by centrifugation and washed with ethanol several times, which was then dried in an oven at 70 °C. Sulfur powder (5 mg) and dried ZIF-8@TA-Zn were placed in two porcelain boats. The porcelain boat with sulfur powder was put in the front end of the port of tubular furnace while the porcelain boat with ZIF-8@TA-Zn was put in the middle. Then, the boats were heated to 800 °C at a heating rate of 5 °C min−1, kept at 800 °C for 2 h under a flowing N2 atmosphere, and finally cooled to room temperature naturally to obtain ZnS@NS-MCM. Then, the ZnS@NS-MCM was etched thoroughly with 0.5 M HCl aqueous solution at room temperature, which was followed by repeated filtrations and washing with lots of water to yield Zn1/NS-MCM catalysts.

The Zn1/N-MCM catalyst was prepared with the same treatment, except sulfur powder was not introduced during the annealing process.

The Zn1/NS-HCM catalyst was prepared with the same treatment, except extending the etching time of TA to 15 min.

Electrochemical Measurements.

All the electrochemical experiments were conducted on the CHI 760E electrochemical workstation using a standard three-electrode system with graphite rod and Ag/AgCl electrode as counter and reference electrodes, respectively. The working electrode was prepared as follows: 5 mg HER catalysts were dispersed into 0.5 mL of ethanol/Nafion (9:1, v/v) solution to get a uniform suspension by ultrasonication. Then, a certain amount of the above suspension was dropped on glassy carbon rotating disk electrode (diameter 3 mm) and dried at room temperature. LSV with a scan rate of 5 mV s−1 was conducted in 0.5 M H2SO4 solution. All potentials were quoted against a reversible hydrogen electrode and all polarization curves were corrected by the iR contribution within the cell.

Calculation Details.

The DFT calculations were implemented on Vienna Ab-initio Simulation Package (VASP) code (40, 41). The interaction of core and electrons was treated by projector augmented wave pseudopotential with a cut-off energy of 450 eV and the exchange-correlation function was described by generalized gradient approximation of Perdew–Burke–Ernzerhof method (42–44). A 20-Ångström-thick vacuum layer was used to eliminate the interaction between two adjacent slabs. It was considered that geometry optimization was convergence when the force of each atom less than 0.02 eV Å−1. The DFT-D3(BJ) method was used to account for the vdW-dispersion energy-correction (45). The Brillouin zone was sampled on the Gamma-centered Monkhorst-Pack grids, and the k-point was set to be 4 × 4 × 2 for all the models. According to the experimental fine structural characterization, two-layer models of N4(layer 1)–Zn–S(layer 2) and N3S(layer 1)–Zn–N(layer 2) anchored on the graphene were constructed, where the center metal maintains six coordination, to decipher the dynamic mechanism of HER on these precisely designed active sites. The data processing was assisted by VASPKIT (46), QVASP (47), and VESTA software. The Gibbs free energy difference (ΔG) between initial and final states was defined as (48):ΔG = ΔE + ΔZPE -TΔS,

where E, ZPE, T, and S represent the energy from DFT calculation, zero-point energy, temperature (298.15 K), and entropy, respectively.

Supplementary Material

Appendix 01 (PDF)

Click here for additional data file.

Dataset S01 (XLSX)

Click here for additional data file.

This research was supported by King Abdullah University of Science and Technology.

Author contributions

Y.L., M.R., Y.H., and H.Z. designed research; Y.L., C.C., and G.Z. performed research; and S.Z., F.W., X.Z., Z.W., S.W., and W.Z. analyzed data; and Y.L. and H.Z. wrote the paper.

Competing interests

The authors declare no competing interest.

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information.

Supporting Information

This article is a PNAS Direct Submission. D.-J.L. is a guest editor invited by the Editorial Board.
==== Refs
1 D. Liu , Atomically dispersed platinum supported on curved carbon supports for efficient electrocatalytic hydrogen evolution. Nat. Energy 4 , 512–518 (2019).
2 I. T. McCrum, M. T. M. Koper, The role of adsorbed hydroxide in hydrogen evolution reaction kinetics on modified platinum. Nat. Energy 5 , 891–899 (2020).
3 H. Zhang , Dynamic traction of lattice-confined platinum atoms into mesoporous carbon matrix for hydrogen evolution reaction. Sci. Adv. 4 , 6657 (2018).
4 Z. W. Seh , Combining theory and experiment in electrocatalysis: Insights into materials design. Science 355 , eaad4998 (2017).28082532
5 H. Li , Corrigendum: Activating and optimizing MoS2 basal planes for hydrogen evolution through the formation of strained sulphur vacancies. Nat. Mater. 15 , 364 (2016).
6 D. Deng , Catalysis with two-dimensional materials and their heterostructures. Nat. Nanotechnol. 11 , 218–230 (2016).26936816
7 M. Li , Single-atom tailoring of platinum nanocatalysts for high-performance multifunctional electrocatalysis. Nat. Catal. 2 , 495–503 (2019).
8 Y. Zhou , Peripheral-nitrogen effects on the Ru1 centre for highly efficient propane dehydrogenation. Nat. Catal. 5 , 1145–1156 (2022).
9 Y. Chen , Enhanced oxygen reduction with single-atomic-site iron catalysts for a zinc-air battery and hydrogen-air fuel cell. Nat. Commun. 9 , 5422 (2018).30575726
10 H. Fei , Atomic cobalt on nitrogen-doped graphene for hydrogen generation. Nat. Commun. 6 , 8668 (2015).26487368
11 Y. Liu , Electrosynthesis of chlorine from seawater-like solution through single-atom catalysts. Nat. Commun. 14 , 2475 (2023).37120624
12 F. Li , Identifying the structure of Zn-N2 active sites and structural activation. Nat. Commun. 10 , 2623 (2019).31197162
13 H. Shang , Engineering unsymmetrically coordinated Cu-S1N3 single atom sites with enhanced oxygen reduction activity. Nat. Commun. 11 , 3049 (2020).32546781
14 P. Zhai , Engineering single-atomic ruthenium catalytic sites on defective nickel-iron layered double hydroxide for overall water splitting. Nat. Commun. 12 , 4587 (2021).34321467
15 W. Chen , Single tungsten atoms supported on MOF-derived N-doped carbon for robust electrochemical hydrogen evolution. Adv. Mater. 30 , 1800396 (2018).
16 H. Fei , Microwave-assisted rapid synthesis of graphene-supported single atomic metals. Adv. Mater. 30 , 1802146 (2018).
17 Y. Mun , Versatile strategy for tuning ORR activity of a single Fe-N4 site by controlling electron-withdrawing/donating properties of a carbon plane. J. Am. Chem. Soc. 141 , 6254–6262 (2019).30920818
18 C. Hu , Near- and long-range electronic modulation of single metal sites to boost CO2 electrocatalytic reduction. Adv. Mater. 35 , 2209298 (2023).
19 J. Jiao , Copper atom-pair catalyst anchored on alloy nanowires for selective and efficient electrochemical reduction of CO2. Nat. Chem. 11 , 222–228 (2019).30664719
20 Y. Jia, X. Yao, Defects in carbon-based materials for electrocatalysis: Synthesis, recognition, and advances. Acc. Chem. Res. 56 , 948–958 (2023).36989384
21 J. Chen , Promoting CO2 electroreduction kinetics on atomically dispersed monovalent Zn(I) sites by rationally engineering proton-feeding centers. Angew. Chem. Int. Ed. Engl. 61 , e202111683 (2022).34608726
22 W. Ren , Isolated diatomic Ni-Fe metal-nitrogen sites for synergistic electroreduction of CO2. Angew. Chem. Int. Ed. Engl. 58 , 6972–6976 (2019).30920151
23 H. Wang , Coordination engineering in single-site catalysts: General principles, characterizations, and recent advances. Chem 10 , 1–38 (2024).
24 Y. Zheng , Hydrogen evolution by a metal-free electrocatalyst. Nat. Commun. 5 , 3783 (2014).24769657
25 Y. Jiao, Y. Zheng, K. Davey, S. Z. Qiao, Activity origin and catalyst design principles for electrocatalytic hydrogen evolution on heteroatom-doped graphene. Nat. Energy 1 , 16130 (2016).
26 Z. F. Huang , Hollow cobalt-based bimetallic sulfide polyhedra for efficient all-pH-value electrochemical and photocatalytic hydrogen evolution. J. Am. Chem. Soc. 138 , 1359–1365 (2016).26777119
27 H. Hu, L. Han, M. Yu, Z. Wang, X. W. Lou, Metal-organic-framework-engaged formation of Co nanoparticle-embedded carbon@Co9S8 double-shelled nanocages for efficient oxygen reduction. Energy Environ. Sci. 9 , 107–111 (2016).
28 Y. Z. Chen , From bimetallic metal-organic framework to porous carbon: High surface area and multicomponent active dopants for excellent electrocatalysis. Adv. Mater. 27 , 5010–5016 (2015).26193083
29 L. Yu, H. Hu, H. B. Wu, X. W. Lou, Complex hollow nanostructures: Synthesis and energy-related applications. Adv. Mater. 29 , 1604563 (2017).
30 Y. Tong , A bifunctional hybrid electrocatalyst for oxygen reduction and evolution: Cobalt oxide nanoparticles strongly coupled to B, N-decorated graphene. Angew. Chem. Int. Ed. Engl. 56 , 7121–7125 (2017).28523861
31 C. Tang , Coordination tunes selectivity: Two-electron oxygen reduction on high-loading molybdenum single-atom catalysts. Angew. Chem. Int. Ed. Engl. 59 , 9171–9176 (2020).32196867
32 Y. Liang , Co3O4 nanocrystals on graphene as a synergistic catalyst for oxygen reduction reaction. Nat. Mater. 10 , 780–786 (2011).21822263
33 M. Wang , Atomically dispersed Fe-heteroatom (N, S) bridge sites anchored on carbon nanosheets for promoting oxygen reduction reaction. ACS Energy Lett. 6 , 379–386 (2021).
34 Q. Li , Fe isolated single atoms on S, N codoped carbon by copolymer pyrolysis strategy for highly efficient oxygen reduction reaction. Adv. Mater. 30 , 1800588 (2018).
35 Q. Yang, C. C. Yang, C. H. Lin, H. L. Jiang, Metal-organic-framework-derived hollow N-doped porous carbon with ultrahigh concentrations of single Zn atoms for efficient carbon dioxide conversion. Angew. Chem. Int. Ed. Engl. 58 , 3511–3515 (2019).30569535
36 P. Song , Zn single atom catalyst for highly efficient oxygen reduction reaction. Adv. Funct. Mater. 27 , 1700802 (2017).
37 W. Wu , Activation of MoS2 basal planes for hydrogen evolution by zinc. Angew. Chem. Int. Ed. Engl. 58 , 2029–2033 (2019).30600886
38 Q. Yang , Single carbon vacancy traps atomic platinum for hydrogen evolution catalysis. J. Am. Chem. Soc. 144 , 2171–2178 (2022).34995077
39 S. L. Zhang , Engineering platinum-cobalt nano-alloys in porous nitrogen-doped carbon nanotubes for highly efficient electrocatalytic hydrogen evolution. Angew. Chem. Int. Ed. Engl. 60 , 9068–19073 (2021).
40 G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54 , 11169–11186 (1996).
41 G. Kresse, J. Furthmüller, Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Phys. Commun. 6 , 15–50 (1996).
42 P. E. Blochl, Projector augmented-wave method. Phys. Rev. B 50 , 17953–17979 (1994).
43 G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 59 , 1758–1775 (1999).
44 J. P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made simple. Phys. Rev. Lett. 77 , 3865–3868 (1996).10062328
45 S. Grimme, S. Ehrlich, L. Goerigk, Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 32 , 1456–1465 (2011).21370243
46 V. Wang, N. Xu, J. C. Liu, G. Tang, W. T. Geng, VASPKIT: A user-friendly interface facilitating high-throughput computing and analysis using VASP code. Comput. Phys. Commun. 267 , 108033 (2021).
47 W. Yi, G. Tang, X. Chen, B. Yang, X. Liu, qvasp: A flexible toolkit for VASP users in materials simulations. Comput. Phys. Commun. 257 , 107535 (2020).
48 J. K. Nørskov , Origin of the overpotential for oxygen reduction at a fuel-cell cathode. J. Phys. Chem. B 108 , 17886–17892 (2004).
