Microwave-pulse assisted synthesis of tunable ternary-doped 2D molybdenum carbide for efficient hydrogen evolution

Chemicals

All chemicals were commercially acquired and utilized as received without additional purification. The chemical compounds Na₂MoO₄·2H₂O (CAS: 10102-40-6, purity: 99%, analytical reagent grade) and Na₂S·9H₂O (CAS: 1313-84-4, purity: 99.9%, metal-based) were procured from Shanghai Maclin Biochemistry Co., LTD., China. Additionally, Sinopharm Chemical Reagent Co. Ltd. in China provided C₆H₁₂O₆ (CAS: 14431-43-7, analytical reagent grade), H₂C₂O₄ (CAS: 6153-56-6, analytical reagent grade), CH₄N₂O (CAS: 57-13-6, analytical reagent grade), and NaH₂PO₄ (CAS: 7558-80-7, analytical reagent grade).

Synthesis of 2D P,N,S-Mo₂C

Initially, 6 mg NaH₂PO₄ (phosphorus source) and 12 mg Na₂S·9H₂O (sulfur source) were added to 1 mL deionized water under the condition that the non-metallic atomic ratio P:N:S = 1:1:1. Then, 60 mg C₆H₁₂O₆ (carbon source), 206 mg Na₂MoO₄·2H₂O (molybdenum source), 60 mg H₂C₂O₄ (blowing agent) and 1.5 mg CH₄N₂O (nitrogen source) were added, fully mixed, and natural dried for 6 h to obtain a very uniform mixture. The evenly mixed drugs were transferred into a quartz tube with an outer diameter of 16 mm, inner diameter of 12 mm, wall thickness of 2 mm, and length of 60 mm. The tube was placed in a Grant G90F25CN3LN microwave oven for the reaction. First, nitrogen was injected into the reaction chamber for 10 min at a flow rate of 0.8 L/min. Subsequently, the sample was pulse heated three times for 1 min each with an output power of 900 W, under continuous nitrogen flow. After heating, the nitrogen flow was maintained until the apparatus cooled naturally to room temperature. Finally, the samples were taken out, rinsed with distilled water, and desiccated in a vacuum oven at 80 °C for 24 h to obtain 2D P,N,S-Mo₂C.

Synthesis of other doped samples

The same method was used for preparing other doped samples, with the following modifications based on the dopant: A nitrogen-doped 2D Mo₂C (2D N-Mo₂C) sample was obtained by adding 1.5 mg CH₄N₂O as a nitrogen source. A phosphorus-doped 2D Mo₂C (2D P-Mo₂C) sample was obtained by adding 6 mg of NaH₂PO₄ as a phosphorus source. A sulfur-doped 2D Mo₂C (2D S-Mo₂C) sample was obtained by adding 12 mg of Na₂S·9H₂O as a sulfur source. A nitrogen and phosphorus-doped 2D Mo₂C (2D P,N-Mo₂C) sample was obtained by adding 1.5 mg CH₄N₂O as a nitrogen source and 6 mg NaH₂PO₄ as a phosphorus source, respectively. A phosphorus and sulfur-doped 2D Mo₂C (2D P,S-Mo₂C) sample was obtained by adding 6 mg NaH₂PO₄ as a phosphorus source and 12 mg Na₂S·9H₂O as a sulfur source, respectively. A nitrogen and sulfur-doped 2D Mo₂C (2D N,S-Mo₂C) sample was obtained by adding 1.5 mg CH₄N₂O as a nitrogen source and 12 mg of Na₂S·9H₂O as a sulfur source, respectively. All steps are conducted under specific conditions to ensure the balance between the 2D structure and precise doping.

Material characterization

The structural composition of the samples was analyzed using X-ray diffraction (XRD) with Cu Kα radiation on an Ultima IV X-ray powder diffractometer and X-ray photoelectron spectroscopy (XPS) using a Thermo Scientific K-Alpha instrument. Their morphologies were examined using high-resolution field emission scanning electron microscopy (FESEM) on a JSM-7800 F instrument (Japan) and transmission electron microscopy (TEM) on a JEM 2100F instrument (Japan). The thickness of 2D P,N,S-Mo₂C was determined by atomic force microscopy (AFM) using a Bruker Dimension Icon instrument (Germany).

Electrochemical performance

All electrochemical experiments are conducted using rotating disk electrodes (RDEs) of the AFMSRCE model from PINE Instruments in the United States. The electrochemical experiments employed a three-electrode setup with a CHI Electrochemical Station (Model CHI660E, China). As the working electrode, a mirror-polished glassy carbon electrode with a diameter of 0.2 cm^-2 was utilized, while the reference electrodes included Hg/HgO (1 M KOH) and Ag/AgCl (0.5 M H₂SO₄), and the counter electrode was a carbon rod. In a solvent mixture comprising 0.2 mL of deionized water, 0.8 mL of isopropanol, and 20 μL of Nafion (5%, mass fraction), 5 mg of catalyst powder was dispersed. Following that, the mixture underwent ultrasonication in an ice bath for an hour to achieve a homogeneous ink. Afterward, a mirror-polished glassy carbon electrode was drop-cast with 20 μL of the ink, with a catalyst loading of approximately 0.55 mg·cm^-2. The potential, referenced to a Hg/HgO electrode, was converted to the potential versus the reversible hydrogen electrode (RHE) using E(RHE) = E(Hg/HgO) + E(Hg/HgO vs. SHE) + 0.059 pH. The potential, measured against a Ag/AgCl electrode was converted as: E(RHE) = E(Ag/AgCl) + E(Ag/AgCl vs. SHE) + 0.059 pH. The electrolyte is placed in a nitrogen atmosphere for at least 30 min before measurement. Using linear scanning voltammetry at a scan rate of 5 mV∙s^-1, polarization curves were recorded at room temperature. The double-layer capacitance (C_dl) was measured by testing cyclic voltammetry (CV) curves at scan rates ranging from 20 to 200 mV∙s^-1. The electrochemical surface area (ECSA) was calculated based on the C_dl of the samples, determined by CV.

The frequency range for electrochemical impedance spectroscopy (EIS) testing was 0.1 to 10⁵ Hz with an amplitude of 10 mV. Long-term durability tests were performed at 10 mA·cm^-2. The mass activity (MA, A·g^-1) values of the catalysts were calculated from the catalyst loading (m) and measured current density (i). MA = i/m.

Material preparation and characterization

Non-metallic atom P, N, and S doping is a common, simple, and effective strategy to explore the effects of different electronegativity and atomic radius on the band gap, bonding properties, strength, and electron affinity, thus improving the electronic configuration, active sites, and conductivity. The changes in ionic radius, electronegativity, electron affinity, and ionization potential enable adjustments to the electronic structure. Doping with non-metallic heteroatom strategy introduces non-metallic heteroatom, changes electronegativity, electron affinity and ionization potential, and significantly affects the electron structure and local spatial configuration. For example, The P, N and S atoms have larger radii than the C atom, which can promote electron delocalization due to the formation of weaker covalent bonds. Utilizing the transient high-energy and high-temperature advantages of microwave pulses allows for rapid initiation and termination of the reaction, thereby enabling precise doping of foreign atoms into the host material. The ten-second interpulse pause is consistent with the rapid quenching process, and it is this rapid temperature change that makes the precise doping of non-metallic heteroatoms possible. This approach efficiently mitigates side reactions caused by gradual temperature gradients through rapid heating and cooling performance. This occurs because the quick temperature changes prevent prolonged exposure to intermediate temperature ranges, where unwanted reactions could occur. As a result, the overall reaction proceeds more smoothly and yields the desired products with higher purity. This enhancement in reaction selectivity and efficiency is particularly advantageous in complex chemical processes where precise control over reaction conditions is crucial for the desired outcome. The instantaneous temperature change of microwaves makes it possible to regulate heteroatom doping and 2D structure simultaneously. Because of this property of microwaves, non-metallic elements, which are inherently prone to doping reactions, can almost entirely participate in the reaction at a certain doping level without leakage. Thus, the precise doping of binary/ternary is realized and the 2D structure is constructed in one step.

Considering these factors, the experiment was devised. Figure 2 illustrates the schematic of the microwave-pulse-assisted synthesis of tunable Single/Binary/ternary-doped 2D Mo₂C. This setup can operate under various gas conditions, ensuring the requirements of diverse microwave experiments. The detailed configuration of the experimental apparatus is depicted in Supplementary Figure 1. Sodium molybdate tetrahydrate is used as the metal precursor, glucose as the carbon source, oxalic acid as the blowing agent, urea, sodium phosphate and sodium sulfide 9-hydrate as dopants, while ensuring that the atomic ratio about of P:N:S:Mo is 1:1:1:17. It is worth noting that weighing the proportion of metal and non-metal atoms is not strictly consistent with the proportion calculation, because non-metal atoms may partially generate gas outflow during the reaction process, but the microwave pulse method will minimize the extent of this process. The mixture is ground uniformly and placed in a quartz crucible, which is then transferred to a custom microwave reactor surrounded by carbon nanotubes (CNTs). Nitrogen gas is introduced into the system, with its flow meticulously regulated by a flow meter. Real-time monitoring of temperature fluctuations within the reaction zone is achieved using an infrared detector. Throughout the microwave-induced reaction process, the high dielectric constant of CNTs induces dielectric loss, facilitating the rapid and uniform generation of thermal energy within the CNT region. This uniform heating condition is conducive to rapidly introducing non-metal elements and the expedited formation of 2D structures. After the microwave operation, the temperature drops rapidly, resulting in rapid cooling of the reactants, ensuring precise control of the P, N and S atom content. After three microwave pulses, highly tunable non-metallic doped 2D Mo₂C was synthesized in one step striking a balance between kinetics and thermodynamics. In terms of kinetics, the 2D Mo₂C materials doped with single, binary and ternary elements were successfully synthesized using microwave pulses. Regarding thermodynamics, 2D nanosheets were also successfully generated through this method. Utilizing the attributes of real-time controllable microwave pulses facilitates identifying products generated at different stages during the reaction process. Moreover, this capability enables on-site analysis of microwave-induced reactions. Three-dimensional (3D) optical microscopy (OM) was used to characterize the samples with 1, 2, and 3 pulses (T₁, T₂, and T₃, respectively). The results showed three stages of reactants: (1) Polymerization; (2) Bubbling; (3) Burst^[37,41-43]. In the initial stage, glucose in the reactants undergoes thermal decomposition and polymerizes into a gel-like polymer during the first pulse. Simultaneously, due to the thermal decomposition of oxalic acid, gas is generated, forming bubbles within the resulting polymer. During the second stage, the temperature inside the reaction chamber remains high for a period, increasing bubbles within the polymer. The release of a large number of bubbles causes the polymer to be blown into a 3D structure. In the third stage, with the extension of reaction time, the gas released by the high-temperature reaction disrupts the 3D bubble structure, generating a large number of 2D nanosheets. The optical images in the figure show that the sample undergoes nearly a seven-fold volume expansion compared to the precursor material after the reaction. Detailed preparation steps are described in the experimental section.

Figure 2. Schematic diagram of the influence of microwave pulse strategy on the morphology evolution and doping regulation of 2D Mo₂C. 2D Mo₂C: Two-dimensional molybdenum carbides.

The temperature change during the reaction was monitored using an infrared detector. As shown in Figure 3A, the reaction zone temperature rapidly reached a relatively stable temperature of 872 °C within 20 s; due to the absorption and reflection of microwaves by CNTs, significant heat conduction is generated. The inserted image displays a physical photograph corresponding to the discharge phenomenon of microwave pulses at 10 s. Due to the excellent capability of CNTs in absorbing microwave radiation, which can be efficiently converted into thermal energy, when exposed to a microwave field, their thermal conductivity aids in the rapid absorption and propagation of microwave energy throughout the entire nanostructure of the nanotubes. Secondly, the inherent structure of CNTs induces pronounced dielectric losses, leading to molecular vibrational excitation and electron friction within the microwave field, thereby culminating in heat generation. This dielectric loss mechanism enables CNTs to effectively absorb microwave energy and convert it into thermal energy. Additionally, the nanoscale dimensions of CNTs endow them with a substantial surface area-to-volume ratio, advantageous for both absorption and dispersion of microwave energy. This nanoscale architecture further facilitates rapid heating and uniform distribution of thermal energy by CNTs within the microwave field, promoting rapid high-temperature carbonization of reactants and rapid doping of non-metallic elements. Once the microwave pulse ceases, the reactants can rapidly decrease to around 271 to 601 °C within 4 s, facilitating rapid cooling of the products, akin to quenching. The experimental device uses pulse heating; each heating cycle lasts 60 s, then pauses for 10 s, and this sequence is repeated three times. In essence, microwave reactions undergo three discernible stages. This is facilitated by the material's exceptional capability to absorb microwaves to quickly raise the desired reaction temperature within 7 s for each microwave pulse. The rapid heating and cooling characteristics reduce the likelihood of side reactions induced by temperature gradients, facilitating precise doping of heteroatoms. The morphology of the samples after the reaction was then characterized. Scanning electron microscopy (SEM) characterization of the catalyst morphology was performed [Figure 3B], revealing that the obtained P,N,S-Mo₂C exhibits a large 2D nanosheet structure with a smooth and flexibly curled surface, with lateral dimensions reaching up to 30 µm. This significantly increases the electrochemical active sites and surface area, conducive to enhanced electrocatalytic performance. The other six types of single/binary-doped Mo₂C also displayed apparent 2D structures, and the morphology was basically the same as that of 2D P,N,S-Mo₂C [Supplementary Figure 2], indicating that doping did not affect the 2D structure of the sample [Supplementary Figure 3]. Using Atomic Force Microscopy (AFM), the thickness of the P,N,S-Mo₂C nanosheets was measured; analysis revealed a thickness of 2 nm [Supplementary Figure 4]. TEM images in Figure 3C also confirmed the ultra-thin thickness of the P,N,S-Mo₂C nanosheets. The corresponding high-resolution transmission electron microscopy (HRTEM) images [Figure 3D], exhibit clear lattice fringes with a spacing of 0.26 nm, corresponding to the (100) plane of Mo₂C, further indicating the successful synthesis of the Mo₂C phase. Additionally, the selected area electron diffraction (SAED) pattern in the upper right corner of the HRTEM image showed that P,N,S-Mo₂C is a single crystalline hexagonal phase. Compared to traditional heating methods, the microwave pulse heating and cooling process ensures rapid carbonization while simultaneously facilitating the doping process, avoiding side reactions during doping and favoring the formation of single-crystal structures. The element mapping and corresponding elemental mappings, along with the Energy Dispersive X-ray Spectroscopy (EDX) spectra of 2D Mo₂C doped with heteroatoms, show a uniform distribution of elements [Figure 3E-G]. This confirms that the microwave-pulse-assisted strategy effectively achieved precise doping of non-metal elements P, N, and S into 2D Mo₂C, while also constructing the 2D structure in a single step.

Figure 3. Morphology characterization. (A) Real-time temperature changes of samples heated by microwave pulses. The inset image is 10 s microwave discharge diagram; (B) SEM images of 2D P,N,S-Mo₂C; (C) TEM images of 2D P,N,S-Mo₂C; (D) The HRTEM image of 2D P,N,S-Mo₂C and the insert image is the corresponding SAED pattern; (E) EDS elemental mapping scanning from TEM of single doping 2D Mo₂C; (F) EDS elemental mapping scanning from TEM of binary doping 2D Mo₂C; (G) EDS elemental mapping scanning from TEM of ternary doping 2D Mo₂C. 2D: Two-dimensional; Mo₂C: molybdenum carbides; SEM: scanning electron microscopy; TEM: transmission electron microscopy; HRTEM: high-resolution transmission electron microscopy; SAED: selected area electron diffraction; EDS: energy-dispersive X-ray spectroscopy.

Leveraging the unique attributes of microwave pulse assistance, we can characterize the structural evolution of the samples at different stages of the reaction. The XRD data were refined using General Structure Analysis System (GSAS) software [Figure 4A]. The parameters of the synthesized catalyst were calculated, and the changes related to atomic doping were revealed. As shown in Supplementary Table 1, the lattice parameters increase from 3.0124 Å, 3.0124 Å, 4.7352 Å to 3.0150 Å, 3.0150 Å and 4.7476 Å, respectively. The lattice size of 2D P,N,S-Mo₂C is larger than pristine Mo₂C. This is because the atomic radius of C is different from that of P, N, and S atoms, suggesting that various types of non-metallic heteroatoms have been successfully incorporated into the lattice of Mo₂C. As shown in Supplementary Figure 5, the XRD data for samples subjected to one, two, and three microwave pulses clearly demonstrate the structural evolution of 2D P,N,S-Mo₂C. After the first pulse, the XRD spectrum between 15° and 30° shows a broad diffraction peak, possibly due to the thermal decomposition of glucose resulting in a large amount of carbon. Following the second pulse, the XRD features are predominantly those of monoclinic MoO₂ (PDF: 01-076-1807)^[44]. The formation of Mo₂C single crystal phase is observed after the third pulse. The diffraction peaks of the obtained 2D P,N,S-Mo₂C at 34.3, 37.9, 39.3, 52.1, 61.5, 69.5, 72.3, 74.6, and 75.5, when compared with the collection code (023823) from the Inorganic Crystal Structure Database (ICSD), correspond to the hexagonal phase Mo₂C (PDF: 00-035-0787) planes of (100), (002), (101), (102), (110), (103), (200), (112), and (201), respectively^[45]. Doping does not cause obvious change of crystal structure, suggesting that the microwave-pulse-assisted synthesis of 2D P,N,S-Mo₂C does not produce any by-products.

Figure 4. Structural characterization. (A) Difference patterns for the Rietveld refinement from the XRD of 2D P,N,S-Mo₂C; (B-E) High-resolution XPS spectra of Mo 3d, p 2p, N 1s and S 2p of 2D P,N,S-Mo₂C; (F) Summary of the proportion of each heteroatom in 2D doped-Mo₂C. XRD: X-ray diffraction; 2D: two-dimensional; Mo₂C: molybdenum carbides; XPS: X-ray photoelectron spectroscopy.

Subsequently, XPS was employed to investigate the oxidation states of elements in the non-metal-doped 2D Mo2C samples and the bonding of surface elements in depth. As depicted in Figure 4B, the high-resolution Mo 3d spectrum can be resolved into four doublets. The two prominent peaks at 228.6 and 232.0 eV correspond to Mo²⁺ (Mo 3d_3/2 and Mo 3d_5/2), indicating Mo₂C bonds. Concurrently, the presence of Mo³⁺ peaks at 229.0 and 232.8 eV suggests potential P−Mo and N−Mo bonding. The presence of Mo³⁺ indicates that P or N occupies part of the C position in Mo₂C, which may optimize its electronic properties and reduce the desorption energy of the Mo−H bond. The valence states of Mo⁴⁺ (230.4 and 233.8 eV) and Mo⁶⁺ (233.1 and 236.2 eV) may be associated with the inevitable surface oxidation of Mo exposed to air^[46]. In the C 1s spectrum, the peak at 284.0 eV, attributed to the Mo−C bond, further confirms the formation of the Mo₂C phase. The peaks at 288.8, 286.6, 285.4, and 284.7 eV correspond to O−C=O, C=O, C−O and C−C bonds, respectively^[47,48] [Supplementary Figure 6]. The presence of oxygen groups (O−C=O) is beneficial for enhancing the wettability of the catalyst to the electrolyte^[49]. The peaks at 129.4 and 130.4 eV in the P 2p spectrum correspond to the P 2p_3/2 and p 2p_1/2, respectively. [Figure 4C]; all these peaks are associated with the P−Mo bonds after doping. The peak at 133.3 eV corresponds to the P−C bond, while the peak at 134.2 eV corresponds to the P-O signal, likely due to oxidation upon exposure to air^[46,50]. This provides evidence for successful doping of P heteroatoms into Mo₂C. Figure 4D presents the high-resolution spectrum of N 1s. Peaks at 394.7, 398.36, 399.48, and 401.26 eV are ascribed to Mo-N, pyridinic N, pyrrolic N, and graphitic N, respectively^[46,51]. The presence of C-N indicates successful incorporation of N atoms into the Mo₂C phase via microwave thermal pulse, with N incorporation potentially influencing the catalytic activity of Mo₂C. High-resolution S 2p spectrum [Figure 4E] reveals double peaks at 162.2 and 163.3 eV, corresponding to S 2p_3/2 and S 2p_1/2, respectively, indicating successful S doping into Mo₂C. Additionally, double peaks at 163.9 and 165.1 eV correspond to S 2p_3/2 and S 2p_1/2 of the C−S−C bond, while oxidized S is observed at 169.0 eV^[46,50]. Thus, the XPS results confirm successful co-doping of non-metal atoms P, N, and S into Mo₂C, which will further modulate the electronic structure of the electrocatalyst, playing a significant role in the catalytic activity of Mo₂C. An important point to note is that the atomic ratios of P, N and S remain the same after statistical analysis of the composition of 2D Mo₂C doped by non-metallic atoms, each accounting for about 3% ± 0.09% of the total composition [Figure 4F]. This is basically consistent with the proportion of atoms weighed before the reaction, indicating little loss of non-metallic atoms under the high-energy impact of microwave pulses. This shows that the microwave pulse method not only successfully achieves the design of 2D structures but also can precisely control the percentage of atoms, a feat difficult to achieve with traditional methods. It is worth noting that the precise regulation of non-metallic atoms by microwave pulses is not limited to equal doping; if the ratio of P to S in the initial reactant is changed to 2:1, it is also feasible to achieve a 2:1 ratio of P:S atoms in the final product of the reaction. Meanwhile, due to the small proportion of non-metallic atom doping, the number of non-metallic atoms involved in doping is far less than the number of sites that can be replaced, so the competition between doping atoms in this process can be ignored. Additionally, as demonstrated in Supplementary Figures 7 and 8, the high-resolution spectra for each element in 2D P-Mo₂C, 2D N-Mo₂C, 2D S-Mo₂C, 2D P,N-Mo₂C, 2D P,S-Mo₂C, and 2D N,S-Mo₂C further substantiate the high applicability of non-metal doping in 2D Mo₂C using the microwave-pulse method.

Hydrogen evolution performance test

By designing a similar doping atom ratio, the influence of P, N, S doping on electron configuration of non-metallic atoms with different electronegativity and atomic radii can be analyzed. Electrochemical testing can systematically investigate the HER performance mechanism dependent on various electron configurations. In acidic media, HER can be divided into two steps^[52]: firstly, the adsorption of H⁺, known as the Volmer reaction (H₃O⁺ + e^- → H^*_ads + H₂O), followed by the desorption step, known as the Heyrovsky reaction (H^*_ads + H₃O⁺ + e^- → H₂ + H₂O), or the recombination step, known as the Tafel reaction (H^*_ads + H^*_ads → H₂). Therefore, a large amount of electron and ion transfer is involved, and 2D materials can provide an abundant active site for HER reaction and rapid electron and ion transfer speed [Figure 5A].

Figure 5. Synthesis of doped 2D Mo₂C in 0.5 M H₂SO₄ electrocatalytic HER by microwave pulse method. (A) Schematic diagram of HER reaction for rapid electron and ion transport 2D P,N,S-Mo₂C nanosheets; (B) Polarization curve; (C) Overpotentials at 10 mA·cm^-2; (D) The corresponding Tafel slope; (E) Plotting current density variation against scan rate to fit linear regression allows for the estimation of C_dl; (F) Mass activity of various spinels calculated at -0.1 V vs. RHE; (G) Summary of statistical electrochemical properties; (H) Chronopotentiometry test at the current density of 10 mA·cm^-1; (I) The accelerated durability tests of 2D P,N,S-Mo₂C. 2D: Two-dimensional; Mo₂C: molybdenum carbides; HER: hydrogen evolution reaction; RHE: reversible hydrogen electrode.

Figure 5B shows the polarization curves of 2D Mo₂C, 2D P-Mo₂C, 2D N-Mo₂C, 2D S-Mo₂C, 2D P,N-Mo₂C, 2D P,S-Mo₂C, 2D N,S-Mo₂C and 2D P,N,S-Mo₂C s in 0.5 M H₂SO₄. To provide a benchmark for comparison, the electrocatalytic activity of a commercial Pt/C catalyst was measured under identical conditions. The HER potential corresponding to a current density of 10 mA·cm^-2 is plotted in Figure 5C. Notably, At a current density of 10 mA·cm^-2, the overpotential of 2D P,N,S-Mo2C is 58 mV, which is close to that of Pt/C (50 mV) and surpasses 2D P-Mo₂C (86 mV), 2D N-Mo₂C (88 mV), 2D S-Mo₂C (90 mV), 2D P,N-Mo₂C (67 mV), 2D P,S-Mo₂C (74 mV) , 2D N,S-Mo₂C (83 mV) and much better than 2D Mo₂C (114 mV). Due to the doping of non-metallic atoms replacing some carbon atoms and the differences in electronegativity and atomic size between carbon and non-metallic atoms, the intrinsic electronic structure of carbon is altered, creating new active sites and ultimately enhancing catalytic activity^[53-55]. These changes break the symmetry of the electric neutrality and/or spin distribution, resulting in charge/spin redistribution that favors the adsorption of HER intermediate species and promotes HER catalytic reactions^[55]. At the same time, co-doping with multiple non-metallic dopants not only increases the defect concentration but also generates synergistic effects between different non-metallic dopant atoms^[56,57], which can form P−N bonds to regulate the electronic structure with strong synergistic coupling effects, while N and S dopants have synergistic effects as σ-donors and π-inverse-donors, which help improve the adsorption efficiency, weaken the hydroxide bonds, and promote the generation of H₂. These effects significantly enhance the catalytic activity of 2D P,N,S-Mo₂C^[58-60]. The Tafel slope can be used as an important parameter to judge the kinetic properties of HER. Figure 5D provides the Tafel values corresponding to the polarization curves. The Tafel slope of the 2D P,N,S-Mo₂C sample is 48 mV·dec^-1, still superior to 2D P-Mo₂C (61 mV·dec^-1), 2D N-Mo₂C (61 mV·dec^-1), 2D S-Mo₂C (65 mV·dec^-1), 2D P,N-Mo₂C (57 mV·dec^-1), 2D P,S-Mo₂C(58 mV·dec^-1), and 2D N,S-Mo₂C (59 mV·dec^-1), and significantly better than 2D Mo₂C (68 mV·dec^-1). Incorporating appropriate amounts of electron-rich dopants, such as phosphorus, into the Mo₂C lattice mitigates the density of vacant d orbitals, thereby weakening the robust Mo−H bonds within Mo₂C, and promoting the H^*_ads desorption (Heyrovsky/Tafel) step promotes HER kinetics^[61,62]. In order to further understand the underlying mechanism of enhancing HER activity with non-metallic doped Mo₂C catalyst, the electrochemical active area (ECSA) was estimated by testing the C_dl. The CV curves of samples with eight samples are measured in the voltage range of 0.1~0.2 V (vs. RHE) are shown in Supplementary Figure 9. As depicted in Figure 5E, 2D P,N,S-Mo₂C exhibits a C_dl of 28.58 mF·cm^-2, much larger than those of 2D P-Mo₂C (14.12 mF·cm^-2), 2D N-Mo₂C (13.76 mF·cm^-2), 2D S-Mo₂C (12.71 mF·cm^-2), 2D P,N-Mo₂C (20.83 mF·cm^-2), 2D P,S-Mo₂C (18.34 mF·cm^-2) , 2D N,S-Mo₂C (15.72 mF·cm^-2) and 2D Mo₂C (10.10 mF·cm^-2). This indicates that 2D P,N,S-Mo₂C possesses more catalytically active sites than other catalyst components, which can be ascribed to difference in atomic radius formed by P, N and S non-metallic elements doped with 2D Mo₂C, which may lead to lattice distortion or formation of defective structures. These distortions and defects can provide more active sites and alter the surface electronic structure, thereby promoting electrocatalytic reactions^[63]. The parameter of MA serves as a crucial metric for characterizing the intrinsic Hydrogen Evolution Reaction (HER) efficiency of catalysts. Figure 5F shows the MA value at -0.1 V vs. RHE. Notably, The MA of 2D P,N,S-Mo₂C was 83.3 A·g^-1, 8.74 times higher than that of 2D Mo₂C (9.53 A·g^-1). The observed variations in apparent MA among these catalysts indicate that the heightened activity is attributed to the internal electronic structure modulation by non-metallic atoms P, N, and S. As illustrated in Figure 5G, a more intuitive reveals that the 2D P,N,S-Mo₂C exhibits superior performance across various metrics compared to other samples, notably surpassing the 2D Mo₂C. This further substantiates the enhanced electrocatalytic properties of the 2D P,N,S-Mo₂C. Moreover, the EIS measurement results align with the trends in the polarization curves [Supplementary Figure 10]. Compared with other electrocatalysts, The radius of 2D P,N,S-Mo₂C is the smallest, indicating that 2D P,N,S-Mo₂C has the smallest charge-transfer resistance (R_ct). The R_ct values of 2D P-Mo₂C, 2D N-Mo₂C, 2D S-Mo₂C, 2D P,N-Mo₂C, 2D P,S-Mo₂C, 2D N,S-Mo₂C, 2D Mo₂C and 2D P,N,S-Mo₂C were 37, 41, 43, 19, 20, 28, 47 and 16 Ω, indicating that the addition of P, N and S atoms effectively reduces the electrode resistance, resulting in better conductivity and charge transfer rate of 2D P,N,S-Mo₂C, thereby improving the electrochemical HER. This shows that the doping of P, N and S atoms with 2D Mo₂C can effectively reduce the electrode resistance. These results are consistent with the Tafel plots. Additionally, long-term stability is a key parameter for evaluating HER catalysts and their electrocatalytic activity. Figure 5H illustrates the long-term stability of seven samples at a current density of 10 mA·cm^-2 measured by chronoamperometry. The results show that after 22 h of uninterrupted operation in the alkaline electrolyte, the potential of the sample has almost no change, showing excellent stability. Subsequently, the C_dl after the long-term stability test was characterized [Supplementary Figure 11], and it was found that there was no significant change before and after the 2D P,N,S-Mo₂C test, while 2D Mo₂C had a large decline. The corresponding CV curve is shown in Supplementary Figure 12. In addition, in the accelerated durability test, the scanning rate was 100 mV·s^-1, and 3,000 CV curves were recorded before and after scanning; the displacement of the polarization curve is negligible compared to the initial curve, consistently demonstrating excellent HER performance [Figure 5I].

Additionally, the electrocatalytic HER performance of Pt/C, 2D Mo₂C, 2D P-Mo₂C, 2D N-Mo₂C, 2D S-Mo₂C, 2D P,N-Mo₂C, 2D P,S-Mo₂C, 2D N,S-Mo₂C and 2D P,N,S-Mo₂C was also explored in 1 M KOH electrolyte. Figure 6A presents the polarization curves of the prepared samples and the Pt/C catalyst in an acidic medium. The sample achieves a current density of 10 mA·cm^-2 with only a 70 mV overpotential, second only to Pt/C (52 mV). This value is 6~61 mV, lower than unary and binary doping Mo₂C and pure Mo₂C samples [Supplementary Figure 13]. In contrast, the performance of 2D P,N,S-Mo₂C was improved greatly, consistent with the results measured in acidic media. The corresponding Tafel slopes of the polarization curves are shown in Figure 6B. The Tafel slope corresponding to the polarization curve is shown in Figure 6B. The Tafel slope of 2D P,N,S-Mo₂C is 54 mV·dec^-1, again superior to 2D P-Mo₂C (67 mV·dec^-1), 2D N-Mo₂C (68 mV·dec^-1), 2D S-Mo₂C (70 mV·dec^-1), 2D P,N-Mo₂C (60 mV·dec^-1), 2D P,S-Mo₂C (61 mV·dec^-1) , 2D N,S-Mo₂C (64 mV·dec^-1) and much better than 2D Mo₂C (71 mV·dec^-1). It shows that 2D P,N,S-Mo₂C also has good reaction kinetics in an alkaline environment. In addition, CV curves of seven samples were measured in the potential range of 0.1 to 0.2 V (vs. RHE) [Supplementary Figure 14]. On this basis, the C_dl of 2D P,N,S-Mo₂C is estimated to be 21.30 mF·cm^-2, also the highest compared to other samples [Figure 6C]. The samples were similarly tested under alkaline conditions at -0.1 Vvs. RHE for their MA values. Notably, the MA of 2D P,N,S-Mo₂C at 36.9 A·g^-1 is 3.48 times higher than that of 2D Mo₂C (10.6 A·g^-1) [Figure 6D]. As shown in Figure 6E, compared with other samples, 2D P,N,S-Mo₂C is the best in various tests, significantly exceeding 2D Mo₂C, consistent with the trend in acidic media. Under alkaline conditions, the long-term stability of eight samples was also evaluated using chronopotentiometry at a current density of 10 mA·cm^-2. The experimental results are consistent with those measured under acidic conditions. After 22 h of uninterrupted operation, the sample potential and electrochemical properties were almost unchanged [Figure 6F]. C_dl was also tested after long-term cyclic stability, and the results showed that 2D P,N,S-Mo₂C also had excellent stability in alkaline [Supplementary Figure 15]. The corresponding CV curve is shown in Supplementary Figure 16. Also, in the accelerated durability test, the polarization curve displacement is negligible compared with the initial curve, and the excellent HER performance is always maintained [Figure 6G]. As shown in Supplementary Figure 17A, XRD showed that 2D P,N,S-Mo₂C exhibits no significant change before and after the screening; SEM images of 2D P,N,S-Mo₂C after the durability test show that it still maintains a 2D form and no aggregation is observed [Supplementary Figure 17B]. In addition, HRTEM images of 2D P,N,S-Mo₂C after stability testing show that the (100) face of 2D P,N,S-Mo₂C has lattice fringes with a distinct spacing of 0.26 nm [Supplementary Figure 17C]. Finally, as shown in Figure 6H, we summarized the overpotentials and Tafel slopes of doped Mo₂C-based materials in acidic and alkaline electrolytes in recent years. It is noteworthy that 2D P,N,S-Mo₂C shows the best HER performance, making it promising for practical applications. For more information, see Supplementary Tables 2 and 3.

Figure 6. Synthesis of doped 2D Mo₂C in 1 M KOH electrocatalytic HER by microwave pulse method. (A) Polarization curve; (B) The corresponding Tafel slope; (C) Plotting current density variation against scan rate to fit linear regression allows for the estimation of C_dl; (D) Mass activity of various spinels calculated at -0.1 V vs. RHE; (E) Summary of statistical electrochemical properties; (F) Chronopotentiometry test at the current density of 10 mA·cm^-2; (G) The accelerated durability tests of 2D P,N,S-Mo₂C; (H) The 2D P,N,S-Mo₂C was compared with the overpotential and Tafel of other recently reported doped Mo₂C-based electrocatalysts. 2D: Two-dimensional; Mo₂C: molybdenum carbides; HER: hydrogen evolution reaction; RHE: reversible hydrogen electrode.