Immobilization of cobalt oxide nanoparticles on porous nitrogen-doped carbon as electrocatalyst for oxygen evolution

Shusheng Li, Rui Kuang, Xiangzheng Kong, Xiaoli Zhu,*, Xubao Jiang

1 College of Chemistry and Chemical Engineering, University of Jinan, Jinan 250022, China

2 College of Traffic Civil Engineering, Shandong Jiaotong University, Jinan 250357, China

Keywords:Porous polyurea N-doped carbon-Co3O4 hybrid Oxygen evolution Catalyst Electrolysis Electrochemistry

ABSTRACT Highly efficient and robust electrocatalysts have been in urgent demand for oxygen evolution reaction(OER).For this purpose, high-cost carbon materials, such as graphene and carbon nanotubes, have been used as supports to metal oxides to enhance their catalytic activity.We report here a new Co3O4-based catalyst with nitrogen-doped porous carbon material as the support, prepared by pyrolysis of porous polyurea (PU) with Co(NO3)2 immobilized on its surface.To this end, PU was first synthesized, without any additive, through a very simple one-step precipitation polymerization of toluene diisocyanate in a binary mixture of H2O-acetone at room temperature.By immersing PU in an aqueous solution of Co(NO3)2 at room temperature,a cobalt coordinated polymer composite,Co(NO3)2/PU,was obtained,which was heated at 500 °C in air for 2 h to get a hybrid, Co3O4/NC, consisting of Co3O4 nanocrystals and sp2-hybridized N-doped carbon.Using this Co3O4/NC as a catalyst in OER, a current density of 10 mA?cm-2 was readily achieved with a low overpotential of 293 mV with a Tafel slope of 87 mV?dec-1,a high catalytic activity.This high performance was well retained after 1000 recycled uses,demonstrating its good durability.This work provides therefore a facile yet simple pathway to fabrication of a new transition metal oxides-based N-doped carbon catalyst for OER with high performance.

1.Introduction

With the constant increase in global energy demand and the growing environmental pollution,the development of consecutive,eco-friendly, and cost-effective energy resources becomes an urgent issue [1–3].Water splitting is considered as one of the critical solutions, because the strategy offers mass-production of high-purity hydrogen and oxygen, and is a clean alternative in comparison to fossil fuels [1,4–7].However, this process is seriously limited by the sluggish kinetics of oxygen evolution reaction(OER) owing to high overpotentials and high energy requirement by consequence [8,9].Traditional noble metal catalysts, including Ir, Ru or their oxides and alloys, have been identified as the most active OER catalysts[10–12].However,the high cost and low availability make their commercial applications largely limited[8,13,14].Thus, some non-noble metals have been used for OER catalysts,in which Co has been shown to be a promising candidate[15–18].With regard to the carbon materials used as the supports for these catalysts, graphene [19], carbon nanotubes [15], and mesoporous nanocarbon [2,16] are commonly studied, which are featured by their sophisticated preparations and high-cost.Therefore, developing inexpensive and efficient alternatives, without compromising the high performance currently reached with oxides and alloys, is of great interest, and has been a great challenge in OER [14,20–22].It is to note that polyurea (PU) is a class of nitrogen-rich polymers, often synthesized by step-growth polymerization of diisocyanates with diamines or through their reaction with water [23].Porous PU materials were easily synthesized this way by reacting toluene diisocyanate with H2O or diamine [24–27] and used for dye and metal ion removals and enzyme immobilization [28–31].This synthesis process is of the following features: (1) Low-cost monomers and mild synthesis conditions, which makes PU production easy on large scale; (2)Controllable cross-link density through the choice of multifunctional isocyanate or amine monomers, leading to designable porous structure;(3)High specific surface area,which is beneficial to mass transfer and molecular diffusion.

We report here a new type of catalyst for OER,in which porous PU was immersed in an aqueous solution of Co(NO3)2to get PU coordinated with Co2+, namely Co(NO3)2/PU.By heating it at 500 °C in air for 2 h, a composite material, Co3O4/NC consisting of Co3O4nano-crystallites and sp2-hybridized N-doped graphite-like carbon, was then obtained.The hybrid Co3O4/NC was used as OER catalyst, and the results demonstrate that its electrocatalytic property is superior to most of the reported Co3O4-based catalysts.To the best of our knowledge, there has no reported study to use PU as matrix materials for OER catalysts.Therefore,this study provides a facile strategy and opens a new direction to developing transition metal oxide nanoparticles (NPs) based catalysts for OER with high efficiency and high durability.

2.Experimental

2.1.Synthesis of PU, Co(NO3)2/PU and Co3O4/NC

PU was synthesized according to a reported process [24,28].Typically, a mixture of H2O-acetone (90 g) with a mass ratio of 3/7 was located into a round bottom flask at room temperature(25 °C).Under stirring at 300 r?min-1, toluene diisocyanate (10 g,TDI, Beijing Keju Chemical Materials Co.Ltd., industrial grade)was added dropwise in the flask at 20 ml?h-1.The polymerization was allowed to continue for 2 h after TDI addition completed, followed by filtration to separate the solid, rinsing twice by acetone,and dried at 70 °C to get a white PU powder.

To get PU coordinated with Co2+,a known mass(0.05–0.20 g)of Co(NO3)2?6H2O (AR, Aladdin) was dissolved in water (6 ml).PU(0.10 g) was added to get the aqueous mixtures with Co(NO3)2/PU mass ratios at 0.5, 1, and 2.The mixture was stirred at room temperature for 3 h, solid Co(NO3)2/PU composite was obtained by water removal through water evaporation and dried at 60 °C.

The composite Co(NO3)2/PU thus obtained was transferred into a tube furnace (BTF-12000-5, Anhui Best Equipment Co.Ltd.) and heated up to the desired temperature between 300 °C to 600 °C at a heating rate of 5 °C?min-1, and kept at that temperature for 2 h in air,followed by cooling down to room temperature at a ramp of 2 °C?min-1to get the hybrid catalyst Co3O4/NC.

2.2.Structure characterization of Co(NO3)2/PU and Co3O4/NC

PU and the composite Co(NO3)2/PU were subjected to Fourier transform infrared analysis(FTIR, TENSOR 27,Bruker Optiks).Porous morphology of the catalyst (Co3O4/NC), its precursor before pyrolysis and that of PU support were observed by field-emission scanning electron microscope (SEM, FEG250, FEI QUANTA) and by high resolution transmission electron microscopy (TEM, H-8100, Hitachi).Energy dispersive X-Ray spectroscopy (EDX) data were also collected along with SEM for selected samples.Their porous property was determined through Hg intrusion (AutoPore IV 9500, Micromeritics).X-ray photoelectron analysis was also done for Co3O4/NC (XPS, PHI 5000 Versa Probe, ULVAC-PHI).Powder Xray diffraction(XRD)test was done using a diffractometer(SMART APEX CCD,Bruker Optiks,Cu-Kα radiation,λ=0.15418 nm).Pyrolysis of Co(NO3)2/PU was followed by Thermogravimetry analysis(TGA,SDTA851,Mettler)for selective samples,from room temperature to 600 °C at a rate of 10 °C?min-1.

2.3.Electrode preparation

A glassy carbon electrode(Φ =4 mm,S =0.1256 cm2)was polished successively with gold sand-paper and Al2O3slurry,followed by rinsing with distilled water.Co3O4/NC (4 mg), Nafion (30 μl),water (0.72 ml) and ethanol (0.25 ml) were mixed up and sonicated for 15 min to get a homogeneous dispersion of Co3O4/NC in the mixture of water–ethanol (concentration, 4 mg?ml-1).The dispersion was used to make a coating on the freshly polished glassy carbon electrode with a catalyst loading of 0.22 mg?cm-2.The coated electrode was dried in air at room temperature.

2.4.Electrochemical measurement

Electrochemical measurement was carried out with a threeelectrode system on a KOH solution (1.0 mol?L-1) on a CHI660B electrochemical workstation (Shanghai Chenhua Limited, China)at room temperature.Hg/HgO, Pt wire (5 mm × 5 mm × 0.1 m m), and a glass carbon disk (4 mm in diameter) were used as the reference electrode, counter electrode and the working electrode,respectively.The linear sweep voltammogram (LSV) of the OER activity was scanned in the range of 0 to 1.0 V with a scan rate of 5 mV?s-1.Electrochemical impedance spectroscopy (EIS) analysis was conducted in KOH solution(1.0 mol?L-1)between 0.01 and 105Hz with an amplitude of 5 mV.The measured potentials versus Hg/HgO were converted to the reversible hydrogen electrode(RHE)scale according to the Nernst equation, Eq.(1):

whereEHg/HgOis the potential experimentally measured against Hg/HgO reference.Overpotential(η)of OER was calculated by Eq.(2):

Electrochemical double layer capacitance (Cdl) at the solid/liquid interface, proportional to the electrochemical active surface area, was determined by cyclic voltammogram (CV) in non-Faradic capacitance current range of 1.35–1.45 V versus RHE.The capacitive current changed linearly with scan rate in good agreement with Eq.(3) here below:

where,Jis the density of the capacitive current (mA?cm-2); v the scan rate (mV?s-1); andCdlthe slope of the straight line of the capacitive current (J) versus the scan rate (v) [32].

The turnover frequency(TOF)for each active site was calculated by Eq.(4) [33]:

where,Ais the geometric area of the testing electrode;the number 4 is the mole of electrons consumed for evolving one mole of O2;Fthe Faradic constant (96,485 C?mol-1);J, the current density,obtained at defined overpotential of the electrochemical process in KOH solution (1.0 mol?L-1); andm, the number of active sites(mol, equal toA× Γ0), is extracted from the linear relationship between the oxidation peak currents and scan rates by Eq.(5):

where,nis the number of electrons transferred;Γ0the surface concentration of active sites(mol?cm-2);RandTthe ideal gas constant and the absolute temperature, respectively.

3.Results and Discussion

3.1.Synthesis and characterization of Co3O4/NC

The hybrid catalyst of Co3O4on N-doped carbon composites,Co3O4/NC, was prepared in three steps.In the first step, porous PU was prepared by a very easy one-step precipitation polymerization of TDI in a binary solvent of H2O-acetone without any additive at room temperature, yielding the polymer chains consisted of methyl phenylene(from TDI)and carbamido(urea)(Fig.S1 in Supplementary Material).PU as-prepared was characterized to have a specific surface area of 162 m2?g-1and a pore volume of 5.65 cm3-?g-1with a wide pore size distribution.Its porous morphology is given in Fig.1(a) by field-emission SEM.Details of the synthesis and characterization are easily available [24,28] and not given here.

In the second step,PU was immersed in an aqueous solution of Co(NO3)2to have its N and O atoms coordinated with Co2+,leading to Co2+immobilized on PU surface, Co(NO3)2/PU, in a similar way to the complex coordination with Cu2+and Pd2+[25,34,35].Generally,4-or 6-coordination can be formed with Co2+at the center.In the present case for Co2+immobilization on PU surface, it is very likely that a multifunctional complexation of 4-coordination is occurring due to the rigid structure of PU chains,i.e.one Co2+coordinated with four oxygen atoms,four nitrogen atoms,or their combination, as depicted in Fig.S2.This Co2+immobilization onto PU was easily observed by a distinct color change from white (PU)to light pink for Co(NO3)2/PU.The SEM photo of the composite Co(NO3)2/PU is given to show that the porous morphology of PU was not changed after Co2+immobilization(Fig.1(b))and featured by a continuous and interconnected fibrous network.PU and the composite Co(NO3)2/PU were subjected to FTIR analysis (Fig.S3).A strong absorbance at 3296 cm-1,owing to N—H stretching vibration, was observed on the PU spectrum, along with its typical stretching vibration of C=O at 1653 cm-1, that of C—N at 1212 cm-1, and the plane bending vibration of N—H at 1592 cm-1,all as expected[24,25].In comparison to PU,the absorbance peak of N-H stretching became very broad in Co(NO3)2/PU,owing to most likely partial coordinating of the N atoms with Co2+.In addition, a new peak at 3162 cm-1appeared and was assigned to the stretching vibration of N—H bonds coordinated with Co2+.In accordance, besides the stretching vibration of C=O at 1653 cm-1in PU,a new absorbance peak at 1604 cm-1appeared in Co(NO3)2/PU.All these red-shifts of the peaks indicate a decrease in electron density around N—H and C=O bonds because of Co2+coordination with N and O atoms of PU in Co(NO3)2/PU.

The composite Co(NO3)2/PU was heated at 500 °C in air for 2 h for pyrolysis, by which the PU component in the composite was decomposed into a porous N-doped carbon matrix, and Co(NO3)2oxidized into Co3O4nano-crystallites, forming the desired hybrid material Co3O4/NC.SEM examination showed that Co3O4/NC, prepared with Co(NO3)2/PU mass ratio of 1, was also porous(Fig.1(c)), similar to that in the composite Co(NO3)2/PU before pyrolysis.Hg intrusion test indicated that Co3O4/NC was full of macropores and mesopores with a specific surface area of 44.50 m2?g-1, a pore volume of 1.08 cm3?g-1, and a porosity of 44.60% (Table S1, Fig.S4).The porous parameters of Co3O4/NC derived from varied Co(NO3)2/PU mass ratio and at different pyrolysis temperature were also listed in Table S1,showing a decreased specific surface area with increased mass ratio of Co(NO3)2/PU and the pyrolysis temperature, which must have a significant impact on the electrocatalyst.Abundant small nanoparticles (NPs) were uniformly embedded in the matrix, with their typical size of 20 nm, as seen from high resolution TEM picture (Fig.1(d)).EDX showed that C, N, O and Co elements were clearly present (Fig.2(a)),and evenly distributed on PU surface(Fig.S5).A rough estimation on the composition of Co3O4/NC, obtained from Co(NO3)2/PU mass ratio of 1, gave 56.79% of C, 12.35% of N, 17.30 % of O and 13.56% of Co by their atom numbers (Table S2), along with the EDX results of other samples derived from varied Co(NO3)2/PU mass ratio.With increase in this mass ratio, Co content increased in accordance,this may be another important factor for electrocatalyst activity.XRD test revealed the presence of the peaks at 31.33°, 36.85°, 44.77°, 59.32° and 65.23° (Fig.2(b)), which were assigned to crystal planes of the types (2 2 0), (3 1 1), (4 0 0),(5 1 1), and (4 4 0) in Co3O4(JCPDF No.42-1467), respectively[36], confirming that the NPs embedded in the matrix were actually Co3O4nanocrystals.Similar peaks were observed in XRD spectra of Co3O4/NC derived from Co(NO3)2/PU mass ratio of 0.5 and 2,but not in the spectra of Co3O4/NC(from Co(NO3)2/PU mass ratio of 1)pyrolized at 300°C(Fig.S6),indicating that Co(NO3)2was hardly oxidized to Co3O4under this low temperature.TEM images clearly revealed that the NPs in black spots were closely anchored on the matrix(Fig.1(d)),the lattice fringe space of 0.243 nm was in good accordance with the(3 1 1)plane of the spinel Co3O4phase[37,38].

Fig.1. SEM photos of PU prior to Co(NO3)2 deposition(a),that of Co(NO3)2/PU with mass ratio 1(b),that of Co3O4/NC after pyrolysis at 500°C(c);TEM image of Co3O4/NC(d)and the corresponding high resolution image (inset).

Fig.2. EDX spectra of Co3O4/NC obtained from Co(NO3)2/PU (1/1 by mass) pyrolysis at 500 °C (a), and its XRD spectra (b) before (up) and after stability test of 1000 cycles(down).

TGA test for the pyrolysis of Co(NO3)2/PU demonstrates that,besides the very first weight loss (～10.91%) around 173 °C(Fig.S7), likely due to the moisture loss and partial dehydration of Co(NO3)2?6H2O, there appeared two losses, one between 173 °C and 221 °C owing to formations of Co(OH)2and CoOx(including Co3O4),in agreement with reported decomposition pattern of the salt [39–41].The last and the most distinct weight loss(～45%) from 221 to 550 °C was mainly caused by PU decomposition, occurring from 275 °C to 336 °C [24].Nevertheless, a small portion of this loss was also contributed by Co3O4formation,which is occurring up to 550°C or 620°C[41].The same crystal structure of Co3O4was obtained regardless of the thermal decomposition temperature, though the size and morphology were slightly changed [40].

XPS was also done for Co3O4/NC hybrid obtained from Co(NO3)2/PU mass ratio of 1.As expected, all the four elements (Co,C, N, O) were detected as seen from the survey spectrum(Fig.S8).To further illustrate their binding energy and valence state, the spectrum was deconvolved separately for each element(Fig.3), with all the peaks assigned and listed in Table S3.The deconvolved spectrum of Co 2p (Fig.3(a)) has six peaks at 782.3 and 798.5 eV, assigned to 2p3/2and 2p1/2of Co2+, at 780.5 and 796.4 eV,assigned to 2p3/2and 2p1/2of Co3+,as well as two satellite peaks(Sat)at 786.5 and 802.9 eV owing to Co2+[38].These results indicate the coexistence of Co2+and Co3+in the hybrid catalyst Co3O4/NC.From the areas of these peaks, the ratio of Co3+/Co2+was estimated to be about 2.0 [42,43], matching well with that expected in Co3O4.

The spectrum of O 1s was deconvolved into three peaks (Fig.3(b)), the one at 531.0 eV was believed to arise from the typical Co—O bonding in Co3O4, while the other two (532.1 and 533.4 eV) were commonly assigned to adsorbed water and hydroxyl groups(—OH)[38,44,45].The surface hydroxyl groups are reactive intermediates in this type of catalytic process, and have an important effect on OER reaction rate, generating a positive effect on the OER catalysis[46].C 1s spectrum was deconvolved into four peaks (Fig.3(c)).The one at lower energy side (284.4 eV) was assigned to sp2-hybridized graphitic C=C bond [47]; The following two (285.0 & 285.6 eV) were attributed to sp2C—N and sp3C—N bonding [47,48]; and the smallest peak at higher energy side at(288.5 eV) was often attributed to C=O bond [44].N 1s spectrum was also deconvolved, giving three peaks (Fig.3(d)): Co—N at 399.2 eV,N—H at 400.3 eV and C=N at 398.4 eV [49–51].All these results, together with those from XRD, SEM and TEM, confirmed that the Co3O4/NC hybrid was composed of Co3O4nanocrystals and sp2-hybridized N-doped carbon matrix.

3.2.OER performance of Co3O4/NC

In order to optimize the catalyst, the precursor was prepared using different Co(NO3)2/PU mass ratio (0.5, 1, 2), and each pyrolized at different temperature (300 °C, 400 °C, 500 °C and 600 °C), making 12 samples of Co3O4/NC hybrid in total.Their OER performance was examined through LSV at a scan rate of 5 mV?s-1.The results are displayed in Fig.S9,including those from pure Co3O4and its blend with N-doped carbon,i.e.pure PU pyrolized at 500 °C, and the corresponding data are listed in Table 1.

The data in Table 1 indicates that, OER activity of the catalyst Co3O4/NC was closely dependent on pyrolysis temperature and precursor composition,Co(NO3)2/PU.For instances,for the samples prepared with Co(NO3)2/PU mass ratio of 0.5,Tafel slope and overpotential, the two key parameters for a catalyst, were decreasing regularly with pyrolysis temperature from 300 °C up to 500 °C,where the lowest Tafel slope (89 mV?dec-1) and overpotential(309 mV)were observed.With further increased pyrolysis temperature to 600°C,both Tafel slope(103 mV?dec-1)and overpotential(342 mV)started to increase,i.e.the optimal pyrolysis temperature was 500 °C, below which the catalyst performance was improved with temperature increase, while above which the performance was declined.This effect of pyrolysis temperature was seen for all three sets of the samples prepared with different Co(NO3)2/PU mass ratio, which enabled one to conclude that the best pyrolysis temperature was 500°C.By comparing the samples prepared with different Co(NO3)2/PU mass ratios, it was quite evident that the samples prepared at Co(NO3)2/PU=1 delivered the best performance, regardless of the pyrolysis temperature used (Fig.4(a)).Therefore, the optimized preparation for the hybrid catalyst is to use Co(NO3)2/PU mass ratio of 1 for the precursor with pyrolysis at 500 °C (Fig.4(a)), and 293 mV was the lowest overpotential required to achieve the same current density.

Table 1Comparison of OER activity of Co3O4/NC catalysts prepared under different conditions

Fig.3. XPS spectra of Co 2p (a), O 1s (b), C 1s (c), and N 1s (d) of Co3O4/NC.

The difference in catalyst activity of Co(NO3)2/NC may be attributed to combined action of the porous structure (Table S1) and Co3O4loading in Co3O4/NC.With Co(NO3)2/PU mass ratio increased from 0.5 to 1 and to 2, the specific surface area decreased slightly from 46.23 m2?g-1to 44.50 and rapidly to 35.26 m2?g-1, and Co loading increased from 7.34% to 13.56% and to 21.86% in accordance(Table S2).It is obvious that,larger specific surface area will provide more active sites for catalysts.The lower Co3O4loading in Co3O4/NC will lead to less active site, giving the catalyst a lower OER activity.At the same time,it is easy to conceive that a too high Co3O4loading on a low surface would make Co3O4nanocrystals aggregated, while too low Co3O4loading (owing to low Co(NO3)2/PU ratio in the precursor) would lead to lower number of active sites, reducing therefore the catalytic activity of the catalyst in both cases.This demonstrates that there is an optimal ratio for the precursor Co(NO3)2/PU, at which the specific surface area of the N-doped carbon matrix after pyrolysis is well matched with the amount of Co3O4nanocrystalsin-situformed.For this present system, optimized Co(NO3)2/PU ratio is unity.

Fig.4. OER performance of Co3O4/NC prepared with different Co(NO3)2/PU mass ratio and pyrolized at 500 °C, with their corresponding Tafel slope given as inset (a);Dependence of capacitive current at 1.40 V vs.RHE (b) (All tested in 1.0 mol?L-1 KOH).

To demonstrate the effectiveness and benefit of the hybrid catalyst Co3O4/NC,parallel OER tests through LSV were done for pure Co3O4alone,obtained by calcinations of Co(NO3)2at 500°C,and for its simple blend with N-doped carbon matrix obtained by pyrolysis of pure PU at 500°C,at Co3O4/NC mass ratio of 1(Fig.S9(d)).The results confirmed that, to achieve the same current density of 10 mA?cm-2, much higher overpotentials, 420 mV for pure Co(NO3)2and 456 mV for the blend, were required (Table 1, the two lines at bottom).It is obvious that the performance of Co3O4alone,and its simple blend with N-doped carbon matrix was much poorer in comparison to all the hybrid samples prepared 500 °C and those obtained from Co(NO3)2/PU=1.The significant enhancement of OER catalytic performance in Co3O4/NC must arise from the porous structure and high conductivity of the NC support.It is to point out that, in comparison to reported cobalt-based OER catalysts,this OER performance(293 mV at 10 mA?cm-2)is comparable to or significantly better off.To reach the same current density (10 mA?cm-2), an overpotential of 325 mV was needed for Ni-Co binary oxide catalyst [52], 330 mV for CoP-CNT [36],370 mV for Zn/Co hydroxyl sulfate [53], 320 mV for Ni-Co metal oxide [54], and finally 298 mV for Co3O4nanocubes supported on N-doped graphene[55].In two other reports under similar catalysis conditions, lower overpotentials were effectively achieved: in one study, a slightly better result,i.e.an overpotential of 280 mV at a current density of 10 mA?cm-2, was obtained from Co3O4nanocubes supported on N-doped graphene[56];in another study,an overpotential of 235 mV was achieved for an optimized catalyst of Co3O4/NC nanosheet[57].However,the preparation of the catalyst was truly sophisticated, with multi-steps of long lasting at high temperature.For instance, for the N-doped graphene supported Co3O4nanocubes [56], graphene oxide (GO) was first prepared through Hummer’s Method, by stirring the reaction mixture at 60 °C for 24 h.This GO was then reduced by aqueous solution of hydrazine hydrate under stirring for 12 h at 80 °C,followed by high speed centrifugation at 10,000 r?min-1and washing and drying at 80°C for 10 h.The reduced GO was N-doped with melamine under stirring for 24 h.The obtained N-doped-reduced GO was heated at 800°C for about 4 h under an argon atmosphere.After that, the product was used as the substrate for synthesizing the Co3O4/NC catalysts by dispersing it along with cobalt acetate in a binary solvent of ethanol–water and heated in a Teflon coated stainless steel autoclave at 120°C for 12 h.The desired catalyst was finally obtained after drying at 80°C in an oven for 10 h.The overpotential achieved was slightly lower, 280 mV versus 293 mV at the same current density of 10 mA?cm-2.In another more recent study, where an overpotential of 235 mV was indeed achieved for a catalyst of type Co3O4/NC nanosheet [57].For this catalyst,a metal organic framework (MOF) with laminar architecture and functionalized with amine group (CoMOF-NH2) was required,which was firstly synthesized byin-situtransformation of Co2(-OH)3Cl array at 100 °C for 24 h (this array was synthesized ahead in a PTFE autoclave by treating CoCl2?6H2O in the presence of cetyltrimethylammonium bromide in a mixed solution of deionized water and methanol at 180 °C for 24 h at least).The product was rinsed with deionized water several times, and placed in a 60 °C oven for 6 h.The outcome laminar was pyrolized under an argon atmosphere at high temperature (300–600 °C) for at least 4 h.In contrast to these reports, it is easy to see that the preparation of this present catalyst is significantly facile and simple, with low-cost materials and easy operation, while the performance is very similar to or better than the above mentioned catalysts, a great merit of this work.

Pyrolysis temperature made also a difference to the catalytic activity of the catalyst because of its important effect on the morphologies and chemical composition of the hybrid materials.To illustrate this effect, porous property of the hybrid catalysts, prepared with Co(NO3)2/PU=1 and pyrolized at different temperature, was examined using Hg intrusion test (Fig.S4) and the morphology observed under SEM(Fig.S10).It was easily seen that,for the samples pyrolized at 300°C to 500°C(Fig.S10,Fig.1(c)),the materials were highly porous with specific surface area of 68.32 m2?g-1and 44.50 m2?g-1(Table S1).In contrast,for the samples pyrolized at 600 °C, this porous structure practically disappeared, with only Co2O3NPs remained (Fig.S10(c) & (d)), giving a lower specific surface area of 18.37 m2?g-1(Table S1).Not only the pores of the size around 200 nm completely vanished, those around 11 μm were also largely reduced (Fig.S4(d)).The highly porous structure, observed in the samples pyrolyzed at 500 °C(Fig.1(c))or lower in the carbon matrix(Fig.S10(a)&(b)),was collapsed at this high temperature (Fig.S10(c) & (d)).In contrast, for the simple blend of Co3O4/NC at a mass ratio of 1,with Co3O4from Co(NO3)2calcination and NC from PU pyrolysis, both at 500 °C(Fig.S10(e),(f)),the Co3O4nanoparticles were just stuck to the surface of NC,and aggregated together,which is significantly different from the catalyst prepared by pyrolysis of the catalyst precursor,Co(NO3)2/PU,where Co3O4nanoparticles were homogeneously distributed in the hybrid catalyst,as seen in Fig.1 and in Fig.S10.This must be the very reason for the observed low OER activity from the simple blend Co(NO3)2/NC.

To gain further insight into the superior OER activity of Co3O4/NC,Cdlof the hybrid Co3O4/NC was compared to that of Co3O4NPs using CV tests to estimate their electrochemical active surface area.CV values were collected in the non-Faradic capacitance current region of 1.35–1.45 Vvs.RHE (Fig.S11(a) & (b)).The linear change of the capacitive currents (J) versus scan rate is given in Fig.4(b).The obtainedCdlfor Co3O4/NC (69.8 μF?cm-2)was about 2.9 times of that for Co3O4NPs(24.2 μF?cm-2),suggesting that Co3O4/NC have a much higher electrochemical active surface and more accessible active sites for OER catalysis.Notably,during the OER process, the surface of the catalyst was easy to restructure and to form a hydroxyl oxide shell, which must have played a major role in the catalytic process.

Fig.5. EIS spectra of Co3O4/NC prepared with Co(NO3)2/PU mass ratio of 1 and pyrolized at different temperature (a); Chronoamperometric response for Co3O4/NC at a constant overpotential of 450 mV (b) and the LSV curves recorded for Co3O4/NC (b, inset) pyrolized at 500 °C before and after 1000 voltammetry cycles.

EIS technique was often employed to gain deep insight into the kinetics of OER process, and lower internal and charge-transfer impedances mean high electron conductivity [58].EIS analysis was also carried out for OER kinetic process for the samples prepared with Co(NO3)2/PU mass ratio of 1 (Fig.5(a)).As observed,the results indicate a sharp decrease in resistance with pyrolysis temperature change from 300 °C to 500 °C, and rebounded up thereafter, leading to the lowest resistance for the sample pyrolized at 500 °C, in good agreement with the data in Table 1.This was attributed to the enlarged contact area of the Co3O4nanocrystals with the electrolytes and therefore a reduced energy barrier[2].The resistance rebound for the catalyst prepared at 600 °C(Fig.5(a)) must be caused by the disappearance of the porous structure shown in Fig.S10(c), yielding reduced contact area for the catalyst with the electrolytes, and so was the catalytic activity by consequence.

To further illustrate the intrinsic OER behavior of Co3O4/NC,TOF was estimated using Eq.(4) at a constant overpotential 293 mV,using the optimized catalyst in Table 1,i.e.Co3O4/NC prepared with Co(NO3)2/PU mass ratio at 1 and pyrolized at 500 °C.To this end,the number (m=A× Γ0, Eq.(4)) of surface catalytic sites was required, and obtained from the slope of the linear relationship,defined by Eq.(5), between the oxidation peak value and the scan rate (Fig.S12(a) or (b)).This slope was 0.1455 for Co3O4/NC, and 0.0675 for pure Co3O4(by Co(NO3)2calcination at 500 °C,Fig.S12(c)).The corresponding TOF value was 0.348 s-1at overpotential of 293 mV for Co3O4/NC, and 0.093 s-1for pure Co3O4(Fig.S12(d)).The TOF value of the hybrid catalyst Co3O4/NC was always distinctly higher than that of pure Co3O4, once the overpotential was above 200 mV, indicating an obviously increased catalytic activity of the hybrid catalyst.

All these results suggest a synergistic effect between Co3O4NPs and their NC support.The existence of Co atoms in higher oxidation state (Co3+) improves the electrophilicity of adsorbed oxygen(Co3+–O), and therefore enables the formation of Co3+–OOH through nucleophilic attack by OH.This is believed to stimulate the deprotonation of OOH species through the inductive effect of electron-withdrawing to harvest O2[59].Moreover, thein-situformed N-doped carbon is also very important to increase the conductivity of the hybrid, to reduce the particle size of oxides and to anchor them to form intimate interparticle interface.Generally speaking, a decrease in resistance indicates an enhanced ability of charge transport between Co3O4nanocrystals and the porous matrix of N-doped carbon, and therefore a better performance of the electrocatalyst.

The durability is also a key parameter for OER catalyst evaluations.The chronoamperometric response demonstrates that the hybrid catalyst Co3O4/NC showed a stable current within 25 h(Fig.5(b)).This property was also evaluated through LSV test after cyclic voltammetric sweeps between 0.00 to 0.70 V(vs.Hg/HgO)at 100 mV?s-1.The LSV curve after 1000 cycles retained practically the same as that in the initial test (Fig.5(b)), and the morphology of Co3O4/NC remained unchanged as ascertained by SEM and TEM(Fig.S13).The XRD spectra of Co3O4/NC are exactly the same before and after the reaction(Fig.2(b))for all characteristic peaks,and the elements(Co,C,N,O)atoms ratio of Co3O4/NC from EDX(Table S4)are practically the same.These results demonstrated that this catalyst is of high reusability, thanks to a strong coordination and electrostatic interaction between Co3O4NPs and the N-doped carbon matrix, efficiently protecting Co3O4NPs from dissolution in alkaline solution.

4.Conclusions

Polyurea (PU) was prepared through a very simple one-step precipitation polymerization of TDI in a binary solvent of H2Oacetone.Using this PU as support,Co3O4was immobilized by coordination adsorption to get a composite Co3O4/PU, followed by pyrolysis at 500 °C to get a hybrid electrocatalyst for OER, Co3O4/NC, consisting of Co3O4NPs and sp2-hybridized N-doped porous carbon.Excessive growth of Co3O4NPs was prevented owing to the coordinated immobilization of Co3O4nano-crystallites on the matrix.Among different catalysts prepared under different conditions, Co3O4/NC prepared with Co(NO3)2/PU at a mass ratio of 1 and pyrolyzed at 500 °C exhibited the best OER catalytic performance: a current density of 10 mA?cm-2was achieved with a low overpotential of 293 mV with a Tafel slope of 87 mV?dec-1.In consideration of the simplicity and easiness of the synthesis,this high activity obtained at such low overpotential demonstrates that this catalyst Co3O4/NC is a promising candidate for OER.The present strategy provides a new pathway to preparation of new OER catalysts of high 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.

Acknowledgements

This work was financially supported by the Natural Science Foundation of Shandong Province, China (grant numbers ZR2021MB112, ZR2019MB031, ZR2020QB065); Natural Science Foundation of Guangdong Province, China (grant number 2020A1515110374); and by Science and Technology Bureau of Jinan City, Shandong Province, China (2021GXRC105).

Supplementary Material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.cjche.2021.12.021.