Rapid solvent-evaporation strategy for three-dimensional cobalt-based complex hierarchical architectures as catalysts for water oxidation

It is a challenging task to seek a highly-efficient electrocatalyst for oxygen evolution reaction (OER) of water splitting. Non-noble Co-based nanomaterials are considered as earth-abundant and effective catalysts to lower overpotential and increase polarization current density of OER. In this work, we reported, for the first time, a “rapid solvent-evaporation” strategy for the synthesis of three-dimensional (3D) cobalt complex hierarchical architectures constructed by two-dimensional (2D) nanosheets. The 3D structured cobalt complexes have excellent performances in catalyzing OER with lower onset potential, overpotential, Tafel slope and better stability than commercial IrO2. Superior electrochemical performances would be beneficial from the unique 3D structure. This extremely simple method for 3D Co complex with good OER activities makes the complex be promising commercial OER catalyst to replace earth-rare and expensive IrO2.


Synthesis of SE-100 and SE-150.
Comparatively, the bottles with the mixing solution of urea and cobalt acetate were put in drying oven at 100 °C and 150 °C for 1.5 h for slow evaporation. The obtained pink products were washed alternately with deionized water and ethanol, dried overnight at 80 °C and designed as SE-100 and SE-150, respectively.
Synthesis of cc-air and cc-n 2 . CC-air and CC-N 2 were prepared through heat annealing treatment of 3D CC-G at the temperature of 600 °C for 3 h in the air or N 2 atmosphere with a heating rate of 2 °C/min. characterizations. Scanning electron microscopy (SEM) was performed on Hitachi S-4800 at 10 kV.
Transmission electron microscopy (TEM) images were obtained by using a JEOL JEM-2100 transmission electron microscope operating at 200 kV. Powder X-ray diffraction (XRD) patterns were collected by using a Bruker D8 ADVANCE diffractometer with CuKα radiation (λ = 1.5418 Å). Infrared spectra were obtained on Fourier transform infrared spectroscopy (FT-IR, Nicolet 6700, Thermo Company). Thermogravimetric analysis of the powders were preformed in air on Pyris Diamond TG/ DTA (Perkin-Elmer). The flow rate of air was set at 120 mL/min and the temperature was increased from 30 to 800 °C at a rate of 10 °C/min. X-Ray photoelectron spectroscopy (XPS) was collected on an ESCALab MKII X-ray photoelectron spectrometer, using non-monochromatized Al Kα X-ray as excitation source. The binding energies were corrected for specimen charging by calibrating the C 1s peak to 284.6 eV. Element analysis (EA) was characterized by element analyzer (German Heraeus Company, CHN-0-Rapid). The percentage composition of cobalt was analyzed by Inductively Coupled Plasma (America PE Company, Optima 5300 DV). electrochemical measurements. The electrochemical measurements were performed in a conventional three electrode cell using a CHI760D (Shanghai Chenhua, China) electrochemical workstation with catalyst-coated glassy carbon (GC) as the working electrode, a Pt wire as the counter electrode and a saturated Ag/AgCl electrode as the reference electrode. The catalyst ink was prepared by blending 5 mg of the catalyst with 50 μL of Nafion solution (4 wt %) and 950 μL of water/isopropanol solution (3:1) via sonication. An amount of 5 μL of the dispersion was transferred onto the GC electrode with a catalyst loading of about 0.18 mg/cm 2 . Then, the prepared catalyst electrode was dried at room temperature. With this electrode as the working electrode, electrochemical measurements were conducted in 1 M KOH solution. Polarization curves were obtained at a scan rate of 1 mV/s. Accelerated degradation measurement was conducted for 3000 cyclic voltammetry (CV) cycles at a scan rate of 50 mV/s. In all measurements, the reference electrode was calibrated with respect to reversible hydrogen electrode (RHE). All Polarization data were without iR-corrected.

Results and Discussion
Figures 1a and S1 briefly depict the facile synthesis process of the 3D Co complex. Figure 1b,c show SEM images of the product, revealing that the sample consists of plate-like 3D superstructures. The 3D plate is assembled by a great number of 2D nanosheets, which are curled and perpendicular to the plate surface, resulting in a lot of porous channels. TEM images in Fig. 1d,e also confirm the 3D porous superstructure of the product constructed by thin 2D nanosheets. Figure S2 shows the adsorption/desorption isothermals and the pore size distribution of the obtained Co-complex architecture, which clearly confirm the mesoporous structure with the pore sizes in the range of 3~12 nm. Since the solvent evaporation process is very quick and there is no enough time for slow crystallization in solution, no obvious diffraction peaks can be detected in the XRD patterns (Fig. 2a), suggesting the amorphous nature of the obtained Co product. Fourier transform infrared spectra (FTIR) is employed to analyze the information of functional groups. Figure 2b shows the FTIR spectra of the synthesized products and pure urea. It is obviously observed that both CC-B and CC-G have similar curves to urea, suggesting the remaining existence of urea molecules in the product, which would serve as ligands to coordinate with Co 2+ . Newly emerged peak at 2184.8 cm −1 indicates the coordination bonding between Co 2+ and ligands (urea). Compared to that of urea, stretching vibration of C-N in the product nearly disappears, suggesting possible interaction between Co 2+ and amino group that weakens the vibration of C-N. The specific peak at 1746.4 cm −1 in CC-G may result from coordination bond of Co-O by the hydrolysis of CC-B and water. Energy dispersive spectrum (EDS) in Fig. 2c further proves the co-existence of multiple elements including C, O, N and Co (Si from substrate and Au from gold sputtering), confirming the obtained product could be a Co-based complex with the organic ligand. Elemental mappings (Fig. 2d~h) verify that the product is made up of C, O, N and Co in accordance with the EDS analysis, and these elements are uniformly dispersed in the product, suggesting that the product obtained through the fast solvent evaporation process is not cobalt oxides or hydroxides. It can be concluded by combining the XRD, IR and EDS results that the as-prepared product would be amorphous Co complex through the coordination between Co 2+ and urea.
Since the OER mainly occurs on the surface of catalyst, it is necessary to study the surface properties with X-ray photoelectron spectroscopy (XPS). All of the binding energies in XPS were corrected for specimen charging by referencing to the C1s peak (set at 284.6 eV). Co, C, N and O were all detected from the overall XPS spectrum of the 3D Co complex in Fig. 3a. Figure 3b shows the fine-scanned C1s spectrum, in which the peaks of C1s locates at around 288.9 eV and 285.4 eV contributing to the carbonyl and C-N. Figure 3c shows that the sample has a obvious peak located at 399.9 eV, which should be attributed to the Co-N according to the NIST XPS data. The Co2p XPS spectrum is shown in Fig. 3d, in which the peaks of Co 2p 3/2 and Co2p 1/2 are at around 781.0 eV and 796.9 eV, respectively, with a ∆E 3/2-1/2 of 15.9 eV, indicating that the state of cobalt could be Co 2+ . The O1s XPS spectrum in Fig. 3e shows three peaks locating at 532.7 eV, 531.7 eV, and 531 eV, which indicate the existence www.nature.com/scientificreports www.nature.com/scientificreports/ carbonyl oxygen and hydroxy oxygen. Inductively coupled plasma spectroscopy (ICP) and elemental analyzer (EA) are regarded as efficient means to characterize the content of elements. The ICP measurement demonstrates that the mass fraction of Co in Co complex is about 42.75%, and EA measurement shows the approximate atomic ratio of C:N:H is 1:1:3. Combining the existing urea, hydroxy, and possible carbonate according to the XPS and FTIR data, the structural formula of the Co-based complex can be inferred as Co 2 (CN 2 H 4 O)(OH) 2 CO 3 , which has a Co content of 43.06% in accordance with the ICP and EA results. Figure 3f presents the thermogravimetric analysis (TGA) and differential scanning calorimeter (DSC) curves of the Co complex under air atmosphere. After the decomposition and the oxidation reaction of Co 2+ to Co 3+ , the sample has a remaining mass of 57.8% of initial mass, which is in accordance with the result of ICP and demonstrates the rationality of the proposed structural formula of the Co-based complex.
The formation of 3D Co 2 (CN 2 H 4 O)(OH) 2 CO 3 architecture goes through the following two steps: Firstly, the free Co ions react with urea and CO 3 2− produced by the decomposition of urea to form blue bulk Co complex with constant evaporation of solvent at 200 °C under open system, and secondly, the blue bulk Co complex can be transferred to green 3D Co complex architecture assembled by nanosheets through hydrolysis process by washing  www.nature.com/scientificreports www.nature.com/scientificreports/ the blue Co complex with deionized water. The first step is the key for the formation of the 3D Co 2 (CN 2 H 4 O) (OH) 2 CO 3 structure. The rapid solvent evaporation under high temperature prevents the total decomposition of the used urea. The remained urea can serve as ligand to coordinate with Co 2+ in the following procedures. At the same time, the decomposition of part of urea results in the formation of CO 3 2− to balance the positive charges. The Co complex tends to form dense bulk in the concentrated system, then the 2D nanosheets produced by hydrolysis directly grow on the blue bulk for final 2D nanosheets-assembled 3D architecture. So, it is highly necessary to set high temperature condition for the rapid evaporation of solvent for the formation of 3D architecture. Conversely, relatively slow solvent-evaporation experiments at 100 °C and 150 °C were also conducted. Figure S3 shows the SEM images of the obtained products at 100 °C (SE-100) and 150 °C (SE-150), which both present disordered distribution of bulk nanoplates. XRD patterns (Fig. S4a) confirm that they are also amorphous in nature and the corresponding XPS spectra (Fig. S4b) demonstrate that the state of cobalt are also Co 2+ . However, the most of urea, which serve as ligand to form complex in RSE system, gradually decomposes into NH 3 and CO 2 in the relatively long time under low temperatures. Therefore, the 2D nanosheets-assembled 3D Co 2 (CN 2 H 4 O) (OH) 2 CO 3 complex cann't be realized under low temperatures because of the absence of ligands.
The OER electrocatalytic activity were investigated in 1 M KOH aqueous solution using a standard three-electrode system. As shown in Fig. 4a, initial linear sweep voltammetry (LSV) curves recorded with the Co complex (CC) reveals a relatively lower overpotential of 360 mV at a current density of 10 mA/cm 2 than that of commercial IrO 2 (400 mV) under the same conditions. Meanwhile, to highlight the superior OER performance, Co 3 O 4 (CC-air) and Co 3 C x (CC-N 2 ) nanostructures were also synthesized by calcining the 3D Co complex under air or N 2 atmosphere. XRD patterns, IR, XPS spectra and SEM images shown in Figs S5 and S6 clearly demonstrate the complete transformation from Co complex to Co 3 O 4 and Co 3 C x . The polarization curves of Co 3 O 4 (CC-air) and Co 3 C x (CC-N 2 ) nanostructures are also displayed in Fig. 4a, whose onset potentials and overpotentials are higher than those of Co complex or IrO 2 . The Tafel slope derived from polarization (Fig. 4b) was used to evaluate OER kinetics. Co complex has an approximately equivalent Tafel slope to commercial IrO 2 , indicating the excellent OER activity at low overpotential. It is worth mentioning that the obtained Co 3 O 4 has a lowest Tafel slope, contributing to an even higher enhancement in OER activity at high potential. Worse OER performance of magnetic CoC x may be caused by the deletion of oxygen-containing functional groups under high temperature condition, which benefits the absorption of H 2 O molecules on the catalyst. LSV curves after 100 CV cyclings at a scan rate of 0.1 V·s −1 were utilized to compare the stability of 3D Co complex and IrO 2. Figure 4c shows the almost completely coincident LSV curves, indicating the excellent stability for Co complex. The corresponding SEM images, XRD pattern and IR spectrum of the catalyst after 100 OER cycles shown in Fig. S8 demonstrate that the microstructure and composition of Co complex are highly stable. In contrast, the LSV of IrO 2 after 100 www.nature.com/scientificreports www.nature.com/scientificreports/ CV cyclings has a big deviation. Figure S9a shows the LSV curves of 3D Co complex, IrO 2 and amorphous SE-100 after 1000 CV cyclings, which indicates the 3D Co complex has a remarkably superior performance for OER than IrO 2 or SE-100 in the long-term cycle testing, though it has a fine deviation compared to initial cycle that might result from possible fall off of catalyst on the surface of electrode. AC impedance spectra for Co complex and IrO 2 displayed in Fig. 4d indicate the low electron transfer impedance (R ct ) for Co complex . Impedance data demonstrate that charge transfer and ions diffusion are not advantages versus that of IrO 2 . It is well accepted that the electrochemically active surface area (ECSA) of the catalysts can be estimated from measurements of the electrochemical double-layer capacitance (C dl ) 12,14 . The double layer capacitance (C dl ), which can calculated by using cyclic voltammetric method, is proportional to ECSA. The bigger the effective electrochemical surface area, the more of the active sites. Figure 5a shows the CVs of the CC products, which reveals that the sample has an obvious increase in current density according to different scan rates. The calculated EDLC of CC is 12.25 mF/ cm 2 (Fig. 5b), superior to the reported OER electrocatalysts 12,14 . The result indicates that the number of electrochemically active sites for water oxidation significantly increases due to the special structure. Compared with the control samples, CC has a highest specific surface area of 142.9 m 2 /g (Table S1). It is believed that the initial OER process in alkaline conditions involves adsorbed OH and O species on the active sites of catalysts based on the following scheme 32 . Step (1) and (2) play an important role to determine the overall OER rate, while reactions (3), (4) for O 2 production are inversible and fast 33 . Therefore, the adsorption energy of OH − greatly affects the OER process. 2D nanosheets arrays with abundant hydrophilic functional groups have high specific surface area for the contact of catalyst and OH − or H 2 O. Meanwhile, amorphous structure supplied abundant defect sites for OER [34][35][36] . More importantly, the self-assembly of 2D nanosheets for stable 3D architecture that avoids the undesirable re-stack and condense of 2D nanosheets can remarkably improve the duration performance of catalyst for OER.

conclusions
In summary, we firstly proposed an extremely facile route called "rapid solvent-evaporation strategy" to prepare 3D Co complex architectures constructed by 2D nanosheet blocks as highly efficient catalyst of OER. The new-type Co-based catalyst exhibited excellent electrocatalytic activity of OER with a low onset potential and a lower overpotential of 360 mV at a current density of 10 mA·cm −2 than that of commercial IrO 2 . Abundant active sites and hydrophilic group of 2D nanosheets, amorphous and stable 3D structure, cooperatively achieved the superior catalyzing OER performance. It is expected that the 3D Co complex prepared by simple, green and low-cost method would be as promising OER catalyst for large-scale application to replace expensive and earth-rare IrO 2 . Meanwhile, it is believed that through the simple "rapid solvent-evaporation" strategy other novel 3D structures with great application potential could be obtained in the future.