Co-Doping Inducing d-Electron Delocalization in α-MnO₂ for High-Performance Zinc-Ion Batteries

Abstract

Element doping technology is widely recognized as an effective strategy for high-performance MnO₂-based cathode materials. While this approach improves the electronic and ionic conductivity of MnO₂, it is often accompanied by the introduction of oxygen vacancies. This synergistic effect poses challenges for precisely investigating the effect of doping elements on the d-electron configuration of the Mn site and establishing atomic-level structure-activity relationships for high-energy aqueous zinc-MnO₂ batteries. In this paper, the rational design of d-electron configurations in the Mn site has been achieved through simple Co doping in α-MnO₂ (CMO). Experimental results confirm that the introduction of Co can delocalize the d-electrons of the Mn site and increase the ratio of e_g (d_z² and d_x²_−y²) occupancy. Consequently, the charge transfer resistance, electrode polarization, and Zn²⁺ diffusion coefficient of the CMO-2 cathode have been greatly optimized. Thus, the as-prepared electrode delivers a high specific capacity of 287.4 mAh g⁻¹ at 1 A g⁻¹, with a capacity retention rate of 92.8% and a corresponding remaining capacity of 199.7 mAh g⁻¹ after 700 cycles. This study paves the road for the designation and construction of high-energy MnO₂ cathodes with optimized electronic structures for advanced aqueous zinc ion batteries.

1. Introduction

The urgent need for high-performance and scalable energy storage has triggered a great deal of research into battery technologies beyond lithium-ion systems. Aqueous zinc-ion batteries (AZIBs) have gained considerable attention due to the advantages of non-flammable aqueous electrolytes, minimal environmental impact, and significantly lower material costs [1,2,3,4]. MnO₂ is a promising cathode candidate owing to its high theoretical capacity (≈308 mAh g⁻¹), multiple accessible oxidation states, natural abundance, and low cost [5,6,7]. The electrochemical viability of MnO₂ for AZIBs applications was first conclusively demonstrated by Kang and colleagues in 2011, which established the foundation for subsequent research efforts in this field [8]. However, the application of MnO₂ cathodes is still hindered by the critical limitations of unsatisfactory electronic and ion conductivity. Worse still, the redox processes of MnO₂ involve complex phase transformations and dissolution issues in which Mn(IV) reduction produces unstable Mn(III) that disproportionates into Mn(IV) and Mn(II), resulting in permanent structural degradation and material loss [9,10]. Plenty of works have been reported focusing on nanostructuring and composite engineering strategies, while the capacity and cycling stability of MnO₂ are still insufficient [11,12]. The key to developing high-performance MnO₂-based materials should be emphasized on the rational design of their intrinsic electronic configuration, which is essential for fundamentally improving electronic and ionic transport kinetics, as well as establishing a clear atomic-level understanding of the structure–performance relationship [13,14,15].

The electronic configuration of MnO₂ is intrinsically related to the delocalization degree of d-electrons and the occupancy ratio of the e_g (d_z² and d_x²_−y²) and t_2g orbitals (d_xy, d_xz, and d_yz) of the Mn site [16,17,18]. Element doping to achieve more delocalized electrons and a high occupancy ratio of the e_g orbitals is regarded as an appealing strategy for high electrochemical performance of MnO₂ [19,20,21]. It should be noted that the abundant oxygen vacancies and expanded lattice spacing can be widely observed during the doping process, and the enhanced electrochemical performances are always attributed to the synergistic effect between dopants and oxygen vacancies, which is, however, an important obstacle to the precise design of the d-electron configuration of MnO₂ and a better understanding of the mechanisms for enhanced electrochemical performance [22,23,24,25]. The research, conducted by Professor Gong Yun’s team [26] from Chongqing University, prepared vanadium-doped manganese dioxide K_0.41V_0.26MnO_2−x·0.5H₂O (VMnO). It was found that V doping effectively suppressed the Jahn-Teller effect through orbital hybridization and electron transfer, stabilizing the low-valent Mn. Meanwhile, oxygen vacancies disrupted the local symmetry of the Mn central coordination environment, facilitating electron transfer along the path of Mn-O-V, thereby enhancing electronic conductivity. Recently, Feng’s group [27] proposed a Mg-doped α-MnO₂ (MMO) with deficient oxygen vacancy by precisely controlled hydrothermal reaction conditions to reveal the mechanisms of Mg doping in boosting MMO’s electrochemical performance. The electron/ion transfer kinetics for Zn²⁺ storage had been improved, and the as-prepared electrode can deliver a capacity of 311 mAh g⁻¹ at 0.6 A g⁻¹. Nevertheless, the d-electron configuration of MnO₂ remains inadequately characterized, and the work mechanism of the dopants on the specific capacity and the kinetics of MnO₂ should be deeply discussed from an electronic perspective.

In this work, a Co-doped α-MnO₂ (CMO) with deficient oxygen vacancy is prepared to optimize the d-electron configuration of MnO₂ with delocalized d-electrons spin states and higher e_g occupancy. The selection of Co²⁺ (0.65 Å) as a dopant was based on its near-identical ionic radius to both Mn³⁺ (0.645 Å) and Mn⁴⁺ (0.53 Å), ensuring less impact on the crystal structure of α-MnO₂. The X-ray photoelectron spectroscopy (XPS), electron paramagnetic resonance (EPR), and density functional theory (DFT) calculations have been conducted to comprehensively investigate the d-electron configurations of Mn sites. The structural characterization and EPR data suggested that CMO-2 featured stable crystal structures and a deficient oxygen vacancy. XPS and theoretical calculations demonstrated that CMO-2 had an optimized e_g occupancy ratio of 44.15%, higher than that of MnO₂ (40.09%). Band structure analysis indicates that CMO-2 exhibits a narrower bandgap and a significantly elevated density of states (DOS) near the Fermi level. This transformation in electronic structure fundamentally enhances its conductivity, thereby optimizing electron transport efficiency within the electrode. Leveraging these advantages, the CMO-2 electrode achieved a high specific capacity of 287.4 mAh g⁻¹ at a current density of 1 A g⁻¹. Its energy density and power density reached 183.9 Wh kg⁻¹ and 659.7 W kg⁻¹, respectively, with overall performance surpassing the vast majority of reported MnO₂-based AZIBs materials. In principle, an effective strategy of engineering the d-electron configurations has been presented in this work to construct a lattice-stable and kinetically optimized cathode towards advanced AZIBs.

2. Results and Discussion

The structure of the prepared CMO is shown in Figure 1a. In a typical preparation process, an appropriate amount of KMnO₄ (Shentai Chemical Reagent Co., Ltd., Tianjin, China) and MnSO₄ (GFCR Technology Reagent Co., Ltd., Tianjin, China) were added to deionized water to form a uniform solution. By gradually increasing the concentration of Co(NO₃)₂·6H₂O (GFCR Technology Reagent Co., Ltd., Tianjin, China), CMO-1, CMO-2, and CMO-3 were successfully prepared through a one-step hydrothermal method. As shown in Figure 1b,c, CMO and bare MnO₂ showed a nanorod structure, and the morphologies are not changed by Co doping. (Figure S1a,c). The samples are further analyzed by a microscope (TEM). As shown in Figure 1d and Figure S1b,d, the surface of the CMO-1, CMO-2, and CMO-3 is smooth, and no obvious defects or cracks are found, which is basically consistent with α-MnO₂ (Figure 1f). Figure 1e,g (HRTEM images) demonstrates clear streaks with the crystal lattice fringe spacing of 0.258 and 0.253 nm, corresponding to the (211) crystal face of CMO-2 and α-MnO₂, indicating that the nanorods have good crystallinity inside. In addition, the uniform distribution of Co, Mn, and O elements in the sample confirmed the successful preparation of CMO, as shown in Figure S1e.

XRD is further utilized to investigate the crystal structure of CMO and α-MnO₂. As shown in Figure 2a, the diffraction peaks of all materials are well matched to those of JCPDS No. 44−0141 MnO₂, demonstrating that Co doping does not produce a new crystal structure. Peaks at 12.42°, 17.84°, 28.64°, and 37.2° belong to the (110), (200), (310), and (211) faces of MnO₂, respectively. It can be observed that the XRD patterns of CMO-1, CMO-2, and CMO-3 exhibit a slight left shift in peak positions and minor variation in peak sharpness compared to that of α-MnO₂, which can be ascribed to the embedding of Co²⁺ [28]. However, more attention should be focused on the fact that there is no significant peak deviation that can be detected for all these samples, and the characteristic peaks of (110), (200), and (310) facets are still located at similar positions, demonstrating that Co doping negligibly affects the crystal structure of MnO₂. To further verify the Co-doped crystal structure, the Raman spectra are carried out for the as-prepared materials, and the results are displayed in Figure 2b. The α-MnO₂ featured a typical 2 × 2 tunnel structure consisting of [MnO₆] octahedra, and the positions of the characteristic peaks of the four samples remained basically unchanged. The bands centered at ~579 and 633 cm⁻¹ are related to the symmetric Mn-O vibrations, while the peak at ~383 cm⁻¹ can be indexed to the Mn-O bending vibrations [29,30]. In addition, a slight decrease in peak intensities at ~633 cm⁻¹ is detected, which can be attributed to a decreased average Mn oxidation state and a strong interaction between Co and oxygen in the [MnO₆] octahedra. Figure 2c illustrates the crystalline structure of CMO, which basically consists of the [MnO₆] octahedral of α-MnO₂, while partial Mn was substituted by Co. The above results indicate that Co²⁺ is successfully incorporated into the MnO₂, and the tunnel structure consisting of [MnO₆] octahedral is not destroyed by Co doping.

The content of dopants also has a significant influence on the electronic configuration of the materials. Previously, efforts have reported that a high ratio of Mn/O in MnO₂-based materials would lead to an increased average valence of Mn, suggesting a higher e_g occupancy and more delocalized d-electrons [31,32,33]. To accurately determine the element content of these materials, XPS is conducted, and the as-calculated results are displayed in Figure 2d. The Co content of CMO-1, CMO-2, and CMO-3 is 3.81, 4.22, and 4.24%, respectively. Notably, CMO-2 showed the highest Mn/O ratio of 55.82/39.96, while the Mn/O content of CMO-1 and CMO-3 were 54.63/41.56 and 54.39/41.37%, respectively, which is consistent with the results obtained from inductively coupled plasma optical emission spectrometry (ICP-OES), demonstrated in Table S1. Compared with CMO-2, a further increase in the precursor concentration shows a limited impact on the Co content, but reduces the ratio of Mn/O in CMO-3, which can be ascribed to the excessive amount of precursors. All these results indicated that the CMO-2 can be regarded as the optimized candidate for Zn²⁺ storage, and further detailed characterizations and electrochemical performance will be focused on the sample.

In manganese-based cathode materials, the introduction of oxygen vacancies represents a core strategy for optimizing their electrochemical performance. This process is typically achieved through elemental doping or reduction. Nevertheless, in order to more accurately study the intrinsic effect of the doping element on the electronic structure of α-MnO₂, the oxygen vacancies in the material should be minimized as much as possible. Our EPR measurements (Figure 3a) reveal no characteristic signals of oxygen vacancies, demonstrating that bare MnO₂ and CMO-2 both featured deficient oxygen vacancy characteristics, indicating that CMO-2 can serve as an excellent candidate material for investigating the influence of Co doping on the d-electron configuration of the Mn site.

To further investigate the electron configurations of the cathodes, XPS spectra of CMO-2 were analyzed. Figure 3b is the Mn 2p spectrum consisting of Mn 2p_3/2 at 642.1 eV and Mn 2p_1/2 at 653.8 eV, with an energy difference of 11.7 eV. Compared with bare MnO₂, the Mn 2p position of CMO-2 slightly shifts to the lower binding energy, which can be attributed to the doping by other metal ions. The Mn 2p_3/2 orbital has been fitted into two characteristic peaks at ~642.41 eV and ~644.8 eV, corresponding to Mn³⁺ and Mn⁴⁺, respectively. Meanwhile, the Mn 2p_1/2 orbital can be well fitted into peaks located at 653.6 eV (Mn⁴⁺) and 655.8 eV (Satellite peak) [16]. XPS analysis indicates that the Mn³⁺/Mn⁴⁺ ratio in CMO-2 is 3.7, higher than the 3.3 observed in bare MnO₂. This result confirms that the introduction of cobalt induces a shift in manganese sites from Mn⁴⁺ to Mn³⁺, which suggests that the enhanced delocalization of electrons and increased d-electron cloud density of the Mn site have been successfully achieved in CMO-2. The Co 2p spectrum in Figure S2 presents typical peak positions of Co³⁺, corresponding to the Co 2p_3/2 and Co 2p_1/2 signals with a spin-energy separation of 15.1 eV [34]. As for the O 1s spectra, as shown in Figure 3c, there are three distinct peaks at~529.7 eV, ~531.3 eV, and 535.5 eV, attributable to the Mn-O-Mn, Mn-O-H, and H-O-H bonds, respectively [35]. The major O 1s peak for the sample shifts to a lower position, indicating a change in the coordination configuration of Metal-O (Mn, Co), probably due to the formation of the Co-O bonds [36]. By comparing the integrated areas of the peaks of α-MnO₂ and CMO-2, the percentage of Mn-O-H is increased from 0.24 to 0.29 after the modification of the Co atom, which can be attributed to the combination of Co with O within the MnO₂ lattice [37,38].

The density functional theory (DFT, by VASP5.4.4 (VASP Software GmbH., Vienna, Austria)) calculations are employed to elucidate the influence of Co doping on the d-electron configurations of the Mn site. As revealed in Figure 3d, the charge density difference (CDD) results indicate that the distortion degree in CMO-2 is greater than in the bare MnO₂, in which a strengthened response of electrons can be observed. Notably, the interfacial charge state of the Mn atoms connected to the Co atom shifts from a slight charge accumulation to a pronounced charge depletion, indicating that the doping of Co significantly enhances the d-electron delocalization of Mn. Figure 3e presents the contribution of different d orbitals of the Mn atom to DOS (pDOS), listed in Table S2. Notably, the CMO-2 possesses a ratio of eg (d_z² and d_x²_−y²) occupancy around 44.15%, higher than the 40.09% for α-MnO₂, indicating clearly an optimized d-electron configuration and an enlarged d-electron cloud density in CMO-2, which agrees well with the XPS and CDD results. Impressively, CMO-2 presents a decreased bandgap near the Fermi level from ~0.62 eV to ~0.41 eV, attributable to the more delocalized electrons [39]. All the above results demonstrate that the precise control of experimental conditions enables the incorporation of Co without introducing additional oxygen vacancies, while effectively optimizing the d-electron configuration of Mn. CMO-2 possesses appealing structural advantages at both the crystallographic and electronic levels, including negligible oxygen vacancies, higher Mn content, enhanced d-electron delocalization, higher e_g orbital occupancy, and reduced bandgap, suggesting a promising potential for electrochemical Zn²⁺ storage.

The electrochemical performance of α-MnO₂, CMO-1, CMO-2, and CMO-3 was investigated as a cathode with zinc foil as an anode and 2 M ZnSO₄/0.2 M MnSO₄ aqueous solution as electrolyte. The corresponding galvanostatic charge/discharge (GCD) curves at 1 A g⁻¹ between 0.8 and 1.9 V are demonstrated in Figure S3. A comparative analysis of these cathodes reveals that Co doping can significantly enhance the specific capacitance and reduce polarization voltage simultaneously, and CMO-2 demonstrates the highest specific capacity of 217.4 mAh g⁻¹ and the lowest polarization potential of ~173 mV at 1 A g⁻¹, which is consistent with the previous characterizations and discussions about the electronic and crystal structure.

To investigate the zinc storage behavior of CMO-2, cyclic voltammetry (CV) was employed across the 0.8–1.9 voltage range at a scan rate of 0.3 mV s⁻¹. Figure 4a reveals that CMO-2 exhibits redox characteristics similar to those of α-MnO₂. The peaks observed near 1.32 V and 1.21 V correspond to the insertion/extraction reactions of H⁺ and Zn²⁺, respectively. The redox peaks of the CMO-2 displayed a larger integrated area, enhanced current response, and decreased peak potential separation compared with the α-MnO₂ electrode, indicating an enhanced charge storage capacity and improved reaction kinetics and reversibility.

The rate performance of CMO-2 and α-MnO₂ has been systematically evaluated. As shown in Figure 4b,c, the specific capacity of CMO-2 is 236, 233, 219, 209 mAh g⁻¹ from 0.3 to 1 A g⁻¹, respectively, while that of bare MnO₂ is only 187, 124, 112, 98 mAh g⁻¹. Even at 8 A g⁻¹, CMO-2 can still achieve a capacity of 65 mAh g⁻¹, significantly superior to the 13 mAh g⁻¹ of bare MnO₂ (Figure 4d). When the current density returns to 1 A g⁻¹ again, the capacity of α-MnO₂ is about 45 mAh g⁻¹, while the capacity of CMO-2 can still reach 200 mAh g⁻¹. The enhancement of H⁺ intercalation kinetics through cobalt doping enables the CMO-2 cathode to maintain a longer voltage plateau than bare MnO₂, even at high current densities, specifically within the voltage range above 1.3 V [27]. The long-cycle stability of the prepared samples is tested at 1 A g⁻¹. As shown in Figure 4e, the CMO-2 can deliver the highest capacity of 287.4 mAh g⁻¹ with 92.8% capacity retention after 700 cycles, whereas the specific capacity of bare MnO₂ is only around 54.49 mAh g^-1. When the current density is raised to 4 A g⁻¹, the CMO-2 electrode delivered an average capacity of 107 mAh g⁻¹ in Figure S4, and the capacity can be maintained above 100 mAh g⁻¹ after 500 cycles, which is much better than that of α-MnO₂ (25/15 mAh g⁻¹). As shown in Figure S5, the CMO-2 cell can deliver an energy density of 183.9 Wh kg⁻¹ and a power density of 659.7 W kg⁻¹, which are superior to most of the previously reported works. The commonly used electrolyte for aqueous zinc-ion batteries is ZnSO₄ and MnSO₄, and when manganese-based materials are used as the cathode for aqueous zinc-ion batteries, the voltage window is generally within the range of 0.8–2.0 V. As can be seen from the Figure S5, the specific capacity range of these materials at 1A g⁻¹ is 60–180 mAh g⁻¹, which is significantly lower than that of CMO-2. CMO-2 maintains a capacity of around 200 mAh g⁻¹ after 700 cycles at 1A g⁻¹, and the highest specific capacity during cycling can reach 287.4 mAh g⁻¹.

For a better understanding of energy storage mechanisms, ex situ XRD is carried out at 0.2 A g⁻¹ to investigate the crystalline composition cycling from P1 to P8. As shown in Figure 5a–c, the diffraction peaks of zinc hydroxide sulfate (ZSH) exhibit high reversibility during charge–discharge cycles. The signal initially appears at point P1 and intensifies as discharge progresses to P2; during the charging phase, starting from P3, it gradually weakens until it completely disappears at point P5 (1.9 V). Furthermore, the brief single plateau in the initial discharge curve, coupled with the coexistence of MnOOH and zinc manganese oxide (ZMO) diffraction peaks at point P2, collectively reveals the mechanism of synergistic H⁺ and Zn²⁺ co-intercalation into the crystal lattice during the first discharge cycle. During the second cycle (P5 to P8), a continuous weakening of the ZMO diffraction peak signal is observable [22]. Meanwhile, the slight peak shift can be observed in the XRD patterns of ZSH and ZMO during cycling. The bulk crystal structure of α-MnO₂ remains stable throughout the cycling process, with its diffraction peaks consistently present, while simultaneously confirming that the Zn²⁺ insertion/extraction reaction is only partially reversible [40,41]. The morphologies of the CMO-2 cathode during the test are displayed in Figure S6. Numerous layered sheets can be clearly observed when discharged to 0.8 V, then disappear after fully charging to 1.9 V, attributing to the reversible forming of ZSH species. The fibrous morphology of the α-MnO₂ has been well maintained, however, revealing the optimized structural stability.

A systematic analysis of the relationship between scan rate and peak current has laid the foundation for a deeper understanding of its energy storage kinetics mechanism [38]. As shown in Figure 5d and Figure S7, the cyclic voltametric curves of CMO-2 and α-MnO₂ within the range of 0.1 to 1 mV s⁻¹ exhibit excellent peak stability. Although peak shifts occur with increasing scan rates due to enhanced polarization, the fundamental shapes of the redox peaks remain preserved. As shown in Figure 5e and Figure S8, the calculated b-values for CMO-2 (0.60–0.62) are higher than those of α-MnO₂ (0.58–0.60), indicating that they have similar kinetic behaviors, while CMO-2 shows an increase in the proportion of capacitive contribution. High capacitive contribution of 87.34% for CMO-2 electrode at 1.0 mV s⁻¹ can be obtained (the inset of Figure 5e), which is superior to the 76.15% for bare MnO₂ (Figure S9). The proportion of capacitive contributions progressively increased during the test for both cathodes (Figure 5f and Figure S10); however, the capacitive percentage of CMO-2 was persistently higher than that of α-MnO₂, demonstrating a capacitive-dominated charge storage mechanism at high scan rates and thus endowing the CMO-2 with an outperformed electrochemical performance. The electrochemical impedance spectroscopy (EIS) is utilized to comparatively analyze the key impedance parameters of CMO-2 and α-MnO₂ cathodes, as displayed in Figure 5g. The CMO-2 cathode shows a smaller diameter of the semicircle in the high-frequency region and possesses a lower charge transfer resistance of 122 than that of MnO₂ (181.5), indicating that Co²⁺ doping has a positive effect on the electrochemical kinetics of the system. Furthermore, the galvanostatic intermittent titration technique (GITT) is carried out to investigate the ion diffusion in the as-designed cathodes [42]. As shown in Figure 5h and Figure S11, CMO-2 displayed a higher ion diffusion coefficient, ranging from 10^−13.42 to 10^−10.21 cm² S⁻¹, than that of bare MnO₂ (10^−14.28 to 10^−11.14 cm² S⁻¹), indicating the H⁺ and Zn²⁺ insertion processes were simultaneously boosted by Co doping [43,44,45]. Such improved Zn²⁺ transport kinetics of CMO-2 may also be related to its decreased adsorption energy of Zn²⁺, as depicted in Figure 5i.

3. Conclusions

In summary, we have proposed an oxygen-vacancy-free Co-doping strategy, utilizing a one-step hydrothermal method to synthesize bare MnO₂ and MnO₂ with different cobalt doping ratios. This approach aims to rationally design the d-electron configuration of manganese (Mn) centers. The introduction of Co endows the CMO-2 with better lattice stability, higher e_g occupancy, and stronger electron delocalization, which not only guarantees a high capacitive performance but also benefits the improvement of the electron/ion transport kinetics of active materials. All the above merits jointly boosted the energy and power densities of the devices. As anticipated, the composite electrodes exhibit an excellent specific capacity of up to 287.4 mAh g⁻¹ at 1 A g⁻¹, simultaneously with a satisfactory rate and cyclic performance. The synergistic effect of Mn and Co in the CMO-2 provides a high energy density (183.9 Wh kg⁻¹) and power density (659.7 W kg⁻¹). We believe that the strategy presented here displays an important research direction for the high-performance AZIBs.