English

• The increasingly severe environmental issues and energy crises have prompted humans to seek alternative solutions that can replace traditional fossil fuels [1,2]. Sodium-ion batteries, due to abundant sodium resources and excellent low-temperature performance, are expected to develop into an alternative energy storage system for lithium-ion batteries [3]. Currently, iron-based compounds with low cost and environmental friendliness, are attracted extensive attention [4-6]. Iron carbodiimide FeNCN, as a new kind of anode materials, exhibits excellent electrical conductivity due to N═C═N poly-electron structures, which can easily combine with empty orbit of transition metals to form σ-π bonds [7-9]. Moreover, long N═C═N linear structures construct a large number of broad hexagonal tunnels, which facilitate rapid diffusion of sodium ions [10-12]. All these characteristics make FeNCN a very promising anode material for sodium-ion batteries [5,13,14].

  However, FeNCN, as a kind of organometallic framework compounds with poor thermal stability but high synthesis enthalpy, is difficult to synthesize at low temperature [15-18]. The traditional synthetic process of FeNCN is strict and tedious, which slows down its research development [15,19]. Researchers propose a one-step in-situ pyrolysis method to successfully synthesize FeNCN, which greatly simplifies synthesis process and improves fabrication efficiency of FeNCN [10]. In addition, FeNCN, as a conversion-type anode material, is prone to expansion during the fast sodium embedding process, resulting in a short cycle life at high current [7,8]. Therefore, how to improve its structural stability during fast charging is an urgent problem that needs to be solved at present. In the previous study, it is reported to construct an anchor-like structure by combining FeNCN with carbon materials to stabilize the structure of FeNCN and extend its cycling life [7]; other researchers proposes to improve the rate performance of FeNCN by controlling its crystallinity [10]. Although these methods have greatly improved electrochemical performance of FeNCN, it is still a challenge to improve both ion transport rate (rate performance) and structure stability (cycling life) simultaneously. In addition, the detailed conversion mechanism, specifically whether it involves an intercalation-conversion mechanism, is remains unclear. The impact of its microstructure on the fast charging performance and storage mechanism of FeNCN still needs further exploration.

  Current strategies to enhance the fast-charging capability of sodium-ion batteries primarily involve two key approaches: (1) Incorporating highly conductive materials to optimize electron transport efficiency; (2) Reducing the size to increase the proportion of sodium ion stored sites on the surface area, which is called surface-dominated storage process (surface-pseudocapacitance [20,21]). However, the increased surface area will cause a low initial coulombic efficiency [6]. Bruce Dunn and his co-workers put forward intercalation-pseudocapacitance, which occurs in the entire bulk material with facile ion diffusion channels [20,22-25]. Therefore, significant enhancement in capacitive storage can be achieved by increasing the number of fast ion diffusion channels [24,26,27]. As is established, due to the anisotropy of the crystal, different exposed crystal facets have diverse adsorption stability of sodium ions and various diffusion energy barriers, which are crucial for sodium storage performance [28-32]. Active facets with stable adsorption ability and abundant broad channels to diffusion will significantly improve the intercalation pseudocapacitive sodium ion storage [33-35]. Therefore, it is critical to control the crystal orientation toward large exposure of active facets to improve the sodium ion diffusion rate [28,35,36].

  To systematically explore the facet-performance relationship, we rationally designed and synthesized two distinct FeNCN samples with controlled crystallographic orientations of FeNCN sheets (S-FeNCN) exhibiting predominant (002) facets, and polyhedral FeNCN (P-FeNCN) with dominant {010} exposed planes. Comprehensive electrochemical characterizations were conducted to elucidate their sodium ion storage behavior and underlying reaction mechanisms. The experimental results demonstrate that P-FeNCN exhibit superior electrochemical performance with a capacity of 574 mAh/g and 95.8% capacity retention over 300 cycles, outperforming S-FeNCN sheets (82.8 mAh/g, 22.1% retention). Notably, this different electrochemical performance originates from their distinct charge storage mechanisms: P-FeNCN with oriented {010} crystal planes demonstrates pseudocapacitive-dominated Na⁺ storage, whereas S-FeNCN with (002) dominant facets shows diffusion-limited intercalation kinetics. DFT calculations confirm that the {010} planes possess lower adsorption energy (−0.14 eV) and reduced diffusion energy barrier (0.168 eV), making them the dominant active facets [37]. The large exposure of {010} active facets promotes pseudocapacitance behavior. This finding indicates that the sodium storage mechanism and performance can be tailored by regulating the exposed crystal facets of FeNCN, which provides new perspective for improving the performance of other anode materials.

  Fig. 1a illustrates the morphology-controlled synthesis process of two different FeNCN architectures with tunable crystal facet exposure. The sheet-like FeNCN (named S-FeNCN) in Figs. 1b-d with largely exposed (002) facets was synthesized through a solvothermal-assisted molten-salt pyrolysis method at 500 ℃ for 4 h; while the interconnected polyhedral FeNCN (named P-FeNCN) in Figs. 1e-g with preferential {010} facet exposure was obtained via a simple solid-state pyrolysis route at 550 ℃ for 2 h. The interaction of synthesis parameters including precursor, reaction temperature, and processing duration modulates the inherent surface energies of FeNCN crystallographic facets, thereby redirecting their growth kinetics. This facet-specific energy strategy controls anisotropic crystal development, ultimately yielding crystals with preferentially exposed facet orientations. And the polyhedrons exhibit tightly packed layered structure with preferential [001] oriented growth direction (Figs. 1g and h), proved by the following HRTEM (Figs. 1l-n). S-FeNCN exhibits (002) facet dominated sheet structure with an interplanar spacing of 2.74 Å and an angle of 120° between (100) and (010) facets (Figs. 1i and j). A typical diffraction pattern of hexagonal crystals indicates S-FeNCN sheets grow along the 〈010〉 direction (Fig. 1k). Differently, P-FeNCN displays interplanar distance of 4.85 Å corresponding to (002) facets of FeNCN in Fig. 1m. The corresponding SAED patterns and HRTEM image both confirm the oriented growth of P-FeNCN along the [001] direction, which is significantly different with the orientation growth of S-FeNCN. Different facets have various distribution of surface atom and ion transport tunnels, which may have a significant effect on the adsorption and diffusion rate of sodium ions [28,33]. The EDS mapping in Fig. 1o confirms uniform dispersion of Fe, N, and C throughout P-FeNCN.

  Figure 1

  

  Figure 1.  (a) Schematic illustration of preparation route for P-FeNCN and S-FeNCN. SEM images of (b-d) S-FeNCN and (e-h) P-FeNCN. Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM), the selected area electron diffraction (SAED) images for (i-k) S-FeNCN and (l-n) P-FeNCN. (o) STEM image of P-FeNCN with the corresponding Fe, N and C elemental mapping images.

  DownLoad: Full-Size Img PowerPoint

  Fig. S1 (Supporting information) shows X-ray diffraction (XRD) patterns of S-FeNCN and P-FeNCN. Both samples belong to the same hexagonal phase of FeNCN (CSD No. 419223) with distinct relative peak intensities. P-FeNCN exhibits a significantly enhanced texture coefficient ratio of (002)/(100) of 4.08 (Figs. S1a and b), much higher than that of 2.02 of S-FeNCN, implying oriented growth along the [002] direction in P-FeNCN. Fourier-transform infrared (FTIR) spectroscopy was conducted to confirm the presence of N═C═N functional group in both samples (Fig. S1c). The characteristic absorption bands at 2080 and 650 cm^-1 represent asymmetric stretching and bending vibration of the N═C═N group. The small peak around 600 cm^-1 is associated with Fe-N [18,22]. The strong absorption band at 1310, 1420, 1570 and 1640 cm^-1, which are assigned to typical stretching vibration modes of N═C—N [38,39]. The moderate absorption peak at 810 cm^-1 is attributed to aromatic CN_x hetero-cycles. These results indicate the presence of minor C₃N₄ by-product in both S-FeNCN and P-FeNCN, which is generated from precursor pyrolysis of urea and melamine. Notably, the C₃N₄ content remains comparable between both samples (< 8 wt% shown in TGA, Fig. S2a in Supporting information). Nitrogen adsorption-desorption isotherms of P-FeNCN and S-FeNCN were shown in Fig. S2b (Supporting information). Brunauer-Emmett-Teller (BET) results indicate that the specific surface area of P-FeNCN and S-FeNCN are 5.41 and 35.32 m²/g, respectively. The low surface area of P-FeNCN is ascribed to the [001]-oriented densely stacking layered crystallites, thereby minimizing exposed edge sites. XPS measurement was performed to understand the difference between two samples (Figs. S1d-f). C 1s spectra could be divided into three subpeaks of C—C (284.8 eV), C—N (286.3 eV) and C═N (288.4 eV). The C—N peak at 286.3 eV is associated with the existence of C—N═C in C₃N₄ [21]. The N 1s spectra display three subpeaks, corresponding to Fe-N (397.4 eV) in FeNCN [18], pyridinic N(N—C) (398.7 eV) in C₃N₄ and N═C (400.2 eV) in FeNCN. The fraction of N═C═N functional groups in P-FeNCN is higher than that in S-FeNCN, which may be attributed to the different atomic arrangement on various facets of FeNCN. The Fe 2p spectra of two samples display similar shape, reflecting the similar state of Fe element in two samples.

  To further clarify the sodium ion storage mechanism of FeNCN, we conducted in-operando XRD for P-FeNCN during the initial charge-discharge process for the first time (Fig. 2a). The broad peak between 20° and 25° is attributed to Super P carbon in the in-situ cell. The peak at 38° is corresponding to Be. In the initial stage of the reaction, the (002), (100), and (101) peaks persist until 0.40 V. During this stage, the corresponding capacity-voltage curve shows a long plateau, contributing a high percentage of capacity. This process involves the intercalation of sodium ions into the crystal planes of FeNCN. Owing to the large interplanar spacing and broad tunnel structure of FeNCN, the interplanar spacing in XRD does not change significantly during sodium ion intercalation. As the discharge voltage continues to decrease, the characteristic peaks of FeNCN gradually weaken and eventually disappear, indicating that a conversion reaction occurs at this stage. As charging continues and the voltage gradually increases, no obvious peaks of FeNCN can be found, suggesting that the products after charging remain amorphous or exhibit poor crystallinity. This agrees well with corresponding HRTEM images (Fig. 2b). After discharge, the products could be identified as Fe and Na₂NCN nanocrystals according to the interplanar spacing. The sodium storage mechanism of FeNCN is summarized as follows: during discharge process, sodium ions intercalate into the FeNCN tunnel to form Na_xFeNCN between 0.50 V and 0.40 V, and the structure of FeNCN remains intact with no change in interplanar spacing. Then, as the discharging continues, the Na_xFeNCN structure begins to convert into poorly-crystallized Na₂NCN and Fe nanocrystals. When charging begins, the reverse reaction occurs, with Na₂NCN and Fe reacting to ultimately form FeNCN and Na⁺ (Fig. 2c). After charging, the crystallinity of FeNCN significantly decreases (Fig. 2d). Consequently, FeNCN follows an intercalation-conversion mechanism. Further cycling performance is tested by setting cut-off voltage at 0.4 V to avoid the conversion stage. P-FeNCN could extend its cycling life up to 500 cycles with a capacity retention of 89% (Fig. S8 in Supporting information). As FeNCN has a low activation energy of −1.84 eV for conversion reaction, the rate-limiting step crucially resides in the sluggish Na⁺ solid-state diffusion during intercalation [10,23]. Enhancing bulk diffusion kinetics thus becomes the key to optimizing FeNCN's fast-charging performance.

  Figure 2

  

  Figure 2.  (a) Time-resolved operando XRD patterns. HRTEM of the P-FeNCN electrode: (b) After fully discharged at 0.01 V and (c) fully recharged at 3.0 V. (d) Schematic diagram of sodium ion storage in FeNCN.

  DownLoad: Full-Size Img PowerPoint

  In order to evaluate the sodium ion storage performance of two samples, the cyclic voltammetry (CV) curves were recorded to disclose the detailed redox reaction process (Figs. 3a and d). In the first cathodic scan, two distinct peaks could be observed in P-FeNCN. The peak at 0.37 V is related to the insertion reaction to form Na_xFeNCN. The strong peak at 0.27 V may be associated with conversion reaction from Na_xFeNCN to Fe metal, sodium ion inserted into the by-product of C₃N₄, and the formation of solid electrolyte interphase (SEI) film since the peak intensity significantly decays in the following cycles [2,4,8,13]. Accordingly in anodic scan, the minor peak of 0.076 V is the reversible extraction of sodium ion from C₃N₄ [40]. The prominent peak at 1.36 V is attributed to the multi-step reversible conversion reaction of Fe to FeNCN [5,10]. In the subsequent scans, the cathodic peak of 0.27 V shifted to the higher voltage of 0.86 V, which could be attributed to the significant change of morphology and crystallinity after first discharge process. Similarly, the anodic peak of 1.36 V transferred to 1.44 V. The strong peak at 0.86 V and 1.44 V could be almost overlapped in 2^nd and 3^rd cycle, indicating the high structure stability after first cycle for P-FeNCN. For S-FeNCN, the initial cathodic peak at 0.54 V relates to Na⁺ insertion. Its rapid intensity attenuation from the 1^st to 3^rd cycle indicates poor structural stability. In addition, the gap between discharge voltage of 0.62 and charge voltage of 1.44 V is 0.82 V, which is much larger than 0.58 V for P-FeNCN composites (0.86 V ~discharge voltage, 1.44 V ~ charge voltage). This indicates the low polarization of P-FeNCN and enhanced electrochemical behavior. The corresponding charge-discharge curves present a long platform around 0.35 V which associates with the insertion of sodium ion (Fig. 3b). The slope from 0.37 V to 0.01 V is related to redox reaction. This agrees well with the CV curves. The first discharge and charge capacity for P-FeNCN is 706.9 and 573.7 mAh/g with an initial coulombic efficiency of 81.1%. In contrast, the first discharge and charge capacity for S-FeNCN is 600.5 and 356.2 mAh/g with a low coulombic efficiency of 59.3% (Fig. 3e). The high coulombic efficiency of P-FeNCN is related to low surface area of 5.41 m²/g for P-FeNCN (Fig. S2b).

  Figure 3

  

  Figure 3.  Electrochemical properties of S-FeNCN and P-FeNCN electrodes for sodium ion batteries (SIBs): (a, d) CV curves. (b, e) Charge/discharge curves. (c) Rate performance. (f) EIS spectra. (g) Cycling performance and coulombic efficiency at current density of 1.0 A/g. (h) Diffusion coefficients of the Na-ion. (i, l) CV curves at different scan rates. (j, m) CV curve of FeNCN at 0.8 mV/s marked in green with the pseudocapacitance-contributed part marked in orange and (k, n). The capacitive contribution at different scan rates of P-FeNCN and S-FeNCN electrode.

  DownLoad: Full-Size Img PowerPoint

  The rate performance of P-FeNCN was further tested at different current densities. As displayed in Fig. 3c, P-FeNCN anode delivers much superior capacity of 592.5, 587.1, 562.3, 526.4, 481.4 and 371.8 mAh/g at 0.1, 0.2, 0.5, 1.0, 2.0 and 5.0 A/g. When the current returns to 0.1 A/g, the capacity recovered to 605.5 mAh/g. In contrast, S-FeNCN delivered only 360.9, 339.4, 294.4, 260.2, 220.7 and 146.9 mAh/g at increasing current density. In order to further investigate the cyclic stability of the electrode material, the electrode is tested for 300 cycles at 0.5 A/g. As shown in Fig. 3g, P-FeNCN maintained 574 mAh/g after 300 cycles with a capacity retention of 95.8%. In contrast, S-FeNCN electrode exhibits a poor cycling stability with capacity dropping from 376.2 mAh/g (first charge capacity) to 82.8 mAh/g after 300 cycles with a retention rate of 22.1%. To investigate the contribution of C₃N₄ byproduct to the total capacity for P-FeNCN and S-FeNCN, the cycling test of C₃N₄ was carried out at 0.1 A/g. As shown in Fig. S4 (Supporting information), C₃N₄ only delivers 23.5 mAh/g, which can be negligible compared to 592.5 mAh/g of P-FeNCN. The above test results indicate P-FeNCN exhibits superior sodium storage performance including high initial coulombic efficiency of 81.1% and high specific capacity of 574 mAh/g after 300 cycles with 95.8% capacity retention (Fig. S9 in Supporting information). The morphologies of two electrodes after long cycling further confirm high structure stability of P-FeNCN (Fig. S10 in Supporting information). The outlines of FeNCN polyhedrons remained distinguishable. Additionally, the cycling performance of P-FeNCN was also evaluated in ester-based electrolytes (EC/DEC). A dramatic capacity decay occurs after 30 cycles, indicating structural instability of the SEI layer formed in ester electrolytes (Fig. S11 in Supporting information).

  Further studies focused on the kinetic behavior of sodium storage with electrochemical impedance spectroscopy (EIS) and the galvanostatic intermittent titration technique (GITT). The electrochemical impedance (EIS) Nyquist plot consists of a compressed semicircle at the high frequency region representing charge transfer resistance and a linear Warburg-like slope at the low frequency region corresponding to sodium ion diffusion rate. According to the fitted equivalent circuit diagram of Fig. 3f, the charge transfer resistance (R_ct) of P-FeNCN (52.5 Ω) and S-FeNCN (55.1 Ω) are comparable after 100 cycles at 0.5 A/g. Notably, the profound difference between them is the phase angles for the Nyquist plots. P-FeNCN exhibits a more-vertical slope, indicating a faster diffusion rate [41]. GITT was further employed to quantify the sodium ion diffusion rate of both samples (Fig. S3). The ion diffusion rate is calculated by following equation [9]:

  $ D=\frac{4}{\pi \tau}\left(\frac{m_B V_M}{M_B S}\right)^2\left(\frac{\Delta E_S}{\Delta E_t}\right)^2 $

  (1)

  where τ refers to relaxation time; and m_B, V_M, and M_B are corresponding to the mass, molar volume, molar mass of active materials FeNCN, S is defined as effective electrodes/electrolyte interfacial area. ΔE_s/ΔE_t is the ratio of the steady potential change between two pulses and the potential difference caused by constant current pulse in a single pulse. According to Eq. 1, the obtained Na⁺ diffusion coefficient (D_Na⁺) of P-FeNCN is much higher than that of S-FeNCN during the whole charge/discharge process (Fig. 3h). The highest sodium ion diffusion rate of P-FeNCN could reach to 3.8 × 10^–11, which is related to the unique structure of P-FeNCN with highly enriched {010} facets.

  To further explore the relation between rate performance and storage mechanisms, CV curves of S-FeNCN and P-FeNCN with different scanning rates (0.2–1.0 mV/s) were recorded (Figs. 3i and l). For P-FeNCN, there are two distinct pairs of redox peaks around 0.28/1.01 V (peak1 and peak1') and 0.76/1.51 V (peak 2 and peak 2′) in Fig. 3i, corresponding to intercalation-dominated reaction and conversion-type redox process proved by in-situ XRD and TEM in Fig. 2. Conversely, the CV curves of sheet-like S-FeNCN with different scan rates shows severely broadened peaks with low current, proving sluggish solid-state diffusion kinetics occurring in the S-FeNCN electrodes (Fig. 3l). The different shape of CV curves further confirms the facet exposure has a significant effect on sodium ion storage kinetics. To deeply understand the sodium ion storage kinetics of S-FeNCN and P-FeNCN, corresponding calculation were carried out according the equation i = av^b [20]. The b value could be obtained from the slope of ln(peak current, i) versus ln(scan rate, v) plot. When the b value approaches 0.5, the storage process is governed by diffusion; while b value is close to 1, reflecting capacitance-controlled process [22]. As shown in Figs. S5 and S6 (Supporting information), the b values of two pairs of peaks are 1.03/1.00 and 1.08/0.92, which are all close to 1 for P-FeNCN, indicating capacitive mechanism. For S-FeNCN electrode, b values are 0.35/0.40 and 0.59/0.57, reflecting a diffusion-controlled sodium ion storage process. The contribution percentage of capacitive behavior could be further quantify by the following formula [21]:

  $ i(\mathrm{~V})=k_1 v+k_2 v^{1 / 2} $

  (2)

  where k₁v and k₂v^1/2 represent pseudocapacitance and diffusion controlled process, respectively [6]. Accordingly, the pseudocapacitance percentage for P-FeNCN and S-FeNCN was calculated to be 88.0% and 33.4% at 0.8 mV/s, respectively (Figs. 3j and m). Notably, the pseudocapacitance contributes 78.1% of total capacity for P-FeNCN even at ultralow scan rate of 0.2 mV/s (Fig. 3k), than 18.5% of S-FeNCN (Figs. 3k and n) revealing that the pseudocapacitive storage of P-FeNCN arises from a synergistic combination of surface and intercalation capacitance [42,43].

  To get a profound understanding on sodium ion storage mechanisms, DFT methods were employed [37,44,45]. The surface adsorption energies of the (002) and {010} crystal planes were calculated according to the adsorption energy calculation model shown in Fig. S7 (Supporting information) [33,46]. The simulating results showed that the adsorption energies of (002) and {010} crystal plane are −0.14 and −2.32 eV, respectively. Results show adsorption energies of −0.14 eV for (002) and −2.32 eV for {010}, indicating stronger sodium-ion stabilization on {010} planes (Figs. 4a-c). This agrees well with our previous speculation that {010} facets distribute N═C═N multiple electron functional groups with higher electronegativity, which tend to adsorb positively charged sodium ion. As shown in Figs. 4d and e, the differential charge density map reveals atomic-scale charge redistribution during Na⁺ adsorption on {010} and (002) facets of FeNCN. The sodium ion induces a charge transfer from the ion to the surface of FeNCN. The Fe and N atoms of FeNCN exhibit significant charge density changes, which accumulate more electrons. There are numerous N═C═N functional groups distributing on the {010} surface (Fig. 4d). As a result, Na⁺ adsorbing at {010} facets triggers more obvious charge redistribution than (002) facet, suggesting enhanced stability between Na⁺ and Fe/N atoms on {010} facets. Crystallographic facet engineering greatly affects sodium ion transport kinetics in FeNCN. In S-FeNCN hexagonal sheets with predominantly exposed (002) planes, sodium ions diffuse along [001] direction, DFT reveals this pathway incurs a high diffusion barrier of 0.827 eV (Figs. 4g and h), arising from winded and narrow diffusion tunnels with short diameter of 2.16 Å (Fig. 4h). In contrast, polyhedral P-FeNCN exposes six symmetrically equivalent {010} facets (Figs. 4a and b), which establish broad and linear diffusion highways facilitated by expanded interlayer spacing of 4.85 Å (Fig. 4f). This architecture reduces the [010] oriented diffusion barrier to 0.168 eV (Fig. 4g), which can be attributed to the geometric advantage of hexagonal channel cross-sections with diameter of 3.12 Å (Fig. 4f), and continuous electronic facilitation of N═C═N group. Therefore, it is energy-convenient for sodium ions to diffuse through the (010) plane and then move along the broad channels in < 010 > directions. Considering theoretical calculation results, {010} planes feature both lower binding energy and migration barrier (< 010 > direction), confirming them as active facets enabling rapid sodium storage.

  Figure 4

  

  Figure 4.  Theoretical simulations of Na storage and diffusion in FeNCN crystal structure; (a, c) FeNCN with different exposed facets. (b) Crystal structure of FeNCN. (d, e) Differential charge density map for {010} and (002) facets (yellow for electron accumulation and blue for electron depletion, respectively). (f, h) Typical models of Na atom diffusion paths along [010] and [001] direction. (g) Na diffusion energy profiles through the different paths.

  DownLoad: Full-Size Img PowerPoint

  Finally, to assess the practical application potential of the FeNCN anode material, we assembled a sodium-ion full battery with commercial Na₃V₂(PO₄)₃ as cathode and FeNCN as anode. Fig. S14a (Supporting information) shows the galvanostatic charge-discharge (GCD) profiles of the Na₃V₂(PO₄)₃ half cell, FeNCN half cell, and the Na₃V₂(PO₄)₃//FeNCN full cell, demonstrating their respective voltage ranges and confirming the practical applicability of the full cell. The Na₃V₂(PO₄)₃//FeNCN full cell exhibits a reversible capacity of 236.8 mAh/g at a current density of 0.1 A/g with an energy density of 475 Wh/kg. The full cell could power an LED message board, further confirming its practical application potential as a negative electrode material for SIBs.

  Based on the discussion above, schematic diagram illustrates the Na⁺ storage mechanisms of FeNCN are governed by different exposed facets of {010} and (002) (Fig. 5). Various exposed facets have a significant different effect on the sodium ion storage performance between S-FeNCN and P-FeNCN. In P-FeNCN anode, exposed {010} planes contain broad tunnels facilitating rapid Na⁺ diffusion. Additionally, there are a great deal of N═C═N multi-electron structure on {010} crystal planes, which attracts Na⁺ to adhere and stored on the electrode surface. Therefore, the {010} crystal plane is the active facets with a low adsorption energy and a low diffusion barrier for sodium ion to immigrate [38]. The low adsorption energy and fast diffusion rate facilitate intercalation-pseudocapacitive mechanisms. In contrast, sheet-like S-FeNCN has a large exposure area of the (002) crystal plane. The diffusion tunnels along [001] direction are narrow and tortuous, resulted in an elevated diffusion energy barrier and a high adsorption energy. This restricts the high diffusion rate of sodium ion along [001] direction, forcing charge storage to proceed through solid-state diffusion-dominated mechanism in S-FeNCN.

  Figure 5

  

  Figure 5.  Schematic diagram of Na⁺ storage mechanism of FeNCN crystal structure with different exposed facets of {010} and (002).

  DownLoad: Full-Size Img PowerPoint

  In summary, we successfully synthesized FeNCN with different morphologies through facet engineering: polyhedral P-FeNCN dominated by {010} facets and sheet-like S-FeNCN with preferentially exposed (002) facets. Sodium-ion storage performance tests demonstrate that P-FeNCN anodes with preferentially exposed {010} facets deliver superior rate capability (372 mAh/g at 5 A/g) and enhanced cyclability (95.8% capacity retention after 300 cycles), outperforming (002) oriented S-FeNCN (22.1% capacity retention after 300 cycles). More interestingly, kinetics analysis indicates that FeNCN with different exposed facets exhibits diverse storage mechanisms: {010} facets promote pseudocapacitive contribution up to 88.0% at 0.8 mV/s; while (002) facets enforce diffusion-controlled kinetics. In-situ XRD and DFT simulation further disclose the relationship between morphology and properties. FeNCN follows intercalation-conversion storage mechanism and solid-state diffusion rate is the rate determining rate in storage process. DFT simulation demonstrates that the active {010} dominated surfaces enable rapid Na⁺ transport via hexagonal tunnels with 0.168 eV barrier along 〈010〉 directions. This work demonstrates that tailoring exposure of active crystallographic facets in FeNCN can alter the sodium storage mechanism. This strategic facet engineering can also be applied into other crystalline electrode materials to improve the sodium storage performance.

  Declaration of competing interest

  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.

  CRediT authorship contribution statement

  Hui Qi: Writing – original draft, Project administration. Chaozheng He: Writing – review & editing, Supervision, Project administration, Funding acquisition. Chenfei Song: Methodology, Investigation. Juncui Gao: Investigation, Data curation. Qing Gao: Software, Investigation. Weipeng Luo: Data curation. Ze Zhang: Formal analysis. Haoyu Liu: Data curation. Xiaojing Yuan: Supervision, Formal analysis. Wenfeng Wu: Software. Bohang Zhao: Software, Investigation. Lina Kong: Formal analysis. Yayi Cheng: Software. Ling Guo: Investigation.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 52002305), Natural Science Basic Research Program in Shanxi Province of China (Nos. 202403021221184, 202403021222281), Natural Science Basic Research Plan in Shaanxi Province of China (Nos. 2025JC-YBMS-478, 23JK0424), College Students' Innovation Program of Taiyuan Normal University (No. CXCY2443).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111591.

  
  
• 1. [1]

     I. Mohan, A. Raj, K. Shubham, et al., J. Energy Storage 55 (2022) 105625.

  2. [2]

     Z. Song, G. Zhang, X. Deng, et al., Nano-Micro Lett. 14 (2022) 53. doi: 10.1127/pala/2021/0111

  3. [3]

     Q. Fu, F. Song, C. Mu, et al., Energy Storage Mater. 74 (2025) 103960.

  4. [4]

     C. Liu, B. Wang, Z. Song, et al., Adv. Funct. Mater. 34 (2024) 2312905.

  5. [5]

     E. Zhang, S. Wu, X. Sun, et al., Chem. Eng. J. 453 (2023) 139966.

  6. [6]

     H. Qi, L. Cao, J. Li, et al., ACS Appl. Mat. Interfaces 8 (2016) 35253–35263. doi: 10.1021/acsami.6b11840

  7. [7]

     P. Guo, L. Cao, R. Wang, et al., Adv. Funct. Mater. 30 (2020) 2000208.

  8. [8]

     H. Qi, C. Zhao, J. Huang, et al., Nanoscale 14 (2022) 780–789. doi: 10.1039/d1nr06705b

  9. [9]

     H. Li, F. Wu, P. Guo, et al., Nano Lett. 24 (2024) 8525–8534. doi: 10.1021/acs.nanolett.4c01091

  10. [10]

      H. Qi, Y. Hou, W. Wang, et al., Dalton Trans. 51 (2022) 8223–8233. doi: 10.1039/d2dt00924b

  11. [11]

      A.J. Corkett, O. Reckeweg, R. Pöttgen, et al., Chem. Mater. 36 (2024) 9107–9125. doi: 10.1021/acs.chemmater.4c01615

  12. [12]

      H. Bourakhouadar, J. Hempelmann, J. van Leusen, et al., J. Am. Chem. Soc. 146 (2024) 26071–26080. doi: 10.1021/jacs.4c06162

  13. [13]

      Z. Song, G. Zhang, X. Deng, et al., Adv. Funct. Mater. 32 (2022) 2205453.

  14. [14]

      K. Chen, M. Fehse, A. Laurita, et al., Angew. Chem. 59 (2020) 3718–3723. doi: 10.1002/anie.201914760

  15. [15]

      M.T. Sougrati, A. Darwiche, X. Liu, et al., Angew. Chem. Int. Ed. 55 (2016) 5090–5095. doi: 10.1002/anie.201600098

  16. [16]

      A. Eguía-Barrio, E. Castillo-Martínez, X. Liu, et al., J. Mater. Chem. A 4 (2016) 1608–1611.

  17. [17]

      T. Li, W. Zhao, H. Bi, et al., J. Power Sources 467 (2020) 228252.

  18. [18]

      Z. Lv, W. Dong, B. Jia, et al., Chemistry 27 (2021) 2717–2723. doi: 10.1002/chem.202003684

  19. [19]

      X. Liu, M. Krott, P. Müller, et al., Inorg. Chem. 44 (2005) 3001–3003. doi: 10.1021/ic050050a

  20. [20]

      Z. Zhao, Y. Wu, R. Hu, et al., Adv. Funct. Mater. 33 (2023) 2307794.

  21. [21]

      Y. Cheng, J. Huang, F. Yu, et al., Chem. Eng. J. 481 (2024) 148737.

  22. [22]

      N. Liu, X. Wu, L. Fan, et al., Adv. Mater. 32 (2020) e1908420.

  23. [23]

      M.R. Pallavolu, A.N. Banerjee, N. Roy, et al., Chem. Eng. J. 498 (2024) 155240.

  24. [24]

      M. Wang, H. Wang, Q. Zhang, et al., ACS Nano 18 (2024) 25813–25825.

  25. [25]

      G. Zeng, U. Ali, M. Sun, et al., J. Colloid Interface Sci. 653 (2024) 46–55.

  26. [26]

      Y. Yang, H. Zhu, J. Xiao, et al., Adv. Mater. 32 (2020) e1905295.

  27. [27]

      W. van den Bergh, H.N. Lokupitiya, N.A. Vest, et al., Adv. Funct. Mater. 31 (2020) 2007826.

  28. [28]

      G. Zhang, X. Yin, D. Ning, et al., Angew. Chem. Int. Ed. 64 (2025) e202415450.

  29. [29]

      F. Li, Z. Sun, M. Dong, et al., Chem. Eng. J. 480 (2024) 147940.

  30. [30]

      H. Chen, N. Sun, Q. Zhu, et al., Adv. Sci. 9 (2022) e2200023.

  31. [31]

      Y. Liu, S.P. Jiang, Z. Shao, Mater. Today Adv. 7 (2020) 100072.

  32. [32]

      J. Feng, X. Zhang, Y. Xu, et al., Energy Storage Mater. 46 (2022) 278–288.

  33. [33]

      C. Zhao, M. Xi, J. Huo, et al., Chin. Chem. Lett. 34 (2023) 107213.

  34. [34]

      Q. Mao, R. Gao, Q. Li, et al., Chem. Eng. J. 382 (2020) 122978.

  35. [35]

      J. Sun, G. Lian, L. Jing, et al., Nano Res. 15 (2022) 4108–4116. doi: 10.1007/s12274-021-4060-1

  36. [36]

      X. Xu, R. Zhao, B. Chen, et al., Adv. Mater. 31 (2019) e1900526.

  37. [37]

      J. Huo, H. Wei, L. Fu, et al., Chin. Chem. Lett. 34 (2023) 107261.

  38. [38]

      Y. Zhao, X. Zhu, Q. Zhang, et al., ACS Nano 18 (2024) 32556–32568. doi: 10.1021/acsnano.4c09448

  39. [39]

      L. Wang, Y. Chen, C. Zhang, et al., ACS Appl. Mat. Interfaces 16 (2024) 4581–4591. doi: 10.1021/acsami.3c14826

  40. [40]

      S. Wang, Q. Wang, W. Liu, et al., Appl. Surf. Sci. 640 (2023) 158321.

  41. [41]

      G. Li, Y. Tang, Y. Wang, et al., Angew. Chem. Int. Ed. 137 (2025) e202417602.

  42. [42]

      H.J. Kim, J.S. Yeon, H.R. Park, et al., ACS Appl. Mater. Interfaces 15 (2023) 9350–9361. doi: 10.1021/acsami.2c21034

  43. [43]

      S.E. Bak, W. Chung, M.A. Abbas, et al., ACS Appl. Energy Mater. 5 (2022) 5508–5512. doi: 10.1021/acsaem.2c00648

  44. [44]

      C. He, R. Sun, L. Fu, et al., Chin. Chem. Lett. 33 (2022) 527–532.

  45. [45]

      H.J. Monkhorst, J.D. Pack, Phys. Rev. B 13 (1976) 5188–5192.

  46. [46]

      C. He, Q. Zhang, J. Huo, et al., Chin. Chem. Lett. 33 (2022) 3281–328.

• 

  Figure 1  (a) Schematic illustration of preparation route for P-FeNCN and S-FeNCN. SEM images of (b-d) S-FeNCN and (e-h) P-FeNCN. Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM), the selected area electron diffraction (SAED) images for (i-k) S-FeNCN and (l-n) P-FeNCN. (o) STEM image of P-FeNCN with the corresponding Fe, N and C elemental mapping images.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Time-resolved operando XRD patterns. HRTEM of the P-FeNCN electrode: (b) After fully discharged at 0.01 V and (c) fully recharged at 3.0 V. (d) Schematic diagram of sodium ion storage in FeNCN.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Electrochemical properties of S-FeNCN and P-FeNCN electrodes for sodium ion batteries (SIBs): (a, d) CV curves. (b, e) Charge/discharge curves. (c) Rate performance. (f) EIS spectra. (g) Cycling performance and coulombic efficiency at current density of 1.0 A/g. (h) Diffusion coefficients of the Na-ion. (i, l) CV curves at different scan rates. (j, m) CV curve of FeNCN at 0.8 mV/s marked in green with the pseudocapacitance-contributed part marked in orange and (k, n). The capacitive contribution at different scan rates of P-FeNCN and S-FeNCN electrode.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Theoretical simulations of Na storage and diffusion in FeNCN crystal structure; (a, c) FeNCN with different exposed facets. (b) Crystal structure of FeNCN. (d, e) Differential charge density map for {010} and (002) facets (yellow for electron accumulation and blue for electron depletion, respectively). (f, h) Typical models of Na atom diffusion paths along [010] and [001] direction. (g) Na diffusion energy profiles through the different paths.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Schematic diagram of Na⁺ storage mechanism of FeNCN crystal structure with different exposed facets of {010} and (002).

  下载: 全尺寸图片 幻灯片
•