Enriching surface oxygen vacancies of spinel Co₃O₄ to boost H₂O adsorption for HER in alkaline media

Abstract

Oxygen-defective Co₃O_2.64S_0.33 (namely CoO_0.88S_0.11) particulate spheres (PS) were prepared via the solid-state sulfurization of spinel Co₃O₄ at 300 °C in an Ar atmosphere. The introduction of sulfur (S) atoms was found to enrich the surface oxygen vacancies (V_so) remarkably in the as-derived CoO_0.88S_0.11 PS relative to pristine Co₃O₄ and CoS₂, thus offering more active sites for catalytic reactions. Theoretical calculations were also performed to justify that the enrichment of V_so is favorable to boost the adsorption of H₂O molecules. As a result, these CoO_0.88S_0.11 PS with exposed (111) facets demonstrated enhanced electrocatalytic performance in H₂ evolution reaction (HER) with a low overpotential (OP) of 83 mV to realize a current density of 10 mA cm⁻², and a high turnover frequency (TOF) of 1.589 H₂ s⁻¹ was obtained in 1 M KOH aqueous electrolyte.

---

1. Introduction

Cobalt oxides and sulfides, such as Co₃O₄, CoO, CoS, Co₃S₄, Co₉S₈, and CoS₂, have been widely explored in photo/electrochemical energy storage and conversion because of their intrinsic merits of low cost, high availability, and excellent chemical stability.^1,2 For example, Co₃O₄ has been intensively employed as an anode material in lithium ion batteries due to its high theoretical capacity and high reversibility.^3–5 As a typical p-type semiconductor, CoO is one of the most studied photocatalysts, which can utilize solar energy for water splitting by engineering the band edges.^6,7 Compared to their oxide counterparts, cobalt sulfides usually exhibit higher electroactivity and better electric conductivity, making them more promising in metal ion batteries,^8–10 supercapacitors,^11–13 and electrocatalysis.^14–16 Nano engineering and structural design are considered effective strategies to optimize the physical and chemical properties of cobalt oxides/sulfides,¹⁷ because the specific surface area, porous texture, electron transport, and chemical stability of the functional materials can be readily tuned and manipulated, so as to realize high specific capacities, prolonged cycle life, or excellent catalytic rates.¹⁸ However, the intrinsic properties of the respective cobalt oxides/sulfides are still not fully explored due to the poor understanding of the mechanisms for charge distribution and transfer in the electrochemical/catalytic process.

The electrocatalytic performance of particulate catalysts is highly related to atomic defects, such as oxygen vacancies.^19,20 For example, V_so are reported to influence the adsorption of zinc ions on the material surface, whereby the calculated Gibbs free energy of Zn²⁺ can be altered.²¹ A recent work has shown that V_so played a complicated role in the PEC process, where V_so can increase the carrier concentration to enhance charge transfer and suppress bulk recombination.²² Engineering of V_so in cobalt oxides has been attractive for electrocatalysis. Wang et al. demonstrated that the presence of V_so in Co₃O₄ has led to the formation of more Co²⁺ active centers on the Co₃O₄ surface, which is significant to improve the catalytic performance.²³ In a later work by the same group, they found that V_so are favorable to promote the reconstruction/deprotonation of intermediate Co–OOH˙, providing a strong evidence for the role of V_so in electrocatalytic oxygen evolution reaction (OER).²⁴ Ma et al. prepared defective Co₃O_4−x by calcining SBA-15 to downshift the conduction bands of the Co 3d orbitals, thus improving the electron transfer for electrocatalytic OER.²⁵ It has been reported that the substitution of chalcogen elements can be used to generate oxygen vacancies in a recent research,²⁶ and the enrichment of oxygen vacancies could be attributed to the extrusion of O by S atoms with a larger atomic radius. By surface engineering of Co₃O₄, V_so were generated by a deoxygenation process to optimize the band edges and electronic structures, which can facilitate charge transfer during the catalytic process. Though various strategies for V_so have been developed and many exciting results have been reported, controllable enrichment of V_so in spinel Co₃O₄ by partial S substitution has been rarely performed so far.

In this work, CoO_0.88S_0.11 PS with enriched V_so were prepared via the solid-state sulfurization of spinel Co₃O₄ PS at 300 °C in the presence of sulfur (S) powder in Ar gas. Specifically, the generation and preservation of V_so were realized by the introduction of appropriate S species in addition to deoxygenation in Ar gas. As a result, the as-synthesized CoO_0.88S_0.11 PS with exposed (111) facets exhibited enriched V_so, improved electronic conductivity, and a higher ECSA, relative to pristine Co₃O₄ and CoS₂ that was derived from complete sulfurization. When used as electrocatalysts for HER, these CoO_0.88S_0.11 PS demonstrated a low OP of 83 mV to realize a current density of 10 mA cm⁻² with a small Tafel slope of 80 mV dec⁻¹ (in 1 M KOH aqueous solution), which are much better than those of pristine Co₃O₄ (168 mV, 196 mV dec⁻¹) or CoS₂ (96 mV, 119 mV dec⁻¹). In addition, CoO_0.88S_0.11 PS can yield a high TOF of 1.589 H₂ s⁻¹, suggesting a great potential for practical electrocatalytic application. Density-functional theory (DFT) was also employed to calculate the adsorption energies of H₂O molecules at the (111) facets for all the samples to verify the experimental results. CoO_0.88S_0.11 with enriched V_so exhibited the largest H₂O adsorption energy of 0.4 eV (absolute value) among all the samples. This work has provided a new path of V_so engineering for metal oxide electrocatalysts to enhance HER.

2. Experimental section

2.1 Synthesis of the Co₃O₄ particulate spheres (PS)

The cobalt based precursor (CBP) PS were synthesized via a solvothermal method in an isopropanol (IPA) solution. In a typical synthesis procedure, 2 mmol Co (NO₃)₂·6H₂O was firstly dissolved in 30 mL of IPA by magnetic stirring for 10 min. Next, 1 mL of ethylene glycol (EG) was added to the above solution and the mixture was stirred for another 10 mins. Finally, 3 mg of polyvinyl pyrrolidone (PVP, average molecular mass ∼10 000 g mol⁻¹) was dissolved into the mixture by magnetic stirring for another 10 min. The obtained transparent lavender solution was transferred into a Teflon lined sealed stainless-steel autoclave and heated at 180 °C for 6 h. The light purple product was washed by a rinse-centrifugation process with ethanol several times. The collected CBP powder was dried in an air-flow oven at 60 °C for 24 h. Co₃O₄ PS were obtained by annealing the as-obtained CBP PS at 200 °C in a muffle furnace at a heating rate of 1 °C min⁻¹ for 2 h.

2.2 Synthesis of CoO_0.88S_0.11 PS

The collected Co₃O₄ powder (100 mg) was placed at the downstream side of a tube furnace, and sulfur powder (200 mg) was placed at the upstream side of the tube furnace. Then, Co₃O₄ was heated to 300 °C (heating rate of 3 °C min⁻¹) and reacted for 5 h under an argon (Ar) atmosphere. After the reaction, the tube furnace was allowed to cool down naturally to ambient temperature to collect the CoO_0.88S_0.11 PS product. In the control experiments, cobalt sulfide (CoS₂) PS were obtained using the same sulfurization process of Co₃O₄ powder but at a higher temperature of 400 °C. Partial sulfurization was also performed at 200, 250, and 350 °C using the same procedure to produce the Co₃O₄-200, 250, and 300 samples, respectively, which were specifically subjected to ESR measurements.

2.3 Materials characterization

All the samples were characterized by scanning electron microscopy (SEM, Quanta FEG 250) equipped with energy dispersive X-ray spectroscopy (EDX), transmission electron microscopy (TEM, Titan G2 60-300) equipped with an elemental mapping system, X-ray diffraction (XRD, Rigaku D/Max 2500, CuK_α = 1.5406 Å), and X-ray photoelectron spectroscopy (XPS, ESCALAB250Xi). Room-temperature electronic spin resonance (ESR) spectra were measured using an ESR spectrometer (Bruker A300) at 300 K. The frequency of microwave used is 9.853 GHz with a power of 10.8 mW. The magnetic center field is 3510 G with a sweep width of 100 G. The module amplitude is 1 G with a time constant of 1.25 mS and a sweep time of 19.456 s.

2.4 Electrochemical measurements.

(1) Alkaline electrolyte: all the electrocatalytic measurements were carried out using an IVIUM V38120 electrochemical workstation in a three-electrode cell with a 1 M of potassium hydroxide (KOH) aqueous solution as the electrolyte. The saturated calomel electrode (SCE) and graphite rod served as the reference and counter electrodes, respectively. The working electrodes were prepared as follows: 18 mg of catalyst powder, 2 mg of polyvinylidene fluoride (PVDF), and 0.2 mL of N-methyl-2-pyrrolidinone (NMP) were mixed together and then ground for 10 minutes to obtain a homogeneous slurry. The slurry was then pasted homogeneously onto a piece of nickel foam (1 cm × 2 cm, coating area ∼1 cm², mass loading ∼ 4 mg). Then the as-prepared working electrode was dried in a vacuum oven at 60 °C for 12 h.

(2) Acidic electrolyte: the electrochemical experiments were tested in standard three electrode system using an IVIUM V38120 electrochemical workstation, and the electrolyte was an aqueous solution of 0.5 M sulfuric acid (H₂SO₄). The saturated calomel electrode (SCE) and graphite rod served as the reference and counter electrodes, respectively. The working electrodes were prepared as follows: 4 mg of catalyst and 30 μm of Nafion solution were added to 1 mL of water–ethanol mixture solution (volume ratio of 3 : 1), and the solution was ultrasonicated for 1 h to form a homogeneous ink. Then 5 μL of the ink was dropped onto a glassy carbon electrode (GCE) with a 3 mm diameter.

Linear sweep voltammetry (LSV) was performed at a scan rate of 5 mV s⁻¹ in a potential window of 0 to −0.5 V versus reversible hydrogen electrode (RHE). Electrochemical impedance spectroscopy (EIS) was performed at the frequency from 0.1 MHz to 0.01 Hz with an amplitude of 5 mV. The current variation with time (i–t curve) under constant voltage was measured for 12 h. To simplify the evaluation, a current density of 10 mA cm⁻² was used to compare the OPs for all the samples. Electrochemical surface area (ECSA) measurements were performed within a non-faradaic voltage window ranging from −0.1 to − 0 V (vs. SCE). The lattice parameters of all the samples were obtained from PXRD to calculate the unit lattice volumes. The TOFs were then estimated based on the lattice volume, ECSA, and exchange current density using the following equation:²⁷

where the total hydrogen turnovers can be calculated from the current density collected from the LSV curves, the absolute value of current density is the value at zero overpotential, and active site density is the number of active sites per real surface area calculated from unit lattice volume.

2.5 Theoretical calculations

The calculations were conducted using density functional theory (DFT) with the Perdew–Burke–Ernzerhof (PBE) form of generalized gradient approximation functional (GGA).² The Vienna ab initio simulation package (VASP)^28,29 was used to calculate the exchange–correlation interaction of Co (3d⁷4s²), O (2s²2p⁴), S (3s²3p⁴), and H(1s¹) valence electrons with the plane wave energy cutoff at 400 eV. The gamma-centered K-point meshes were set as 3 × 3 × 3 for Co₃O₄ and CoS₂ bulk, and 2 × 2 × 1 for Co₃O₄(111) surface, CoO_0.88S_0.11(111) surface, and CoS₂(111)surface. The energy (converged to 1.0 × 10⁻⁶ eV per atom) and force (converged to 10⁻³ eV Å⁻¹) were set as the convergence criteria for geometry optimization.

We have obtained the XRD results of Co₃O₄ (JCPDS no. 74-2120: a = b = c = 0.8400 nm, space group Fd3̄m), and CoS₂ (JCPDS no. 89-1492: a = b = c = 0.5506 nm, space group PA3̄), and thus the basic crystal models of Co₃O₄, CoO_0.88S_0.11, and CoS₂ can be illustrated. The (111) facets of all the samples were identified from the XRD and HRTEM images. In all the optimization for structural calculations, the atoms of the bottom layers were fixed, while the positions of the other atoms were allowed to relax. A vacuum layer as large as 15 Å was used along the c direction normal to the surface to avoid periodic interactions. Oxygen vacancies are formed by deleting oxygen atoms. In the Co₃O₄ (111) surface model, there are 40 Co atoms, 48 O atoms, and 16 H atoms (for the bottom passivation). If one oxygen atom is deleted, there is one oxygen vacancy, and the percentage of oxygen vacancy is 100 × (1/48)%. If two oxygen atoms are deleted, there are two oxygen vacancies, and the percentage is 100 × (1/24)%. For the defective CoS₂ (111), one of the surface S atom was removed to simulate the S vacancy.

The chemisorption energies of H₂O on the (111) facets are described as follows:

where E_{H2O/catalytic(111)} is the total energy of H₂O on the (111) surface; E_H2O is the total energy of free H₂O and E_{catalytic(111)} is the total energy of the (111) surface. E_{H2O/catalytic(111)} and E_{catalytic(111)} were calculated with the same parameters. E_H2O was calculated by setting the isolated adsorbate in a box of 10 Å × 10 Å × 10 Å.

Based on the crystalline structures, the calculations of adsorption energies were carried out for the (111) facets with Co–O or Co–S terminals. Adsorption sites for perfect surfaces, one, and two units of O or S vacancies were built to simulate the surfaces with no defect or enriched defects.

3. Results and discussion

A schematic of the transformation from spinel Co₃O₄ to CoO_0.88S_0.11 with enriched V_so by S atom substitution via solid-state sulfurization is displayed in Fig. 1a. At an elevated temperature, sublimation of sulfur powder takes place and the S species enter the spinel structure of Co₃O₄, and then replace partial O atoms at the solid–gas interface. As a result, more O (R = 0.074 nm) atoms are squeezed out of the crystalline structure due to the much larger atomic radius of S (R = 0.102 nm), thus forming CoO_0.88S_0.11 with enriched oxygen vacancies at the surface. It should be noted that the crystalline structure is not altered after introduction of S species because a slight surface sulfurization is employed herein (recall experimental details).

Fig. 1 Schematic crystalline structures of spinel Co₃O₄ and the derived CoO_0.88S_0.11 with enriched surface oxygen vacancies by S substitution (a), and diagrams of H₂O adsorption sites (b–d) for Co₃O₄ (111) with one unit of O vacancy (b), CoO_0.88S_0.11 (111) with two units of O vacancies (c), and CoS₂ (111) with one unit of S vacancy (d).

It has been reported that the presence of surface defects is important to promote the intrinsic properties of catalysts. As for electrocatalytic water splitting, adsorption of H₂O molecules at the catalyst surface is the dominant step in the HER process. Thus, DFT calculations were then performed to compare the H₂O adsorption energies of the (111) facets (will be discussed in XRD and HR-TEM later) for all the three samples. The detailed calculation parameters are described in the experimental part, and corresponding simulation models are presented in Fig. S1 (ESI†). Fig. 1b–d show the diagrams of the adsorption sites (ads) of Co₃O₄ (111) with Co–O terminals, CoO_0.88S_0.11 (111) with Co–O terminals, and CoS₂ (111) with Co–S terminals, respectively. Specifically, each Co atom at site 1 of Co₃O₄ (111) is bonded with four O atoms, and contains one unit of O vacancy, as shown in Fig. 1b. Similarly, each Co atom at site 1 of CoO_0.88S_0.11 (111) in Fig. 1c is bonded with two O and one S atoms, and contains two units of O vacancies (denoting enriched V_so). Meanwhile for CoS₂ (111), each Co atom at site 1 is bonded with three S atoms, and contains one unit of S vacancy. The calculation results have revealed that the adsorption energies (ΔE_ads) at sites 1 for Co₃O₄, CoO_0.88S_0.11, and CoS₂ are −0.25, −0.40, and −0.15 eV, respectively. Significantly, CoO_0.88S_0.11 (111) with one unit of O vacancy also shows a larger ΔE_ads of −0.39 eV, compared to its Co₃O₄ and CoS₂ counterparts (more calculations shown later). A larger absolute number of ΔE_ads suggests a more stable adsorption of H₂O molecules at the catalyst surface, which is favorable to expedite the HER process. These calculations have preliminarily implied the importance of surface sulfurization with enriched V_so for improving electrocatalytic performance.

Following our theoretical and experimental methods, cobalt based materials were then synthesized and characterized. Cobalt based precursor (CBP) PS were prepared using a solvothermal process (SEM images shown in Fig. S2, ESI†). Next, the as-synthesized CBP PS were transformed to Co₃O₄ through a heat treatment at 200 °C in air, which were subsequently used for partial or complete sulfurization. As the typical sample of this work, CoO_0.88S_0.11 PS were examined by SEM and EDX, as shown in Fig. 2. From the SEM results, monodispersed spheres with uniform particle size (∼2 μm) can be observed (Fig. 2a and b). A magnified SEM image in Fig. 2c reveals that a particle is composed of nanosheets with a porous texture. The overall morphology of the PS is well retained after partial sulfurization, compared to SEM results of Co₃O₄ PS (Fig. S3a–c, ESI†). A complete sulfurization from Co₃O₄ to CoS₂ has led the conversion of nanosheets to aggregated particles, as shown in Fig. S3d–f (ESI†). Then, the EDX information of the typical sample in Fig. 2d was captured from the area indicated by the yellow rectangle in Fig. 2c, which confirms the presence of Co, O, and S elements (atomic ratio is roughly at 1 : 0.88 : 0.11, hence denoted as CoO_0.88S_0.11) in the PS. In addition, elemental mappings were performed to find out the distributions of Co, O, and S, as shown in Fig. 2e–g. It can be seen that Co and O signals are almost overlapped, while S shows a slightly larger area than Co and O, suggesting a partial surface sulfurization of the pristine Co₃O₄ particles.

Fig. 2 SEM images (a–c), EDX (d), and elemental mappings (e–g) of the as-prepared CoO_0.88S_0.11 PS.

The detailed structural, compositional, and electronic features were further investigated by TEM and ESR. Fig. 3a displays an individual CoO_0.88S_0.11 PS, in which a hollow interior can be clearly observed. The magnified TEM image in Fig. 3b indicates that PS are composed of porous nanosheet structures, suggesting a high surface area of the material. Fig. 3c shows the HRTEM image which was taken from the area indicated by a yellow oval in Fig. 3b. The lattice spacings of 0.57 and 0.28 nm can correspond to Co₃O₄ (111) and (220), respectively.³⁰ Notably, the spacing of (111) was enlarged from the usual ∼0.46 nm to 0.57 nm, which should be caused by the substitution of O by S with a larger atomic radius. In addition, a lattice distortion of ∼17° is observed for the (111) planes, demonstrating that the introduction of S heteroatoms may have deformed the periodic crystalline structure of Co₃O₄. The TEM and HR-TEM images were also obtained and presented for pristine Co₃O₄ and CoS₂ (Fig. S4, ESI†), which clearly reveal the high exposure of Co₃O₄ (111) and CoS₂ (111). To confirm the existence of V_so, ESR tests were carried out for pristine Co₃O₄, CoS₂, and the partially sulfurized Co₃O₄ samples at different temperatures, respectively. The ESR spectra in Fig. 3d clearly show that CoO_0.88S_0.11 (Co₃O₄-300) delivers a much higher signal than Co₃O₄ or CoS₂ at g = 2.004, which indicates the appearance of oxygen vacancies.³¹ Unexpectedly, CoS₂ exhibited only a bit higher signal than pristine Co₃O₄, which suggests that complete substitution of O with S may have changed the crystalline structure, thus causing the oxygen vacancies to vanish. In addition, it is interesting to find that lower sulfurization temperatures (Co₃O₄-200 and 250) failed to generate a large number of V_so, whereas higher temperature (Co₃O₄-350) was not favorable to preserve the O vacancies.

Fig. 3 TEM (a and b) and HRTEM (c) images of CoO_0.88S_0.11 PS, and the ESR results (d) of pristine Co₃O₄, CoS₂, and the partially sulfurized Co₃O₄ at different temperatures (Co₃O₄-300 is CoO_0.88S_0.11); the signal at g = 2.004 corresponds to the O vacancies.

The XRD patterns of all the three samples are displayed in Fig. S5 (ESI†). All the peaks recorded from 10–80° (Fig. S5a, ESI†) for Co₃O₄ can be assigned to spinel Co₃O₄ (JCPDS no. 74-2120),³² while all the identified peaks for CoS₂ should be assigned to cubic CoS₂ (JCPDS no. 89-1492).³³ For CoO_0.88S_0.11, only peaks of Co₃O₄ can be matched, indicating that the crystalline phase is not altered under such a slight sulfurization. However, slight shifts can still be distinguished in the diffraction peaks for all the XRD peaks of CoO_0.88S_0.11. For example, as shown in the magnified region in Fig. S5b (ESI†), the peak of Co₃O₄ (311) was shifted from 36.86° to 36.78° after partial sulfurization, which should be caused by the replacement of the O positions by S atoms. In addition, it can be calculated that the unit cell parameter of CoO_0.88S_0.11 is 0.8098 nm, which has been reduced compared to that of Co₃O₄. To further examine the chemical composition and status of the typical sample CoO_0.88S_0.11, XPS was carried out and the results are shown in Fig. 4. The survey spectrum in Fig. 4a confirms again the presence of Co, O, and S. The paired peaks of 779.7 and 795.7 eV shown in Fig. 4b can be attributed to Co 2p_3/2 and Co 2p_1/2, respectively, corresponding to the characteristic spin orbitals of Co³⁺.³⁴ Co³⁺ has shifted toward lower binding energies after partial sulfurization compared to pristine Co₃O₄ (781.1 and 796.4 eV, respectively, Fig. S6a, ESI†), which is consistent with a previous work that demonstrates oxygen defects in Co₃O₄.³⁵ In addition, the presence of two satellite peaks centered at around 786 and 803 eV, respectively, should be attributed to the Co²⁺ oxidation state, which indicates that part of Co³⁺ has been reduced to Co²⁺ during the formation of V_so.³⁶ There are two binding energies for O 1s (Fig. 4c), which can be assigned to Co–O (529.3 eV),³⁷ and O⁻ (531.2 eV) related to oxygen vacancies,³⁸ respectively. Both Co–O and O⁻ shifted to lower binding energies compared to those of pristine Co₃O₄ (530.7 and 532 eV, respectively, Fig. S6b, ESI†). The XPS peaks located at 162.3 and 164 eV (Fig. 4d) should be ascribed to S 2p_3/2 and S 2p_1/2, respectively.³⁹ The broad peak at 168.6 eV can be assigned to S–O bond associated with SO_x.⁴⁰ Fig. S7 (ESI†) shows the XPS results of the CoS₂ sample, in which strong peaks of S–O bonds (oxidized S, 168–170 eV) can be distinguished. Combining the results of TEM, XRD, and XPS analyses, we have demonstrated that S species were introduced into the lattice of Co₃O₄ without altering the crystalline phase. However, S atoms may replace O positions in the Co₃O₄ crystal structure, and thus induce the generation of V_so with lattice distortion, which is essential to increase the intrinsic catalytic activity.

Fig. 4 XPS survey spectrum (a), Co 2p (b), O 1s (c), and S 2p (d) of the as-prepared CoO_0.88S_0.11 PS.

The HER properties were evaluated for all the samples using a standard three-electrode configuration in 1 M KOH. In an alkaline electrolyte, the HER process can proceed as the following Volmer–Heyrovsky and Volmer–Tafel reactions (1)–(3):⁴¹

H₂O + e⁻ → H* (Volmer step) (1)

H* + H₂O + e⁻ → H₂ (Heyrovsky step) (2)

H* + H* → H₂ (Tafel step) (3)

The reactions start with the adsorption of H₂O molecules on the catalyst surface to generate H* via the Volmer-step reduction process. Next, H₂ molecules will be formed by the electrochemical bonding of H₂O with H* (Heyrovsky step), or through the chemical combination of two H* (Tafel step). By increasing the adsorption energy and reducing the dissociation energy barrier, the HER process can be efficiently facilitated in the presence of V_so.⁴² In this work, H₂O adsorption energies in the presence of O vacancies in different samples have been studied using DFT calculations (more results and discussion later).

Fig. 5a shows the LSV curves of all the samples. The OPs at a current density of 10 mA cm⁻² for Co₃O₄, CoS₂, CoO_0.88S_0.11, and Pt/C are recorded as 168, 96, 83, and 23 mV, respectively. Additionally, an OP of 329 mV is required for CoO_0.88S_0.11 to realize a current density of 100 mA cm⁻², which is the best performance among all the three samples (all the data are iR corrected). Tafel slopes are used to describe the catalytic kinetics of electrocatalysts, and a smaller Tafel slope usually represents faster kinetics. Tafel slopes of 196, 119, 80, and 32 mV dec⁻¹ were calculated for samples Co₃O₄, CoS₂, CoO_0.88S_0.11, and Pt/C, respectively (Fig. 5b). The LSV curves of the partially sulfurized samples obtained at 200, 250, and 350 °C were also collected, using which the OPs were calculated to be 176, 151, and 113 mV, respectively (Fig. S8, ESI†). To evaluate the stability of the materials in an acidic electrolyte, electrochemical measurements were also carried out in a sulfuric acid aqueous solution (0.5 M H₂SO₄), and results have been shown in Fig. S9 (ESI†). The OPs at a current density of 10 mA cm⁻² for the samples Co₃O₄, CoS₂, and CoO_0.88S_0.11 are determined to be 205, 157, and 116 mV (Fig. S9a, ESI†), respectively, while Fig. S9b (ESI†) shows that the Tafel slopes of Co₃O₄, CoS₂, and CoO_0.88S_0.11 are 180, 121, and 87 mV dec⁻¹, respectively. As a comparison, a summarization table (Table S1, ESI†) listing similar HER electrocatalysts reported in the previous work is also provided, revealing that the electrocatalytic performance of this work is comparable to those reported in recent years.

Fig. 5 LSV curves (a), Tafel plots (b), EIS (c), and ECSA (d) of the as-prepared Co₃O₄, CoS₂, and CoO_0.88S_0.11 samples. LSV and Tafel slope of Pt/C electrode are also shown as a comparison.

To further study the difference in the HER process of all the three samples, EIS and ECSA measurements were performed and the results were plotted. It can be seen from Fig. 5c that all the Nyquist plots show similar semicircles, from which transfer resistances of 6.2, 4.0, and 3.3 Ω can be calculated for Co₃O₄, CoS₂, and CoO_0.88S_0.11, respectively. The inset in Fig. 5c shows an equivalent circuit, which contains constant phase element (CPE), electrolyte resistance (R_s), and charge transfer resistance (R_ct). The smallest R_ct value of CoO_0.88S_0.11 suggests its best electronic conductivity among all the samples, which verified its best HER performance. Next, CV tests (Fig. S10, ESI†) were conducted in a non-faradaic potential region (−0.1 to 0 V, vs. SCE) to estimate the ECSA for the samples (Fig. 5d). Accordingly, the double-layer capacitances (C_dl) for Co₃O₄, CoS₂, and CoO_0.88S_0.11 are estimated to be 22.6, 34.1, and 44.9 mF cm⁻², respectively. The ECSA is considered positively proportional to C_dl, and thus CoO_0.88S_0.11 holds the largest ECSA among all the samples. Catalytic materials with higher surface areas usually show better electrocatalytic performance, because they can provide larger numbers of active sites for catalytic reactions.^43,44 Turnover frequency (TOF) indicates the numbers of gaseous H₂ molecules evolved in the HER process at a unit time, which is regarded as the most important parameter to reflect the intrinsic catalytic properties. Herein, the TOFs of all the three samples were calculated and compared according to a calculation method described in a previous work.²⁷ CoO_0.88S_0.11 yielded a TOF of 1.589 H₂ s⁻¹, which is much higher than those of Co₃O₄ (0.307 H₂ s⁻¹) or CoS₂ (0.806 H₂ s⁻¹).

Electrocatalytic stability is considered important for practical application. Durability tests were performed with respective OPs to maintain a current density of 10 mA cm⁻² for all the three samples. The obtained results (Fig. 6) show that the current densities of both Co₃O₄ and CoS₂ decrease upon catalytic time, but CoS₂ keeps a higher current than Co₃O₄. Remarkably, CoO_0.88S_0.11 has shown a much more stable current density after 2 h compared to the other samples, which suggests that it possesses the best durability among all the three samples. Apparently, a quick current decay of Co₃O₄ was observed after 2 h, which should be due to its poor conductivity, and the mass loss of the electrode materials during the prolonged cycles.⁴⁵ After 12 h, current densities of 8.91, 7.62, and 6.33 mA cm⁻² can be retained for samples CoO_0.88S_0.11, CoS₂, and Co₃O₄, respectively, indicating the lowest current decay of 10.9% for CoO_0.88S_0.11.

Fig. 6 i–t curves obtained under the respective OPs of the three samples in 1 M of KOH aqueous solution. The inset schematically shows a HER process enhanced by enriched V_so of CoO_0.88S_0.11.

The inset in Fig. 6 shows a schematic illustration of the HER process that has been promoted by enriched V_so of CoO_0.88S_0.11. The Co atoms serve as the main active sites for catalytic reactions, where water molecules are reduced to generate H₂. Additionally, V_so present at the lattice of the spinel structure (Co atom and V_so indicated by the solid circle) can act as extra active centers for absorbance of H₂O and subsequent reduction. Thus, electrocatalytic H₂ evolution can be strengthened by such an enhancement. Furthermore, the LSV curves of all the three samples after stability tests were also obtained, as shown in Fig. S11 (ESI†). The results indicate that the catalytic activities of the samples Co₃O₄ and CoO_0.88S_0.11 are improved after durable tests (reduced OPs for η₁₀₀), while a decay of catalytic properties can be observed for CoS₂, which is probably due to its relatively low chemical stability.

The electrocatalytic results have shown that CoS₂ is superior over Co₃O₄. To further study the role of V_so in the H₂O adsorption process, DFT calculations were then extended to simulate the perfect (111) facets without any defects for all the three samples. Fig. 7 showcases the ΔE_ads results at sites 1 for Co₃O₄ (111) with Co–O terminals (Fig. 7a), CoO_0.88S_0.11 with Co–O terminals (Fig. 7b), and CoS₂ (111) with Co–S terminals (Fig. 7d), respectively. The absolute values of adsorption energies are in the order of (Co₃O₄, 0.18 eV) < (CoO_0.88S_0.11, 0.23 eV) < (CoS₂, 0.28 eV). Hence, CoS₂ (111) exhibits the most stable adsorption for H₂O molecules among all the three samples based on a perfect crystalline structure mode. However, our previous ESR results (recall Fig. 3d) have confirmed a remarkable V_so enrichment in the CoO_0.88S_0.11 sample, compared to the pristine Co₃O₄ or CoS₂. Then the ΔE_ads at site 1 was also calculated for CoO_0.88S_0.11 (111) with one unit of O vacancy. Interestingly, the absolute value of ΔE_ads was increased from 0.23 to 0.39 eV, which is much higher than that of CoS₂. There results strongly validated that the presence and enrichment of vacancies can play a very significant role in the adsorption of water molecules.

Fig. 7 Diagrams of H₂O adsorption sites for the (111) facets without any defects (a, b, and d) for Co₃O₄ (a), CoO_0.88S_0.11 (b), and CoS₂ (d), respectively, and that of CoO_0.88S_0.11 (111) with one unit of O vacancy (c).

In addition, further calculations for adsorption sites 2 with perfect lattices have obtained the ΔE_ads values of −0.03, −0.27, and −0.66 eV for Co₃O₄, CoO_0.88S_0.11, and CoS₂, respectively, which suggests that CoS₂ is a very good electrocatalytic candidate for HER. However, in fact it is very difficult to remove all the defects within the crystalline structures. For CoS₂ (111), the ΔE_ads of site 1 decrease to −0.15 eV when there is one unit of S vacancy, compared to −0.28 eV obtained without any defects. Therefore, a proper S substitution of Co₃O₄ by partial and slight sulfurization to enrich surface defects reported herein is a rational strategy to optimize the intrinsic catalytic properties.

4. Conclusions

In summary, CoO_0.88S_0.11 electrocatalysts with enriched V_so were prepared via a partial sulfurization process of spinel Co₃O₄ with exposed (111) facets. The introduction of S atoms into the crystalline lattice of spinel Co₃O₄ has led to the generation and preservation of lattice distortion with V_so. The as-formed CoO_0.88S_0.11 PS with V_so exhibit a higher ECSA and better electronic conductivity compared to pristine Co₃O₄ and CoS₂. The DFT calculations have verified that enriched V_so can significantly increase the adsorption energies of H₂O during the HER process. In virtue of the presence of V_so, high ECSA, and good charge transfer properties, CoO_0.88S_0.11 PS have demonstrated a low OP (η₁₀ = 83 mV) with an enhanced TOF (1.589 H₂ s⁻¹) for HER when used as an electrocatalyst for water splitting. This research work provides a new path for surface defect engineering of cobalt based materials with improved physical and chemical properties as efficient electrocatalysts.

Author contributions

T. Zhu conceived the idea and designed the experiments; J. Pan carried out the experiments, and collected and analyzed the data; Y. Xiao, A. Q. Pan, and S. Q. Liang provided assistance in analyzing the data; T. Zhu wrote the manuscript and S. Q. Liang provided assistance in the revision; T. Zhu and S. Q. Liang provided financial support to this work.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgements

This work was financially supported by the start-up grant of Central South University (No. 202045001), Innovation-Driven Project of Central South University (No. 2019CX028), the Huxiang Assembly Program for High-level Talents (Department of Science and Technology of Hunan province, China, No. 2018RS3018), and the National Natural Science Foundation of China (Key Program, No. 51932011). The authors also thank Prof. Lei Han from Hunan University for his valuable comments and discussion.

References

1. X. Y. Yu and X. W. Lou, Adv. Energy Mater., 2018, 8, 1701592.
2. X. F. Lu, S. L. Zhang, E. B. Shangguan, P. Zhang, S. Y. Gao and X. W. Lou, Adv. Sci., 2020, 7, 2001178.
3. Y. G. Li, B. Tan and Y. Y. Wu, Nano Lett., 2008, 8, 265–270.
4. Z. S. Wu, W. C. Ren, L. Wen, L. B. Gao, J. P. Zhao, Z. P. Chen, G. M. Zhou, F. Li and H. M. Cheng, ACS Nano, 2010, 4, 3187–3194.
5. D. L. Wang, Y. C. Yu, H. He, J. Wang, W. D. Zhou and H. D. Abruna, ACS Nano, 2015, 9, 1775–1781.
6. L. B. Liao, Q. H. Zhang, Z. H. Su, Z. Z. Zhao, Y. N. Wang, Y. Li, X. X. Lu, D. G. Wei, G. Y. Feng, Q. K. Yu, X. J. Cai, J. M. Zhao, Z. F. Ren, H. Fang, F. Robles-Hernandez, S. Baldelli and J. M. Bao, Nat. Nanotechnol., 2014, 9, 69–73.
7. W. L. Shi, F. Guo, C. Zhu, H. B. Wang, H. Li, H. Huang, Y. Liu and Z. H. Kang, J. Mater. Chem. A, 2017, 5, 19800–19807.
8. T. Chen, Z. W. Zhang, B. R. Cheng, R. P. Chen, Y. Hu, L. B. Ma, G. Y. Zhu, J. Liu and Z. Jin, J. Am. Chem. Soc., 2017, 139, 12710–12715.
9. H. C. Li, H. C. Yang, Z. H. Sun, Y. Shi, H. M. Cheng and F. Li, Nano Energy, 2019, 56, 100–108.
10. B. L. Guan, S. Y. Qi, Y. Li, T. Sun, Y. G. Liu and T. F. Yi, J. Energy Chem., 2021, 54, 680–698.
11. Z. Y. Gao, C. Chen, J. L. Chang, L. M. Chen, P. Y. Wang, D. P. Wu, F. Xu and K. Jiang, Chem. Eng. J., 2018, 343, 572–582.
12. C. Li, J. Balamurugan, N. H. Kim and J. H. Lee, Adv. Energy Mater., 2018, 8, 1702014.
13. H. Y. Wang, Y. Yang, Q. H. Li, W. Lu, J. Q. Ning, Y. J. Zhong, Z. Y. Zhang and Y. Hu, Sci. China-Mater., 2021, 64, 840–851.
14. S. J. Deng, Y. Zhong, Y. X. Zeng, Y. D. Wang, X. L. Wang, X. H. Lu, X. H. Xia and J. P. Tu, Adv. Sci., 2018, 5, 1700772.
15. H. J. Xu, J. Cao, C. F. Shan, B. K. Wang, P. X. Xi, W. S. Liu and Y. Tang, Angew. Chem., Int. Ed., 2018, 57, 8654–8658.
16. L. H. Li, L. Song, H. Guo, W. Xia, C. Jiang, B. Gao, C. Wu, T. Wang and J. P. He, Nanoscale, 2019, 11, 901–907.
17. X. Wang, Y. Chen, Y. J. Fang, J. T. Zhang, S. Y. Gao and X. W. Lou, Angew. Chem., Int. Ed., 2019, 58, 2675–2679.
18. J. Joo, T. Kim, J. Lee, S. I. Choi and K. Lee, Adv. Mater., 2019, 31, 1806682.
19. S. Liu, C. Chen, Y. Zhang, Q. Zheng, S. Zhang, X. Mu, C. Chen, J. Ma and S. Mu, J. Mater. Chem. A, 2019, 7, 14466–14472.
20. H. Zhang, J. Zhang, Y. Li, H. Jiang and C. Li, J. Mater. Chem. A, 2019, 7, 13506–13510.
21. T. Xiong, Z. G. Yu, H. J. Wu, Y. H. Du, Q. D. Xie, J. S. Chen, Y. W. Zhang, S. J. Pennycook, W. S. V. Lee and J. M. Xue, Adv. Energy Mater., 2019, 9, 1803815.
22. Z. L. Wang, X. Mao, P. Chen, M. Xiao, S. A. Monny, S. C. Wang, M. Konarova, A. J. Du and L. Z. Wang, Angew. Chem., Int. Ed., 2019, 58, 1030–1034.
23. L. Xu, Q. Q. Jiang, Z. H. Xiao, X. Y. Li, J. Huo, S. Y. Wang and L. M. Dai, Angew. Chem., Int. Ed., 2016, 55, 5277–5281.
24. Z. H. Xiao, Y. C. Huang, C. L. Dong, C. Xie, Z. J. Liu, S. Q. Du, W. Chen, D. F. Yan, L. Tao, Z. W. Shu, G. H. Zhang, H. G. Duan, Y. Y. Wang, Y. Q. Zou, R. Chen and S. Y. Wang, J. Am. Chem. Soc., 2020, 142, 12087–12095.
25. H. Y. Zeng, M. Oubla, X. P. Zhong, N. Alonso-Vante, F. Du, Y. Xie, Y. H. Huang and J. W. Ma, Appl. Catal., B, 2021, 281, 119535.
26. A. Karmakar, K. Karthick, S. Kumaravel, S. S. Sankar and S. Kundu, Inorg. Chem., 2021, 60, 2023–2036.
27. B. Dutta, Y. Wu, J. Chen, J. Wang, J. K. He, M. Sharafeldin, P. Kerns, L. Jin, A. M. Dongare, J. Rusling and S. L. Suib, ACS Catal., 2019, 9, 456–465.
28. O. Bokanowski, I. Schindler and H. Zidani, Nonlinear Anal.: Theory Methods Appl., 1999, 35, 1073–1090.
29. E. J. Baerends, Theor. Chem. Acc., 2000, 103, 265–269.
30. Y. Y. Lu, W. W. Zhan, Y. He, Y. T. Wang, X. J. Kong, Q. Kuang, Z. X. Xie and L. S. Zheng, ACS Appl. Mater. Interfaces, 2014, 6, 4186–4195.
31. A. Janotti and C. G. Van de Walle, Appl. Phys. Lett., 2005, 87, 122102.
32. Y. Z. Liu, X. W. Chi, Q. Han, Y. X. Du, J. Q. Huang, X. H. Lin and Y. Liu, Nanoscale, 2019, 11, 5285–5294.
33. D. Ma, B. Hu, W. D. Wu, X. Liu, J. T. Zai, C. Shu, T. T. Tsega, L. W. Chen, X. F. Qian and T. L. Liu, Nat. Commun., 2019, 10, 3367.
34. J. J. Zhu, K. Kailasam, A. Fischer and A. Thomas, ACS Catal., 2011, 1, 342–347.
35. Z. C. Wang, W. J. Xu, X. K. Chen, Y. H. Peng, Y. Y. Song, C. X. Lv, H. L. Liu, J. W. Sun, D. Yuan, X. Y. Li, X. X. Guo, D. J. Yang and L. X. Zhang, Adv. Funct. Mater., 2019, 29, 1902875.
36. Y. C. Wang, T. Zhou, K. Jiang, P. M. Da, Z. Peng, J. Tang, B. A. Kong, W. B. Cai, Z. Q. Yang and G. F. Zheng, Adv. Energy Mater., 2014, 4, 1400696.
37. L. W. Chen, X. Zuo, S. J. Yang, T. M. Cai and D. H. Ding, Chem. Eng. J., 2019, 359, 373–384.
38. W. B. Hu, Y. Liu, R. L. Withers, T. J. Frankcombe, L. Noren, A. Snashall, M. Kitchin, P. Smith, B. Gong, H. Chen, J. Schiemer, F. Brink and J. Wong-Leung, Nat. Mater., 2013, 12, 821–826.
39. N. Yao, P. Li, Z. R. Zhou, R. Meng, G. Z. Cheng and W. Luo, Small, 2019, 15, 1901993.
40. Y. J. Fang, D. Y. Luan, Y. Chen, S. Y. Gao and X. W. Lou, Angew. Chem., Int. Ed., 2020, 59, 2644–2648.
41. Y. P. Lei, Y. C. Wang, Y. Liu, C. Y. Song, Q. Li, D. S. Wang and Y. D. Li, Angew. Chem., Int. Ed., 2020, 59, 20794–20812.
42. Z. Cai, Y. M. Bi, E. Y. Hu, W. Liu, N. Dwarica, Y. Tian, X. L. Li, Y. Kuang, Y. P. Li, X. Q. Yang, H. L. Wang and X. M. Sun, Adv. Energy Mater., 2018, 8, 1701694.
43. Q. C. Wang, Y. P. Lei, Y. C. Wang, Y. Liu, C. Y. Song, J. Zeng, Y. H. Song, X. D. Duan, D. S. Wang and Y. D. Li, Energy Environ. Sci., 2020, 13, 1593–1616.
44. Y. P. Lei, Q. C. Wang, S. J. Peng, S. Ramakrishna, D. Zhang and K. C. Zhou, Adv. Energy Mater., 2020, 10, 1902115.
45. Y. L. Tong, H. Q. Liu, M. Z. Dai, L. Xiao and X. Wu, Chin. Chem. Lett., 2020, 31, 2295–2299.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1ma00566a

---