Citation

Zhang Zhijia, Sun Shihao, Chen Yuefang, Wei Yanhao, Zhang Mengmeng, Li Chunsheng, Sun Yan, Zhang Shaofei, Jiang Yong. Epitaxial growth of Cu_2-xSe on Cu (220) crystal plane as high property anode for sodium storage[J]. Chin. Chem. Lett., 2024, 35(7): 108922  

Zhijia Zhang^a,*, Shihao Sun^a, Yuefang Chen^a, Yanhao Wei^a, Mengmeng Zhang^a, Chunsheng Li^b,c,*, Yan Sun^b,c, Shaofei Zhang^d,*, Yong Jiang^a     

Resource shortage of lithium is becoming a major issue that will impede the development of energy storage devices [1-3]. Inspired by the similar physicochemical properties with lithium, sodium (Na) has aroused strong research interest in the application of Na-ion battery (SIB) [4-8]. The practical feasibility of SIBs, however, is still limited by the sluggish reaction kinetics and poor cycling stability, mainly attributes to repeated insertion/extraction of high radius Na⁺ (1.02 Å) together with serious structural collapse of the electrodes [9-11]. Hard carbon as a primary anode material has been implemented in industrial area, while the low specific capacity (around 350 mAh/g) cannot meet the high-energy-density requirement [12-14]. Hence, developing alternative anode materials with high theoretical capacity, fast reaction kinetics and excellent cyclic stability are benefit for the advancement of SIBs but still challenging [15,16].

Regarding reported anode candidates, transition metal chalcogenides (TMCs) with merits of tailored morphology and multitudinous charge storage mechanism are greatly popular as promising alternatives to hard carbon [17-19]. Among various TMCs anodes, metal selenides, especially for copper selenide, can reserve Na⁺ through both intercalation and multiple electron conversion mechanism. It means that it can produce high reversible specific capacities [20]. For example, Xu et al. designed a Cu₂Se electrode by selenizing metal-organic frameworks (MOF) [21]. It delivered a high reversible specific capacity of 207 mAh/g at the current density of 1.0 A/g. To further improve the electrochemical performance, Xiao et al. prepared a porous Cu₂Se architecture with high capacity of 426 mAh/g at 0.1 A/g, and high capacity retention of 90% at a current density of 20 A/g [22]. However, the critical issues including the poor electronic conductivity and sluggish surface ion transport are still pending.

Compared to Cu₂Se, the non-stoichiometric Cu_2-xSe are rich in crystal defect [23]. The disorderly distributed Cu ions among Se atoms makes Cu_2-xSe a p-type semiconductor, which provides much higher electronic conductivity than stoichiometric Cu₂Se. Furthermore, the larger space of the layered structure in Cu_2-xSe could accommodate more Na⁺, thus produce high reaction reversibility and fast kinetics. For example, Li et al. synthesized Cu_2-xSe nanocrystals with mesoporous structure by solvothermal reaction. Benefitting from the mesoporous architecture and enriched defects, it could maintain a high specific capacity of 212.4 mAh/g at 5 A/g after 3000 cycles [24]. However, most of those Cu_2-xSe materials were powder-like, which should be used by coating the mixture slurry (containing active materials, polymer binders and carbon black) on current collectors [25,26]. It can inevitably induce increased ohmic resistance and poor specific surface area properties [27,28]. For that matter, a free-standing Cu_2-xSe electrode with nanoarchitectures is of great significance to solve above obstacles [29,30]. The specific surface area properties and reaction kinetics can be easily promoted by anchoring Cu_2-xSe on a free-standing and 3D conductive enhancement, especially in a thick electrode. Yet issues arise in thick electrode design because of prolonged diffusion distances and poor mechanical stability [31,32]. Therefore, there is also an urgent demand to develop integrated 3D Cu_2-xSe anodes by a simple preparation process for high-energy SIBs.

In this paper, the sandwich-like nanoporous Cu_2-xSe@Cu thick electrode was in situ constructed by selenization treatment at room temperature based on flexible nanoporous Cu skeleton, which was fully exposed (220) crystal surface during dealloying CuMn/Cu/CuMn process. Density functional theory (DFT) calculation result verifies that the adsorption energy of Se atoms on Cu(220) plane is stronger than that of on Cu(111) plane. Such a 3D nanoporous architecture not only provides abundant active sites and mechanical stability, but also accelerates reaction kinetics, resulting in advanced electrochemical properties in SIBs. The optimized Cu_2-xSe@Cu-0.15 composite electrode shows high areal specific capacities of 950.6 µAh/cm² at 50 µA/cm². Meanwhile, the Cu_2-xSe@Cu-0.15 electrode displayed a long-term stability of 457.6 µAh/cm² over 500 cycles at 500 µA/cm². This work highlights scientific insights for the exploration of metal selenides by a facile method for SIB.

Fig. 1 shows the preparation process for the flexible Cu_2-xSe@Cu composite electrode, involving the successive dealloying, annealing and selenation. XRD was first used to analyze the phases evolution from CuMn@Cu alloy to Cu_2-xSe@Cu electrode (Fig. S1 in Supporting information). After carrying out dealloying, the alloy peaks in NPA-Cu shift to right obviously with peaks location at 43.32°, 50.45° and 74.12° due to the absolutely dissolution of Mn, and can be ascribed as fcc structural Cu (JCPDF No. 04–0836) (111), (200) and (220) planes. In H—NPA-Cu sample, the peaks represented as Cu are well-crystallized because of heat annealing. Noteworthy, the H—NPA-Cu has an obvious preferred orientation, and more (220) crystal planes are exposed. The Rietveld refinement result (Fig. S2 in Supporting information) was employed, and the calculated R_wp value is as low as 4.61%. The preferred orientation of H—NPA-Cu could be attributed to the texture generated by rolling the alloy. Since the previous studies demonstrated the activity of Cu can affect the selenidation process, the crystal facet modification on different Cu planes is revealed here. The hydroxyethylthioselenide, H—NPA-Cu (220) and Cu (111) crystal plane were used as models, and the adsorption energy of hydroxyethylthioselenide on different crystal planes were calculated by DFT using VASP. Figs. 2a and b show the optimized reaction paths of HOCH₂CH₂SSe⁻ with H—NPA-Cu (220) and Cu (111) crystal planes, respectively. As shown in Fig. 2c, the calculated energies of intermediate complexes at step 2 are −1.97 eV and −3.26 eV for Cu (111) and H—NPA-Cu (220), respectively. Clearly, the H—NPA-Cu (220) crystal planes show stronger adsorption of HOCH₂CH₂SSe⁻than Cu (111), which could accelerate the formation of intermediate complexes. The energy of the intermediate complex increases slightly when the transition state is formed, but the product energy is more negative than the intermediate complex, meaning that the reaction is still a spontaneous process. The product energy obtained by the reaction of hydroxyethylthioselenide with H—NPA-Cu (220) crystal plane is more negative, indicating that the product is in a more stable state of thermodynamics.

Figure 1

	Download: JPG larger image
Fig. 1. Schematic representation of the preparation of copper process of the flexible Cu2-xSe@Cu composite electrode.

Figure 2

	Download: JPG larger image
Fig. 2. Configurations of four different stages in-situ selenization process: (a) Cu (111), (b) H—NPA-Cu (220). (c) Energy profiles for reactions of CH2OHCH2SSe− with Cu (111) and H—NPA-Cu (220).

The XRD pattern of Cu_2-xSe@Cu was shown in Fig. 3a. The peaks at 2θ of 26.75°, 44.60°, 71.59°, and 88.39° are close to fcc Cu₂Se (JCPDS No. 88–2043), which can be indexed as (111), (220), (311) and (511) planes, respectively. It is worth noting that the peaks of Cu at 2θ of 42.87°, 49.98°, 73.20°, and 88.85° are shift to lower diffraction angles after selenidation, showing that the introduction of Se²⁻ increased the interplanar distance of Cu. The observation of Cu_2-xSe and Cu verified the successful formation of Cu-core@Cu_2-xSe-shell structure, which is benefit for enhancing electronic conductivity.

Figure 3

	Download: JPG larger image
Fig. 3. (a) XRD pattern of Cu2-xSe@Cu composite electrode. XPS spectra of Cu2-xSe@Cu composite electrode: (b) Full survey, (c) Cu 2p, and (d) Se 3d.

XPS was used to further characterize the elements valence distribution of Cu_2-xSe@Cu. Cu and Se signals were facilely observed in the XPS survey spectrum (Figs. 3b–d). Cu 2p_3/2 and Cu 2p_1/2 can be separately divided into two peaks. The peaks near at 932.62 and 952.34 eV could be ascribed to Cu⁺ 2p_3/2 and Cu⁺2p_1/2, respectively. While the peaks located at 934.40 and 954.17 eV are corresponded to Cu²⁺ 2p_3/2 and Cu⁺ 2p_1/2. Furthermore, the peaks at 940.71 eV, 943.48 eV, and 962.15 eV are confirmed as characteristic satellite peaks of Cu²⁺. The peaks with binding energy located at 53.96 eV and 54.76 eV can be indexed as Se²⁻ 3d5/2 and Se²⁻ 3d3/2, respectively (Fig. 3d). A broad and weak peak at 58.79 eV is observed, which is originated from the surface oxidation of Se species. The absence of Cu⁰ and Se⁰ indicates that the Cu ligament was fully covered by Cu_2-xSe.

The microscopic morphology and structural evolution of the nanoporous anode was observed by SEM. Due to the desolution of Mn from CuMn layer, the NPA-Cu shows a bicontinuous nanoporous structure with pore size distributing near at 50 nm (Figs. S3a and c in Supporting information). To accelerate the infiltration of the reaction solution and favor the homogeneous growth of Cu_2-xSe on the whole ligament, we further caried out annealing treatment to broaden the pore/ligament size. As shown in Figs. S3b and d (Supporting information), the pore size is expanded to around 500 nm, and the copper ligament is coarsened to about 700 nm. The pore size distribution tested in BET is consistent with the SEM results (Fig. S4 in Supporting information). Fig. 4a shows the cross-section morphology of the Cu_2-xSe@Cu composite electrode with thickness of ~100 µm. The loose layer above and below Cu foil could still be seen, indicating its integrity after in-situ growth of Cu_2-xSe. The copper foil act as not only current collector to enhance the electronic conductivity but also as supporter (Fig. 4a). The Cu_2-xSe@Cu composite electrode inherited good flexibility and mechanical stability from NPA-Cu (Fig. 4a and Fig. S5a in Supporting information) or H—NPA-Cu (Fig. S5b in Supporting information). The high-resolution SEM images of Cu_2-xSe@Cu composite electrode are shown in Figs. 4b and c. Compared to H—NPA-Cu sample, the pore size became slightly smaller and the surface became rougher, caused by the in-situ growth of Cu_2-xSe. The surface morphologies of the Cu_2-xSe@Cu composite electrodes with Se mole proportion of 0.05, 0.1, and 0.15 (Cu_2-xSe@Cu-0.05/0.1/0.15) are shown in Fig. S6 (Supporting information). Clearly, the Cu_2-xSe species exhibited spherical shape with differentiate particle size distributions. It exhibits a loose dense distribution of Cu_2-xSe spheres (~1 µm) for Cu_2-xSe@Cu-0.05 sample. Then the particle reduces to smaller sizes along with increase the proportion of Se (600 nm for Cu_2-xSe@Cu-1.5). In addition, the Se content inside the Cu_2-xSe@Cu anodes with three different ratios was characterized by EDX (Figs. S7a–c in Supporting information), and the internal selenium content increases apparently along with the increasement of Se during selenidation process.

Figure 4

	Download: JPG larger image
Fig. 4. SEM image of Cu2-xSe@Cu composite electrode: (a) Cross section (optical image is shown in the inset); (b, c) Amplification diagram of cross section. TEM Cu2-xSe@Cu composite electrode; (d) Low-resolution image; (e, f) HRTEM image; (g) EDS maps of the Cu and Se distributions.

The microstructure of the Cu_2-xSe@Cu-1.5 composite electrode was further characterized by TEM. Fig. 4d shows that the Cu ligament is surrounded by Cu_2-xSe (~80 nm). The selected area electron diffraction (SAED) pattern (insert of Fig. 4d) with diffraction rings manifests that Cu_2-xSe@Cu is a mixed polycrystalline structure. The interplanar spacings of 0.320 (Fig. 4e) and 0.183 nm (Fig. 4f) in high-resolution TEM images could be ascribe to the (220) crystal plane of Cu_2-xSe and (200) crystal plane of Cu, respectively. The high-angle annular dark field (HAADF) image with its compositional mappings of Cu and Se (Fig. 4g) depict the homogeneous distribution of Se and Cu with a typical core shell structure.

The electrochemical performance of flexible Cu_2-xSe@Cu anodes are revealed by assembling CR2032 half-cells. Firstly, the sodium-ion storage behavior of the Cu_2-xSe@Cu-0.15 composite electrode was investigated using CV at scan rate of 0.1 mV/s (Fig. 5a). In the first discharge process, the two peaks around 1.89 V and 1.57 V, corresponding to the insertion of Na⁺ into Cu_2-xSe lattice with the formation of Cu and Na_xCuSe phases. The broad peak at 0.62 V maybe caused by the complete conversion of Na_xCuSe to Na₂Se, the SEI film formation. In the initial anodic scan, the intense anodic peaks could be observed at 1.62, 1.84 and 2.09 V, which respectively corresponds to the conversion of Na₂Se and Cu to Na_xCuSe and Cu_2-xSe alloy step by steps. In the following scans, the CV patterns display the same tendency and almost completely overlap, illustrating that the conversion reaction is reversible and the stability is good. Fig. 5b shows the galvanostatic charge-discharge profiles of Cu_2-xSe@Cu-0.15 with 50 µA/cm² from the 1^st to 50^th. All detected charge-discharge plateaus could well correspond to the redox peaks in the CV curves, proving a multistep conversion reaction.

Figure 5

	Download: JPG larger image
Fig. 5. Electrochemical performance for Cu2-xSe@Cu composite electrode in SIBs: (a) CV curves and (b) galvanostatic charge-discharge profiles of Cu2-xSe@Cu-0.15. (c) Rate capability and (d) cycling performance of Cu2-xSe@Cu-0.05/0.1/0.15 electrodes. (e) Long cycling performance and (f) capacity comparison of the Cu2-xSe@Cu-0.15 electrodes.

The cycle and rate performance of Cu_2-xSe@Cu-0.05/0.1/0.15 electrodes are also performed in this work. The discharge area specific capacities of Cu_2-xSe@Cu-0.15 electrode are 950.6, 868.3, 701.6, 511.1, 370.4, and 217.6 µAh/cm² at current densities of 50, 100, 250, 500, 1000, and 2500 µA/cm², respectively, which were significantly superior to Cu_2-xSe@Cu-0.05 and Cu_2-xSe@Cu-0.1 electrodes (Fig. 5c). When the current density returned to 50 µA/cm², the capacity could be maintained at 762.6 µAh/cm², demonstrating an outstanding reversibility. The Cu_2-xSe@Cu-0.05 and Cu_2-xSe@Cu-0.1 electrodes exhibit a discharge area specific capacity of 472.4 µAh/cm² and 601.7 µAh/cm² after 50 cycles, respectively (Fig. 5d). In comparison, the specific area capacity of Cu_2-xSe@Cu-0.15 could achieve 754.2 µAh/cm². Even at 500 µA/cm², Cu_2-xSe@Cu-1.5 anode could still maintains the highest reversible capacity of 322 µAh/cm² after 500 cycles, better than Cu_2-xSe@Cu-0.05 (143.4 µAh/cm²) and Cu_2-xSe@Cu-0.10 (200.5 µAh/cm²) anodes (Fig. 5e). The microstructure of Cu_2-xSe@Cu-0.15 electrode after 500 cycles was further characterized by SEM (Fig. S8 in Supporting information). The dense Cu_2-xSe layer on the electrode surface disappeared well and exposed more channels, comparing to the initial sample, so that the electrolyte can smoothly enter the deep of electrode. The internal bicontinuous nanoporous structure of the Cu_2-xSe@Cu-0.15 electrode is extremely stable (Fig. S8a). Compared with similar electrodes, Cu_2-xSe@Cu-0.15 anode exhibits higher energy density, especially under low current conditions (Fig. 5f).The high performance of Cu_2-xSe@Cu-0.15 is owing to both of intrinsic nature of Cu_2-xSe and high stable nanoporous framework, which facilitates the Na⁺ diffusion.

The electrochemical behavior was evaluated by kinetic investigation to further understand the high performance of Cu_2-xSe@Cu-0.15 electrode (Figs. S9a–d in Supporting information). The currents increase and potentials slightly shift with the increasing scan rates from 0.1 mV/s to 1 mV/s, illustrating fast charges and ions diffusion in Cu_2-xSe@Cu electrode. In SIBs, the diffusion-controlled process (battery behavior) and pseudocapacitive-controlled process (capacitive behavior) determines the Na⁺ charge storage capacity. The capacitive contribution increases apparently from 53% to 91.05%, demonstrating the excellent ion accessibility (Figs. S9c and d in Supporting information). The high-rate performance could be ascribed to the bicontinuous nanoporous structure.

The electrochemical impedance spectroscopy (EIS) was further investigated with the electrode charged to 1.6 V after 5 cycles (Figs. S9e and f in Supporting information). The Nyquist plots present two semicircles, corresponding to the resistance of Na⁺ diffusion through the SEI film (R_f) and charge-transfer resistance (R_ct), respectively. In addition, the ohmic resistance (R_s), containing separator paper resistance and electrode resistance, is only 6.7 Ω. The equivalent circuit is shown in insert image of Fig. S9e. The fitted result shows that Cu_2-xSe@Cu-0.15 exhibited a lower R_ct of 257.4 Ω than Cu_2-xSe@Cu-0.05 (384.5 Ω) and Cu_2-xSe@Cu-0.1 (530.6 Ω). Benefiting from the unique nanoporous skeleton, the charge transfer resistance can be reduced and Na⁺ transfer efficiency can be enhanced in the Cu_2-xSe@Cu-0.15 composite electrode.

In summary, we introduce a facile strategy to in-situ grow Cu_2-xSe on (220) facets-exposed Cu by direct selenization of a nanoporous Cu skeleton. The Cu_2-xSe@Cu-0.15 anode possessed excellent electrochemical properties, including high reversible capacity of 950.6 µAh/cm² at 50 µA/cm². The high electrochemical properties attribute to the following factors: Firstly, the sandwiched Cu can synergy with nanoporous Cu skeleton to enhance the electronic conductivity and maintain the material integrity; Secondly, the nanoporous Cu_2-xSe frameworks on both sides provide sufficient spaces to accommodate Na⁺ diffusion. This novel strategy provides a new way to synthesize the flexible integrated electrodes with high storage capacity for sodium ions.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

This work was financially supported by the National Natural Science Foundation of China (Nos. 52271011, 52102291, 52101251). We would like to thank the Analytical & Testing Center of Tiangong University for Transmission Electron Microscope work.

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108922.