Magnetic and Optical Field Multi-Assisted Li-O2 Batteries with Self-Regulated Charge and Discharge

The photo-assisted lithium-oxygen (Li-O 2 ) system emerged as an important direction for future development by effectively reducing the large overpotential in Li-O 2 batteries. However, the advancement is greatly hindered by the rapidly recombined photoexcited electrons and holes upon the discharging and charging processes. Herein, we make a breakthrough in overcoming these challenges by developing a new magnetic and optical eld multi-assisted Li-O 2 battery with 3D porous NiO nanosheets on the Ni foam (NiO/FNi) as a photoelectrode. Under illumination, the photogenerated electrons and holes of the NiO/FNi photoelectrode play a key role in reducing the overpotential during discharging and charging, respectively. By introducing the external magnetic eld, the Lorentz force acts oppositely on the photogenerated electrons and holes, suppressing the recombination of charge carriers. The magnetic and optical eld multi-assisted Li-O 2 battery achieves an ultra-low charge potential of 2.73 V, a high energy eciency of 96.7%, as well as a good cycling stability of 200 h. This external magnetic and optical eld multi-assisted technology paves a new way of developing high-performance Li-O 2 batteries and other energy storage systems.


Introduction
Rechargeable Li-O 2 batteries have aroused ever-increasing research interest in high-energy-density storage systems because of their high theoretical speci c energy of 3500 Wh kg −1 1-4 .However, the large overpotential of the batteries caused by sluggish ORR and OER kinetics (oxygen reduction and evolution reaction) is still a critical challenge to be surmounted (Fig. 1a) [5][6] . To well resolve the aforementioned obstacles, a variety of strategies have been introduced, including the construction of electrocatalyst materials and soluble redox mediators [7][8][9] . Although the charging voltage has been signi cantly reduced, electrocatalysts materials would expedite the parasitic decomposition of electrolyte during charging [10][11] . Furthermore, soluble redox mediators migrate to the anode for reduction by electron shuttle, leading to low e ciency and instability of Li metal anode [12][13] . Therefore, it is necessary to explore new ways to intrinsically promote the formation and decomposition of Li 2  Recently, incorporating green and renewable solar energy to improve the reaction kinetics of ORR and OER in Li-O 2 battery is recognized as one of the promising options (Fig. 1b) [14][15][16] . Wu et al. rst reported a photo-assisted Li-O 2 battery with the dye-sensitized TiO 2 photoelectrode, which e ciently utilize the photovoltage to scavenge the Li 2 O 2 product 17 . Zhou et al. further employed a graphic carbon nitride (g-C 3 N 4 ) as a cathode, achieving an ultra-low charge potential (1.96V) 18 . However, the rapid recombination rate of the photoelectrons and holes generated in semiconductors is still a key obstacle in the photocatalyst-involved Li-O 2 battery. Conventionally, various state-of-art semiconductor heterojunctions, such as Schottky junctions, p-n junctions, and z-scheme heterostructures have been constructed to suppress the recombination of charge carriers [19][20][21][22] . Yet, the synthetic methods for heterostructures are complicated and rigorous. Accordingly, the noncontact and environmental-friendly external-magneticeld-tuned approach can be used as an e cient strategy 23 . It has been reported that the application of the magnetic eld in the solar cell shows a signi cantly improved carrier separation and photoelectron conversion e ciency, which is ascribed to a deviation of the charge movement with a vertical force to the direction of movement in the magnetic eld plane 24 . Hence, it is feasible and signi cate to apply an external-magnetic-eld into a photo-assistant Li-O 2 battery to enhance ORR and OER kinetics of the battery.
Based on the above understanding, we rst report a novel prototype of magnetic and optical eld multiassisted Li-O 2 battery with a 3D porous NiO/FNi photoelectrode (Fig. 1c). Bene ted from the abundant electron-hole pairs generated in the photoelectrode under the optical eld, the di cult formation/decomposition of Li 2 O 2 during the discharge/charge process in conventional Li-O 2 battery is greatly accelerated. When a magnetic eld (MF) was introduced into the photo-assisted Li-O 2 battery system, the Li-O 2 battery could work stably at ultra-low Li 2 O 2 oxidization voltage, which is ascribed to the well-suppressed electrons and holes. Notably, the charge voltage of the battery has achieved an ultra-low charge voltage ~2.73 V and a high stable cycling performance. Further tests were performed to discuss the interactions of the built-in electric elds, magnetic eld and light, and its effects on the battery performance enhancement. The concept of the magnetic and optical eld multi-assisted Li-O 2 battery offers an effective strategy to store the solar energy and introduces a universal method in energy storage systems.

Results
Structural analysis of the NiO/FNi cathode. Porous NiO nanosheets were in-situ deposited on the Ni foam (NiO/FNi) and employed as a photoelectrode in the magnetic and optical eld multi-assisted Li-O 2 battery. The schematic diagram of the synthesis strategy for the NiO/FNi cathode is shown in Fig. 2a, which involves two steps, i.e., hydrothermal treatment and subsequent annealing at high temperature 25 .
The powder X-ray diffraction (PXRD) patterns of the nal products are shown in Fig. 2b and all the diffractions are well indexed to cubic NiO. The X-ray photoelectron spectroscopy (XPS) further con rms the composition of the NiO/FNi cathode. In Fig. 2c, the peak located at 531.3 eV is assigned to the O of NiO-based material, and the peak at 532.7 eV is ascribed to the defective O. The peaks appear at 853.8 and 861 eV are attributed to the Ni 2p 3/2 main peak and its satellite, and the 872.8 and 879.5 eV corresponded to Ni 2p 1/2 main peak and its satellite (Fig. 2d) 26 . The energy dispersive spectrum (EDS) element mappings reveal that the Ni and O atoms distribute uniformly across the NiO/FNi cathode ( Supplementary Fig. 1). The microstructure of NiO/FNi cathode is investigated by the scanning electron microscopy (SEM, Fig. 2e), which displays the dense NiO nanosheets grown on Ni foam ( Supplementary   Fig. 2 and 3), as well as the color change from light to dark is displayed (inset). The enlarged SEM image of the NiO/FNi cathode clearly shows the homogeneously aligned and highly porous structure on a large scale (Fig. 2f). It is believed that the interconnected nanosheets can not only provide abundant open space and electroactive surface for electron transfer and electrolyte diffusion, but also enough area for light adsorption. Transmission electron microscopy (TEM) image in low-magni cation demonstrates the mesoporous structure of the NiO nanosheets, which can maximize the active surface and enable the achievement of excellent battery performance (Fig. 2g). The high resolution-TEM (HR-TEM) image shows a lattice fringe spacing distance of 0.148 nm, consistent with the (220) planes of face-centered cubic NiO, which is also in consistency with the NiO composition from the PXRD pattern 27 . Furthermore, the nitrogen absorption-desorption isotherms suggest the interconnected porous structure with a high speci c area of 144.77 m 2 g -1 (Fig. 2h). Utilizing the advantages of maintaining good FNi macropores, NiO/FNi cathode could provide a low-resistance pathway for electron and mass transfer via the freestanding structure.
Photoelectrochemical behavior of NiO/FNi. The optical properties of NiO/FNi photoelectrode were investigated with a conventional three-electrode cell. The Mott-Schottky plots of NiO/FNi show a negative slope, which is typical for p-type semiconductors (Fig. 3a), indicating that the photo-induced holes are the dominant charge carriers. As can be seen from the ultraviolet-visible (UV-vis) absorption spectroscopy, the solution. The onset potential of NiO/FNi with illumination for ORR reaches up to -0.28 V vs. Ag/AgCl, suggesting that the effectively improved ORR catalytic performance is originated from the light (Fig. 3c).
The photo-assisted ORR process on NiO/FNi is further con rmed by its current responses in Ar atmosphere with and without illumination in Supplementary Fig. 5. To better understand the reaction kinetics of the ORR process, Tafel slopes were calculated based on the LSV curves. The lower Tafel slope of 510 mV dec -1 for NiO/FNi under light indicates that illumination plays a signi cant role in promoting the reaction kinetics of ORR (Fig. 3d). Additionally, during the OER process, the NiO/FNi under light also exhibits obviously favorable performance comparing to that in dark (Fig. 3e), demonstrating the positive effect of light during OER process. The smaller Tafel slope of NiO/FNi under light (259 mV dec -1 ) than that in dark (429 mV dec -1 ) further veri es the positive role on OER process (Fig. 3f). As shown in Supplementary Fig. 6, the EIS results show the reduced charge-transfer resistance of NiO/FNi under light compared to the initial electrode, leading to an increased conductivity, which is also consistent with the above results.
Photocatalytic properties of NiO/FNi under magnetic eld. According to the previous work, Lorentz force can effectively act on photocatalysis process, due to the external MF can produce opposite force on the photoelectrons and holes 19,21 . Thus, we applied NiO/FNi in the photocatalysis process with MF and without MF (MF and NMF, respectively) to understand the in uence of MF on the photocatalytic activities.
Firstly, the nearly unchanged Mott-Schottky slopes of NiO/FNi under MF (Fig. 4a) imply that the MF has a slight in uence on the intrinsic charge carrier density. The LSV curves for ORR at MF and NMF condition nearly show the same trend, suggesting that the Lorentz force cannot facilitate the photocatalytic process without the charge carriers generated via photoelectric conversion (Fig. 4b). Fig. 4c displays the photocatalytic ORR performance of the NiO/FNi under the MF and NMF conditions by using the magnetic photocatalyst setup (inset). The photogenerated current density obviously increased, suggesting that the presence of magnetic eld plays a key role in enhancing the photocatalytic performance. In the presence of the external MF, electrons move under the function of the Lorenz force. Besides, driven by the Lorenz force, the free electrons in the system form a polarized distribution that generates a motion-induced electromotive force, which is probably the key reason for the enhancement of the photocatalytic performance. I-t curves were recorded under NMF and MF conditions to analyze the recombination rate of the charge carriers (Fig. 4d). The most stable and highest photocurrent response of the NiO/FNi under light and MF can be assigned to the newly separated photo-charges under MF conditions, which is related to the electromagnetic induction current.
To deeply understand the interaction between the photoelectrode and the magnetic eld, the magnetic hysteresis loop (M-H) of NiO/FNi was measured at room temperature. As observed in Fig. 4e, M-H curve exhibits a typical ferromagnetic property with small coercivity and saturated under a magnetic eld of 10 KOe, indicating NiO/FNi can be manipulated by an external magnetic eld. According to the above experiment results, it is proposed that the electrons and holes can be deviated when they move in MF, leading to better separation of the charge carriers. The generated induced electromotive force under MF can interact with the photogenerated carriers through the mutual attraction of positive and negative charges, contributing to favorable charge separation. The schematic diagram clari ed the in uence of the MF on the NiO nanostructure under light (Fig. 4f). Herein, COMSOL was further performed to schematically illustrate the in uence of the magnetic eld on the NiO/FNi (Fig. 4g). Comparing with the Lorentz force of with the electric force of the effects of the magnetic eld on the charge carriers can be discussed by referring to the electric eld 19 . It can be clearly seen that a parallel and homogenous electric eld presented in the NiO nanosheets (NSs) based on the electromagnetic induction (Fig. 4h). Under magnetic eld, the Lorentz force acts on the photo-induced electrons and holes due to the induced current difference between two sides, driving them to opposite directions, thus improving their e cient spatial separation and inhibiting recombination.
Electrochemical performance of the optical eld assisted Li-O 2 battery. Applying NiO/FNi as the reversible photocathode, an optical eld assisted Li-O 2 battery was constructed to investigate the effect of solar energy on the batteries (Fig. 5a). Due to the conversion of light energy to electricity, light-emitting devices can display higher brightness under illumination. Fig. 5b shows the galvanostatic charge and discharge curves of Li-O 2 battery with NiO/FNi, the illumination increases the discharge potential (from 2.59 V to 2.69 V) and concurrently decreases the charge potential (from 3.74 V to 2.92 V). Meanwhile, the optical eld assisted Li-O 2 battery displays promoted energy e ciency from 69.3% to 92.1%, indicating the positive effect of light on both the input and output electric energy conversion. One the one hand, the photogenerated electrons and holes on the surface of NiO/FNi in the optical eld can act on the discharging and charging processes, respectively. On the other hand, the deposition and morphology of the discharge products could be adjusted in the optical eld 29 . The discharge products obtained in the dark densely coated on the nanosheets structure, which would hinder the contact between the electrolyte and the active sites, leading to large polarization during the subsequent charging process ( Supplementary  Fig. 7a). In sharp contrast, the lm-like product deposited on the surface of the cathode in the optical eld, which can maintain the active sites after the discharge process and are bene cial for the decomposition process ( Supplementary Fig. 7b). Whether under dark or light, the discharge product Li 2 O 2 was mostly decomposed after the 1st recharged ( Supplementary Fig. 7c, d). However, the undecomposed Li 2 O 2 accumulated on the cathode surface in the dark after 20th recharged, while the smooth cathode surface is still exposed in the optical eld ( Supplementary Fig. 8). These results suggesting that the reversible decomposition of Li 2 O 2 was more effective in the optical eld.
When suffering the periodic ON-OFF cycles, the Li-O 2 battery under light exhibits reproducible and continuous photo-responsiveness during the discharge and charge processes (Fig. 5c). The results suggest that the quick photoexcitation process and fast mass transfer under light could suppress the polarization of the battery. Moreover, the optical eld assisted Li-O 2 battery presents better rate capability than that in dark ( Supplementary Fig. 9). The voltage gaps between charge and discharge process are maintained within 1.85 V even at high current density of 1.2 mA cm -2 and recovered to 0.36 V when returned to 0.01 mA cm -2 after 40 cycles at different current densities, which further implying the facilitated effect of illumination on the rate capability ( Supplementary Fig. 10). The oxygen reduction reaction (ORR) curves on the rotating disk electrode show improved kinetics under light, which is responsible for the excellent catalytic performance (Fig. 5d). The photoelectric conversion effect of NiO/FNi mainly acts on reducing the charging voltage and increasing discharge voltage upon illumination, which is corresponding to the position of CB and VB of NiO/FNi (Fig. 5e). During discharge, O 2 is reduced on the NiO/FNi surface involving the electron participation, thereby promoting the ORR.
Under illumination, the photoelectrons from NiO/FNi transfer to the Li anode through the external circuit and hence reduce the Li + to Li metals during charging. Meanwhile, holes can be collected on the NiO/FNi surface, and then utilized to oxidize the discharge product. Thus, the charge potential of NiO/FNi can be compensated by the generated photoelectric potential. In short, light plays a key role in achieving the energy storage and conversion from light to electrochemistry.  Fig. 11). These results implied that more photogenerated electron-hole pairs existed in the batteries because the rapid recombination of the photo-carriers can be greatly hindered under the MF condition. Even at high current density of 1.2 mA cm -2 , the Li-O 2 battery under MF (5 mT) and light also shows the expected low charge voltage, which can be recovered to 2.78 V when returned to 0.01 mA cm -2 after 40 cycles at different current densities (Fig. 6b). To further understand how MF facilitate the ion diffusion and improve the battery kinetics, EIS of photo-assisted Li-O 2 battery with NiO/FNi under MF was measured (Fig. 6c) (Fig. 6d). The in uence of the MF on the photo-assisted Li-O 2 battery is proposed and illustrated in Fig. 6d (inset). For a typical NiO/FNi photoelectrode, photogenerated electron-hole pair can easily recombine because the formed excited state obeys the conservation laws for energy and momentum, which results in limited participation of the photogenerated carriers in the charge transport. A potential or force to drive the positive and negative carriers towards opposite directions would signi cantly enhance the charge carrier density. MF was present during the photoelectrochemical process, according to the left-hand rule, both the moving electrons and holes experience a Lorentz force vertical to the direction of movement. Moreover, the forces acting on the negative electron and positive hole of the electron-hole pair are in opposite directions, leading to the deviation of the electrons and holes in opposite directions. Therefore, the photo-induced electron-hole pairs in moving semiconductor particles can be separated in MF, even at the initial generation of an electron-hole pair.
To further examine the stability of the magnetic and optical multi-assisted Li-O 2 battery upon cycling, the morphology changes of the recharged O 2 cathodes after the 30th cycle were examined. SEM image was carried out to investigate the discharge products under MF (5 mT) and light. The lm-like morphology is similar to that under MF (0 mT) and light, indicating that MF has negligible effect on the morphology of the discharge products ( Supplementary Fig. 12). The discharge product is evidenced by the Li 1s signal at 54.7 eV in the XPS spectra of NiO/FNi under MF (0 mT) and MF (5 mT) light, which reveals that Li 2 O 2 is the only discharge product after the 30th discharged ( Supplementary Fig. 13). After the 30th recharged, XRD patterns demonstrate that the Li 2 O 2 can be oxidized completely during the charging progress (Fig.   6e). As shown in the SEM image, partial residues are still remained on the cathode after the 30th recharge under MF (0 mT) and light (Fig. 6f). This continuous accumulation of the undecomposed products upon cycling would hinder the transportation of the oxygen, electrons and lithium ions within the cathode during subsequent cycling. In sharp contrast, the lm-like discharge products have disappeared, and the homogenously nanosheets structure of the whole cathode is almost fully recovered, even after the 30th recharge under MF (5 mT) and light, implying good structure reversibility and stability of the magnetic and optical multi-assisted Li-O 2 battery (Fig. 6g). The Li 1s signal at 54.7 eV in the XPS spectra of NiO/FNi further reveals the reversible Li 2 O 2 decomposition under MF (0 mT and 5 mT) light after the 30th recharged (Fig. 6h). Moreover, gas chromatography was performed on the batteries to demonstrate the gas evolution on recharge. As shown in Supplementary Fig. 14

Discussion
In summary, we have introduced the magnetic eld into the photo-assisted Li-O 2 battery system for the rst time to improve its ORR and OER kinetics. It is revealed that under illumination, the photogenerated electrons and holes can signi cantly act on the charging and discharging process, respectively. The photogenerated electrons participate in promoting the reduction of O 2 during the discharging process, Material characterization. The crystallographic information and chemical composition of the as-prepared products were established by a Rigaku D-Max 2550 diffractometer using Cu Kα radiation. The morphological investigations were carried out on a JEOL JSM-6700F scanning electron microscope. Microstructures were characterized by high-resolution transmission electron microscopy (HRTEM, a JEM-2200FS electron microscope). Nitrogen adsorption-desorption isotherms were measured at 77 K with a Micromeritics 2020 analyzer. The speci c surface area was calculated with the Brunauer-Emmett-Teller (BET) equation, and the pore size distribution was calculated from the adsorption curve by the Barrett-Joyner-Halenda (BJH) method. X-ray photoelectron spectroscopy (XPS) was recorded on an ESCALAB 250 spectrometer. The UV-vis spectra were acquired with a U-4100 UV-vis spectrometer under the diffusere ection model using an integrating sphere (UV 2401/2, Shimadzu). The uorescence spectra of the samples were taken by the photoluminescence (PL) spectro uorometer (FLUOROMAX-4 spectrophotometer) with an excitation at 300 nm light. The magnetic properties of the samples were measured using a vibrating sample magnetometry (SQUID-VSM magnetic measurement system, USA

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.