Electrocatalytic performances of g-C3N4-LaNiO3 composite as bi-functional catalysts for lithium-oxygen batteries

A low cost and non-precious metal composite material g-C3N4-LaNiO3 (CNL) was synthesized as a bifunctional electrocatalyst for the air electrode of lithium-oxygen (Li-O2) batteries. The composition strategy changed the electron structure of LaNiO3 and g-C3N4, ensures high Ni3+/Ni2+ ratio and more absorbed hydroxyl on the surface of CNL that can promote the oxygen reduction reaction (ORR) and oxygen evolution reaction (OER). The composite catalyst presents higher activities than the individual components g-C3N4 and LaNiO3 for both ORR and OER. In non-aqueous Li-O2 batteries, CNL shows higher capacity, lower overpotentials and better cycling stability than XC-72 carbon and LaNiO3 catalysts. Our results suggest that CNL composite is a promising cathode catalyst for Li-O2 batteries.

electron-filled e g orbit, and therefore, provides the favorable bond energy of M (B site)-O bond for ORR and OER 28 . In addition, the catalytic activity of LaNiO 3 could be readily enhanced through increasing the Ni 3+ /Ni 2+ ratio by replacing part of Ni ions with Mg or Fe ions. g-C 3 N 4 has a large number of removable electronic between layers, and cavities encircling with some pyridine-like nitrogen 29 . So it has a high ability to complex with abundant metal ions and adjust its conjugation structure subsequently. Therefore combining g-C 3 N 4 with LaNiO 3 can theoretically improve the electron transfer and catalytic activity of the CNL composite. However, to the best of our knowledge, g-C 3 N 4 have never been composited with LaNiO 3 to explore the enhanced the catalytic activity for both ORR and OER.
In this work, we synthesized a composite catalyst g-C 3 N 4 -LaNiO 3 (CNL), and demonstrated that the CNL has significantly enhanced electrocatalytic activity for OER and ORR compared with individual component g-C 3 N 4 or LaNiO 3 in an alkaline electrolyte. Furthermore, CNL exhibited improved round-trip efficiency and cycling stability in Li-O 2 batteries. These results suggest that the CNL is a promising bi-functional catalyst for oxygen oxidation/evolution reactions.

Results and Discussion
Characterization of CNL catalyst. Figure 1 depicts the synthetic process of CNL catalyst and the molecular structure of the catalytic spots of CNL catalyst. Figure 2 displays the XRD patterns of pure g-C 3 N 4 , LaNiO 3 and the hybrid catalyst CNL with different weight percent of g-C 3 N 4 . It can be seen that the pure g-C 3 N 4 has two distinct peaks. The strong peak at 27.39°, corresponding to the (002) peak with in-planar distance of 0.326 nm (JCPDS 87-1526), is attributed to the long-range in-planar stacking of aromatic units 30 . Another peak at 13.08°, with a much weaker intensity, can be assigned to (100) diffraction peak 20 , which corresponds to a distance of 0.686 nm, belongs to tri-s-triazine units, due to the interlayer stacking. After the LaNiO 3 being composited with g-C 3 N 4 , the CNL hybrids show a coexistence phases of both g-C 3 N 4 and LaNiO 3 , although the characteristic peak of g-C 3 N 4 was not obvious for the 5 wt.% and 10 wt.% CNL hybrids.
For the further information of the materials microstructure, Fig. 3 shows the SEM images of g-C 3 N 4 , LaNiO 3 and 10 wt.% CNL. Pure g-C 3 N 4 shows an aggregated layered structure with smooth surface. The agglomerate size is several micron meters. LaNiO 3 particles are hundreds nanometers large, more than one order of magnitude smaller than g-C 3 N 4 . As to the image of 10 wt.% CNL, LaNiO 3 particles are deposited on the surface of g-C 3 N 4 .
In this study we used the X-ray photoelectron spectra to research the physical and chemical characteristic of the composite catalyst surface. Figure 4a shows the O 1s XP-spectra of g-C 3 N 4 , LaNiO 3 and CNL catalysts. The O 1s XP-spectra contains four distinct species. Two low binding energy peaks located at 528.6 eV and 530.1 eV are ascribed to the lattice oxygen, O 2− species in lanthanum oxide and nickel oxide 31 . The peaks located at 531.5 eV and 532.2 eV are from lanthanum and nickel hydroxides 32,33 . With the increase of g-C 3 N 4 , the peaks  at low binding energy become weaken, and the peaks at high binding energy become stronger, indicating that the g-C 3 N 4 compositing LaNiO 3 can promote the adsorption of hydroxyl on the surface of the composite catalyst and the covalency of Ni-O bond become stronger 34 . It has been demonstrated that the rate of O 2 2− /OH − displacement and OH − regeneration dominate the ORR kinetics, and the great covalence of the Ni-O bond can promote the ORR kinetics 30 . Furthermore, OER is either through chemisorbed OOH − formation or oxidation of surface OH − . The rate-determining step of the OER on perovskite oxides is governed by the concentration of hydroxide species that participate in the formation of the O-O bond in hydroperoxide. Thus, the high lattice hydroxide concentration on the surface can promote the HOO − forms, and consequently has a positive effect on OER activity.  The N 1s XPS spectrum of the g-C 3 N 4 shown in Fig. 4b can be fitted into three peaks at 398.8, 400.5 and 404.7 eV, corresponding to pyridine-like N, pyrrole-like N, and oxidized N, respectively 35,36 . After adding g-C 3 N 4 , a new peak emerged at 398.2 eV, which also corresponds to pyridine-like N. The reason why it was lower than 398.8 eV, can be explained by the existence of chemical bonds in the g-C 3 N 4 -LaNiO 3 heterojunction.
Ni 2p 3/2 was deconvoluted into two peaks. The binding energy of 854.3 eV was due to Ni 2+ , and 856 eV was due to Ni 3+ as shown in Fig. 4c 31 . The peak intensity of the lower oxidation state (Ni 2+ ) is weakening and the higher oxidation state (Ni 3+ ) is enhancing with the increase of g-C 3 N 4 in the composite. It suggests that through heat treating LaNiO 3 could be doped with nitrogen-atoms forming LaNiO 3−x N x , which promoted the formation of Ni 3+ on the surface of the composite catalyst. According to previous report, the existence of Ni 3+ in LaNiO 3 are beneficial to both ORR and OER activities of the material 37 .
Electrocatalytic performance. The typical polarization curves for the ORR was investigated in O 2 sat- . And the half-wave potential of 10 wt.% CNL is − 0.35 V, ~350 mV higher than the 2 nd high value of 5 wt.% CNL. In additional that the 10 wt.% CNL also exhibits the largest diffusion-limiting current, and the limiting current was 4.36 mA cm −2 , much higher than 2.82 mA cm −2 of LaNiO 3 . Furthermore, for kinetic analysis the current-potential curves at various rotation speeds were also analyzed with the Koutecky-Levich equation: This equation has been widely used to analyze the ORR reaction kinetics [38][39][40] , where i d is the tested disk currentdensity, n is the number of electron transfer in the ORR, k is the Boltzmann constant, F is the Faraday constant, , and n could be calculated from the slope. Figure 5b shows the Koutecky-Levich plots of XC-72, g-C 3 N 4 and CNLs based on ORR polarization curves at − 0.6 V with rotation speed of 2000, 1600, 1200 and 900 rpm. All curves exhibit good linear characteristic, indicating the first-order kinetics characteristics of the ORR. According to equations (3 and 4), the n value could be calculated as shown in Table 1. The n value of 10 wt.% CNL is about 3.86, indicating the ORR was a four-electrons process. In comparison with LaNiO 3 catalyst, as-synthesized 10 wt.% CNL material delivers a better ORR activity and kinetics. The higher electrocatalytic activity is due to the high Ni 3+ /Ni 2+ ratio in 10 wt.% CNL, which could promote adsorption/desorption of oxygen and thus the ORR activity in an alkaline electrolyte. Therefore in this work, the composite strategy has demonstrated to be highly effective to improve the ORR catalytic activity of the components.
The catalytic activity of the synthesized CNL for OER was also investigated by the RDE technique by measuring the polarization curves in O 2 saturated 0.1 M KOH solution from 0.0 V to 1.2 V, at a scan rate of 5 mV s −1 and a rotation speed of 1600 rpm. As shown in Fig. 5c, all the synthesized composite catalysts exhibit much better catalytic activity for OER than XC-72 carbon and LaNiO 3 . The OER onset potential for LaNiO 3 is 0.70 V and the OER diffusion-limiting current density at 1.2 V is 20.2 mA cm −2 . For a comparison, the 10 wt.% CNL has a more This improved property of 10 wt.% CNL is mainly due to the enhanced adsorption of hydroxyl on the surface of CNL composite. The hydroxyl participates in the formation of HOO − , and the increase of the HOO − concentration leads to the high OER electrocatalytic activity. As to the elexctrocatalytic activity of g-C 3 N 4 , due to the stronger electro-negativity of N than C, the electron will transfer from C to adjacent N, while the π -conjugation causes the electrons to return to the adjacent Cp 2 orbital 41,42 . These donation and back-donation processes not only help the O 2− adsorption by the strong chemical bond between O and C, but also promote O 2 desorption on the adjacent C atoms, resulting in enhanced ORR and OER activity. The reason that ORR and OER activities of the 20 wt.% CNL and 30 wt.% CNL are lower than these of 10 wt.% CNL, can be interpreted by the worsen electron conductive environment due to too much non-conductive in the composites.
XRD was used to confirm the formation and decomposition of Li 2 O 2 on the 10 wt.% CNL cathode during the charge-discharge process in non-aqueous Li-O 2 batteries. Figure 6 shows XRD patterns of the 10 wt.% CNL cathode at pristine,1st discharge and 1st recharge condition. After 1st discharge, compared to the XRD pattern of the pristine cathode, some diffraction peaks at 34.7, 40.3, 58.3 and 70.5° appear, which can be ascribed to Li 2 O 2 formation during the discharging process. After the 1st charge process, the diffraction peaks of Li 2 O 2 disappear, indicating that the formed Li 2 O 2 in the 1st discharge is decomposed during the charging process. This XRD pattern evolution demonstrates the rechargeability of this 10 wt.% CNL cathode.
We further investigated the catalyst activity of the composites in non-aqueous Li-O 2 batteries. The first discharge and charge profiles of XC-72, LaNiO 3 and 10 wt.% CNL in the Li-air batteries were compared at a current density of 50 mA g −1 (Fig. 7a). In particular, the initial discharge capacity of the 10 wt.% CNL is 5500 mAh g −1 , higher than that of LaNiO 3 (4600 mAh g −1 ) and XC-72 carbon (3600 mAh g −1 ). Also, it can be clearly seen that the discharge voltage plateau of the 10 wt.% CNL catalyst was ~2.8 V, higher than that of the LaNiO 3 (~2.7 V) and the XC-72 carbon (~2.7 V). The charge voltage plateau of the 10 wt.% CNL was ~4.0 V, lower than that of LaNiO 3 (~4.3 V) and XC-72 carbon (~4.4 V). The smaller overvoltages of CNL suggests the higher catalytic activity than LaNiO 3 and XC-72. In additional, 10 wt.% CNL catalyst presented 65 cycles between voltage window of 2.6 V to 4.7 V as shown in Fig. 7b. While the performance of the LaNiO 3 and XC-72 carbon catalyst are only up to 32 and  In conclusion, LaNiO 3 and g-C 3 N 4 composite was made as the bi-functional catalyst for oxygen reactions. The composite material delivers batter ORR and OER activities than individual LaNiO 3 and g-C 3 N 4 , the reason is that the CNL catalyst has a high Ni 3+ /Ni 2+ ratio that can promote adsorption/desorption of oxygen in an alkaline electrolyte, and more absorbed hydroxyl on the surface of CNL that can promote HOO − formation. LaNiO 3 and g-C 3 N 4 composite is for the first time used as the bi-functional catalysts for lithium air batteries. The composite presents enhanced specific capacity and cycle stability. The results suggest CNL composite can be a potential bifunctional catalyst for the Li-O 2 batteries, and the composite strategy is worth of further investigation.
Bulk g-C 3 N 4 sheets were synthesized by pyrolysis method. Briefly, 5 g melamine were heated at 550 °C for 2 hours with a heating rate of 5 °C min −1 . The resultant agglomerates were ground for 1 hour in a mortar, then further heated at 550 °C for 2 hours. Then the as-prepared bulk g-C 3 N 4 were obtained.
Synthesis of CNL composite material. The CNL were synthesized as follows: 0.04 g g-C 3 N 4 were dispersed in 50 mL ethanol in a beaker, and treated by ultrasonic for 30 min to fully disperse g-C 3 N 4 . Then 0.4 g LaNiO 3 NPs were added into the g-C 3 N 4 suspension, and the mixture were further dispersed by the ultrasonic treatment for 30 min, and then heated under reflux for 6 hours at 90 °C with stirring. Then evaporated the alcohol completely with stirring, a precipitate was obtained. The precipitate was collected and heated at 550 °C for 2 hours in nitrogen atmosphere. Then the resultant compound were ground for 1 hour in the mortar, to make the final 10 wt.% CNL composite material (the weight percentage number imply the g-C 3 N 4 ). By this method, CNL catalysts in different mass ratios (5 wt.% CNL, 20 wt.% CNL and 30 wt.% CNL) were also prepared.
Characterization of materials. The crystal phase and purity of the as-prepared powders were determined by X-ray powder diffraction with Cu Kα radiation (λ = 0.154056 nm) from 10° to 90°. The structures of the samples were determined by field-emission scanning electron microscope (SEM) performed on a Hitachi SU8020 at an acceleration voltage of 3 kV. X-ray photoelectron spectra (XPS) data were obtained on a Rigaku D/MAX2500V X-ray photoelectron spectrometer with an exciting source of Al Kα (1486.6 eV).
Electrochemical measurements. Electrochemical characterizations of the samples were performed on a rotating-disk electrode (RDE, ATA-1B) with a three-electrode cell. A platinum electrode was used as the counter electrode. A saturated calomel electrode was used as the reference electrode. A glassy carbon electrode was used as the working electrode. 0.1 M potassium hydroxide (KOH) solution was used as the electrolyte. For CV, ORR and OER studies, the 1 mg composite catalyst powder was mixed with 2.5 mg XC-72 carbon to ensure sufficient electronic conductivity. The as-prepared catalyst and 30 μL of 5 wt.% Nafion solutions were dispersed in 1 mL de-ion water/isopropanol solution (the volume ratio of de-ion water: isopropanol is 5:1), with an ultrasonic treatment for 40 min to make a homogeneous ink. The glassy carbon electrodes (3 mm diameter, 0.071 cm 2 area) were polished on a clean polishing cloth with the 0.05 μm alumina slurry, rinsed and dried. Then we pipeted 2.8 μL of the suspension drop-cast and loaded it on the glassy carbon, which was dried slowly inside a closed container. The glassy carbon electrode loaded with catalyst was immersed into a N 2 -saturated KOH electrolyte for the steady-state CVs. The N 2 was switched to O 2 and purged for another 30 min. After that the ORR polarization curve was measured from 0 V to − 1.0 V. Then, switched the O 2 to N 2 and purged for 30 min. Subsequently, the OER polarization curve were examined by a voltage scan from 0.0 V to 1.2 V. Electrochemical data were collected with an Autolab electrochemical workstation.
For Li-O 2 battery studies, the battery were fabricated in a glove box which was filled with pure argon gas. A lithium foil was used as the anode, 1 M LiCF 3 SO 3 in tetraethylene glycol dimethyl ether (TEGDME) as the electrolyte, glass fiber filter paper as separator, carbon-paper-supported catalysts as the cathode electrode and two piece of nickel foam (1 mm thick) as the cathode current collector 43,44 . The oxygen electrodes were prepared by mixing the as-prepared CNL catalysts, XC-72 carbon, 5 wt.% polyvinylidene fluoride (PVDF) and solvent N-methyl-2-pyrrolidone (NMP), and the mass ratio of CNL catalyst: XC-72 carbon: 5 wt.% PVDF is 1:2:3. The uniform slurry was coated onto the carbon-paper by a brush and dried in vacuum at 80 °C for 12 hours. The ready-made Li-O 2 battery were transferred into a sealed bottle which was filled with pure oxygen and linked with the multichannel battery testing system (LAND CT 2001A). Galvanostatic discharge of the battery were studied in a potential range of 4.5-2.0 V. The first discharge cycle performances were conducted at the current density of 50 mA g −1 . The following cycle performances were conducted at the current density of 250 mA g −1 , and the capacity was limited to 500 mAh g −1 . It should be noted that the current density and capacity were all calculated based on the mass of catalyst loaded on the cathode.