Xiang Ding*a,
Cuixian Hua,
Yong Fana,
Yue Linb,
Jinyang Liuc,
Yibing Yangb,
Liangwei Liub,
Haibao Mab,
Yi Xiaob and
Lili Han*b
aCollege of Chemistry and Materials Science, Fujian Normal University, Fuzhou 350007, China. E-mail: dingx@finu.edu.cn
bState Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou 350002, China. E-mail: llhan@fjirsm.ac.cn
cCollege of Physics and Energy, Fujian Normal University, Fuzhou 350007, China
First published on 2nd August 2024
A P2-Na0.67MnO2 (NMO) layered cathode has the merits of high capacity and easy preparation. Nevertheless, the severe phase transition (P2–O2) and inferior electronic/ionic conductivities seriously limit its practical application. Herein, a synergistic doping with Al3+ and Zn2+ into Mn-sites is employed to prepare a series of P2-Na0.67Mn1−x−yZnxAlyO2 cathodes. In situ XRD and in situ EIS measurements profoundly verify the suppressed phase transition, enhanced structural reversibility and even faster charge transfer during the charge–discharge cycle, respectively. Moreover, DFT calculations confirm the much stronger metallicity of the NMO3 electrode than that of NMO, derived from the more likely transition capability of mobile free electrons in the vicinity of the Fermi level. As a result, a more stable structure and stronger conductivity are obtained. The optimum P2-Na0.67Mn0.85Al0.05Zn0.1O2 (NMO3) can display more superior electrochemical performance, including working capacity (181.85 mA h g−1@0.1C), cycling stability (83%@93.87 mA h g−1@600 cycles@5C) in half-cells and energy density (334.5 W h kg−1) in NMO3//hard carbon full-cells, thereby showing extraordinary application potential in advanced sodium-ion batteries.
To address the above issues,14,15 the bulk lattice doping strategy has been extensively employed, mainly consisting of alkalis (e.g. Li+), alkaline-earth metals (e.g. Ca2+ and Mg2+) and multi-valent ions (e.g. Zn2+, Fe3+, Co3+, and Ti4+) incorporated into the Mn site, respectively. It can adjust the electronic and spatial structures by creating a multi-electronic environment and providing pillar effects, thus boosting the bulk electron conductivity, Na+ diffusion efficiency and structural stability. Ling et al. reported Na0.67Cu0.1Mn0.9O2 that manifested a suppressed Jahn–Teller effect as well as comparatively less phase transition. It can deliver a high capacity (222.7 mA h g−1@10 mA g−1; 1.5–4.2 V) with capacity retention (76%@1 A g−1@300 cycles).16 Zhang's group introduced electrochemically active Fe3+ into the Mn-site to enhance the Na+ diffusion kinetics. The optimized Na2/3Ni1/3Mn7/12Fe1/12O2 exhibited long-term cycling stability with a fading rate of 0.05% per cycle over 300 cycles.17 Most recently, a series of co-doping studies have been performed and turned out to be far more efficient for optimizing the electrochemical properties and structures than the corresponding individual-doping ones.18 Generally, co-doping has emerged as the preferable optimization technique due to the more comprehensive regulation on each lattice site, which can lead to increased delocalized electron density, ionic radius and electronegativity rooted in the introduced ions.19 Luo et al. proposed a Cu/Fe co-doped Na0.67Mn0.92Fe0.04Cu0.04O2 cathode that can deliver superior discharge capacity (90.68%@110.5 mA h g−1@100 cycles@5C).20 Huang et al. adopted Zn/Mg dual-doped P2-Na0.67MnO2 which demonstrates a high performance of 67.2 mA h g−1@10C.21 Hence, the synergistic effect of co-doping into the Mn-site is always more effective in performance optimization.
In this work, a series of Zn/Al co-doped P2-Na0.67Mn1−x−yAlxZnyO2 are well-designed by regulating different dopant amounts. A range of in situ and ex situ characterization techniques and DFT calculations systematically elucidate the Zn/Al co-doping synergistic effect, occurrence of of suppressed P2–O2 phase transition and much enhanced structural stability and bulk electric/ionic conductivities during cycles. Consequently, the optimal P2-Na0.67Mn0.85Al0.05Zn0.1O2 (NMO3) can deliver superior discharge capacity (181.85 mA h g−1@0.1C) and cycling stability (83%@93.87 mA h g−1@600 cycles@5C) in half-cells and sufficient energy density (334.5 W h kg−1) in NMO3//HC full-cells. This efficacious optimization tactic of collaborative doping is above rubies for designing pragmatic layered P2-phase materials for SIBs.
Fig. 1 (a) Schematic structure of NMO and NMO3; (b) XRD patterns; Rietveld refinement of (c) NMO and (d) NMO3; (e) Zn 2p, (f) Mn 2p and (g) Al 2p. |
In general, Al3+/Zn2+ co-doping can expand the interplanar spacing, thereby promoting the diffusion kinetics of Na+. Furthermore, the valence state of elements is determined by XPS measurements (Fig. 1e–g). The binding energies located at 1043.8 and 1020.8 eV are ascribed to the 2p1/2 and 2p2/3 of Zn2+ (Fig. 1e), respectively. The peaks situated at 643.6 and 654.5 eV and 641.7 and 653.6 eV are attributed to the 2p1/2 and 2p2/3 of Mn3+ and Mn4+ (Fig. 1f), respectively. This proves the existence of mixed valence states of Mn3+/4+ in the NMO3 sample. Besides, the binding energies located at 74.18 and 73.18 eV are assigned to the 2p1/2 and 2p2/3 of Al3+ (Fig. 1g). Hence, the valence states of Mn, Zn and Al are +3/+4, +2 and +3 in the NMO3 sample.
To explore the morphology of these P2-phase materials, SEM and TEM images are obtained (Fig. 2a, b and S2†). The NMO3 material presents a regular hexagonal lamella with diameter of 3 μm as shown in the SEM image (Fig. 2a). The thickness of such a lamella is measured from the cross section to be 1 μm from the TEM image (Fig. 2b). The HRTEM image displays a lattice spacing of 0.207 nm that corresponds to the (103) crystal plane in the P2-phase NMO3 layered structure26 (Fig. 2c). The SAED patterns clearly exhibit the (010), (100) and (110) lattice planes in the hexagonal P63/mmc space group, revealing excellent crystallinity (Fig. 2d). Moreover, the EDS-mapping images indicate the uniform distribution of Na, Mn, O, Al and Zn elements in the lamellar NMO3 material (Fig. 2e–i). The sparse bright spots of Al and Zn elements are due to the minor doping content compared to Na, Mn, and O elements.
Fig. 2 (a) SEM, (b) TEM, and (c) HRTEM images, (d) SAED patterns and (e–i) EDS-mapping images of the NMO3 material. |
The electrochemical properties in half-cells are tested during 2.0–4.2 V as illustrated in Fig. 3a–e. The NMO3 sample is capable of delivering an initial discharge capacity of 181.85 mA h g−1 (Fig. 3a), higher than that of the others (55.9, 39.25, 18.25, and 12.3 mA h g−1) (Fig. S3†). The redox peaks in the 1st, 2nd and 3rd CV curves (Fig. 3b and S4†) are consistent with the charge–discharge voltage plateaus. In the 1st charge process, the oxidation peaks at about 2.33 V and 3.094 V can be ascribed to the Mn3+ → Mn4+ redox, along with the extraction of Na+ from interlayers. Subsequently, in the 1st discharge process, the reduction peaks located at 2.0 V and 3.023 V are assigned to the Mn4+ → Mn3+ redox,10,11 accompanied by the Na+ insertion process. These redox polarization values in CV curves for NMO3 (0.33 V) are far smaller than those of NMO (0.446 V), NMO1 (0.44 V), NMO2 (0.437 V) and NMO4 (0.429 V), respectively, revealing much better electrochemical reversibility. Besides, the modified NMO3 sample displays superior rate capabilities (e.g. 124.64 mA h g−1@1C; 94.87 mA h g−1@5C) compared with the others (Fig. 3c). The cycling performances at 0.1C are demonstrated in Fig. 3d. The NMO3 material shows the most competitive cycling stability (89%@181.85 mA h g−1@100 cycles@0.1C) as expected. Furthermore, to investigate their long-term cycling abilities, 600 charge–discharge cycles at 5C are carried out (Fig. 3e). The NMO3 cathode can deliver a capacity of 94.87 mA h g−1 initially with 83% capacity retention during 600 cycles. To the best of our knowledge, such an NMO3 material is superior to those in selected latest reported studies (Fig. 3f). On increasing the loading mass to 5 mg cm−2 as shown in Fig. S7,† the NMO3 sample can still deliver a superior initial discharge capacity of 179.29 mA h g−1, rate capabilities (e.g. 123.71 mA h g−1@1C; 90.95 mA h g−1@5C), and cycling stability (93%@179.29 mA h g−1@30 cycles@0.1C), thereby showing good application potential. Moreover, the DNa+ of these samples is evaluated by GITT tests (Fig. 3g, h and S5†). The DNa+ of NMO3 ranges from 10−9.6 to 10−8.1 during a cycle, while they are 10−10.5 to 10−9 for the NMO sample, respectively. As shown in Fig. 3g, h and S5,† the calculated Na+ diffusion coefficients (DNa+) of NMO1, NMO2, NMO3, and NMO4 samples vary from 10−10.4 to 10−9, 10−10.3 to 10−8.4, 10−9.6 to 10−8.1 and 10−9.9 to 10−8.1, respectively. Hence, co-doping of Zn2+ and Al3+ cations indeed boost the DNa+ compared to that of bare NMO, and the highest DNa+ of NMO3 indicates the fastest reaction kinetics so as to offer superior rate capability.
Fig. 3 Charge–discharge at 0.1C (a); CV curves (b); rate (c); cycling at 0.1C (d); long-term cycling at 5C (e); performance comparison27–31 (f); GITT and corresponding DNa+ (g and h) for NMO and NMO3. |
In order to clarify the structural evolution process behind the improved electrochemical performance of the NMO3 sample, in situ XRD patterns are obtained (Fig. 4a and b). In the charge process (2.0–4.2 V), the diffraction peaks of (002), (100), (102), (104) and (103) gradually shift towards smaller 2θ degrees by about 0.5° (Fig. 4a), respectively, accompanied by Na+ extraction from the Mn–O interlayers and lattice expansion. Subsequently, these diffraction peaks consistently move towards a larger 2θ angle and return to their original positions, along with Na+ insertion into the Mn–O interlayers and lattice contraction. Hence, derived from such Zn and Al co-doping, highly symmetrical diffraction peaks during a cycle symbolize excellent electrochemical reversibility, indicating an effectively suppressed P2 → O2 phase transition. Furthermore, for investigating the evolution of electronic conductivity in a complete electrochemical cycle, in situ EIS curves are obtained (Fig. 4b–e). In the 2nd cycle, the Rct values progressively reduce first (charging-Na+ extraction-lattice expansion) and then increase later (discharging-Na+ insertion-lattice contraction) for the NMO3 sample (Fig. 4c), which is consistent with in situ XRD analysis (Fig. 4a). Besides, the Rct for the NMO sample also undergoes the same trend of change in the 2nd cycle except for larger values than those of NMO3 (Fig. 4b). Moreover, the Rct maintains the same pattern of change as before with increased values for both NMO and NMO3 samples in the 600th cycle (Fig. 4d and e). Meanwhile, it can be clearly observed that the Rct values of the NMO3 sample are always smaller than that for the NMO sample in both the 2nd and 600th cycles (Fig. 4f and g). Such results profoundly clarify the strengthened Na-insertion/extraction kinetics which originated from a larger lattice, more stable structure, and faster ionic/electronic conductivities of the NMO3 material. The ICP-OES result of the electrolyte (1.0 M NaClO4 in PC with 5 vol% FEC) after 600 cycles has displayed that the content of the Mn element is 0.0084 mol, which means the dissolution of Mn2+ (disproportionation reaction: 2Mn3+ ↔ Mn4+ + Mn2+) due to the Jahn–Teller effect.10,11 As a result, there is about 1% Mn dissolution in the electrolyte that deteriorates the cycling performance of the battery, but to a small extent due to the co-doping of Al/Zn.
Fig. 4 In situ XRD patterns (a); in situ EIS curves in the 2nd (b and c) and 600th (d and e) cycles, and corresponding evolution of Rct values (f and g) for NMO3 and NMO samples. |
To explore the evolution of the morphology and structure after long-term cycles, ex situ XRD and SEM are carried out as shown in Fig. 5. The diffraction peaks of (002), (004), (102), (103) and (104) become slightly weaker than the pristine for the NMO3 sample, while they become very weak, and some even disappear for the NMO sample compared with the fresh sample (Fig. 5a). This suggests the much better structural stability for modified NMO3 than bare NMO. Moreover, the morphological evolution process also displays smaller cracks of bulk particles during long-term cycling (2nd → 100th → 600th) for NMO3 than for NMO (Fig. 5b). Besides, EDS-mapping images (Fig. 5c–f and S6†) and EDS line-scan curves (Fig. 5g) also confirm that the Na, Mn, Zn and Al elements are still uniformly distributed on the surface of NMO3 particles, rather than excessive migration or dissolution in electrolyte, revealing inhibited Mn3+ migration and phase transitions. Thus, the structural collaborative doping modification is conducive to both microscopic and macroscopic stabilities in long-term cycling.
Fig. 5 Ex situ XRD patterns (a) and SEM images (b) of NMO and NMO3; EDS-mapping images (c–f) and EDS line-scan curves (g) of NMO3 after 600 cycles. |
To evaluate the practicality of the as-designed NMO3 cathode, NMO3//HC full-cells are assembled as shown in Fig. 6a–e. Fig. 6a shows the operation schematic diagram of full-cells. In detail, NMO3//HC is able to exhibit a competitive working capacity of 165.89 mA h g−1 with an energy density of 334.5 W h kg−1 (Fig. 6b). The energy density of the NMO3//HC full-cell is 331.48 W h kg−1 taking the amount of electrolyte (60 mg) into account. Moreover, it can display reliable cycling stability (84%@140.54 mA h g−1@50 cycles@0.1C) as well as superior rate capabilities (86.38 mA h g−1@2C) as illustrated in Fig. 6c–e. The unexceptional cycling stability is due to the mediocre cycling performance (94.7%@244.58 mA h g−1@50 cycles@0.1C) of the HC anode as shown in Fig. S8,† and we will improve this issue in the future. Hence, such an NMO3 cathode has great potential for commercial application.
Furthermore, for clarifying the structure–activity relationship behind such synergistic doping, DFT calculations are carried out (Fig. 6f–i). Obviously, discrete islands comprising Na 1s, Mn 3d and O 2p orbitals are detected surrounding the Fermi level for all the samples that show characteristics of conductors due to the overlap of conduction bands and valence bands.32 In contrast, two additional discrete islands for Al and Zn 3p bonds are situated between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). Meanwhile, DOS of NMO3 (Fig. 6i) shows the largest electron cloud density located near the Fermi level than that of the NMO (Fig. 6f), NMZO (Fig. 6g) and NMAO (Fig. 6h) materials. It must be noted that the enhanced phase conductivity after Al3+ and Zn2+ doping is consistent with the evolution of Rct values as shown in Fig. 4b–e, thereby manifesting the strongest reaction kinetics in both the bulk and interface. Thus, the introduction of Al3+ and Zn2+ elements into the structure can supply more valence bands and conduction bands with contiguous quantized energies. The free electron transition becomes much easier, thereby exhibiting better conductivity.33
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta04224g |
This journal is © The Royal Society of Chemistry 2024 |