Introduction

The reusable lithium-ion batteries (LIBs), as a significant electrochemical energy storage devices, have aroused worldwide concern for their varied applications such as mobile phone, notebook computer, new energy vehicle, and so on [1,2,3]. For the past decade, people have paid a high attention to improve the performance of LIBs [4,5,6]. As we know, graphite anode materials are used in most commercial LIBs. The graphite does not meet the requirements of future high-performance LIBs because its theoretical specific capacity is only 372 mAh g−1. Hence, there is strong demand for developing new anode materials of LIBs with high reversible capacity, stable cycling life. Recently, silicon-titanium nanocomposites can also be obtained by molten salt electrolysis as anodes for LIBs, which have good electrochemical properties [7]. Nowadays, scientists have discovered several kinds of new anode materials to replace graphite, such as Si [8], Ge [9], Li4Ti5O12 [10], SnO2 [11], CuO [12], and transition metal oxalates (TMOxs) [13].

Recently, great progress has been made in the research of electrochemical properties of TMOxs [13, 14], such as MnC2O4 [15, 16], FeC2O4 [17, 18], CoC2O4 [19,20,21,22,23], NiC2O4 [17, 23], and ZnC2O4 [24]. The large volume change of the active material is a major factor leading to poor cycle performance during charging/discharging. In this aspect, the morphology, structure, or particle size has been deemed to a crucial factor to surmount the disadvantage of LIB anode material [25]. The ribbon-like nanosized FeC2O4 and CoC2O4 as LIB anode materials can give capacities of 700 mAh g−1 and 900 mAh g−1 at 2000 mA g−1, respectively, after 40 cycles [26, 27]. The mesoporous FeC2O4 and CoC2O4 nanostructures prepared by chimie douce precipitation method exhibited smaller size. These results shown higher reversible capacity, more stable cycling, and better rate capabilities compared to those reported previously [18, 28]. Reasonable synthesis methods could be effective means to elevate their electrochemical performances. The SnC2O4/graphene composite was synthesized as the anode material of LIB by microwave-assisted method. The reversible discharge capacity was 657 mAh g−1 after 100 cycles at 200 mA g−1 [29]. Microwave-assisted methods significantly reduce particle size and reaction time, indicating potential application in actual production.

Herein, we report the microwave-assisted methods for the controlled synthesis of CuC2O4·xH2O-based anode materials by choosing suitable synthetic mediums (water and ethyl glycol). The quasi-spherical hydro-CuC2O4·xH2O (hCuOx) and quasi-cotton solvo-CuC2O4·xH2O (sCuOx) were obtained and the as-prepared CuC2O4·xH2O nanostructures exhibited superior reversible specific capacity and cycling stability as the anode materials for LIBs. Their reversible specific capacity of hCuOx and sCuOx can be maintained as high as 1201 and 1008 mAh g−1 at the current density of 200 mA g−1 after 50 cycles, respectively.

Experimental

Materials

All chemicals were analytical grade and without further purification: EG (AR, Aladdin Industrial Corporation, 99%), H2C2O4·2H2O (GR, Aladdin Industrial Corporation, 99.8%), CuCl2·2H2O (AR, Aladdin Industrial Corporation), separator (Celgard 2400), electrolyte (EC:DMC:EMC=1:1:1+1%VC, Nanjing Mojiesi).

Synthesis of CuC2O4·xH2O materials

The microwave reaction is carried out in the EXCEL full-function microwave chemistry working platform (PreeKem Shanghai). CuCl2·2H2O (s, 1.194 g, 7 mmol) was dissolved into 35 mL DI (or 35 mL EG). H2C2O4·2H2O (s, 0.883 g, 7 mmol) was added into the solution. After stirring for 1 h, the suspension was transferred into a 100-mL Teflon liner and heated at 120 °C for 20 min in a microwave system of EXCEL. The products hCuOx and sCuOx were collected by centrifugation. The precipitates were washed with DI and then ethanol (ET) three times and dried at 120 °C under vacuum for 12 h. The schematic diagram for the formation of hCuOx and sCuOx is shown in Scheme 1.

Scheme 1
scheme 1

The synthesis illustration of CuC2O4·xH2O obtained by facile microwave solvothermal reactions and electrochemical performance test

Full size image

Characterization

The powder X-ray diffraction (XRD) data of both hCuOx and sCuOx were collected at room temperature on a Bruker D8 powder X-ray diffractometer equipped with Cu Kα radiation (λ = 1.5418 Å) in the 2θ range 5°–80°. Thermogravimetric (TG) analysis was performed on a SDTQ600 TG–DSC (TA Instruments) with a continuous 10 °C min−1 ramped from room temperature to 800 °C. The morphology of both samples were observed on a scanning electron microscope (SEM, Hitachi S-3400), transmission electron microscopy (TEM, JEM 2100), and high-resolution transmission electron microscopy (HRTEM, H-800). The dynamic light scattering (DLS) measurements were performed in MilliQ water using a Nano-s90 (Malvern Instruments, Worcestershire, UK).

Electrochemical experiments

The working electrodes were prepared with active materials, acetylene black and polyvinylidene fluoride (PVDF) by the proportion of 7:2:1, and the solvent was 1-methyl-2-pyrrolidone. The slurry was smeared on a copper foil (wetting thickness: 150 μm, 1 mg) and dried in a vacuum oven at 80 °C for 12 h. Lithium metal was used as a counter electrode, and a carbon copper foil was coated as a counter current collector, and glass fibers were used as a separator and an electrolyte. The cells’ charge/discharge tests on XINWEI CT-4008 system (Shenzhen Neware Battery, China) with range of 0.01–3.0 V. Cyclic voltammograms (CVs) were measured on electrochemical workstation (CHI-660e, Huachen, Shanghai) with range 0.01–3.0 V at a scan rate of 0.1 mV s−1.

Results and discussion

Characterization of CuC2O4·xH2O

Figure 1a shows XRD patterns of orthorhombic CuC2O4·xH2O (0 < x < 1) (JCPDS 21-07). The widths of all diffraction peaks are quite broad demonstrating. According to the Debye-Scherrer equation [30], the calculated interplanar spacing of (110) faces can be treated as an equivalent for the average size of primary nanocrystals, giving a value of 8.9 and 14.1 nm for hCuOx and sCuOx, respectively. As shown in Fig. 1b, the peaks below 1000 cm−1 correspond to typical Cu–O vibrations, and the bands at 1322–1600 cm−1 are attributed to the C=C vibrations [31]. From the results about Raman spectra (Fig. S1), it can be found that the characteristic main peaks of these two materials are almost identical [32].

Fig. 1
figure 1

a XRD patterns, b FT-IR spectra, c TGA and d DSC of hCuOx and sCuOx

Full size image

TGA/DSC measurements were performed to estimate the weight of crystallized water and the structural stability of CuC2O4·xH2O (Fig. 1c, d). Although it is hard to clearly distinguish thermal and oxygen decomposition at a heating step of 10 ˚C/min-1 , the decomposition of hCuOx and sCuOx were carried out in two steps. The release of crystallized water molecules and decomposition of anhydrous copper oxalate can be described by Eqs. (1) and (2) [20, 33]. From Fig. 1d, the thermal dehydration and oxalate decomposition processes occurred at ca. 260 °C and 320 °C, respectively.

$$ {\mathrm{CuC}}_2{\mathrm{O}}_4\cdotp {\mathrm{xH}}_2\mathrm{O}\to {\mathrm{CuC}}_2{\mathrm{O}}_4+{\mathrm{xH}}_2\mathrm{O} $$
(1)
$$ {\mathrm{CuC}}_2{\mathrm{O}}_4\to \mathrm{Cu}+{2\mathrm{CO}}_2 $$
(2)

The TGA/DSC result suggests that the total weight loss of hCuOx and sCuOx were 58.7% and 60.4%, and the corresponding x values were 0.14 and 0.49, respectively [34].

The surface properties and porosity of the hCuOx and sCuOx were further investigated by N2 adsorption measurements. The Brunauer, Emmett, and Teller (BET) tests were also performed to analyze the pore size distribution. The Barrett, Joyner, and Halenda (BJH) pore size distribution and the type IV N2 adsorption/desorption isotherm curves of hCuOx and sCuOx are exhibited in Fig. 2a, b, respectively. N2 adsorption/desorption isotherms have weak hysteresis loop at the relative pressure of 0.7–1.0 P/Po. As Fig. 3a shows, these average pore diameters for hCuOx and sCuOx are determined at 10.4 nm and 15.4 nm, which indicated mainly mesopores. The BET surface area of sCuOx (50.5 m2 g−1) is larger than that of hCuOx (15.4 m2 g−1). The hCuOx and sCuOx show larger pore sizes compared to the results of other copper compounds and other oxalates listed in Table 1. Furthermore, the large porous is of beneficial help for providing more active sites for quick electron transfer at the electrolyte/electrode interface, which results in huge improvement of electrochemical performance [37].

Fig. 2
figure 2

a BJH pore size distribution and b N2 adsorption-desorption isotherms curves of hCuOx (up) and sCuOx (below)

Full size image
Fig. 3
figure 3

a, b SEM images of hCuOx at different magnifications and c, d SEM images of sCuOx under the same condition

Full size image
Table 1 Comparison of the surface and textural properties of CuC2O4·xH2O materials between this work and the previous reports
Full size table

As shown in Fig. 3, the SEM images clearly present the considerable difference in the morphology between the hCuOx and sCuOx. The hCuOx has a quasi-spherical morphology with diameter of ca. 200–600 nm. The sCuOx has quasi-cotton morphology (Fig. 3c, d) with an average size of ca. 50 nm. The results demonstrate that the reaction mediums can control growth of these materials with different shapes. Compared to the resistance heating method, microwave-assisted heating method has several advantages: (i) uniform particle size; (ii) the particle size of hydrated copper oxalate decreased obviously; (iii) the morphology has changed dramatically [38, 39].

Further investigation of both CuC2O4·xH2O morphologies were performed by TEM experiments. The hCuOx and sCuOx revealed the quasi-spherical shape (Fig. 4a, b) and the quasi-cotton morphology (Fig. 4d, e), respectively. The electron beam was focused on the CuC2O4·xH2O nanoparticles, and TEM observations present the ambiguous lattice fringes that are amplified and shown as inset in Fig. 4b, e. The HRTEM images of hCuOx exhibited in Fig. 4c reveal a distinct set of visible lattice fringes with interplanar spacing of 0.17 nm and 0.23 nm, corresponding to the (020) and (111) planes of orthorhombic CuC2O4·xH2O, respectively. Similarly, the HRTEM image of sCuOx is displayed in Fig. 4f with the (020) and (120) planes corresponding to interplanar spacing of 0.17 and 0.20 nm.

Fig. 4
figure 4

TEM images (a, b) of hCuOx and TEM images (d, e) of sCuOx. Inset is the corresponding lattice fringes of a representative nanoparticle and HRTEM image (c) of hCuOx and HRTEM image (f) of sCuOx

Full size image

As shown in Fig. S2, the dynamic light scattering (DLS) results reveal that the average hydrodynamic sizes of hCuOx and sCuOx are about 712 and 255 nm, respectively. The results are larger than the data (400 and 50 nm) observed in TEM and SEM analyses (Figs. 3 and 4). This is mainly due to the fact that the TEM and SEM images exhibited the actual size, whereas the size measured by the DLS method is the hydrodynamic diameter in the hydrated state [40].

Electrochemical characterization

Cyclic voltammetry (CV) curves are shown in Fig. 5. The CuC2O4·xH2O first degrade to Cu and Cu2O, which are subsequently involved in conversion reactions. The reaction mechanism of TMOxs (such as iron [18], cobalt, and nickel oxalates [23]) can be described via Eq. (3). The incipient reaction of CuC2O4·xH2O with lithium can be roughly described as Eq. (4).

$$ {\mathrm{MC}}_2{\mathrm{O}}_4+{2\mathrm{Li}}^{+}+{2\mathrm{e}}^{-}\to {\mathrm{Li}}_2{\mathrm{C}}_2{\mathrm{O}}_4+\mathrm{M} $$
(3)
$$ {\mathrm{Cu}\mathrm{C}}_2{\mathrm{O}}_4\cdotp {\mathrm{xH}}_2\mathrm{O}+{\mathrm{Li}}^{+}+{\mathrm{e}}^{-}\to {\mathrm{Cu}}_2\mathrm{O}+\mathrm{Cu}+{\mathrm{Li}}_2{\mathrm{CO}}_3+{\mathrm{Li}}_2\mathrm{O}+\mathrm{LiOH} $$
(4)
Fig. 5
figure 5

CV profile at 0–3 V at a scan rate of 0.1 mV s−1: ahCuOx and bsCuOx

Full size image

As shown in Fig. 5a, two cathodic current peaks at 1.42 and 0.86 V in the first cycle are observed. The high peak located at 0.86 V in the first cycle transferred to 1.05 V during the second cycle. Figure 5b presents CV behaviors of sCuOx electrode, which indicates two cathodic current peaks at 1.21 and 0.89 V, then the peak located at 0.89 V in the first cycle shifted to 1.03 V, finally, deracinated in the following cycle. This boils down to the following reasons: (i) the decomposition of electrolyte leads to the formation of an organic layer deposited on the surface of the particles; (ii) lithium-ion intercalation conversion reaction from Li2CO3, Li2O, LiOH, Cu2O, and metallic copper (Cu0) particles as shown in Eq. (4). Meanwhile, an oxidation peak in 1.28 V was noted during the anodic process, same as the reversible oxidation step of Cu0, Cu+ to Cu2+. Figure 5b displays CV behaviors of sCuOx electrode, which manifests the same lithium storage mechanism as that of hCuOx [24, 33].

Figure 6 presents the voltage profile and the cycling behavior of the samples at 200 mA g−1. The hCuOx electrode materials manifested the incipient discharge potential plateau at ~ 1.5 V, then thin slope corresponding to the electrochemical reaction of CuC2O4·xH2O with lithium. From Fig. 6a, initial discharge capacity of hCuOx electrode is 1436.2 mAh g−1 and charge capacity is 819.1 mAh g−1 with a coulombic efficiency of 57% (Fig. 6a). A significant irreversible capacity loss in the first cycle is due to secondary reactions, such as the electrochemical removal process of crystallized water and the decomposition of electrolyte to form solid electrolyte interphase (SEI) layers on the surface of electrode. The prominently decay of capacity to the four full cycles due to electrochemical properties of CuC2O4·xH2O (0 < x < 1). Subsequently, the hCuOx gives discharge capacities of 495.7, 809.6, 1040.8, and 1192.3 mAh g−1 for the 4th, 10th, 20th, and 50th cycles, respectively. Every reversible cycle reveals high Coulomb efficiency about 97.6% (Fig. 6b). In order to explore the effect of charge and discharge on morphology, we monitored the changes in morphology of the two CuC2O4·xH2O materials before cycling test. Fig. S3 gives the SEM images of fresh electrode and the electrode after 50 charge/discharge cycles at the current density of 200 mA g−1. Comparing with the fresh electrode, the electrode after 50 charge-discharge cycles shows some microstructure changes. This indicates that the two electrode materials are relatively stable.

Fig. 6
figure 6

Discharge-charge profiles of a, bhCuOx and c, dsCuOx electrodes operated at 200 mA g−1

Full size image

Compared with the electrochemical data displayed in Fig. 6a, b, sCuOx electrode indicates higher initial discharge capacity of 1688.1 mAh g−1 and slightly lower charge capacity of 634.3 mAh g−1, with a lower initial Coulomb efficiency of 37.6% (Fig. 6c). At the same time, a recovery for the lithium storage capability of sCuOx is relatively low after the first 4 cycles and displays the 4th, 10th, 20th, and 50th cycle discharge capacity of 521.1, 780.7, 992.5, and 1024.1 mAh g−1, respectively (Fig. 6b). Both the initial and 50th discharge capacities of hCuOx/sCuOx electrodes are outclass commercial graphite (372 mAh g−1).

To further explore the cycling stability and rate cycling performances of hCuOx and sCuOx, the discharge capacities at 200 or 500 mA g−1 are shown in Fig. 7a. The discharge-charge curves of hCuOx and sCuOx at 1000 mA g−1 are given in Fig. S4. In the case of hCuOx, it is observed that the capacity fades up to four cycles and starts to increase thereafter. A maximum discharge capacity of 1263.2 mAh g−1 is observed at the 30th cycle and 200 mA g−1 rate, and the capacity gradually decreases with prolonging time. With the increase of current density, the charge/discharge capacity of hCuOx decreased obviously. At a rate of 500 mA g−1, the maximum discharge capacity was 1209.1 mAh g−1 at the 38th cycle, and then the capacity was gradually decayed. After 100 cycles, the cells deliver stable reversible capacity of 1260.4 and 1064.3 mAh g−1 at 200 and 500 mA g−1 rate, respectively. At a current density of 1000 mA g−1 (Fig. S5), the 100th discharge capacities for hCuOx and sCuOx are 780.2 and 624.6 mAh g−1, respectively.

Fig. 7
figure 7

a Cycling performance measured at 200 mA g−1 and 500 mA g−1. b Rate capability at various current densities. c, d EIS of hCuOx and sCuOx

Full size image

In Fig. 7a, at each current density, the higher initial discharge capacity and lower capacity retention of sCuOx are shown. The discharge capacity decreases in the first four cycles and then increases continuously, the same as those of the hCuOx. If the current density is 500 mA g−1, the 100th discharge capacities are 1064.3 mAh g−1 (hCuOx) and 939.3 mA h g−1 (sCuOx). Compared to the previously reported results of CuO-based anode materials and other dehydrated oxides listed in Table 2, the charge/discharge capacities of CuC2O4·xH2O materials are also very excellent.

Table 2 A summary for the lithium storage capability of as-prepared CuC2O4·xH2O nanostructures, dehydrated oxysalts, and CuO-based anode materials reported previously
Full size table

To check the rate capability, the hCuOx and sCuOx electrodes were cycled 10 times from 100 to 2000 mA g−1 (Fig. 7b). The cells were charged and discharged at different current densities of 100, 200, 500, 1000, and 2000 mA g−1 within a voltage range of 0.01–3.0 V. Both hCuOx and sCuOx electrodes undergo an initial reduction followed by a continuous increase in discharge capacity. A comparison of hCuOx (x~0.14) with sCuOx (x~0.49) at different current densities shows that the hCuOx nanostructures have lower initial discharge capacity and larger capacity decay at current densities of 100 and 200 mA g−1. Thus, higher capacity increase ensues, thereafter, higher capacity maintainability at current densities of 500, 1000, and 2000 mA g−1. Finally, the current density is set back to 100 mA g−1 and hCuOx and sCuOx are still found to deliver high discharge capacities of 1220.3 and 1112.4 mA h g−1, and the results indicate the excellent rate capability of hCuOx and sCuOx.

To further evaluate the electrochemical properties of the CuC2O4·xH2O, the Nyquist plot and corresponding equivalent circuit were studied [15, 36]. Figure 7c, d shows the electrochemical impedance spectra (EIS) of hCuOx and sCuOx electrode of fresh cells. A slope line at lower frequency is related to the Warburg resistance of ion diffusion [41] at the medium frequency, which can be ascribed to the charge-transfer resistance (Rct) and the constant phase element (CPE) of the interface of electrode and electrolyte. The lower slope line of frequency is related to the Warburg resistance of ion diffusion at intermediate frequency, which can be attributed to the Rct and the CPE of the electrode and electrolyte interface [35]. For the fresh cell, Rct values of hCuOx and sCuOx are 433.5 and 308.1 Ω, respectively. This is why the initial discharge capacity of sCuOx is greater than hCuOx. After three cycles in the CV measurement, Rct values of hCuOx and sCuOx are 58.9 and 202.1 Ω, respectively. Thus, hCuOx is far better than sCuOx in Rct.

The charge stored in a battery has two components: faradaic and capacitive. To confirm the quantitative relationship between faradaic and capacitive, CV analyses of hCuOx and sCuOx at various scan rates (0.4–1.2 mV s−1) were carried out at 0.01–3.0 V (Fig. 8a, b). The faradaic current is proportional to the square root of voltage sweep rate, while the capacitive current is proportional to the voltage sweep rate [17]. Hence, faradaic contribution and capacitive can be expressed by the following equation [42]:

$$ i(V)={k}_1v+{k}_2{v}^{0.5} $$
(5)

or

$$ i(V)/{v}^{0.5}={k}_1{v}^{0.5}+{k}_2 $$
(6)

where i(V) is the current as a function of voltage, and k1v and k2v0.5 are amount to capacitive effects and diffusion-controlled reaction, respectively. According to Eq. (6), we can calculate the values of k1 and k2; consequently, the capacitive and faradaic current intensities separately. The contributions of faradaic and capacitive at different sweep rates are shown in Fig. 8c, d for 5 cycles. The general trend is that the higher the voltage scan rate, the higher the capacitive current. For hCuOx and sCuOx, the capacitance contribution rate can reach 68.7% and 63.9% at 1.2 mV s−1, which indicates the irreplaceable role of the capacitive contribution in the total capacity. Therefore, the hCuOx and sCuOx have excellent lithium storage capacity as anode materials for lithium ion batteries.

Fig. 8
figure 8

CV curves of ahCuOx and bsCuOx at various scan rates, from 0.4 to 1.2 mV s−1. chCuOx and dsCuOx images express contribution ratio of the capacitive and faradaic at various scan rates

Full size image

Conclusions

In this work, we used a microwave-assisted hydrothermal and solvothermal method to fabricate the hCuOx and sCuOx anode materials’ facile operation and rapid reaction. They have given distinct different morphologies (quasi-spherical and cotton shapes) explaining this surfactant-free method could be applied in green industrial process. SEM/TEM/BET characterizations revealed that hCuOx and sCuOx are nanoscale and mesoporous. The hCuOx and sCuOx manifest maximum discharge capacity of 1436.2 and 1688.1 mAh g−1, respectively. The presence of crystallized water results in a rapid decrease in discharge capacity during the first four discharge cycles, but the remarkable lithium storage capacity of hydrated oxalates makes it impossible to ignore.