Copper–Iron Selenides Nanoflakes and Carbon Nanotubes Composites as an Advanced Anode Material for High‐Performance Lithium‐Ion Batteries

Metal selenides are widely considered as an emerging anode electrode material for lithium‐ion batteries (LIBs). Hence, the present study uses a conductive carbon materials matrix such as carbon nanotubes (CNTs) with copper–iron selenide (CuFeSe2). The composites (CuFeSe2@CNTs) are synthesized by a hydrothermal method and examined for electrochemical performance as anode applications in LIBs. Herein, the CuFeSe2@CNTs show excellent specific capacity of 783.7 and 720.6 mA h g−1 at 0.1 and 1 A g−1, respectively. This composite further exhibits a high‐capacity value compared to only FeSe2 and CuFeSe2 and a reversible capacity of 691 mA h g−1 after 200 cycles at 1 A g−1 current density. Moreover, the homogeneous combination of CuFeSe2 nanoflakes and CNTs provides structural stability that reduces the volume change during lithium‐ion interactions and successively improves the conductivity of the complex, which is advantageous for better ion and electron kinetics during the reaction.


Introduction
Recently, an increase in the use of fissile fuels and emission of greenhouse gasses due to overuse in the energy requirements generated a negative impact on the environment.Also, the increased demand for miniaturized/transferable electronics and automobiles facilitates the development of high-density and compact energy storage systems. [1,2]Lithium-ion batteries (LIBs) have gained tremendous recognition in this regard, as a result of their broad operating voltage range, high energy density, ecofriendly nature, portable energy sources, and long cycle life. [1]][5] The nature and performance of anode materials are critical to the capacity and cycle life of LIBs; therefore, researchers concentrate on developing excellent electrode materials for LIBs to connect the voids in future energy needs.The morphology as well as the intrinsic properties of the materials are essential for the capacity and performance of the batteries. [2]Therefore, it is important to select appropriate materials to prepare structurally designed nanostructures understanding the reaction during the energy storage cycles according to the required energy requirements.
According to the mechanism of intercalation, alloying, and conversion, the anode materials are classified into three different types.The intercalation reaction is a process of storing and releasing lithium ions through the deembedding of vacancies or defects in the crystal structure of the material.But there is also the limitation of low theoretical specific capacity.Graphite is the most typical intercalation-type material, which is safe for avoiding the problem of lithium dendrites. [6,7]In the alloy reaction, lithium ions are stored and released through alloying and dealloying with the host material.This type of material has a lot larger theoretical specific capacity than the intercalation type but suffers from seriously huge volume expansion that can reach 300%-400% during the charging and discharging process.For safety reasons, all alloy-type materials have not been formally applied to commercial lithium-ion batteries.The materials Si, Ge, and Sn monomers with large theoretical weight capacity (1000-4000 mA h g À1 ) and volume capacity (7000-10 000 mA h cm À3 ) are usually used in special structures to suppress severe volume expansion in current research. [6]In the conversion reaction, the chemical bonds in the active material are broken and reconnected with lithium ions, thus forming new chemical bonds. [8]Oxygen-containing compounds, sulfides, selenides, and other compounds generally belong to the conversion-type materials.[11] Recently, transition metal selenide has emerged as a promising new-generation electrode material for lithium-ion batteries because of their high theoretical capacity and controlled morphologies. [12]t present, research on transition metal selenide electrode materials mainly focuses on some transition metal compounds with simple structure and easy synthesis, such as nickel selenides (NiSe x ) exists mainly in the stable states of NiSe, Ni 0.85 Se, Ni 3 Se 2 , and NiSe 2 .The compounds with different ratios have theoretical capacities of varying sizes, but the complex reaction mechanism of the composites during the chemical reaction affects the cycling performance.Among them, the theoretical capacity of NiSe is 399 mA h g À1 , and the lamellar-structured NiSe synthesized by solvothermal method can almost present the theoretical capacity because of the excellent redox kinetic properties. [13]Furthermore, molybdenum selenide (MoSe 2 ) has the layered structure of Se-Mo-Se interlayers with narrowbandgap semi-conductor properties, which are synthesized using the template-directed method depicting a specific capacity of 630 mA h g À1 at 21 mA g À1 . [14]Again, it is observed from different studies that both ZnSe and VSe 2 synthesized by hydrothermal method possess a long cycle stability.However, like many transition metal compounds, transition metal selenides may suffer weak electrical conductivity and severe volume expansion in the electrode material during the cycle chemical reaction, emerging in electrode pulverization and hence reducing the cycle life.Also, the selenides generated by deep discharge are easily soluble in the electrolyte and lead to loss of active material.Therefore, transition metal selenides are mixed with carbon materials and studied for LIB applications. [15,16]SnSe nanocrystals of 7 nm are decorated on CNTs by Jin et al., depicting a reversible capacity of 706.5 mA h g À1 after 300 cycles at a current density of 200 mA g À1 . [17]Tian et al. fabricated FeSe 2 @CNT composites with microspheres-like nanostructures via a hydrothermal method and studied Li storage that showed a specific capacity of 571.2 mA h g À1 at 0.5 A g À1 . [15]Xiao et al. reported a carbon-coated Cu 1.8 Se decahedron composite, which showed a specific capacity of 320.3 mA h g À1 at 1 A g À1 after a long cycle of 1000. [18]Iron selenides (FeSe 2 ) as the research target due to the stable orthorhombic structure and high theoretical specific capacity of 501 mA h g À1 has been already studied and underperforms in terms of ultralong capacity cycling. [19]Therefore, a second metal (copper) is introduced to construct binary metal selenides copper iron selenide (CuFeSe 2 ). [20]CuFeSe 2 is a ternary chalcogenide with properties like nontoxicity, low cost, narrow bandgap of 0.16 eV, and crucial abundance, making it a promising anode material. [21]Again, the carbon materials with high capacities and their composites show better stability and rate capability in comparison to bare transition metal-based materials.A uniform combination of CuFeSe 2 nanoflakes and CNTs needs to be synthesized with high structural stability, which reduces the volume change during the reaction and is advantageous for better ion kinetics.Also, binary metal selenides and their composite with carbon materials are rarely explored for the LIB study as anode electrode materials.
Hence, in the present study, a simple hydrothermal synthesis of CuFeSe 2 materials with high crystallinity and uniformity is reported.Further, functional CNTs with high electrical conductivity are introduced to improve the structural support and provide better charge transport pathways.Herein, the CuFeSe 2 and CNT composite (CuFeSe 2 @CNTs) shows excellent specific capacity of 783.7 and 720.6 mA h g À1 at 0.1 and 1 A g À1 , respectively.The composite further exhibits high-capacity value compared to only FeSe 2 and CuFeSe 2 .The composite depicted a reversible capacity of 691 mA h g À1 after 200 cycles at a high current density of 1 A g À1 .Moreover, the homogeneous combination of CuFeSe 2 nanoflakes and CNTs provides structural strength that reduces and sustains the volume change in the charging and discharging process and also improves the conductivity of the complex, which depicts the advantageous ion and electron kinetics during the reaction.

Material Synthesis
As shown in Figure 1, CuFeSe 2 and carbon nanotube composites (CuFeSe 2 @CNTs) were prepared via a similarly one-step hydrothermal route.Briefly, CuCl 2 (6 mmol), FeCl 2 4H 2 O (6 mmol), Se (12 mmol), multiwalled CNTs (MWCNTs, 60 mg), and citric acid (CA, 20.8 mmol) were used as the starting materials, which were added into 50 mL of distilled water.Then, they were continuously stirred and sonication was carried out for 4 h, followed by the addition of 20 mL hydrazine hydrate (N 2 H 4 H 2 O, 80%) dropwise.Again, the stirring was continued for 1 h; consequently, the mixed solution was poured into a Teflon-lined stainless-steel autoclave with sealing and heated at 180 °C for 24 h.The autoclave was naturally cooled to room temperature, and then the products formed by the reaction were centrifuged and cleaned with ethanol and deionized water several times.Finally, vacuum drying was employed to get the desired materials.Further, the synthesis of FeSe 2 and CuFeSe 2 was carried out by the same process without the addition of CuCl 2 and MWCNTs, respectively.The detailed product information like length (≈10 μm), diameter (5-15 nm), and specific surface area (150-200 m 2 g À1 ) of the MWCNT (CNT MR99) used for this work are given in the Figure S1, Supporting Information.

Materials Characterization and Electrochemical Measurements
X-ray diffraction (XRD, D8 Discover with GADDS) was employed to study the crystalline phase and microstructure.The morphology of the materials was studied using a field-emission scanning electron microscope (FESEM, 0.8 nm at 15KV, 1.6 nm at 1KV, ULTRA PLUS) and high-resolution transmission electron microscopy (HRTEM, UHR, JEOL).Again, the selected-area electron diffraction (SAED, UHR, JEOL) was carried out to study the crystallinity and details about the crystal-plane orientation of the materials.Again, an energy-dispersive X-ray spectrometer (EDS) was employed for elemental mapping.The surface chemical composition was analyzed by X-ray photoelectron spectroscopy (XPS, Al Monochromator, 1486.6 eV, PHI Quantera-II).Thermogravimetric analysis (TGA, Pyris 1 TG analyzer, Perkin Elmer) was carried out at 25-800 °C with a 5 °C min À1 heating rate under air atmosphere.The surface area and pore nature of the materials were analyzed using nitrogen (N 2 ) adsorp-tionÀdesorption isotherms by Micromeritics (ASAP 2425) at 77 K. Fourier-transform infrared spectra (FTIR) were recorded on Cary670 the instrument to observe the vibrational modes of the molecules.
The electrochemical properties of the synthesized materials were analyzed in a Li half cell, where a 2032-type coin cell was used.The slurry for the anode was prepared by mixing the active material, carbon black, and polyvinylidene difluoride in a weight ratio of 8:1:1 and consequently was coated on Cu foil.Li metal and a polypropylene film were used as the counter electrode and separator, respectively.The electrolyte LiPF 6 of 1.15 M in a 1:1 volume mixture of dissolved diethyl carbonate (DEC) and ethylene carbonate (EC) was used.The charge/discharge measurements were conducted on the Land system (WonATech, WBCS 3000) with a voltage range of 0.01-3 V. Cyclic voltammograms (CVs) and electrochemical impedance spectroscopy (EIS) were tested on the Auto-lab electrochemical workstation (PGSTAT302N, Netherlands).For EIS, the frequency range of 0.1 Hz-100 kHz was used.

Results and Discussion
The morphology of the synthesized materials was observed by SEM. Figure 2a-f exhibits the morphological images in both low and high magnifications for bare FeSe 2 , CuFeSe 2, and CuFeSe 2 @CNTs.The base FeSe 2 depicts microspheres with diameters of ≈2-3 μm (Figure 2a) and consists of polyhedral structures with diameters of about 0.3 μm (Figure 2b).CuFeSe 2 at different magnifications shown in Figure 2c,d appears to be nanoflakes-like structures with different sizes, thicknesses around 50 nm, and relatively smooth surfaces.Also, some of the nanosheets are composed of hexagonal stacks, which may be related to its growth process.
For the CuFeSe 2 @CNTs composite, the CuFeSe 2 flakes are uniformly distributed in the network composed of CNTs confirmed from the images shown in Figure 2e,f, where the CNTs provide structural support to mitigate the change in volume at the time of repeated chemical reactions and improve the electrical conductivity of the complex matrix.The elemental mappings in Figure 2g,h depict that the observations of Fe and Se in FeSe 2 , and Cu, Fe, and Se in CuFeSe 2, are equally distributed throughout the structure.
The optimized synthesis procedure/conditions of the CuFeSe 2 @CNTs which is followed are achieved by the systematic material synthesis process.Initially, reaction temperature, time, and CNT concentration are optimized for FeSe 2 .FeSe 2 is synthesized at different temperatures (160, 180, and 200 °C) with a constant reaction time, whereas at lower temperatures (160 °C) insufficient selenization is found because of the high formation of Fe 2 O 3 (Figure S2a, Supporting Information).The respective SEM and the Li-ion storage performance are shown in the Figure S2b,c, Supporting Information, respectively, which confirms 180 °C is the optimal reaction temperature with stable crystallization, high capacity, and capacity retention.Similarly, FeSe 2 is synthesized at different reaction times (12, 24, and 36 h) at a constant temperature.The optimized temperature of 24 h resulted in a stable structure (Figure S3a,b, Supporting Information) with better capacity performance (Figure S3c, Supporting Information).Further for the FeSe 2 @CNT composite, the CNT mass percentage is varied and it is found that the CNT with 30% provides increased conductivity (Figure S4a, Supporting Information), better structural support (Figure S4b, Supporting Information), and high capacity (Figure S4c, Supporting Information), compared to 10% and 50% CNT mass percentage.Again, the CuFeSe 2 structure is optimized by controlling the precursor concentration of Cu and Fe, where a 1:1 ratio of Cu:Fe shows a stable and uniform structure (Figure S5a,b, Supporting Information) with better capacity performance (Figure S5c, Supporting Information).The 1:2 and 2:1 ratio lead to the formation of FeSe 2 in the structure.The above comparisons give detailed structural and rate performance optimization to generate a better CuFeSe 2 @CNT composite material.
Figure 3 depicts the TEM analysis of CuFeSe 2 @CNTs.TEM images in Figure 3a,b confirm the composite structure where CuFeSe 2 is present in a CNT matrix.The magnified TEM images at a specific position provided in the Figure S6, Supporting Information, confirm the composite structure with a CNT diameter of nearly 15 nm.Again, the SAED pattern in Figure 3c shows the different diffraction ring patterns represented in circular rings, confirming the crystalline nature of CuFeSe 2 .The circular rings correlated with the (100), ( 112), (104), and (204) crystal planes of CuFeSe 2 with the interplanar spacing of 0.553, 0.319, 0.247, and 0.195 nm, respectively.The HRTEM image in Figure 3d confirms the lattice d-spacing of 0.319 nm that correlates with the (112) crystal plane of CuFeSe 2 .
The TGA analysis of CuFeSe 2 and CuFeSe 2 @CNTs from room temperature to 800 °C in an air atmosphere is illustrated in Figure 4b for determining the percentage of carbon materials  present in the composite structure.With an increase in temperature, CuFeSe 2 was oxidized to form Fe 2 O 3 and CuO at a temperature of nearly 300 °C and the chemical reaction is followed as Therefore, an increase in weight is observed between the temperatures ranging from 200 to 350 °C in the wt% versus temperature profile.The increase in wt% is higher for CuFeSe 2 @CNTs (11.27 wt%) compared to CuFeSe 2 (5.89 wt%), which may be due to the presence of CNT.A quick weight loss at high temperatures between 420 and 470 °C is noticed, which is due to the combustion of carbon and volatilization/sublimation of SeO 2 in air. [17,25]igher weight loss of about 64.21 and 49.86 wt% is observed for CuFeSe 2 @CNTs and CuFeSe 2, respectively.According to calculations, it is confirmed that the CNT content is about 30%.
Raman spectroscopy of CuFeSe 2 and CuFeSe 2 @CNTs was carried out to verify the presence of carbon in the composite structure, as shown in Figure 4b.The Raman band at 258 cm À1 for both samples has been generally attributed to the A 1g vibrational mode of copper and iron selenides. [26,27]he less-intense peak at 231 cm À1 corresponds to the vibrational mode of the Se-Se bond.Furthermore, the presence of the carbon structure is confirmed by the origin of D-band (1350 cm À1 ), G-band (1576 cm À1 ), and 2D band (2680 cm À1 ) in the CuFeSe 2 @CNT composite.The D-to-G band ratio (I D /I G ) is calculated and found to be 1.12, confirming the presence of defective carbon structure in the CuFeSe 2 @CNTs. [28]Further combining this observation with the SEM and TEM confirms the presence of CNT in the structure.
To confirm the surface chemical composition of the synthesized CuFeSe 2 @CNTs, sample XPS is employed, where the cleaned products are used for characterization.The XPS survey spectra range from 0 to 1400 eV is depicted in Figure 5a, which confirms the presence of C (C 1s), O (O 1s), Cu (Cu 2p), Fe (Fe 2p), and Se (Se 3d) in the structure.Again, deconvoluted Cu 2p spectra (Figure 5b) depict the presence of 2p 3/2 and 2p 1/2 peaks at 932.1 and 951.9 eV, respectively.The satellite peak at 947.5 eV is also observed.The binding energy separation of 19.8 eV is observed and the fitted curve indicates the presence of Cu þ and Cu 2þ . [29]Further, the presence of Se is confirmed from the Se 3d spectrum (Figure 5c), which consists of the Se 3d 5/2 and Se 3d 3/2 peaks at 55.1 and 55.9 eV, respectively.The peak at 60 eV in Se 3d represents the Se-O bond, which is generated due to the surface oxidation of selenium.Fe 2p spectrum confirms the presence of Fe and fit by the spin-orbit splitting (Figure 5d), which shows two asymmetric main peaks Fe 2p 3/2 and Fe 2p 1/2, at 710.7 and 724.1 eV, respectively. [30]Both peaks can further be deconvoluted and confirm the presence of Fe 3þ (2p 3/2 at 712.5 eV, 2p 1/2 at 726.8 eV) and Fe 2þ (2p 3/2 at 710.5 eV, 2p 1/2 at 724.1 eV).The satellite peaks for Fe 2p were also located at 718.6 and 732.7 eV for Fe 2p spectrum.The binding energy separation with 13.6 eV between 2p 3/2 and 2p 1/2 of Fe 2p is consistent with the reported Fe 3þ spectrum. [31]he prepared FeSe 2 , CuFeSe 2 , and CuFeSe 2 @CNTs are tested in Li half cells to understand their electrochemical performance.The CVs of FeSe 2 , CuFeSe 2 , and CuFeSe 2 @CNTs during the initial four cycles at a scan rate of 0.01 mV s À1 and potential range of 0.01-3.0V are shown in Figure 6a-c.Several redox peaks suggest multiple electrochemical reactions and all the samples show three cathodic peaks and two anodic peaks.In the initial cathodic scan of FeSe 2 , the sharp peak at 0.71 V in CV represents the insertion of Li þ into the FeSe 2 and generates Li x FeSe 2, and further, the solid electrolyte interphase (SEI) layer breaks down by the electrolyte and forms FeSe and Li 2 Se. [32]The two peaks for FeSe 2 at 2.0 and 2.2 V originated in the initial anodic scan, confirming the formation of Li x FeSe 2 and FeSe 2 .Similarly, for the CuFeSe 2 and CuFeSe 2 @CNTs, the sharp peak for the first cycle at 1.4 V represents the insertion of Li þ into the CuFeSe 2 lattice to form Li x CuFeSe 2 .For the consequent cycles, the cathodic peaks at 1.44 and 1.87 V are related to the conversion reaction with the two-step reduction of Cu, Fe, and Li 2 Se.The two anodic peaks located at 1.91 and 2.28 V correspond to the generation of Li x CuFeSe 2 and CuFeSe 2 , respectively.The first cycle is generally different from other cycles due to the formation of the SEI layer near the electrode, which leads to a lower Coulombic efficiency.The third and fourth cycles are the same, suggesting that better lithium interaction does not change the structure of the electrode material during the chemical reactions that occur at chargedischarge (CD) cycles.FTIR is carried out after the CV, and cycle tests are shown in the Figure S7a,b, Supporting Information.The peak is observed at ≈510 cm À1 for the sample after CV vanishes after the cycle test, which represents the Li-Se vibrational modes due to the formation of Li 2 Se when Li þ interacts with the CuFeSe 2 lattice, hence confirming the SEI layer formation. [33,34]he reactions for the CD process are given as follows.
Discharge/Lithiation/Reduction process: Charge/Delithiation/Oxidation process: Figure 6d-f shows the CD curves at a current density of 100 mA g À1 in a voltage range of 0.01-3.0V.The CD curves behaviors are well matched with CV anodic and cathodic peaks behaviors.The CD curves for the CuFeSe 2 @CNTs remain similar for the 150th and 200th cycles, confirming the same reactions during the reversible cycles and hence the stability of the electrode.FeSe 2 and CuFeSe 2 depict a change in CD curves and capacity with cycles.The capacity of CuFeSe 2 @CNTs electrode is high compared to the pristine FeSe 2 and CuFeSe 2 at 1st, 50th, 100th, and 200th.
To understand the capacitive and diffusive-type charge-storage behavior, CV was carried out in a potential range of 0.01-3 V at various scan rates ranging from 0.2 to 0.8 mV s À1 in an interval of 0.2.The curve shows three cathodic and two anodic peaks, and the obvious shift in the peak position is observed at high scan rates due to the time-limited interactions (Figure 7a).The current (i) behavior at a fixed potential (V ) for different scan rates (v) can be analyzed to understand the storage mechanism.The following equations are where a and b are two constant parameters.The b-value can be calculated by plotting log v vs. log i, where more specifically the 0.5 value of b signifies the complete diffusion mediated process, and the 1 value of b indicates a capacitive mediated process.If bvalue lies in between 0.5 and 1 (0.5 < b < 1), the mechanism is of both the nature of charge storage and closer to 1, means a higher proportion of the capacitive-controlled process. [35,36]From the CV curves, the log(i) vs log(v) plots for CuFeSe 2 @CNTs composite for different redox peaks are displayed in Figure 7b.The slopes of the fitted straight-line signify the b values and are calculated to be 0.94-0.98 for different peaks, which confirms that both diffusion and capacitive behavior exist in the charge/discharge cycle, and capacitive-type charge storage shows the dominance.Figure 7c depicts the CV curve at scan rate 0.8 mV s À1 with the different charge storage contributions calculated using Dunn's method, which correlates with the b-value.The capacitive contribution is found to be 79.66%,82.18%, 84.21%, and 87.36% at 0.2, 0.4, 0.6, and 0.8 mV s À1 scan rates, respectively (Figure 7d).Further, the increased surface area in the CuFeSe 2 @CNTs composite may lead to high capacitive contributions; therefore, Brunauer-Emmett-Teller measurement is carried out for CuFeSe 2 and CuFeSe 2 @CNTs.The surface area of CuFeSe 2 @CNTs (18.98 m 2 g À1 ) is found to be higher than CuFeSe 2 (14.13 m 2 g À1 ), as shown in Figure S8a, Supporting Information.Again, the calculated Barrett-Joyner-Halenda (BJH) average pore sizes (diameter) for CuFeSe 2 and CuFeSe 2 @CNTs are centered at 15.1 and 33.5 nm, respectively, with pore distribution of 0-140 nm confirming the mesoporosity nature (Figure S8b, Supporting Information).The BJH pore volumes of CuFeSe 2 and CuFeSe 2 @CNTs are calculated as 0.04 and 0.13 cm 3 g À1 , respectively.Hence, the CuFeSe 2 @CNTs inherit a high surface area and pore volume that facilitate better ion diffusion channels for electrochemical reactions, thereby improving contribution of the capacitive-controlled process and boosting the transfer of Li þ ions effectively.
The EIS measurement was carried out for the fresh and aftercycle samples of FeSe 2 , CuFeSe 2 , and CuFeSe 2 @CNTs, to reveal mechanism and electrode/electrolyte interface charge transfer behavior.The Nyquist plot for the samples in the fresh and after cycle stages is presented in the Figure 8a,b, respectively.The fitted equivalent circuits are given in the insets of Figure 8a,b, where R s , R SEI , R ct , and W refer to the bulk/series resistance, SEI resistance, charge transfer resistance, and Warburg impedance, respectively.Generally, the R s and R ct are obtained in the high-frequency region and medium-frequency region of the EIS spectra, respectively.The details about the obtained parameters after fitting the EIS with an equivalent circuit are tabulated in Table 1.
It is observed that the fresh CuFeSe 2 @CNTs gives less R s (1.32 Ω) and R ct (31.5 Ω), compared to FeSe 2 (R s : 2.12 Ω, R ct : 206.5 Ω) and CuFeSe 2 (R s : 1.35 Ω, R ct : 116.4 Ω).This confirms the better charge transfer behavior due to the presence of the CNT in the composite matrix.After the cycle, it is found that R s (FeSe 2 : 2.15 Ω, CuFeSe 2 : 8.5 Ω, CuFeSe 2 @CNTs: 4.2 Ω) value increases, reflecting the increase in solution resistance.Further, the R ct (FeSe 2 : 153.4 Ω, CuFeSe 2 : 106.3 Ω, CuFeSe 2 @CNTs: 28.6 Ω) is reduced for all the samples, that confirms the better electrochemical kinetics and fast charge/electron transport during the charge/discharge process.The low values of R SEI (2.4 Ω) for CuFeSe 2 @CNTs compared to FeSe 2 (21.15Ω) and CuFeSe 2 (19.26Ω) after cycles resulted in superior electrochemical properties and fast charge/electron transport during the Li þ insertion/extraction reaction.Again, the Li þ diffusion coefficients (D) from the EIS were calculated using the following equation.where R, T, n, A, F, C, and σ w are the gas constant, the temperature in Kelvin, the number of electrons per species reaction, the surface area of the electrode, the Faraday constant, the concentration of Li þ , and the Warburg coefficient, respectively.The relation between σ w , real part of impedance Z 0 , and angular frequency (ω = 2πƒ), where ƒ is the frequency, can be written as [37] Z According to the above equation, Z 0 verses ω À1/2 is plotted and fitted linearly to obtain the Warburg coefficient (σ w ) for the fresh and after-cycle samples, which are depicted in Figure 8c,d.The lithium diffusion coefficient D for the fresh samples of CuFeSe 2 @CNTs is found to be 2.65 Â 10 À12 cm 2 s À1 , while for the bare CuFeSe 2 is 1.87 Â 10 À12 cm 2 s À1 .The higher diffusion coefficient of CuFeSe 2 @CNTs confirms faster ion diffusion compared to the bare CuFeSe 2 .Again, the increase in D value of CuFeSe 2 and CuFeSe 2 @CNTs after cycling is observed due to the structural stability after the initial activation.Therefore, this confirms the excellent electrochemical activity of CuFeSe 2 @CNTs composites which is due to the 1) conductive network provided by the CNTs that improves the overall electrical conductivity of the composite and hence accelerates electron transport and 2) CNTs also provide strong mechanical stability, which can alleviate the volume change to improve the cycle stability.
A better understanding of the reaction at the CuFeSe 2 @CNTs electrode during the discharge process is schematically presented in Figure 9a.The detailed processes are also described in the CV discussion part and noted in Equation ( 2)-( 4).The reverse reaction occurs during the charging process (Equation ( 5) and ( 6)). Figure 9b and c shows rate capacity at different current densities ranging from 0.1 to 5 A g À1 and the cycling performance for 200 cycles at 1 A g À1 for FeSe 2 , CuFeSe 2, and CuFeSe 2 @CNTs, respectively, to investigate the capacity recovery and stability.The bare FeSe 2 displays a capacity value of 380 mA h g À1 at 0.1 A g À1 and delivers the discharge capacities of 310 and 269 mA h g À1 at the high current densities 2 and 5 A g À1 , respectively.For CuFeSe 2 the specific capacity is found to be 540 mA h g À1 at 0.1 A g À1 .The average reversible capacities of CuFeSe 2 @CNTs for five cycles at current densities of 0.1, 0.2, 0.5, 1.0, 2.0, and 5.0 A g À1 were 783.7, 748.6, 734.3, 720.6, 683.5, and 571.9 mA h g À1 , respectively.The recovered capacities of CuFeSe 2 @CNTs after a series of cycles were 774.8 mAh g À1 , which is almost the same as the initial discharge capacity.The cycle stability plot depicts the excellent stability of  CuFeSe 2 @CNTs compared to other samples.The reversible capacity of CuFeSe 2 @CNTs remains 691 mA h g À1 after 200 cycles at 1 A g À1 current density, while the discharge capacity of CuFeSe 2 is only 279 mA h g À1 .This shows the excellent Li storage performance of CuFeSe 2 @CNTs over only FeSe 2 and CuFeSe 2 .Moreover, the resultant specific capacity of CuFeSe 2 @CNTs is higher than the already-reported similar-type single-metal selenides and carbon composite materials for LIBs, that is, FeSe@C core-shell, VSe 2 /graphene nanosheets, Cu x Se@C microspheres, and FeSe 2 @CNT microspheres [15,[38][39][40] .A detailed comparison of the Li-storage performance is tabulated in Table 2.The Coulombic efficiency of all the samples is found to be nearly 100%.Further, in the initial and in-between some cycles, Coulombic efficiency is observed to be higher than 100%.This might be due to 1) the irregular amount of Li-ion transport during CD and 2) side reactions during the charging process, which affect the Li-ion intercalation during the discharge process and in turn exceed the Coulombic efficiency more than 100%. [41,42]Further to confirm the stability of the structure, SEM and FTIR are performed before and after the cycle for CuFeSe 2 @CNTs.SEM provided in the Figure S9, Supporting  Ref.

Figure 7 .
Figure 7. a) Cyclic voltammetry curves of CuFeSe 2 @CNTs at different scan rates.b) the plots of log(i) vs log(i p ) c) charge-storage contribution plot at 0.8 mV s À1 , where the shaded area shows the pseudocapacitive contribution, and d) bar diagram of diffusion and pseudocapacitive percentages at different scan rates.

Figure 8 .
Figure 8. Nyquist plot of a) fresh and b) after cycles' samples (insets showing the respective fitted equivalent circuit).The plots of the real part of impedance (Z 0 ) as a function of the inverse square root of angular frequency (ω À1/2 ) at different voltages in the Warburg region for c) fresh and d) after-cycles samples of FeSe 2 , CuFeSe 2 , and CuFeSe 2 @CNTs.

Table 1 .
The fitting results of EIS parameters for FeSe 2 , CuFeSe 2 , and CuFeSe 2 @CNTs samples before and after cycles.

Table 2 .
Comparison of metal selenides and their composite with carbon materials electrochemical performances in LIBs.