Abstract
Develo** active single-atom-catalyst (SAC) for alkaline hydrogen evolution reaction (HER) is a promising solution to lower the green hydrogen cost. However, the correlations are not clear between the chemical environments around the active-sites and their desired catalytic activity. Here we study a group of SACs prepared by anchoring platinum atoms on NiFe-layered-double-hydroxide. While maintaining the homogeneity of the Pt-SACs, various axial ligands (−F, −Cl, −Br, −I, −OH) are employed via a facile irradiation-impregnation procedure, enabling us to discover definite chemical-environments/performance correlations. Owing to its high first-electron-affinity, chloride chelated Pt-SAC exhibits optimized bindings with hydrogen and hydroxide, which favor the sluggish water dissociation and further promote the alkaline HER. Specifically, it shows high mass-activity of 30.6 A mgPt−1 and turnover frequency of 30.3 H2 s−1 at 100 mV overpotential, which are significantly higher than those of the state-of-the-art Pt-SACs and commercial Pt/C catalyst. Moreover, high energy efficiency of 80% is obtained for the alkaline water electrolyser assembled using the above catalyst under practical-relevant conditions.
Similar content being viewed by others
Introduction
Green hydrogen produced via electrocatalytic hydrogen evolution reaction (HER) has been recognized as a promising alternative to mitigate the pressing carbon emission issues1,2. Thus, HER has been explored extensively through both proton reduction in acidic media, and water reduction in alkaline media3,4. In strong acidic media, noble-metal-based materials are usually needed to catalyse proton reduction for HER to avoid electrode dissolutions5,6. In contrast, abundant materials have shown satisfactory stability for water reduction in alkaline electrolyte7,8. Unfortunately, the HER kinetics is typically one to a few orders of magnitude lower in alkaline electrolyte than in acids, even on the state-of-the-art platinum (Pt)-based electrocatalysts9,10,11,12,13. Besides, HER is also much more sensitive to the catalyst surface-structure in alkaline media than in acids14. Nevertheless, enhancing the sluggish kinetics of water reduction in alkaline electrolyte is crucial to reduce the high overpotential and associated energy loss for green hydrogen production15,16,17.
Development of highly active single-atom catalysts (SACs) with maximized atomic utilization efficiency is a promising solution to address the above challenges18,19,20,21,22. The chemical environment of the SACs active-sites, including the ligand identity, coordination number and configuration, can directly influence the electronic state, degrees of freedom and other physicochemical properties of the active-center that are associated to the adsorptions of the reactant/intermediates, and thus further determines the catalytic performance of the SACs23,24,25. Hence, revealing the chemical-environment/catalytic-performance correlations is critical for designing SACs with improved activity26. Currently, several methods have been developed to synthesize SACs with relatively controllable chemical-environments around the active-center, including defect engineering27, annealing24, metal-support interaction28, heteroatom tethering29, cluster/nanoparticle introduction30, etc. While tremendous progress has been made in develo** new materials using the above synthetic strategies, the harsh conditions involved inevitably break the homogeneity of the active-centers within SACs, e.g., the metal loading, the coordination number/configuration, and the structure of supports, leading complications and uncertainties in establishing definitive correlations between the chemical-environment of SACs active-centers and their catalytic performance31,1c, f, the representative Pt single atoms and lattice spacing are pointed out by the pink arrows and organe arrows, respectively. h Far-Infrared and (i) UV-vis DRS spectra for Cl-Pt/LDH and HO-Pt/LDH. j EPR spectra of Cl-Pt/LDH and HO-Pt/LDH under light/dark conditions.
To exchange the Cl ligand, we treated the Cl-Pt/LDH with white light irradiation (3.75 mW cm−2; 30 min) in KOH. While the atomic dispersion of Pt-atoms remains unchanged after the irradiation (Fig. 1f), the intensity of Cl signal mostly diminished, suggesting the successful removal of Cl ligand. In addition, the decreased signal at 329 cm−1 in the Far-Infrared spectra after the irradiation further confirms the loss of the coulombic Pt-Cl bonds (Fig. 1h)40. We tentatively attribute the Cl loss to the anion exchange by hydroxide, yielding the product of HO-Pt/LDH, its structure is later determined by X-ray adsorption spectroscopy. Obvious changes were also observed for the ligand-to-metal charge transfer (LMCT) induced absorption bands (~200–300 nm) in Ultraviolet-visible Diffuse Reflectance spectra (UV-vis DRS) of Cl-Pt/LDH and HO-Pt/LDH (Fig. 1i), while the metal-to-metal charge transfer (MMCT) remains unchanged, indicating that only ligand exchange occurred after the irradiation41. Electron paramagnetic resonance (EPR) measurements were conducted to study the mechanism of Cl removal42. As shown in Fig. 1j, profound radical signals were appeared when exposing the Cl-Pt/LDH to light and the radical scavenger 5,5-dimethyl-1-pyrroline N-oxide (DMPO) spontaneously, which can be attributed to the oxidation of the Cl ligand to Cl radicals by photogenerated holes, which then are captured by DMPO to form DMPO+● radicals43,44. In contrast, the absence of radical signals were observed for samples of Cl-Pt/LDH in dark and HO-Pt/LDH under illumination as expected.
Taken together, we conclude that the Cl− to OH− ligand exchange is induced by visible light illumination. Later, the HO-Pt/LDH sample was re-immersed into KCl solution, as expected, the −Cl axial ligand can be recovered (denote as R-Cl-Pt/LDH) due to the strong binding affinity of halogens on Pt. Worth noting, the loadings of Pt in Cl-Pt/LDH, HO-Pt/LDH and R-X-Pt/LDH (X = F, Cl, Br, or I) are very close based on results of Inductively Coupled Plasma Atomic Emission Spectrometer (ICP-AES) (Supplementary Table 1), suggesting that there are no other detectable structural changes occur besides the axial-ligand exchange. Overall, the facile irradiation-impregnation procedure allows us to precisely manipulate the axial ligand of the Pt-single-sites while maintaining the homogeneity of the Pt-SACs.
Investigating the coordination and electronic structures
X-ray photoelectron spectroscopy (XPS) was employed to analyse the detailed chemical composition of Cl-Pt/LDH and HO-Pt/LDH. The loss of Cl 2p signal (Fig. 2a) after light irradiation provides further evidence for the complete ligand exchange. Meanwhile, negligible changes can be found in both Fe 2p and Ni 2p spectra (Supplementary Fig. 5), indicating that the homogeneity of the Pt-SAC is well retained. As shown in Fig. 2b, the 4f7/2 peaks of Pt-atoms in both Cl-Pt/LDH and HO-Pt/LDH are located between Pt2+ (72.7 eV) and Pt4+ (74.9 eV), indicating the average valence states of these Pt-atoms are between +2 and +445. The slightly higher binding energy of the Pt 4f-electrons in Cl-Pt/LDH than that of the HO-Pt/LDH agrees well with previous results where electron transfer was observed from Pt to the Cl ligand46.
a and Pt 4 f (b) for Cl-Pt/LDH and HO-Pt/LDH. c XANES spectra of Cl-Pt/LDH, HO-Pt/LDH, Pt foil, K2PtCl4 and PtO2 at the Pt L3-edge. d Corresponding FT-EXAFS spectra. Representative EXAFS fitting curve (inset is the magnified local structure) of Cl-Pt/LDH (e) and HO-Pt/LDH (f). The blue, olive, yellow, red, green, and gray spheres refer to Pt, Ni, Fe, O, Cl and H atoms, respectively.
X-ray absorption fine structure (XAFS) spectroscopy was performed to investigate the coordination environment and electronic structure of Pt-single-sites in Cl-Pt/LDH and HO-Pt/LDH. First, the X-ray absorption near edge structure (XANES) spectra of the above samples and the references of Pt foil, K2PtCl4, and PtO2 at the Pt L3-edge were analysed. The white line at the Pt L3-edge (the intensity of which is a measure of the transition of 2p3/2 → 5d3/2 or 5d5/247) was recognized as a reliable indicator for the Pt electronic structure, since the unoccupied states above the Fermi level are essentially the 5d character of Pt48,49. As shown in Fig. 2c, the white lines of both Cl-Pt/LDH and HO-Pt/LDH locate between those of the K2PtCl4 (Pt2+) and PtO2 (Pt4+), suggesting that the valence states of Pt-atoms in Cl-Pt/LDH and HO-Pt/LDH are between +2 and +4, in line with the above XPS results. After the curve fitting using standard references, the valence states of Pt in Cl-Pt/LDH and HO-Pt/LDH are estimated as 2.78 and 2.08, respectively (Supplementary Fig. 6a), illustrating the existence of strong ligand effects on the electronic structure of Pt-SACs.
Fourier-transformed (FT) k2-weighted extended X-ray absorption fine structure (EXAFS) spectra were employed to further analyse the coordination environment of the Pt-SACs. As shown in Fig. 2d, no Pt-Pt interactions at ~2.50 Å can be detected in both Cl-Pt/LDH and HO-Pt/LDH, demonstrating the atomic dispersion of Pt-atoms in these samples, which is consistent with the CO chemisorption results from the Fourier-transformed Infrared Spectroscopy (Supplementary Fig. 6b)20. Besides, two obvious peaks at around 1.60 Å and 2.06 Å are found in Cl-Pt/LDH, which are associated to the Pt-O path in PtO2 and the Pt-Cl path in K2PtCl4, respectively. In contrast, only one Pt-O like peak at ~1.60 Å can be seen in HO-Pt/LDH, indicating that the Cl ligand is fully replaced after irradiation. Based on the wavelet transform (WT) analysis of the EXAFS spectra (Supplementary Fig. 7), Cl-Pt/LDH displays a higher WT maximum at 5.2 Å−1 than that of HO-Pt/LDH at 4.9 Å−1, suggesting that the average coordination number of the Pt-sites in Cl-Pt/LDH is larger than that in HO-Pt/LDH.
To reveal the chemical environment of the Pt-single-sites, we performed quantitative least-square EXAFS curve-fitting analysis for the Pt-SACs. Several typical structures of LDH based SACs are fitted and modeled (Supplementary Fig. 8). Among them, the Pt-site chelated by three surface O-atoms on the LDH layer and one axial ligand opposite of the Fe-atom is considered to be the most possible structure (Fig. 1a and Supplementary Fig. 8a). The fiducial distances between Pt and coordinated atoms were calculated using standard crystal structures of Pt foil, K2PtCl4, and PtO2 (Supplementary Fig. 9-11). The best-fitted results of Cl-Pt/LDH (Supplementary Table 2) suggest that the main peak at 1.60 Å in the FT-EXAFS spectra at the Pt L3-edge can be attributed to the Pt-O first coordination sphere, whereas the minor peak at 2.06 Å can be attributed to the Pt-Cl first coordination sphere (Fig. 2e and Supplementary Fig. 12). Based on the EXAFS fitting parameters for Cl-Pt/LDH at the Pt L3-edge, O-atoms within the first coordination sphere is located at 2.02 Å with a coordination number of 3.02, suggesting a tetrahedron geometry for the Pt-sites. Besides, the Cl-atom at 2.29 Å with an estimated coordination number of 0.93 is proposed to be a vertical ligand on Pt. All things considered, we propose the most possible structure of the Cl-Pt/LDH as shown in the inset of Fig. 2e. Similar analysis on HO-Pt/LDH demonstrates that it only has O atoms in the first coordination sphere with a total coordination number of 3.60, affirming that the Cl ligand is replaced by the hydroxide group (Fig. 2f) with no other changes occurred to the Pt-sites.
Electrocatalytic alkaline HER on Pt-SACs
The electrocatalytic HER on Cl-Pt/LDH is examined in H2-saturated KOH (1.0 M). For comparison, the HER activities of Ni foam, NiFe-LDH, 20% Pt/C deposited on Ni foam (Pt/CNF) and HO-Pt/LDH are also measured under the same conditions. As shown in Fig. 3a, Cl-Pt/LDH exhibits the superior HER activity and ~100% Faradic efficiency to H2 among all the Pt-SACs tested here (Supplementary Fig 13 and Table 3), with modest overpotentials of 25.2 mV, 51.9 mV, and 72.3 mV to achieve the current densities of 10 mA cm−2, 100 mA cm−2, and 200 mA cm−2, respectively, overperforming both Pt/CNF (27.4 mV, 164.9 mV, and 252.0 mV) and HO-Pt/LDH (41.5 mV, 142.5 mV, and 189.5 mV). The negligible HER activities of pristine Ni foam and the SCN− poisoning experiment (Supplementary Fig. 14) both indicate that the high HER activity of Cl-Pt/LDH is originated from the Pt-sites. Further increase in the Pt loading leads to the formation of Pt nanoparticles (Ptnp/LDH) and decrease in activity (Supplementary Fig. 15). The mass activity of Cl-Pt/LDH normalized to the Pt loading at overpotential of 100 mV is estimated as 30.6 A mgPt−1 (Fig. 3b), significantly larger than those of HO-Pt/LDH (6.6 A mgPt−1), Pt/CNF (0.2 A mgPt−1), and the state-of-the-art Pt-SACs reported elsewhere (Pt@DG37, 6.78 A mgPt−1, PtSA/NiO/Ni34, 20.6 A mgPt−1, and N, Pt-MoS250, 20.2 A mgPt−1). Moreover, the turnover frequencies (TOFs) per Pt-site on Cl-Pt/LDH (30.3 H2 s−1) at overpotential of 100 mV are 5.9 and 126 times greater than those of HO-Pt/LDH (5.1 H2 s−1) and the commercial Pt/CNF (0.2 H2 s−1), respectively (Fig. 3c). Steady-state cyclic voltammetry (CV) in full regions of water splitting is performed to demonstrate the superior performance of Cl-Pt/LDH (Supplementary Figs. 16–17). We then normalized the HER activities by the corresponding ECSAs estimated by measuring double-layer capacitance (Cdl) at non-faradic regions, it turns out that Cl-Pt/LDH remains as the most active catalyst (Supplementary Fig. 18).
a HER polarization curves of Ni foam, NiFe-LDH, Pt/C, Cl-Pt/LDH, and HO-Pt/LDH. b Mass activities of the Pt-based catalysts in this work, and the state-of-the-art SACs reported elsewhere: Pt@DG37, PtSA/NiO/Ni34 and N, Pt-MoS250. c TOFs plots of the Pt-based catalysts. Operando Raman spectrum of (d) Pt/CNF, (e) Cl-Pt/LDH and (f) HO-Pt/LDH. g Micro-polarization curves for Pt/C, Cl-Pt/LDHp and HO-Pt/LDHp. (h) EIS Nyquist plots of the Pt-based catalysts. i Plot of potential vs. time for the Cl-Pt/LDH and commercial Pt/C at a constant cathodic current density of 50 mA cm−2. (inset is chronoamperometry stability tests of Cl-Pt/LDH under the condition with or without the light irradiation.) The Pt loading amount of Cl-Pt/LDH, HO-Pt/LDH, Pt/CNF, Cl-Pt/LDHp, HO-Pt/LDHp, and Pt/C is 1.01 × 10−2, 1.01 × 10−2, 0.4, 2.88 × 10−4, 2.88 × 10−4 and 8.00 × 10−3 mgPt cm−2, respectively.
Kinetics and intrinsic activities of the Pt-SACs for alkaline HER
Operando Raman spectroscopy was performed to probe the surface species and their changes in chemical bonding during HER. As shown in Fig. 3d, only Pt-OH peaks at 1062 cm−1 appeared for HER on Pt/CNF51,52. We therefore propose that the desorption of *OH is slow under the catalytic conditions, which agrees with previous observations suggesting that *OH is not a spectator in alkaline HER9,53,54. In contrast, such obvious Pt-OH peaks were not observed on Cl-Pt/LDH (Fig. 3e) under the same conditions. Instead, only subtle peaks at ~1630 cm−1 were found, which can be attributed to the H−O−H bending mode of the adsorbed water, suggesting the Volmer step and the consequent *OH desorption steps are accelerated, resulting in enhanced HER kinetics55,56,57. The subtle peak for adsorbed water exists in the HO-Pt/LDH as well, in line with the enhanced activity compared with that of Pt/C. Note, while minor Pt-OH peaks were also found in HER on HO-Pt/LDH (Fig. 3f), its intensity did not vary along with the changes in overpotential, suggesting the −OH is the axial ligand instead of coming from the water reduction. Overall, we conclude that the improved HER activity on the Pt-SACs is likely originated from the accelerated Volmer step. However, we note that due to the limitations of the operando spectroscopy, we do not exclude the effect of mass transportation during the measurements.
Tafel analysis was carried out to investigate the reaction kinetics of HER. As shown in Supplementary Fig. 19, Cl-Pt/LDH exhibits a low apparent Tafel slope of 24.33 mV dec−1, corresponding to a Tafel step limited HER kinetics. This is counterintuitive since Volmer step is usually determined as the rate-determine-step (RDS) for alkaline HER58,59. We thus believe that the HER on Cl-Pt-LDH is largely limited by the mass transportation of H2 leaving the electrode surface in the above measurements, resulting uncertainties for investigating its intrinsic activities60. Note, the Pt-SACs may be affected by the H2 mass diffusion to a different extent due to their different gas affinities (Supplementary Fig. 20). To evaluate the intrinsic HER activities of the above catalysts, powder-based catalysts (X-Pt/LDHp) were prepared and tested on a rotating disk electrode (RDE, Supplementary Fig. 21) to eliminate the influence of the H2 mass transportation, at least to the best extent. The Tafel slope obtained for Cl-Pt/LDHp increases significantly with decreased catalyst loadings (Fig. 3f and Supplementary Fig. 21), suggesting that the mass transport effect become less dominant with low catalyst loading, at least within the potential window of interest5. Note, similar activity trends remain (Supplementary Fig. 22) even with extremely low catalyst loadings (0.04 mg cm−2), confirming the enhanced intrinsic activity of Cl-Pt/LDH compared to other samples. Under this condition, Cl-Pt/LDHp shows a much lower Tafel slope (75.6 mV dec−1) compared to those of HO-Pt/LDHp (99.8 mV dec−1) and Pt/C (120.0 mV dec−1), indicating that the sluggish Volmer step is significantly boosted on Cl-Pt/LDHp, and the RDS is likely the mixture of Volmer and Heyrovsky steps (Fig. 3g)61. Note that, the difference in the symmetry of the log J vs. E curves between Cl-Pt/LDHp and HO-Pt/LDHp suggest that *OH is an active participant in the hydrogen reactions via Volmer-Heyrovsky mechanism62.
Electrochemical impedance spectroscopy (EIS) was also employed to investigate the HER kinetics. As shown in Fig. 3h, the first semicircle at medium frequency is believed to be associated to the kinetics of Volmer step (Supplementary Fig. 23)63. Apparently, Cl-Pt/LDHp shows a significant lower adsorption resistance (R2 = 21.15 ohm) than HO-Pt/LDHp (R2 = 28.45 ohm), indicating that the water dissociation step is more favored on Pt-sites with Cl as axial ligand15,39,64. In addition, in-situ Cl ion titration were conducted to reveal the importance of the axial Cl-ligand to the HER activity. As shown in Supplementary Fig. 24, along with the increase of irradiation period, the Cl: Pt mole-ratio decreased from 1.08 to 0.05 accompanied by the decline of HER activity, suggesting that −Cl is replaced by hydroxide, consisting with the above XAFS results. The ligand exchange occurred quickly under the catalytic conditions. As shown in the inset of Fig. 3i, the HER activity of Cl-Pt/LDH dropped and reached to a plateau within the first half-hour under illumination, suggesting the completion of the ligand exchange within this short period. In contrast, Cl-Pt/LDH retained its activity when light was blocked from the catalytic system, indicating that the −Cl axial ligand on Pt is relatively stable under the HER conditions. Physical characterizations also suggest that its morphological and electronic structures were retained (Supplementary Figs. 25–26). Moreover, extended stability tests were conducted for both Cl-Pt/LDH and Pt/C at current densities of 50 mA cm−2 (Fig. 3i) and 500 mA cm−2 (Supplementary Fig. 27) under the same conditions. Over the 100-hour testing period, Cl-Pt/LDH exhibited improved durability compared to the commercial Pt/C catalyst, demonstrating the robust chemical structure of the Pt-single-sites under alkaline HER conditions.
Exploring the origin of the axial-ligand effect
We extended the axial-ligand identity using the above irradiation-impregnation procedure to prepare R-F-Pt/LDH, R-Br-Pt/LDH, and R-I-Pt/LDH. HAADF-STEM images and the corresponding elemental map**s (Supplementary Figs. 28–31) confirmed the desired axial-ligand exchange on Pt-sites. XPS spectra (Supplementary Fig. 32) indicate that the valence state changes only occur to the Pt-atoms instead of the NiFe-LDH supports. All in all, we believe the homogeneity of the chemical environment around the Pt-sites of these catalysts remain after the ligand exchange. Note, at first glance (Fig. 4a and Supplementary Figs. 33–34), the alkaline HER activities of the Pt-SACs with different axial-ligands scale with the first-electron-affinity of the halogen-atoms. In the following we tried to explore the physical insights of this axial-ligand effect.
a HER polarization curves of the Pt-SACs with different axial-ligands. Computational models and localized electric field distributions of (b) Cl-Pt/LDH and (c) HO-Pt/LDH. d Calculated energy barriers of water dissociation kinetics and (e) adsorption free energies of H* on the surfaces of the Pt-SACs, and the Pt (111) slab as a reference. f Calculated Pt 5d band structures of the Pt-SACs. g Relationship between the Ediso-ele and ΔGH*. h Relationship between alkaline HER overpotentials (at 10 mA cm−2) and variables. The Pt loading amount of Cl-Pt/LDHp, HO-Pt/LDHp, R-F-Pt/LDHp, R-Cl-Pt/LDHp, R-Br-Pt/LDHp and R-I-Pt/LDHp is 2.88×10−4, 2.88 × 10−4, 2.84 × 10−4, 3.00 × 10−4, 2.92 × 10−4, and 2.92 × 10−4 mgPt cm−2, respectively.
First, EIS measurements were carried out to investigate the detailed HER kinetics on these catalysts65,66. Based on the equivalent circuit (Supplementary Fig. 23), R2 and R3 are associated to the Volmer step and Heyrovsky step, respectively15,39,63,64. On Pt/C, the phase angle of the Heyrovsky step decreased rapidly while the phase angel of the Volmer step remain largely unchanged when increasing the overpotential for HER (Supplementary Fig. 35). In addition, R2 remains larger than R3 at all overpotentials, indicating the RDS for HER on Pt/C is always the Volmer step under the testing condition (Supplementary Fig. 36). On a contrary, phase angels of both the Volmer and Heyrovsky steps on Pt-SACs decrease along with the increase of overpotential, and R2 is mostly smaller than R3, suggesting that the Volmer step has been accelerated on the Pt-SACs consisting with the previous results (Supplementary Fig. 37 and Supplementary Tables 4–9).
First principal simulations were also performed to rationalize the axial-ligand effect on Pt-SACs for alkaline HER. The computational models for X-Pt/LDH (X = −F, −Cl, −Br, −I, and −OH) were built based on the EXAFS fitting results. As shown in Fig. 4b, c and Supplementary Figs. 38, 39, significant charge redistributions were observed at the Pt-X bonding regions, which is likely the origin of the enhanced HER activity. Stepwise reaction barriers for alkaline HER on Pt-SACs were modeled, including the Volmer step of water dissociation, *OH desorption, and the subsequent conversion of H* to H2. For the Volmer step (Fig. 4d), all the halogen-coordinated Pt/LDHs show stronger H2O adsorption ability and larger formation of enthalpy for water dissociation than those on Pt (111), leading to accelerated Volmer step. Among them, Cl-Pt/LDH exhibits the smallest energy barrier (0.073 eV) for the Volmer step. On another hand, the Cl-Pt-sites also exhibit the most optimized H binding energy among the Pt-SACs, leading to the most favored kinetics for the conversion of H* to H2 (Fig. 4e). Note, while Pt (111) shows slightly more optimized H binding energy compared to that of the Cl-Pt/LDH, its water dissociation step is expected to be slow (Fig. 4e), leading to a slower overall HER kinetics. The combined results indicate that the H binding energy is not the sole descriptor for alkaline HER, OH binding energy also plays significant role as it is likely not a spectator for the reaction67,68,69. The projected density of states (PDOS) of the d-orbitals of the Pt-SACs were calculated as well (Supplementary Fig. 40). Subsequently, the average energy levels of the narrow Pt−5d orbitals occupied by isolated electrons (Ediso-ele) were also calculated for Cl-Pt, F-Pt, HO-Pt, Br-Pt, and I-Pt as −0.901, −0.919, −0.977, −1.005, and −1.052 eV, respectively (Fig. 4f). This is in good agreement with the above XANES results, where the highest Pt valence state was observed for Pt in Cl-Pt/LDH owing to the largest first-electron-affinity of Cl ligand. According to the d-band theory, the adsorption properties of the surface intermediates are directly associated to the electronic structure of catalyst70,71. Increased first-electron-affinity of the ligands on Pt-single-atoms is expected to increase the Ediso-ele, and further weaken the hydrogen adsorption. Cl-Pt with the highest Ediso-ele should yield the weakest interaction with H* intermediate, consequently H* is more readily to desorb and form H2 (Fig. 4g). Overall, the computational results agree with our experimental observations, demonstrating that modifying the electronic structure of the Pt-single-sites with axial ligand is an effective strategy in tuning the HER activity.
We further conducted CO strip** experiments to evaluate the *OH binding strength on Pt-SACs, since one could expect that strong Pt-OH bond strength will facilitate the removal of *CO owing to the adsorbate-adsorbate interactions33,52. As shown in Supplementary Figs. 41–42, a volcano-type relationship between alkaline HER activity and Pt-*OH strength was observed, indicating both *OH adsorption and desorption are important for alkaline HER. Taken together, we show that the alkaline HER activity of X-Pt/SACs increase monotonically with the ligand first-electron-affinity, as both *H and *OH are the dominant descriptors during alkaline HER (Fig. 4h).
HER performance in an industrial relevant reactor
To evaluate the applicability of Cl-Pt/LDH as a cathodic catalyst under practical relevant operating conditions, we constructed an anion exchange membrane-based membrane electrode assemblies (MEA) electrolyser by using the pristine NiFe-LDH as the anode. MEA electrolysers based on commercial Pt/C and Ir/C were built and tested for comparison. The current-voltage profiles were obtained under 60 °C (Fig. 5a). Without further optimizations (reactor design, membrane choice and operating conditions, etc.), the electrolyser based on Cl-Pt/LDH / NiFe-LDH exhibits a much lower cell voltage (1.87 V) than those of the Cl-Pt/LDH / Ir/C (1.99 V) and the Pt/C / Ir/C (2.66 V) based electrolyser at current density of 1.0 A cm−2. Under this condition, their corresponding energy conversion efficiencies are 80%, 75% and 56%, respectively (Fig. 5b). The electrolysers were also subjected to water electrolysis for a 20 h-long stability tests at 1.0 A cm−2. As shown in Fig. 5c, the Cl-Pt/LDH / Ir/C electrolyser exhibits good stability with negligible overpotential loss (Supplementary Fig. 43), demonstrating promising potentials for future implementation for green hydrogen production. As shown in Fig. 5d–f, it is obvious that the activation overpotentials of Cl-Pt/LDH / NiFe-LDH or Cl-Pt/LDH / Ir/C are much lower that of Pt/C / Ir/C, leading to the enhanced energy efficiency originated from the superior HER activity of Cl-Pt/LDH.
a LSV curves of the MEA reactors using Cl-Pt/LDH / NiFe-LDH, Cl-Pt/LDH / Ir/C, and commercial Pt/C / Ir/C as the cathodic and anodic electrodes, respectively. b Cell voltage at current density of 1 A cm−2 in Fig. 5a, and the corresponding energy efficiency. c Stability tests of the MEA water electrolysers at the current density of 1 A cm−2 at 60 °C. Breakout of the ohmic overpotential and activation overpotential for (d) Cl-Pt/LDH / NiFe-LDH, (e) Cl-Pt/LDH / Ir/C and (f) commercial Pt/C / Ir/C. The Pt loading of Cl-Pt/LDH, and Pt/C is 1.01 × 10−2 and 0.8 mgPt cm−2, respectively.
In summary, we present Pt-SACs anchored on NiFe-LDH nanoarrays with different axial ligands (−F, −Cl, −Br, −I, and −OH) as highly active alkaline HER electrocatalysts prepared by a facile irradiation-impregnation procedure. Cl-Pt/LDH exhibits the highest HER activity among the Pt-SACs as well as the commercial Pt/C. The enhanced intrinsic activity of Cl-Pt/LDH towards alkaline HER is attributed to the largest first electron affinity of the axial ligand which poses strong electronic effect on the Pt-single-sites. DFT calculations suggest that the introduction of axial ligand on Pt-single-sites can tune the average energy levels of the isolated electrons occupied Pt-d-orbitals, and further the H* and *OH adsorption energies. Finally, Cl-Pt/LDH was also evaluated in a MEA-based alkaline water electrolyser at industrially relevant reaction rates, energy efficiency as high as 80% was obtained, demonstrating its promising application for hydrogen production. Overall, our work highlights the benefits of tailoring the chemical environments around the catalytic center, both the synthetic strategy and the constructive axial-ligand effect reported here can guide future SACs design with improved performance for large-scale green hydrogen production.
