INTRODUCTION
The growing awareness of climate change impacts, rising energy expenses, and a strong demand for eco-friendly, renewable, and sustainable energy technologies are driving global research into energy storage systems (ESSs) and their efficient utilization, such as lithium-ion batteries (LiBs) and supercapacitors (SCs) [1,2]. As an alternative to fossil fuels, LiBs and SCs are widely available and affordable ESSs with a great potential because they can be used as power and backup energy sources for the energy storage devices and the electricity grid in case of supply disruptions due to the weather-dependent intermittency of other energy sources that are sustainable, such as solar, wind, and wave power, etc [2,3]. Due to their high energy density, LiBs have been widely used to power portable electronics, automobiles, and even power stationaries [4,5]. Recently, battery technology has advanced with increased energy density and long cyclic stability, which is beneficial for plug-in hybrid electric vehicles (PHEVs) and electric vehicles (EVs) [4,6].
As a common anode material, graphite faces challenges with its low theoretical capacity of only 372 mAh g-1, which restricts its ability to meet the growing energy density demands of next-generation LiBs [4,7]. Moreover, the poor rate capability and insufficient safety tolerance under high current loads hinder its applicability in next-generation energy systems, including EVs and large-scale ESSs [8–10]. These limitations have driven intense research into alternative anode materials with higher capacity, improved rate capabilities, and enhanced structural robustness.
Transition metal oxides (TMOs), particularly manganese oxides, have attracted significant attention due to their high theoretical capacities (e.g., 756 mA h g-1 for MnO and 936 mA h g-1 for Mn₃O₄), low cost, low operating voltage, natural abundance, and environmental friendliness [11–17]. However, their practical application is hindered by severe volume expansion during lithiation and inherently poor electrical conductivity, which result in rapid capacity fading and low cycle stability [11–13,15].
To overcome these challenges, strategies such as nanostructuring and hybridization with conductive carbon matrices have been widely explored. One widely adopted approach involves the fabrication of nanostructured manganese oxides—including nanoparticles [18], nanotubes [19], nanowires [20], and other nanoscale architectures—which can effectively shorten lithium-ion/electron transport pathways and mitigate internal mechanical strain during cycling due to their reduced dimensions [21,22]. However, these nanostructures inherently possess a high surface area, which tends to promote excessive formation of the solid electrolyte interphase (SEI) [23,24]. This leads to irreversible capacity loss and increased lithium consumption, ultimately compromising long-term battery performance [25,26]. Another promising strategy is the hybridization of manganese oxides with conductive carbon-based materials such as graphene, porous carbon, and carbon nanotubes (CNTs) [13,27–30]. The excellent electrical conductivity and mechanical flexibility of carbonaceous materials help buffer volume expansion and enhance structural stability. Numerous studies have demonstrated that forming a conformal, conductive carbon coating around manganese oxide particles significantly improves their electrochemical properties [12,15,18,22]. This protective carbon shell not only enhances charge transport but also isolates the active material from direct contact with the electrolyte, thereby reducing side reactions and stabilizing the SEI layer [19,26,31]. However, the broader application of graphene and CNTs faces challenges. Graphene sheets tend to irreversibly aggregate due to strong van der Waals interactions [32], which diminishes their performance. Preventing graphene restacking is crucial to fully leverage its high conductivity, flexibility, and theoretical surface area [33,34]. Additionally, CNTs and graphene are relatively expensive to produce, and their integration into electrode materials often involves complex, multi-step processing, which hinders scalability and limits their suitability for cost-sensitive, large-scale energy storage applications [35–38].
Among them, the rational design of mixed-valence MnO/Mn3O4 heterostructures embedded in a carbon network provides a promising route to synergistically improve electrochemical performance by benefitting from multiple redox couples and buffering mechanical strain [39–42]. In general, mixed-phase materials show enhanced electrical conductivity over single-phase materials owing to the presence of defect structures such as oxygen vacancies that enhance charge transfer pathways [43–47]. For example, Song et al. explored a nanostructured mixed-valent manganese oxide film exhibiting an exceptionally high specific capacitance of ~2530 F g-1 at 0.61 A g-1 with an active material loading of ~0.16 mg cm⁻², along with outstanding power density and long-term cycling stability for supercapacitor applications [44]. They attributed the remarkable electrochemical performance to the synergistic effect of the mixed-valence states and porous nanostructure, which collectively promotes fast ion diffusion, enhance double-layer capacitance, and facilitate reversible redox reactions at both Mn and O sites, as confirmed by in situ XAS and DFT calculations. Recently, Chu et al. investigated MnO@Mn3O4 nanoparticles embedded within a nitrogen-doped porous carbon framework (MnO@Mn3O4/NPCF), synthesized via carbonizing pyrolysis of mixed-valent Mn8 clusters [39]. As-prepared MnO@Mn3O4/NPCFs electrode achieved a high reversible capacity, long cycling stability, and excellent rate capability, delivering a capacity of 1500 mAh g-1 at a current density of 0.2 A g-1 over 270 cycles. However, this method requires a prolonged processing time (More than 2 days, even excluding annealing process) to obtain the final sample. Zhou et al. introduced a rapid one-step laser scribing method for the fabrication of laser-induced MnO/Mn3O4 composites integrated with N-doped graphene [40]. The MnO/Mn3O4/N-doped-graphene composites displayed good electrochemical performance, with a high reversible capacity of 992 mAh g-1 at 200 mA g-1, as well as good rate capability of 365 mA g-1 at 2.0 A g-1. However, these methods are typically costly, require multiple processing steps, and often necessitate the addition of external carbon sources from polymers and carbon template, such as carbon fibers, CNTs, and graphene, etc.
Herein, we demonstrate an investigation on the mixed-valence MnO/Mn3O4/C composites, taking advantages of synthetic effects in MnO/Mn3O4/C composites using N,N-dimethylacetamide (DMAc) as a solvent and a carbon source via a facile solvothermal and post-heat treatment method. As an in-situ carbon source, DMAc enables the one-step synthesis and surface modification of Mn3O4/C composites by simultaneously serving as a solvent and carbon precursor during solvothermal process. After post-heat treatment, impressively, by controlling annealing temperature under N2 atmosphere, the crystalline phase of product transforms from Mn3O4/C to MnO/Mn3O4/C. This approach not only eliminates the need for external carbon additives but also facilitates the formation of mixed-valence metal oxide composites, allowing enhanced electronic conductivity and structural stability of electrodes through the formation of DMAc-derived carbon networks. In contrast, incorporating pre-made graphene or CNTs often involves multiple processing steps and higher material costs [32,35,36]. Moreover, the carbon network derived from DMAc can form a continuous and adherent coating, which not only buffers volumetric expansion during cycling but also maintains electrical connectivity and spatial separation between oxide particles. Therefore, it is imperative to adopt a rational design strategy that harnesses synergistic effects in generating carbon matrix and mixed-valence MnO/Mn3O4 heterostructures. In this work, the MnO/Mn3O4/C composite annealed at 500°C demonstrates enhanced capacity, superior rate performance, and excellent cycling stability, owing to its elevated electrochemical activity. The enhanced electrochemical performance highlights the efficient utilization of active materials, accelerated reaction kinetics enabled by the mixed-valence MnOx phases, and the mechanical robustness of the composite structure during repeated charge–discharge processes. Furthermore, the synergistic effect between the mixed-valence MnO/Mn3O4 and the in-situ formed carbon network not only boosts overall capacity but also significantly enhances the electrical conductivity of the electrode.
EXPERIMENTAL
Preparation of MnO/Mn3O4/C composites for LiBs
In a typical procedure to prepare MnO/Mn3O4/C, 0.574 g of manganese (II) nitrate hexahydrate (Mn(NO3)2∙6H2O) was firstly dissolved into 40 mL DMAc solvent for 1h at room temperature. Then, 2.0 mL D.I. water was added in Mn-DMAc solution for 2h at room temperature. Subsequently, the precursor solution was transferred into a 100 mL PTFE-lined stainless container, and subsequently heated at 160°C for 12 h. The dark-brown precipitate at the bottom was separated by centrifugation with a speed of 8000 rpm. The products were washed repeatedly with absolute ethanol and D.I. water for several times, then dried in oven at 80°C for 12 h. Lastly, the product was annealed at 400, 500, and 600°C for 2 h under N2 atmosphere with a ramp rate of 5°C min−1. As-prepared samples from different annealing temperatures of 400, 500, and 600°C under N2 atmosphere were referred to as Mn-N4, Mn-N5, and Mn-N6, respectively. A schematic illustration of synthetic procedure is shown in Fig. 1.
Fabrication of Working Electrode
The electrodes were prepared as follows: a slurry of 80 wt.% powder, 10 wt.% carbon black (Super P), 10 wt.% of the polyvinylidene difluoride (PVdF) binder, and a few drops of 1-methyl-2-pyrrolidinone (NMP) were lightly coated on Cu substrate. The electrodes were then dried at 80°C for 12 h under vacuum. The typical loading mass of active materials on circular discs (ɸ12 mm) of Cu foil was around 0.7–0.9 mg cm-2. Coin cells (CR2032) were fabricated using lithium metal as the counter electrode, A microporous polyethylene (PE) film (Toray) as the separator, and LiPF6 (1.15 M) in ethylene carbonate (EC), dimethyl carbonate (DMC), and ethylmethyl carbonate (EMC) electrolyte, with a 2:4:4 vol ratio as the electrolyte.
Materials Characterization
The crystalline properties and elements present were investigated using X-ray diffraction (XRD; D8Advance A25, Bruker AXS GmbH, USA). The surface morphology of the samples was analyzed using field emission scanning electron microscopy (FE-SEM; MIRA3 LM, TESCAN, Czech). Transmission electron microscopy (TEM; JEOL, JEM–2200EX, Japan), high angle annular dark field imaging (HAADF), and energy–dispersive X–ray spectroscopy (EDS).Thermogravimetric analysis (TGA, TA Instrument, USA) was performed using an analyzer in a nitrogen atmosphere over the temperature range from 30°C to 800°C at a heating rate of 5°C min−1. was 0.09cm−1. The elements were identified using X-ray photoelectron spectroscopy (XPS, PHI 5000 Versa Probe, ULVAC-PHI Inc., Osaka, Japan). Raman spectra were recorded on a Horiba Jobin Yvon LabRam Aramis Raman spectrometer, using a 514.5 nm Ar laser at 0.5 mW power.
Electrochemical Measurements
Cyclic voltammetry (CV) measurements and electrochemical impedance spectroscopy (EIS) were performed using a potentiostat/galvanostat (VSP, Bio-Logic, France), with a three-electrode system at scan rates of 0.1, 0.2, 0.4, 0.6, 0.8, and 1.0 mV s-1 between 0.01 V and 3.0 V (Li/Li+) at room temperature and in the frequency range from 100 kHz to 100 mHz at an open circuit potential, respectively. Galvanostatic charge/discharge (GCD) and galvanostatic intermittent titration technique (GITT) test were carried out using a Maccor 4000 series charger between 0.01 and 3.0. V at 25.0 ± 0.1°C using currents corresponding to various current density of 0.1, 0.2, 0.5, 1.0, and 2.0 A g-1 for rate capability and 0.5 A g-1 for 1000 cycles, respectively.
RESULTS AND DISCUSSIONS
Materials Characterization of MnO/Mn3O4/C composites
SEM images reveal a clear temperature-dependent morphological evolution, as shown in Fig. 2. At 400°C, particles are loosely aggregated with high surface roughness and defect density. Upon annealing at 500°C, a more compact and uniform microstructure forms, characterized by optimal particle fusion and hierarchical texture. Further heating to 600°C results in coarsened, densely packed grains, indicative of overgrowth and reduced porosity. With organic matters from DMAc, the carbonization process is carried out. The carbonization temperature was determined to be 500°C by TGA analysis (Fig. S1). The weight loss below 220°C could be attributed to the evaporation of water [28]. From 220 to 500°C, an obvious weight loss was witnessed, which corresponds to the decomposition of organic components and the formation of carbon coated manganese oxide. The weight increase at temperatures above 500°C can be assigned to the reoxidation of manganese oxides by organic matters [48]. Based on the TGA results (Fig. S1), the Mn3O4/C was annealed from 400 to 600°C under N2 atmosphere to generate the MnO/Mn3O4/C composites. The final products were characterized by XRD.
The X-ray diffraction (XRD) patterns of pristine Mn3O4, Mn3O4–N300, Mn–N4, Mn–N5, and Mn–N6 are presented in Fig. 3a. In the Mn–N4, Mn–N5, and Mn–N6 samples, diffraction peaks at 2θ = 29.0°, 32.3°, 36.1°, and 59.8° are indexed to the (112), (103), (211), and (224) planes of tetragonal Mn3O4 (JCPDS No. 24-0734). Meanwhile, additional weaker reflections appearing at 34.9°, 40.6°, 58.7°, 70.2°, 73.8°, and 87.8° correspond to the (111), (200), (220), (311), (222), and (400) planes of cubic MnO (JCPDS No. 07-0230). These results indicate that samples annealed at temperatures above 400°C undergo partial reduction, forming a MnO/Mn3O4 composite, with Mn3O4 as the dominant phase and MnO as a minor secondary phase. In contrast, the Mn3O4–N300 sample, annealed at 300°C under N2, exhibits only the diffraction peaks associated with Mn3O4 (JCPDS No. 24-0734), without any indication of MnO formation. This suggests that thermal reduction does not occur below 400°C, and the Mn3O4 phase remains structurally intact at this lower temperature. These findings confirm that annealing temperature plays a key role in tuning phase composition, where temperatures over 400°C under N2 atmosphere promote partial conversion to MnO, enabling formation of MnO/Mn3O4/C heterostructures, while annealing at or below 400°C preserves only the Mn3O4 phase.
High-resolution transmission electron microscopy (HRTEM) analysis (Fig. 3) was conducted to further examine the microstructure of the Mn–N5 composite, as shown in Fig. 3b. The elemental mapping images of manganese (Mn), oxygen (O), and carbon (C) (Fig. 3c), acquired via energy-dispersive X-ray spectroscopy (EDX) in scanning transmission electron microscopy (STEM) mode, confirming the homogeneous distribution of all elements throughout the sample. This uniform dispersion supports the successful formation of a well-integrated MnO/Mn3O4/carbon composite. Furthermore, the HRTEM lattice-resolved image and its inset (Fig. 3d) reveal the polycrystalline nature of the composite. The measured interplanar spacings of 0.31 nm and 0.26 nm correspond to the (112) crystal plane of Mn3O4 and the (111) plane of MnO, respectively, confirming the coexistence of both oxide phases. These findings are consistent with the XRD results, further validating the phase composition and crystallographic identity of the MnO/Mn3O4 heterostructure formed during the annealing process.
Raman spectroscopy was employed to investigate the structural evolution of the carbon matrix and confirm the formation of MnO/Mn3O4/C composites across different annealing temperatures. (Fig. 3e and 3f) The corresponding Raman spectra for the three samples are shown in Fig. 3f. As the annealing temperature increased from 400°C to 500°C, the D/G ratio increased, indicating a higher degree of disorder and defect density within the carbon framework [49]. This suggests that annealing at 500°C promotes partial carbonization while introducing structural disorder favorable for enhanced electrochemical activity [14,49]. In contrast, the sample annealed at 600°C exhibited no clearly distinguishable G band, implying a significant reduction in graphitic carbon content. This results from thermal degradation or decomposition of carbonaceous species at elevated temperature, leading to a lower overall carbon content in the composite, as discussed in the following XPS analysis section. In addition to the carbon-related peaks, all MnO/Mn3O4/C samples exhibit distinct Raman-active modes at ~320, 376, and 660 cm−1 (Fig. 3e), which are attributed to the symmetric stretching vibrations of MnO6 octahedra, characteristic of crystalline Mn3O4 [50,51]. A weaker band at 577 cm−1 is also observed, corresponding to Mn–O vibrations within the octahedral framework [52,53]. These vibrational signatures confirm the presence of well-formed Mn3O4 crystalline domains in the composite materials.
X-ray photoelectron spectroscopy (XPS) was conducted to gain deeper insight into the chemical composition and electronic states of the MnO/Mn3O4/C composite. The survey spectrums (Fig. 4 and 5) confirm the presence of Mn, O, and C, indicating successful formation of the ternary composite. High-resolution analysis of the Mn 2p region for the Mn–N5 sample (Fig. 4b) reveals characteristic spin–orbit doublets. The peaks located at 640.4 eV (Mn 2p3/2) and 651.7 eV (Mn 2p1/2) are attributed to Mn2+, while those at 641.7 eV and 653.2 eV correspond to Mn3+. This clearly demonstrates the presence of mixed-valence manganese states (Mn2+/Mn3+) [54,55], consistent with the coexistence of MnO and Mn3O4 phases observed in XRD and HRTEM analyses. The Mn2+/Mn3+ atomic ratio, calculated from the deconvoluted peak areas, is approximately 0.64, which is slightly higher than the theoretical ratio of 0.5 expected for stoichiometric Mn3O4 (Mn2+ [Mn23+]O4). This deviation suggests the presence of a MnO-enriched phase, supporting the conclusion that partial reduction occurred during annealing, resulting in a MnO/Mn3O4 heterostructure. For Mn–N4 (Fig. 4a), the Mn2+ signals appear at 640.15 eV (Mn 2p3/2) and 651.35 eV (Mn 2p1/2), while the Mn3+ components are located at 641.45 eV (Mn 2p3/2) and 653.10 eV (Mn 2p1/2). These peak positions confirm the presence of mixed Mn2+/Mn3+ valence states. In Mn–N6 (Fig. 4c), the Mn2+ peaks are observed at 640.59 eV (Mn 2p3/2) and 651.99 eV (Mn 2p1/2), slightly shifted to higher binding energy compared to Mn–N4. Meanwhile, Mn3+ peaks are located at 641.72 eV and 653.41 eV. This sample shows the highest Mn2+ proportion at 44.38%, indicating a more reduced phase (Fig. 4d), likely due to prolonged annealing. The upshift in Mn2+ binding energy reflects increased lattice stabilization or oxygen-deficient MnO domains, consistent with reduced oxygen vacancy content and diminished Mn3+ redox sites. The Mn2+/Mn3+ atomic ratio progressively increased from 0.640 (Mn-N5) to 0.798 (Mn-N6) (inset of Fig. 4d), revealing a trend toward increased reduction of manganese species with more intense annealing. The N5 sample offers a favorable redox balance, while N6 represents a more MnO-dominant composition, potentially influencing electrochemical performance.
The Mn 3s core-level XPS spectra provide further insight into the manganese oxidation states through multiplet splitting (ΔE). In this study, the ΔE values for Mn–N4, Mn–N5, and Mn–N6 were determined to be 5.7 eV, 5.6 eV, and 5.8 eV, respectively. (Fig. 4e-g) These ΔE values strongly support the Mn 2p and Mn 3s XPS results, confirming a progressive reduction trend from Mn–N5 to Mn–N6 [39,56], driven by thermal treatment and carbon-induced oxygen vacancy formation. The observed increase in Mn 3s multiplet splitting from 5.6 eV (Mn–N5) to 5.8 eV (Mn–N6) reveals a shift toward lower Mn oxidation states, consistent with MnO-rich composite formation, and aligns with Mn2+/Mn3+ ratios obtained from Mn 2p analysis.
The O 1s XPS spectra were deconvoluted into three components: la1ce oxygen (O²⁻) around 529.2–529.7 eV, (Fig. 5) attributed to metal–oxygen bonds in Mn–O/Mn-O-C; oxygen vacancies or defect-related oxygen near 530.1–530.3 eV; and C–O surface functional groups appearing at ~531.4–531.9 eV [57,58]. Notably, the presence of the Mn–O–C bonding feature indicates a strong chemical interaction between the Mn oxide phase and the carbon matrix, suggesting that carbon is chemically integrated into the MnO/Mn3O4 structure [58]. The deconvoluted fitting results for Mn–N4, N5, and N6 are presented in Fig. 5d. The Mn-N5 sample exhibits the highest oxygen vacancy content (48.1%) with a moderate lattice oxygen percentage (38.8%). This indicates an oxygen-deficient surface, promoting enhanced redox activity and improved charge transfer kinetics. Notably, oxygen vacancies associated with the Mn₃O₄ phase can reduce the activation energy required for metallic Mn nucleation during the conversion reaction, thereby facilitating faster reaction kinetics and improved electrochemical reversibility [59,60]. The elevated vacancy level is likely a result of balanced decomposition of surface groups and mild reduction, aligning with the optimal Mn2+/Mn3+ ratio (~0.64). For Mn–N6, despite having the highest lattice oxygen content (46.0%), shows fewer surface adsorbed oxygen species (11.1%), suggesting greater structural stability and more ordered Mn–O framework. Its oxygen vacancy ratio (42.8%) remains significant but slightly lower than Mn-N5. Mn–N4 displays the highest adsorbed oxygen content (17.7%), along with relatively lower oxygen vacancy density (38.9%). This implies that more hydroxyl or water groups are present on the surface, which may cause surface passivation and hinder faradaic reactions.
The deconvoluted C1s spectrum (Fig. 5e–h) shows three peaks at 284.5, 285.6, 286.3, and 287.8 eV, representing sp2 carbon structure, sp3-like defect, C-O, and O-C=O/O-C-O, respectively [29,58,61,62]. In Fig. 5b, Mn–N5 exhibits the highest C–C (sp2) content (60.9%) and the lowest proportion of oxygen-containing functional groups (C–O: 12.6%, C=O: 8.7%), indicating a well-developed graphitic carbon structure (Fig. 5h). This is beneficial for electronic conductivity [63,64]. Mn–N4, in contrast, contains less graphitic sp² carbon (29.4%) and a significantly higher proportion of C–O (19.8%) and C=O (15.7%) species. Mn–N6 shows a moderate C–C (sp²) ratio (38.3%) and the highest C–C(sp³)/defect component (48.2%), suggesting high structural disorder possibly induced by high-temperature annealing (Fig. 5h). Interestingly, despite the higher annealing temperature, the Mn–N6 sample exhibits a lower fraction of sp²-hybridized carbon compared to Mn–N5, as evidenced by the C 1s XPS analysis and Raman spectra. This counterintuitive behavior can be attributed to over-annealing effects at 600°C under N₂ atmosphere. At elevated temperatures, the carbonaceous species derived from DMAc undergo partial thermal decomposition and volatilization, leading to a reduced carbon yield and disruption of small graphitic domains. In addition, residual nitrate-derived species from Mn(NO₃)₂ may promote local carbon gasification reactions, further suppressing graphitic ordering. As a result, the carbon framework in Mn–N6 becomes increasingly disordered, characterized by a higher proportion of sp³-like carbon defects and diminished graphitic domains. This interpretation is supported by the weakened or absent G band in Raman spectra, the increased C–C(sp³) contribution in XPS C 1s fitting, and the lowest BET surface area observed for Mn–N6. These findings indicate that excessive thermal treatment compromises the structural integrity and electronic effectiveness of the in-situ carbon network, underscoring the importance of optimized annealing conditions for achieving conductive carbon frameworks. Also, thermogravimetric analysis (TGA, Fig. S1) reveals a reduced carbon content in Mn–N6 compared to Mn–N5, consistent with partial carbon volatilization and decomposition at elevated annealing temperature (600°C). The decreased carbon yield corroborates the XPS and Raman observations of diminished sp² carbon content.
Electrochemical Performance
As shown in Fig. 6a–c, cyclic voltammetry (CV) measurements were conducted at a scan rate of 0.1 mV s-1 within the potential window of 0.01–3.0 V vs. Li/Li+ for the first four consecutive cycles to investigate the electrochemical behavior of the MnO/Mn3O4/C electrodes. In the first discharge cycle, a broad cathodic peak near 0.8 V is observed, which is typically attributed to the formation of the solid-electrolyte interphase (SEI) layer due to electrolyte decomposition and surface reactions. Below 0.35 V, distinct cathodic peaks appear, corresponding to the irreversible reduction of Mn3O4 and MnO to metallic Mn, along with the formation of Li2O. (MnO + 2Li+ + 2e− → Mn + Li2O, Mn3O4 + 8Li+ + 8e− → 3Mn + 4Li2O) [27,65,66]. Notably, Mn3+ species are reduced at higher potentials than Mn2+, which directly influences the evolution and relative intensity of the cathodic peaks observed in the cyclic voltammetry profiles. This behavior is consistent with the fact that Mn3+ species undergo reduction at higher potentials than Mn2+, thereby governing the cathodic peak evolution during cycling. Moreover, the coexistence of MnO and Mn₃O₄ introduces multiple conversion pathways, which facilitate faster reaction kinetics and improved electrochemical reversibility during repeated lithiation/delithiation processes [39,67]. The anodic peaks at ~1.28 V are assigned to the oxidation of Mn to MnO [27]. The shift of the cathodic peak from ~0.87 V in the first cycle to ~0.35 V occurred in subsequent cycles. In addition, the mixed-valence configuration broadens the effective voltage window for reversible lithium storage, thereby enhancing the overall electrochemical utilization of the active material [39,68]. This indicates an enhancement in the reducibility of MnO to metallic Mn [31], resulting in a transition from initial bulk conversion to a more electrochemically favorable state, where Mn reduction occurs with lower overpotential.
The mechanisms of conversion reactions in the MnO/Mn3O4 can be succinctly described as following Eqs. (1), (2), (3):
The galvanostatic charge-discharge (GCD) curves of MnO/Mn3O4/C composites (Fig. 6d–f) show the corresponding voltage plateaus coincided with CV curves. The initial discharge and charge specific capacities of 1434 and 913.7 mAh g-1 for Mn-N5 are obtained at 0.1 A g-1 (Fig 6e), corresponding to a high initial coulombic efficiency (ICE) of 63.7%. Otherwise, the initial discharge and charge specific capacities of 1350 and 796.9.7 mAh g-1 for Mn-N4 (Fig. 6d) and those of 1686 and 908.9 mAh g-1 for Mn-N6 (Fig. 6f) are achieved at 0.1 A g-1, corresponding to a high initial ICE of 59.0 and 53.9%, respectively.
The performance of the MnO/Mn3O4/C composites were further evaluated by its rate capability and long-term cyclability. Fig. 6g shows that the Mn-N5 can deliver reversible capacities of 957.8, 665.9, 523.6, 411.3, and 308.1 mA h g-1 at 0.1, 0.2, 0.5, 1.0, 2.0 and 5.0 A g-1, respectively. For Mn-N6, the reversible capacities of 908.9, 577.2, 411.5, 311.8, and 212.1 mAh g-1 obtained at a current density of 0.1, 0.2, 0.5, 1.0, 2.0 and 5.0 A g-1, respectively. For Mn-N4, the reversible capacities of 796.9, 476.0, 317.6, 225.9, and 133.7 mAh g-1 obtained at a current density of 0.1, 0.2, 0.5, 1.0, 2.0 and 5.0 A g-1, respectively. From Fig. 6d–f and 6g, the ICE and the rate capacities of the Mn-N5 are higher than those of the Mn-N4 and Mn-N6. This result further evidences the great promise of using the Mn/Mn3O4/C (Mn-N5) as a potential anode material for LiBs. The long-term cyclability measurement in Fig. 6h shows that the as-prepared Mn-N5 sample can deliver a relatively stable reversible capacity of 472.5 mAh g-1 up to 1000 cycles at the current rate of 0.5 A g-1, although a slight fluctuation (~200 cycles), presumably caused by the gradual activation of the active materials, resulting from the formation of SEI or the decomposition of Li2O and electrolyte, which lead to the reduction of charge transfer resistance [69,70]. Rate capability measurements are conducted using freshly assembled or minimally aged cells, where the electrode–electrolyte interfaces are well preserved and internal resistance remains relatively low. Under such conditions, lithium-ion diffusion and electron transport are highly efficient, allowing the electrode to deliver comparatively high capacities even at elevated current densities. In contrast, long-term cycling tests inherently involve progressive interfacial and structural evolution. During extended cycling, irreversible solid–electrolyte interphase (SEI) growth, partial aggregation of metallic Mn nanoparticles, and continuous electrolyte decomposition collectively increase charge-transfer resistance and induce mechanical stress within the electrode. These cumulative effects gradually reduce the fraction of electrochemically active material participating in the conversion reactions.
The cycling performance of the Mn–N5 electrode at a current density of 0.2 A g-1 are presented in Fig. 6i. The electrode delivers an initial reversible capacity of 594 mAh g-1, which is relatively modest. Over the first 30 cycles, the capacity gradually decreases, but then undergoes a progressive increase, reaching 861 mAh g-1 after 250 cycles. Throughout the cycling test, the Mn–N5 electrode maintains a CE close to 100%, except for a slight decrease for first 5 cycles. Remarkably, the capacity after 270 cycles exceeds the theoretical value of MnO (756 mAh g-1) and approaches that of Mn3O4 (937 mAh g-1), demonstrating outstanding long-term electrochemical performance. Interestingly, the gradual increase in capacity observed during the initial ~200 cycles can be attributed to an electrochemical activation process. This activation arises from improved electrolyte penetration, partial reversibility of Li₂O decomposition, and enhanced utilization of previously inaccessible active sites within the mixed-valence MnO/Mn₃O₄ framework, a phenomenon commonly reported for conversion-type anode materials. It should be noted that the reversible capacity of the Mn–N5 electrode exceeds the theoretical capacity of MnO (756 mAh g-1) and approaches that of Mn₃O₄ (937 mAh g-1) during prolonged cycling. Such behavior should not be interpreted as a simple advantage, but rather reflects the complex electrochemical processes inherent to conversion-type anode materials. In addition to the classical conversion reactions of MnO and Mn₃O₄, several supplementary storage mechanisms may contribute to the apparent excess capacity [39,40], including partial reversibility of Li₂O decomposition [17], interfacial lithium storage at the Mn/Li₂O boundaries [65], and redox-active surface oxygen species associated with oxygen vacancies [71]. Furthermore, the formation of polymeric gellike layers derived from electrolyte decomposition can provide additional pseudocapacitive contributions during long-term cycling [72,73]. These processes collectively enhance the measured capacity but are often accompanied by increased interfacial complexity. Therefore, the observed capacity beyond the theoretical limit reflects a combination of bulk conversion, interfacial storage, and surface redox reactions, highlighting the necessity of cautious interpretation when evaluating high capacities in mixed-valence MnO/Mn₃O₄-based electrodes.
The superior reaction kinetics of the Mn–N5 electrode, compared to Mn–N4 and Mn–N6, were systematically evaluated using cyclic voltammetry (CV) at varying scan rates based on Dunn’s method (Fig. 7), along with electrochemical impedance spectroscopy (EIS) measurements. (Fig. 8)
CV tests were conducted at increased scan rates from 0.2 to 1.0 mV s-1 (Fig. 7a–c) to further investigate the diffusion behavior and capacitance contribution of Li ions and electrons in the MnO/Mn3O4/C composite electrodes. As demonstrated in Fig. 7a–c, the shapes of CV curves at different scan rates are similar, indicating the good reversibility of MnO/Mn3O4/C composites. The relationship between peak current (i) and scan rate (v) can be described according to the following equation [74],
According to Fig. 7d–f, the b-values for Mn-N4, Mn-N5 and Mn-N6 are calculated as 0.56, 0.54 and 0.53 for the anodic peak, respectively, demonstrating the combination of diffusion- and capacitance-controlled behavior for the MnO/Mn3O4/C composites [74]. Although pseudocapacitive behavior is not directly associated with conversion reactions in lithium-ion batteries, these results indicate that the presence of mixed-valence Mn sites significantly modifies the local electronic environment, which is highly relevant to the electrochemical behavior of the MnO/Mn₃O₄ system.
The capacity contribution from the capacitance-controlled process can be analyzed quantitatively in the light of equation below,
where k1v and k2v1/2 refer to the surface capacitance-controlled and diffusion-controlled contributions successively. The Mn-N5 electrode exhibits a capacitive contribution ratio of 39.9% at a scan rate of 0.2 mV s-1 (Fig. 7h). For Mn-N4 and Mn-N6, the capacitive contribution ratio of 50.3 and 27.8 % were achieved in Fig. 7g and 7i.
Electrochemical impedance spectroscopy (EIS) was performed to assess the ion diffusion and charge transfer characteristics of the MnO/Mn3O4/C composite electrodes. As shown in Fig. 8, the Nyquist plots for Mn–N4, Mn–N5, and Mn–N6 electrodes were recorded at open circuit voltage (OCV), and the corresponding fitting results using an equivalent circuit model are presented in Fig. 8a. Each Nyquist plot features a semicircle in the high-frequency region, representing the charge transfer resistance (Rct), and a sloped line at medium-to-low frequencies, attributed to the Warburg impedance (Zw), which reflects Li+ ion diffusion.
Interestingly, although Mn-N4 and Mn-N5 exhibit comparable charge-transfer resistance in the Nyquist plots, as shown in Fig. 8a, their textural and surface chemical properties differ significantly. Mn-N4 possesses a higher BET surface area (3.26 m² g-1) and a slightly larger average pore diameter (12.2 nm) than Mn-N5 (1.85 m² g-1 and 9.4 nm, respectively) (Fig. S2), which enhances the electrolyte-we@ed interfacial area and facilitates Li⁺ transport across the electrode–electrolyte interface [75,76]. This enlarged active surface can partially compensate for the intrinsically lower carbon conductivity of Mn-N4, as evidenced by its smaller C–C(sp²) fraction in the C 1s XPS spectra. In contrast, Mn-N5 features a more graphitic carbon network with higher sp² content, enabling more efficient electronic conduction and faster Li⁺ diffusion. These properties are consistent with its higher DLi⁺ values and superior rate and cycling performance, even though the small-signal Rct values appear similar. A more detailed discussion of Li⁺ diffusion behavior will be provided in the following section on GITT analysis.
As highlighted in the inset of Fig. 8b, the measured Rct values for Mn–N4, Mn–N5, and Mn–N6 are 485.2 Ω, 286.2 Ω, and 299.8 Ω, respectively. These results clearly indicate that Mn–N5 exhibits the lowest Rct, demonstrating its superior electron and ion transport properties, which are consistent with its outstanding electrochemical performance.
Although the Mn–N4, Mn–N5, and Mn–N6 samples exhibit similar charge-transfer resistance values prior to cycling, their phase composition and carbon structures differ substantially, enabling meaningful internal comparison without requiring separate MnO-only [17], Mn₃O₄-only [13,16], or Mn₃O₄/C samples [13,30]. Importantly, previous studies have shown that pure Mn₃O₄ anodes generally suffer from rapid capacity fading and poor rate performance due to severe volume expansion and low intrinsic conductivity, typically retaining less than 20–30% of initial capacity after tens of cycles under similar test conditions (see Table S1 in Supplementary Information). In contrast, Mn₃O₄/C composites with sufficient carbon content exhibit significantly improved stability and conductivity, retaining 70–90% of capacity. Mixed MnO/Mn₃O₄ systems reported in the literature also outperform single-phase MnO or Mn₃O₄, attributable to the availability of multiple redox couples and enhanced defect-assisted charge transport [41]. To place our results in context, the Mn–N5 electrode delivers a reversible capacity of 861 mAh g-1 at 0.2 A g-1 and 473 mAh g-1 aGer 1000 cycles at 0.5 A g-1—performance metrics that are competitive with or superior to many reported MnₓOᵧ and MnₓOᵧ/C systems. These comparisons demonstrate that the behavior of the Mn–N4, Mn–N5, and Mn–N6 series already reflects the functional differences typically observed between MnO, Mn₃O₄, and carbon-hybrid controls, confirming that the primary performance differences originate from annealing-guided phase evolution and carbon-network formation within our DMAc-derived system.
To further investigate the lithium-ion diffusion behavior in the MnO/Mn3O4/C composite electrodes, galvanostatic intermittent titration technique (GITT) was employed (Fig. 9). The test involved applying a constant current pulse of 1800 s at 50 mA g-1, followed by an open-circuit relaxation period of 7200 s (Fig. 9c, d). This procedure was systematically repeated for Mn-N4, Mn-N5, and Mn-N6 electrodes. The corresponding GITT results are presented in Fig. 9 and Fig. S3. The lithium-ion diffusion coefficients (DLi+) were calculated using the following equation (4) [77].
where represents the duration of the current pulse, nB denotes the number of moles of active material involved in the reaction, and VM is the molar volume of the electrode material. ΔES is defined as the steady-state potential change measured during the relaxation period after the current pulse, while ΔEτ represents the potential change occurring under the applied current during the pulse duration [78]. Based on the GITT curves, the DLi+ of the Mn-N5 (Fig. 9b and Fig. S3) was found to range from 8.31×10−17 cm2 S−1 to 1.01×10−14 cm2 S−1, which is higher than those of the Mn-N4 (9.89×10−18 cm2 S−1 to 1.11×10−14 cm2 S−1) (Fig. S3a) and Mn-N6 (6.71×10−19 cm2 S−1 to 1.08×10−14 cm2 S−1) (Fig. S3b), indicating superior Li+ ion diffusion kinetics.
The outstanding electrochemical performance of the MnO/Mn3O4/C composite (Mn-N5) stems from a combination of multiple, synergistic factors arising from its integrated structural design and thermal posttreatment. First, the coexistence of MnO and Mn3O4 introduces multiple redox couples (Mn2+/Mn3+), enabling a broader range of reversible lithium storage reactions. This mixed-valence configuration enhances specific capacity by improving electron-transfer kinetics and increasing the availability of electroactive sites during the lithiation/delithiation process. Second, the in-situ carbon derived from the thermal decomposition of DMAc forms a uniform, conductive coating around the MnOx particles. This interconnected carbon network not only facilitates electron transport and maintains interparticle electrical contact, but also suppresses agglomeration and buffers volume changes during cycling—leading to enhanced rate performance and long-term cycling stability. Third, the post-synthesis annealing at 500°C optimally balances crystallinity, defect structure, and oxygen vacancy. This condition results in a porous framework that promotes lithium-ion diffusion, lowers charge-transfer resistance, and maintains structural integrity over extended cycling.
CONCLUSION
In summary, this study presents a rationally designed mixed-valence MnO/Mn₃O₄/C composite anode, synthesized via a simple solvothermal process using N,N-dimethylacetamide (DMAc) as a dual-function solvent and in-situ carbon source. The resulting heterostructure takes advantage of the synergistic effects of Mn2+ and Mn3+ redox couples and carbon network, enabling improved electronic conductivity, structural integrity, and lithium storage capability. Among the as-prepared samples, Mn–N5 (annealed at 500°C under N2) achieved the best balance between crystalline phase formation, oxygen defect density, and conductive carbon network, delivering the highest reversible capacity of 861 mAh g-1 at 0.2 A g-1, retained 309 mAh g-1 at 2 A g-1, and maintained 473 mAh g-1 aGer 1000 cycles at 0.5 A g-1, confirming outstanding cycling durability. EIS analysis showed Mn–N5 had the lowest charge transfer resistance (Rct = 286.2 Ω), attributed to its conductive carbon framework and minimized interfacial impedance caused by reduced surface oxygenated species. XPS and Raman analyses confirmed a favorable Mn2+/Mn3+ ratio and a high sp2 carbon content in Mn–N5, supporting enhanced electron/ion transport. In contrast, Mn–N4 exhibited higher surface passivation and poorer conductivity, while Mn–N6, though more reduced, suffered from excessive structural disorder. Electrochemical testing and kinetic analysis via Dunn’s method and EIS established that the Mn–N5 composite exhibits superior diffusion kinetics and redox reversibility, making it a promising anode material for high-performance lithium-ion batteries.
These findings highlight the pivotal influence of annealing temperature and interfacial chemical engineering in regulating the structural transformation, redox dynamics, and electrochemical performance of mixed-valence MnOx-based electrodes. By enabling the formation of a continuous, conductive carbon framework without relying on costly additives such as CNTs or graphene, this approach presents a scalable and economically viable strategy for advancing high-performance anode materials in lithium-ion batteries. Collectively, this work emphasizes the significance of coupling phase optimization with in-situ carbon matrix design to unlock the full potential of manganese-based systems for next-generation energy storage technologies.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
Supplement1
