One-step synthesis of novel mesoporous three-dimensional GeO2 and its lithium storage properties

Novel mesoporous three-dimensional GeO2 was successfully synthesized by a facile one-step synthesis method followed by mixing with graphene using a spray drying process. The well-dispersed mesoporous GeO2 demonstrates a bean-like morphology (b-GeO2) with a particle size of 400 to 500 nm in length and 200 to 300 nm in diameter, in which mesopores with an average size of 3.6 nm are distributed. The b-GeO2 without any additional conductive surface layer shows a high reversible capacity for lithium storage of 845 mAh g−1 after 100 cycles, with nearly no capacity fading. When graphene was employed to be mixed with GeO2via a spray drying method, the electrochemical performance is further significantly improved. The b-GeO2/graphene composite electrode gives a higher de-lithiation capacity of 1021 mAh g−1, and the capacity retention is measured to be as high as 94.3% after 200 charge–discharge cycles for constant current cycling at 0.2 C, as well as an excellent rate performance, even displaying a reversible capacity of 730 mAh g−1 at a rate of 5 C.


Introduction
Nowadays, rechargeable lithium-ion batteries (LIBs) are the essential part of the most important energy storage devices for portable electronic devices and electric vehicles. [1][2][3] In order to meet the requirements of high energy density as well as high power performance, a worldwide effort has been made to develop novel high-performance electrode materials to keep pace with the fast increasing market demands. The relatively low theoretical specic capacity (372 mAh g À1 ) and poor charging rate capability, in particular at low temperatures, are still the main drawbacks of currently commercialized graphite anode materials. 4,5 Various compounds that form intermetallic phases (so-called "alloys") with lithium have been proposed as alternative anode materials, such as silicon (Si), 6-8 tin (Sn) 9,10 and germanium (Ge). 11 Compared to Si, less attention has been paid so far to Ge due to its higher costs. Nevertheless, its good lithium diffusivity (400 times higher than silicon) and high electronic conductivity (ca. 104 times higher than silicon) enable it as a promising anode material for the next-generation LIBs. Currently, researchers report on different structures of Gebased materials, such as nanotubes, 12 nanowires, 13,14 porous structures 15 and Ge-based composite materials. 11,16 So far, only a few studies focus on GeO 2 as an anode material. GeO 2 is of interest for application as a negative electrode material due to its high theoretical reversible capacity (1125 mAh g À1 based on 4.25 mol Li per mol Ge), low operation voltage (0.7 V to 0 V vs. Li/ Li + ) and higher thermal stability compared to Ge. [17][18][19] The reaction of GeO 2 with Li involves two steps: (1) GeO 2 + 4Li / Ge + 2Li 2 O; (2) xLi + Ge # Li x Ge (x # 4.25). While the rst reaction can be considered as to a large part electrochemically irreversible, resulting in a large irreversible charge loss by formation of Li 2 O, the lithiation/de-lithiation reaction in step 2 can be regarded as reversible, providing the discharge capacity of GeO 2 . The high irreversible capacity during the rst lithiation process, which is related to the formation of Li 2 O, is also wellknown for SnO 2 -based anodes and is one of the main reasons hindering so far the application in commercial lithium-ion batteries. 20 Despite the fact that these metal oxides will have no application in lithium-ion batteries, there is still a vast number of publications on new oxide composites, structures and varieties. 17,[21][22][23][24] The motivation for this work seems to come from the interest in new materials and their synthesis processes rather than from application in electrochemical power sources.
In general, the synthesis of GeO 2 can be achieved by different methods, including hydrolysis reactions from a Ge precursor, chemical vapor deposition (CVD) and sputter processes. Ngo et al. synthesized GeO 2 particles by a sol-gel method, which presented a reversible capacity of around 750 mAh g À1 at a rate of 0.1 C. 25 Feng et al. reported on GeO 2 lms as anode materials which were produced via a reactive radio frequency sputtering process at different temperatures. The results showed that the GeO 2 thin lm with 10 nm thickness possesses the best electrochemical performance, which gives an initial capacity of 930 mAh g À1 with 89% capacity retention aer 100 cycles. 26 Guo et al. reported a GeO 2 /graphene composite anode material prepared via a one-step in situ chemical reduction synthesis method. This material presented a high reversible capacity of 1000 mAh g À1 aer 50 cycles with a capacity retention of 90%. 17 Herein, we present a facile one-step preparation method for mesoporous GeO 2 particles with an average particle size of 500 nm and a bean-like morphology. The bean-like GeO 2 (hereaer abbreviated as b-GeO 2 ) material without any additional conductive surface layer shows a good cycling performance and good rate capability. When graphene was employed to be mixed with GeO 2 via a spray drying method, the electrochemical charge-discharge cycling and rate performance are further signicantly improved.

Preparation of GeO 2 /graphene composite
Graphene oxide (GO) was synthesized from natural ake graphite (325 mesh, Sigma-Aldrich) by a modied Hummers method (see ESI †). 27,28 The GeO 2 /graphene composite was prepared via a spray drying method. In a typical process, 0.1 g GeO 2 was homogeneously dispersed in a 25 mL aqueous GO suspension (1 mg mL À1 ), and then the mixture was transferred into the spray drying device (BÜCHI Mini Spray Dryer B-290), followed by annealing at 500 C for 4 hours in an Ar atmosphere to generate GeO 2 /GSs (GeO 2 /graphene sheets).

Structure and morphology characterization
X-ray diffraction (XRD) measurements were carried out using a Bruker D8 Advance X-ray diffractometer (Bruker AXS GmbH) equipped with a copper target X-ray tube (radiation wavelength: The morphologies of the samples were observed using a scanning electron microscope (Carl Zeiss AURIGA ® , Carl Zeiss Microscopy GmbH). Transmission electron microscopy (TEM, JOEL JEM-100CX) was used to investigate the structure of graphene oxide.
The BET specic surface area and BJH pore diameter distribution have been determined by nitrogen adsorption measurements using an ASAP 2020 (Accelerated Surface Area and Porosimetry Analyzer, Micromeritics GmbH). Before the measurement, the samples were degassed at 120 C until a static pressure of less than 0.01 Torr (0.0133 mbar) was reached.
Thermogravimetric analysis (TGA) was conducted using a TGA Q5000 IR system (TA Instruments). The measurements were carried out in an oxygen atmosphere in the temperature range of 30 C to 800 C at a heating rate of 10 C min À1 .

Electrode preparation, cell assembly and electrochemical investigations
Composite electrodes were prepared using a composition of 80 wt% GeO 2 active material, 10 wt% of conductive carbon black agent C-nergy Super C65 (Imerys Graphite & Carbon) and 10 wt% of sodium carboxymethylcellulose (CMC, Walocel CRT 2000 PA 12) as a binder. Prior to the dispersion of solid compounds, the binder polymer was dissolved in de-ionized water to obtain a 2.0 wt% solution. An appropriate amount of Super C65 was added to the binder solution and the mixture was further homogenized by stirring. Aerwards, a high-energy dispersion step (Ultra-Turrax T25, 1 hour, 5000 rpm) was employed to eliminate agglomerates and homogenize the mixture. The paste was cast on copper foil by a standard labscale doctor-blade technique. The gap of the doctor-blade was set at 120 mm wet lm thickness, leading to an average mass loading of 1.08 mg cm À2 . Aer casting, the tapes were transferred into an oven and dried in air for 1 hour at 80 C. Electrodes with a diameter of 12 mm were cut out and a further drying step was performed under an oil-pump vacuum (<0.1 mbar) at 120 C for 24 hours. Thereaer, the electrodes were stored in an argon lled glove box (UniLab, MBraun) with water and oxygen contents of less than 1 ppm.
Electrochemical experiments were performed using CR2032type coin cells with Celgard 2400 as the separator and highpurity metallic lithium (Rockwood Lithium) as the counter electrode. The electrolyte was 1 M LiPF 6 in a mixture of ethylene carbonate (EC) and diethyl carbonate (DEC) (3 : 7 in weight ratio). The cells were assembled in an argon-lled glove box with oxygen and water contents less than 10 ppm. The electrochemical performance was evaluated on a Maccor 4300 battery test system at 20 C. The cut-off voltage was 0.01 V for the charge process (lithiation) and 1.5 V for the discharge process (delithiation). The specic capacity was calculated on the basis of the total composite weight, and the C-rate was calculated with respect to a theoretical capacity of 1165 mAh g À1 (1 C). In the case of the GeO 2 /graphene composite electrode, the theoretical capacity is 1103 mAh g À1 (where the theoretical capacity of graphene is 540 mAh g À1 . 29 Cyclic voltammetry (CV; 0.01-1.5 V) was performed at a scan rate of 0.02 mV s À1 using a VMP multichannel constant voltage-constant current system (Biologic ® Science Instrument).

Results and discussion
As schematically illustrated in Fig. 1, our synthesis approach of GeO 2 is based on the self-assembly of a germanium precursor chelated by GeCl 4 and ethylene glycol in the presence of aniline. Aniline, which acts as a base, may promote the growth and stabilization of the chelated germanium precursor by depletion of excess of the acid HCl, which is formed during the chelating process. It should be noted that GeO 2 can readily react with alkali and thus it is important to select a weak base, such as aniline, which can not only absorb the acid arising from the reaction, but also prevent the dissolution of GeO 2 . Finally, mesoporous GeO 2 can be spontaneously obtained via the removal of the organic phase using ethanol.
The GeO 2 phase of the b-GeO 2 particles was characterized by powder X-ray diffraction analysis, as shown in Fig. 2. The XRD pattern can be indexed (ICDD# 36-1436) to a hexagonal structure of the P3221 (no. 1544) space group, and no impurity phase was observed. 11 Meanwhile the positions of all peaks are in good agreement with commercial GeO 2 , except that commercial GeO 2 presents sharper peaks. The synthesis process of our work was completely operated at room temperature without further heattreatment, which can well explain the existence of broad peaks of the as-prepared material. Furthermore, the broadened peaks are an indication of a very small crystallite size.
The as-prepared GeO 2 shows a bean-like morphology with a particle size of 400 to 500 nm in length and 200 to 300 nm in diameter (Fig. 3a-d). It can be clearly observed that the as-made GeO 2 particles exhibit a uniform morphology. In addition, the well-dispersed GeO 2 shows no tendency to agglomerate, which will be favorable for its electrochemical performance. Furthermore, b-GeO 2 shows a porous structure with a BET specic surface area of 112 m 2 g À1 , a pore volume of 0.163 cm 3 g À1 and a pore size of about 3.6 nm, as depicted in Fig. 5, which is typical type H3 of a IV isotherm curve, representing the existence of the porous structure, 30,31 in particular displaying mesopores.
In a further step, we take advantage of graphene (a lowdimensional carbon material with high electronic conductivity and excellent mechanical properties; graphene can contribute to the capacity) to further enhance the electrochemical performance of b-GeO 2 . 32,33 A spray drying method was employed to simply mix GeO 2 and graphene (the morphology of graphene oxide is shown in Fig. S1, ESI †). As illustrated in Fig. 4, it can be clearly observed that GeO 2 particles are tightly anchored on or     wrapped within the graphene sheets. Notably, during the mixing process, GeO 2 particles are still homogenously dispersed within or rmly encapsulated by the graphene sheets. To further determine the chemical distribution of the composite, energy dispersive X-ray spectroscopy (EDX) analysis was performed. Fig. S3 (see ESI †) shows the elemental mapping of the corresponding micrograph of the b-GeO 2 /graphene composite, in which carbon is settled everywhere, and GeO 2 was found to be homogeneously distributed within the carbon phase.
The amount of graphene was determined by thermogravimetric analysis, as illustrated in Fig. S2 (see ESI †). The weight loss was observed from 200 C to 700 C; a rapid mass loss takes place between 500 C and 650 C. At 700 C, the total mass loss is around 12%, so that the mass percentage ratio of graphene and GeO 2 is calculated to be 12% and 88%, respectively.
In Fig. 6a, the voltage vs. specic capacity proles for the charge-discharge cycling process of b-GeO 2 at a rate of 0.1 C in the rst cycle (formation cycle) and 0.2 C in the following cycles are presented. With respect to the rst cycle, the electrode gives a lithiation capacity of 2394 mAh g À1 and a high de-lithiation capacity of 1117 mAh g À1 , which is almost equal to the theoretical capacity of GeO 2 (1125 mAh g À1 ). However, the rst cycle efficiency is only about 46.7% (see Table 1), which is quite low compared to currently used graphite-based anode materials. The large capacity loss can be ascribed to the irreversible reaction of GeO 2 during the Li insertion process by formation of Li 2 O. 18 Nevertheless, from the second cycle, the efficiency of the electrode was improved signicantly and rises nearly to 99% aer 50 cycles. The electrochemical lithiation/de-lithiation characteristics of the b-GeO 2 material were determined using cyclic voltammetry (CV), as shown in Fig. 6b. Upon the initial cathodic sweep, a broad reduction peak can be observed at about 0.05 V. It is proposed that the reaction of GeO 2 with Li involves two steps: (1) GeO 2 + 4Li / Ge + 2Li 2 O; (2) xLi + Ge # Li x Ge (x # 4.25). In the case of the rst anodic potential sweep, one sharp oxidation peak around 0.55 V and a small broad peak at 0.65 V are revealed, which correspond to the Li extraction reaction. The subsequent cycles show four reduction peaks at 0.50 V, 0.36 V, 0.18 V and 0.09 V, which can be related to the lithium alloying reaction to form different Li x Ge alloys, as described by eqn (1), 34,35 and only one oxidation peak at about 0.55 V, which are in good agreement with the electrochemical characteristics of metallic Ge.
Lithiation: Ge / Li 9 Ge 4 / Li 7 Ge 2 / Li 15 Ge 4 / Li 22 Ge 5 (1) In order to demonstrate the improved electrochemical performance of b-GeO 2 and its composite electrodes (b-GeO 2 / graphene), commercial GeO 2 powder was studied under the same conditions for comparison. Fig. 7a shows the de-lithiation capacity curves as well as the Coulombic efficiency curves for the charge-discharge cycling of b-GeO 2 , b-GeO 2 /graphene and commercial GeO 2 at a charge-discharge rate of 0.2 C (the rst cycle is conducted at a rate of 0.1 C). Although commercial GeO 2 provides a high initial reversible capacity of 1182 mAh g À1 , a rapid capacity fade occurs from the very beginning of cycling, most likely due to the large volume variation, which results in the pulverization and electronic detachment of the active material. In contrast, the b-GeO 2 electrode reveals a signicantly improved cycle stability, which demonstrates a high reversible capacity of 845 mAh g À1 aer 200 cycles, with nearly no capacity fading (according to the capacity of the 2 nd cycle, see Table 1). The improved cycling performance can be ascribed to the existence of the mesoporous structure, which can offer sufficient inner space to absorb the volume changes of GeO 2 . By mixing GeO 2 with graphene, the initial Coulombic efficiency, cycling performance as well as reversible capacity is further improved. The electrode gives a higher de-lithiation capacity of 1021 mAh g À1 with an initial efficiency of 56.4%, and the capacity retention is measured to be as high as 94.3% aer 200 cycles. Moreover, from the second cycle, it is observed that the reversible capacity of the b-GeO 2 /graphene composite received a remarkable improvement. The coulombic efficiency is above 99.3% from the second cycle. The signicantly improved electrochemical performance of the composite is mainly related to its stabilized whole structure, in which the GeO 2 nanoparticles are rmly anchored on or wrapped within the graphene sheets. Together with the good exibility of the graphene sheets, this composite material might effectively accommodate the big volume changes of GeO 2 during lithiation. In addition, graphene can not only improve the electronic conductivity of GeO 2 and therefore raise its utilization efficiency, but also contribute to the specic capacity of the composite. Fig. 7b shows the rate capability of commercial GeO 2 , b-GeO 2 and b-GeO 2 /graphene anode materials. The b-GeO 2 /graphene electrode delivers higher rate capability than the other electrodes, especially at high specic currents. It shows a high reversible capacity of 730 mAh g À1 even at 5 C, while b-GeO 2 and commercial GeO 2 only maintain 300 mAh g À1 and 180 mAh g À1 , respectively. In particular, the charge-discharge prole of the b-GeO 2 /graphene composite exhibits a stable voltage plateau at about 0.2 V for lithium insertion at the high rate of 5 C (as shown in Fig. S4 †), which is much higher than the metallic lithium potential. In contrast, the voltage of b-GeO 2 and commercial GeO 2 presents a quick drop towards 0 V during lithium uptake, which may result in the formation of metallic lithium or even lithium dendrites on the surface of the anode and therefore results in safety issues. As mentioned above, the excellent rate capability of the GeO 2 /graphene composite benets from its uniformly dispersed mesoporous structure, in which the GeO 2 particles are homogeneously distributed on or within the conductive graphene sheets, which also shortens the Li + diffusion pathways. The open structure is favorable for fast transport of Li + and gives rise to high rate performance.

Conclusions
In summary, a 3D mesoporous GeO 2 material with a unique morphology was designed and developed via a facile and one-step synthesis process. The obtained mesoporous GeO 2 particles present a bean-like morphology (b-GeO 2 ), and display a good electrochemical charge-discharge cycling performance and rate capability for the lithiation/de-lithiation process. When graphene was employed via a spray drying method, the electrochemical performance of the b-GeO 2 /graphene composite was further signicantly improved. The b-GeO 2 /graphene electrode gives a higher de-lithiation capacity of 1021 mAh g À1 at 0.2 C and the capacity retention is measured to be as high as 94.3% aer 200 cycles. Furthermore, the composite displays an excellent rate performance even up to a charge-discharge rate of 5 C, where a reversible capacity of 730 mAh g À1 is obtained. However, the main drawback of GeO 2 , similar to the well-known SnO 2 -based anode materials, is its high irreversible capacity during the rst lithiation due to the irreversible formation of Li 2 O. Nevertheless, there is potential to match these anode materials with Li-rich cathode materials for the application in lithium-ion full cells, in which the excessive Li source from the cathode material may compensate the large Li loss of the anode in the rst cycle or with sacricial metallic Li added to the anode mix. The excellent electrochemical cycling performance, a potential application in lithium-ion full cells and in particular the novel and facile synthesis route to a 3D mesoporous material make the GeO 2 composite an interesting anode material for lithium-ion batteries.  Table 1 Representative values for the de-lithiation capacity (1 st , 2 nd , and 201 st cycle), coulombic efficiency (1 st , 2 nd , and 201 st cycle) and capacity retention (201 st cycle; related to the 2 nd cycle) of commercial GeO 2 , b-GeO 2 and b-GeO 2 /graphene anode materials. Data correspond to the constant current cycling performance in Fig. 6a Cycle b-GeO 2 /graphene b-GeO 2