Electrostatic layer-by-layer self-assembly of 1D α-LiFeO2 with enhanced rate capability and cycling performance

α-LiFeO2 is a promising cathode material for lithium-ion batteries due to its theoretically high specific capacity (282 mAh g−1), abundant nature, low cost of raw materials and environmental friendliness. However, the intrinsic sluggish kinetics and poor electronic conductivity of α-LiFeO2 prevent its practical use. In this work, we introduce a novel electrostatic layer-by-layer self-assembly method using PAH and PSS charged polyelectrolytes to grow in situ Ag nanoparticles on the surface of α-LiFeO2 nanorods to improve the electronic and ionic conductivity in this material. The experimental results show that such tailored design effectively improves the cycling stability and provides the material with a superior rate capability. The Ag-1D α-LiFeO2 material delivers a high discharge capacity of 162.6 mAh g−1 at 0.5 C and a capacity retention of 89.6% after 50 cycles. The excellent electrochemical behavior may be ascribed to synergistic effects which combine the use of Ag NPs, which provide with improved electronic conductivities, and the large specific surface areas given by the 1D morphology of the nanorods, providing increased lithium and electron conduction pathways.


Introduction
Rechargeable lithium-ion batteries (LIBs) have become the predominant power device for portable electronic products, electric and hybrid vehicles and smart grid applications [1,2]. However, with rapid technological developments and serious environmental concerns, it is necessary to explore new cathode materials with high energy density, long cycle life, low toxicity and low production costs [2]. The commercialized LiCoO 2 cathode material provides stable cycling and rate performance, but the unethically sourced cobalt coupled with its toxic nature poses a problem for its future use [3]. Thus, alternative cathode materials are ought to be explored. a-LiFeO 2 , with a disordered rock-salt structure, has been considered as a potential candidate to replace LiCoO 2 due to its high theoretical capacity (282 mAh g -1 ), high elemental abundance, the low cost of the raw materials, environmental friendliness and safety during operation [4][5][6][7][8][9]. Unfortunately, a-LiFeO 2 suffers from an intrinsically poor electronic conductivity and slow lithium-ion diffusion kinetics. To this end, various strategies have been developed to improve the conductivity and lithium-ion diffusivity of a-LiFeO 2 . Generally, nanostructured electrodes with one-dimensional (1D) morphology (e.g., nanorods, nanotubes and nanowires) have improved lithium storage performances due to their architecture, which facilitates electron transportation [10][11][12][13][14]. Furthermore, it has been demonstrated that modification with conductive additives can facilitate electron transport between the particles. Typically, researchers have used carbonaceous materials such as carbon from different organic sources and MWCNTs as conductive additives for a-LiFeO 2 [15][16][17][18]. A different approach has been to use metals such as silver and copper to decorate the surface of the electrode material to improve the electronic conductivity of the composite material [19][20][21]. Silver has been used as a conductive additive in cathode materials such as LiMn 2 O 4 /Ag, LiFePO 4 / Ag, Co 3 O 4 /Ag and Si/Ag composites [22][23][24][25][26][27] given its high electrical conductivity (6.30 9 10 7 S m -1 ). However, it is difficult to precisely control the uniformity and particle size of the additives when using traditional methods such as liquid precipitation, in situ synthesis and silver mirror reactions [28][29][30]. By contrast, layer-by-layer self-assembly (LBL) technology, which is a technique that lies on the deposition of alternating layers of oppositely charged materials is an efficient method that may provide with a uniform particle distribution [31,32].
In this work, we demonstrate a novel, low-cost, simple and scalable method to prepare 1D a-LiFeO 2 modified with Ag NPs based on an electrostatic layer-by-layer self-assembly process using PAH and PSS polyelectrolytes. The results from structural and electrochemical analyses show that the Ag NPs were grown in situ on the surface of the a-LiFeO 2 nanorods. As a result, enhanced ionic and electronic conductivities which enabled improved electrochemical performance in terms of cycling capability and rate performance were observed. This design strategy opens the pathway to the fabrication of new electrode designs with improved electrochemical properties.

Experiment section Synthesis
Synthesis of a-LiFeO 2 nanorods a-LiFeO 2 nanorods were prepared through a selfsacrificial template route using a-FeOOH nanorods [33]. The detailed experimental process for the synthesis of a-FeOOH nanorods is shown in the SI (Methods), and PXRD data can be found in Fig. S1. Then, 0.5 g a-FeOOH, 0.47 g LiOHÁH 2 O and 0.78 g LiNO 3 (Li/Fe = 4:1 molar ratio) were ground together with a mortar and pestle using a small amount of distilled water and anhydrous alcohol. After this, powders were dried at 80°C for 12 h and then sintered at 400°C for 6 h in a N 2 atmosphere using a heating rate of 3°C min -1 . After naturally cooling down to room temperature, the product was washed with distilled water and ethanol and dried at 80°C for 12 h. For comparison, a-LiFeO 2 particles were prepared by a traditional solid-state method using LiOHÁH 2 O and Fe 2 O 3 (Li/Fe = 1.05:1 molar ratio) as precursors. The precursors were sintered at 800°C for 8 h in a muffle furnace with a ramp of 3°C/min and automatically cooled down to room temperature to obtain a-LiFeO 2 . PXRD and SEM data can be found in Fig. S2 and Fig. S3, respectively.
Preparation of Ag-1D a-LiFeO 2 nanocomposite by electrostatic self-assembly Typically, 0.25 g of as-synthesized 1D a-LiFeO 2 were stirred for 1 h in 120 ml of a 0.5 M NaCl solution and after that 0.3 g of PSS (sodium-4-styrene sulfonate) were added and stirred for another 0.5 h. Then, the sample was washed repeatedly with deionized water to remove the excess of PSS. The same procedure was carried out using 0.3 g of PAH (poly(allylamine hydrochloride)), and this was repeated with PSS. As a result, PSS/PAH/PSS-modified 1D a-LiFeO 2 was prepared. The modified 1D a-LiFeO 2 nanorods were placed into 50 mL water with 0.005 g AgNO 3 and 2.0 g polyvinyl pyrrolidone (PVP) and were stirred for 1 h. After this, 20 mL of a 0.025 M NaBH 4 solution were added dropwise into the layer-by-layer modified sample. After stirred for 1 h, the resulting products were washed with distilled water and ethanol repeatedly and dried at 80°C for 12 h. Figure 1 depicts a schematic of the electrostatic selfassembly process for the preparation of Ag-1D a-LiFeO 2 .

Microstructural characterization
SEM (JEOL, JSM-6360LV) was performed using an accelerating voltage of 60 kV to determine the microstructure of the as-prepared a-LiFeO 2 samples. TEM images were obtained using a JEOL, JEM-3010 at an accelerating voltage of 200 kV. TEM mapping images were obtained on an FEI Talos F200X with a field emission gun operating at 200 kV.

Zeta potential measurements
The zeta potentials of the Ag-1D a-LiFeO 2 suspension liquids were determined using a Zetasizer Nano ZS90 equipment.

X-ray Photoemission Spectroscopy (XPS)
X-ray photoelectron spectroscopy (XPS, ESCALAB 250XI) was measured on an Ultra DLD using a monochromic Al X-ray source. The electronic conductivity is measured by a digital four-point probe tester (SX1934 (SZ-82)).

Electrochemical studies
The electrochemical performance of the 1D a-LiFeO 2 samples and pristine material was tested by using CR2025 coin-type cells, which were assembled in an Ar-filled glove box. Electrode slurries were made by mixing 85 wt% active material, 10% acetylene black (conductive additive) and 5% polyvinylidene difluoride (PVDF) (binder) with a mortar and pestle. The mixture was added into an N-methyl pyrrolidone solution that was stirred for 12 h. Then, the slurry was evenly coated onto aluminium foil with the help of a Doctor blade, and the films were dried at 80°C for 30 min, followed by punching them into electrodes with a diameter of 12 mm. The electrodes were further dried at 100°C for 12 h under vacuum before assembling them into coin cells. A loading of ca. 1.3 mg cm -2 was achieved in the electrodes of this study. For coin cell assembly, lithium foil was used as the counter and reference electrode, a Celgard microporous polypropylene membrane was used as a separator, and 1 M lithium hexafluorophosphate (LiPF 6 ) dissolved in ethylene carbonate (EC) and diethyl carbonate (DEC) in a 1:1 volume ratio was used as the electrolyte. Galvanostatic measurements were performed on a Land CT2001A battery testing system at different current densities between 1.5 and 4.8 V (vs. Li ? /Li). Electrochemical impedance spectroscopy (EIS) analysis was conducted on a CS-350 electrochemical workstation with a frequency range of 0.1 Hz to 100 kHz at an amplitude of 5 mV. Cyclic voltammetry (CV) was conducted at a scan rate of 0.1 mV s -1 in the 1.5-4.8 V voltage range. Figure 1 describes the electrostatic self-assembly approach for the synthesis of Ag-1D a-LiFeO 2 , where poly(allylamine hydrochloride) (PAH) and poly(sodium 4-styrenesulfonate) (PSS) were used as positively and negatively charged polyelectrolytes, respectively. The polyelectrolytes were used to homogeneously coat the 1D a-LiFeO 2 based on a selfassembly process. The PAH-PSS couple is characterized by a strong hydrophobic electrostatic interaction between the ammonium group ÀNH þ 3 Â Ã of PAH and the sulfonate group ÀSO 2À 3 Â Ã of PSS, leading to a high charge density [31,32]. A similar approach using poly(diallyl dimethylammonium chloride) (PDDA) and poly(sodium-4-styrene sulfonate) (PSS) polyelectrolyte couples was used to coat a 3D carbon substrate with LiFePO 4 nanoparticles [34].

Results and discussion
During the PSS/PAH/PSS dipping cycles, we observed changes in the surface charge of the 1D a-LiFeO 2 nanorods, going from negative (zeta potential = -7.91 mV) to positive (zeta potential = 9.0 mV) and then back to negative (zeta potential = -16.32 mV) (Fig. 2a), indicating the formation of opposite charges on the 1D a-LiFeO 2 sample. At this point, positively charged Ag ? ions (from a 5.8 9 10 -4 M AgNO 3 solution) were adsorbed onto the surface of 1D a-LiFeO 2 nanorods due to the electrostatic attraction between these and PSS. A solution of NaBH 4 was then used to reduce in situ the Ag ? ions to Ag NPs onto the surface of 1D a-LiFeO 2 nanorods.
PXRD data of the 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 samples are shown in Fig. 2c and d. Both samples show diffraction peaks that can be indexed to the layered structure of a-LiFeO 2 (Space group Fm3m, JCPDS No. 74-2284) and with lattice parameters in agreement with those reported in the literature [35]. The Ag-1D a-LiFeO 2 sample shows an additional peak at 38.4°2h (marked in red), which matches to the most intense reflection of Ag (JCPDS No. 65-2871) [36] and thus confirms the presence of silver nanoparticles in the sample. No diffraction peaks from the a-FeOOH precursor were observed, indicating that these nanorods are fully converted into a-LiFeO 2 Furthermore, PXRD data confirm that the LBL process does not change the crystal structure of the a-LiFeO 2 nanorods, and thus, no side reactions occurred. Figure 2b shows the crystal structure of a-LiFeO 2 .
Micrographs of the samples are shown in Fig. 3ad. Figure 3a shows a typical SEM image of the assynthesized a-FeOOH sample, which consists of nanorods with diameters up to ca. 150 nm and lengths of a few micrometers. The micrographs indicate that the morphology of the a-LiFeO 2 nanorods is retained after the thermal treatment on the a-FeOOH nanorods (Fig. 3b). Similarly, SEM images revealed that there are no changes in the diameter and morphology of the a-LiFeO 2 nanorods after the LBL process (Fig. 3c, d) and showed the presence of Ag nanoparticles of a few nanometers uniformly coated on the surface of the 1D-a-LiFeO 2 .
TEM analysis further confirms the detailed microstructure of 1D a-LiFeO 2 and the distribution of the crystalline Ag nanoparticles onto the surface of 1D a-LiFeO 2 . Figure 4a shows a distinct a-LiFeO 2 nanorod with a diameter of ca. 150 nm, which is fully consistent with the SEM data (Fig. 3b). The HRTEM image in Fig. 4b shows well-resolved lattice fringes, illustrating the high crystalline nature of the 1D a-LiFeO 2 nanorods. The lattice fringe spacing of 0.24 nm is consistent with the (111) crystal planes of a-LiFeO 2 [15]. Figure 4c and d shows high-magnification images that reveal round NPs with diameters of ca. 5 nm attached to the surface of 1D a-LiFeO 2 . The (111) crystal plane with an interplanar distance of 0.255 nm marked in Fig. 4e indicates the presence of silver nanoparticle [37]. The corresponding SAED pattern presented in Fig. 4f shows the diffraction rings formed from scattered dots, indexed from inside to outside as (111), (200), (220), (311), which are in good agreement with the XRD results. Based on TEM evidence, it can be found the Ag NPs are in the nano-range and uniformly distributed on the surface of the 1D a-LiFeO 2 nanorods, suggesting a desirable architecture beneficial for fast electron transportation along the nanorods [30]. TEM-EDS mapping analysis shows a uniformly elemental distribution of O, Fe and Ag in the 1D-a-LiFeO 2 and Ag-1D a-LiFeO 2 nanorods (Fig. 5). Furthermore, the uniform coating of Ag nanoparticles is further confirmed in Ag-1D a-LiFeO 2 , suggesting that Ag nanoparticles can be evenly coated onto 1D a-LiFeO 2 through a simple LBL approach.
X-ray photoelectron spectra (XPS) of the 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 samples are shown in Fig. 6. According to the spectrum in Fig. 6a, Ag-1D a-LiFeO 2 consists of Li, O, Fe and Ag peaks, which is in  Figure 6b shows Fe 2p peaks for both samples locating at 711.04 and 724.56 eV, which are assigned to Fe 2p 3/2 and Fe 2p 1/2 spin-orbit peaks of a-LiFeO 2 , respectively, thus confirming the Fe(III) oxidation state [38]. The two main peaks are accompanied by satellite peaks on their respective higher binding energy sides at about 719.8 and 733.4 eV [38]. No peak shift occurs to a-LiFeO 2 before and after the LBL process, indicating that the LBL process does not change the chemical state of Fe. The high-resolution XPS spectra of the Ag doublet (3d 3/2 and 3d 5/2 ) in the Ag-1D a-LiFeO 2 nanocomposites are shown in Fig. 6c. The doublet is centered at 368.3 and 374.2 eV, which are higher than those of metallic Ag (0) (3d 3/2 = 367.9 eV; 3d 5/2 = 373.9 eV) [39], suggesting a successful reduction of Ag(I) ions to Ag (0) through the LBL process.
The electrochemical performance of the Ag-1D a-LiFeO 2 and 1D a-LiFeO 2 materials as positive for LIBs was evaluated in coin cells using lithium as the counter and reference electrode. Typical charge/discharge profiles in the voltage range 1.5-4.8 V at 0.5 C current rate are shown in Fig. 7a and b. The 1st cycle load profiles are similar on both samples. The voltage increases rapidly from the open-circuit voltage to 4.5 V, followed by a plateau (40 mAh g -1 ) before reaching the cutoff voltage.
During the first discharge, the voltage decreases sharply to 2.75 V and then enters a semi-plateau region that extends to 1.75 V, followed by a plateau in the 1.75-1.5 V region. Subsequent charge/discharge curves were S-shaped, as observed in other a-LiFeO 2 materials [40]. Cyclic voltammetry was carried out to better understand the cycling behavior and redox processes occurring in the Ag-1D a-LiFeO 2 material using a sweep rate of 0.1 mV s -1 between 1.5 and 4.8 V (vs. Li ? /Li) (Fig. S4). We observed an anodic peak at 4.3 V during the first charge process which is consistent with the literature and the load curves (Fig. 7a) [41]. This peak disappears in the following cycles. Furthermore, the cathodic peak at 1.55 V decreased in intensity from the first to the second cycle. All the peaks maintained the same shape and current intensity after the first cycle. Based on this result, we assume that a-LiFeO 2 underwent a structural change during the first cycle and maintained the changed structure after the second cycle. Several works from other researchers [4,5,16,40,42,43] have explained the mechanism of structural evolution during the first and consequent cycles. Sakurai et al. [4,42] suggested that unstable Fe 4? forms during the first charging process and oxidizes the electrolyte to produce Fe 3? ions. It should be noted that not all the Fe 3? oxidizes to Fe 4? and thus the low capacity observed during the first cycle. Upon discharge and consequent cycles, the predominant redox reaction is that coming from the relatively more stable Fe 3? /Fe 2? redox couple. Furthermore, Morales et al. [43] have proposed a structural rearrangement upon lithium removal, where Fe 4? were displaced from octahedral 4a sites to tetrahedral 8c positions. The 1D a-LiFeO 2 sample exhibits a first cycle discharge capacity of 151.2 mAh g -1 and retains 73.7% of this capacity after 50 cycles. By contrast, a discharge capacity of 162.6 mAh g -1 was observed for the Ag-1D a-LiFeO 2 electrode, with a capacity retention of 89.6% after 50 cycles. Coulombic efficiencies close to 100% (except for the first cycle) were observed in both samples, indicating a superior reversible capacity (Fig. 7c). This initially low coloumbic efficiency has been observed in other Li-Fe-O systems [6,42,44]. For reference, the electrochemical performance of the 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 electrodes was tested against as-prepared pristine (P-a-LiFeO 2 ) using the same voltage window at 0.1 C (Fig. 7d). As summarized in Table 1, the capacity of P-a-LiFeO 2 fades from 176 to 109 mAh g -1 after 50 cycles with only a 61.9% capacity retention, while this is 82.2% and 93.7% for 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 electrodes, respectively.
The rate performance of the P-a-LiFeO 2 , 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 electrodes was tested and waterfall plots are shown in Fig. 7e. Samples were cycled at 0.1 C, 0.2 C, 0.5 C, 1 C and 2 C (five cycles each) between 1.5 and 4.8 V and then returned to 0.1 C. Overall, the discharge capacity decreases for all samples with increasing current. The Ag-1D a-LiFeO 2 sample delivers the highest of all current densities compared to the other two samples. Average discharge capacities of 193.5, 176.7, 147.9, 115 and 81.18 mAh g -1 at 0.1, 0.2, 0.5, 1 and 2 C were observed for the Ag-1D a-LiFeO 2 electrode, which contrast to the average discharge capacities of 176.6 (0.1 C), 167.6 (0.2 C), 133.8 (0.5 C), 94.6 (1 C) and 70.5 mAh g -1 (2 C) for 1D a-LiFeO 2 and 175.9 (0.1 C), 127.4 (0.2 C), 109.38 (0.5 C), 73.88 (1 C) and 44.2 mAh g -1 (2 C) for the P-a-LiFeO 2 electrode (Fig. 7f). When the current density returns from 2 to 0.1 C, the average discharge capacities of the P-a-LiFeO 2 , 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 electrodes are restored to 97.6, 104.83 and 138 mAh g -1 , respectively. We attribute the enhanced rate performance of the Ag-1D a-LiFeO 2 sample to the incorporation of Ag NPs on the surface of the nanorods, which allow for improved electronic and ionic conductivity in these materials. This is further confirmed by the impedance analysis (Fig. 8).
To further investigate the effects of Ag on the long cycling performance, 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 were cycled 100 times at a 0.1 C rate (Fig. S5). Our data show that the capacity retention ratio of 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 are 49% and 75.93%, respectively. We attribute this improvement in capacity retention to the improved electronic and ionic conductivity in Ag-1D a-LiFeO 2 .
The Nyquist plots and equivalent circuit of the P-a-LiFeO 2 , 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 electrodes before and after cycling are shown in Fig. S6 and Fig. 8a. The semicircle dimensions of all cells increase after 50 cycles, suggesting an increased charge transfer resistance of the electrodes with cycling [45]. A broader semicircle was observed for the P-a-LiFeO 2 electrode before and after 50 cycles compared to the 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 electrodes. This is also supported by the fitted parameters listed in Tables S1 and Table 2. The electronic conductivity of the 1D a-LiFeO 2 and Ag-1D a-LiFeO 2 electrodes was measured, and the results are shown in Table S2. These indicate that the homogeneous incorporation of the silver nanoparticles can decrease the resistance value in the a-LiFeO 2 nanorods, providing a long charge carrier transportation length. Furthermore, we calculated the lithium-ion diffusion coefficient of Ag-1D a-LiFeO 2 to be 8.24 9 10 -12 cm 2 S -1 , which is about 6 times higher than that for 1D a-LiFeO 2 (1.3 9 10 -12 ). See supplementary information for more details.
Given the reduction of the charge transfer resistance value (R CT ) and the increase of exchange current density i 0 À Á after the modification of Ag nanoparticles, it can be concluded that 1D a-LiFeO 2 combined with acetylene black to enhance the surface electron conductivity of the electrode, provides more active sites for electron and lithium-ion transfer (Fig. 9). Moreover, the introduction of Ag NPs further enhances the electron conductivity and lithiumion diffusion coefficient, thereby contributing to the high cycling stability and rate capability.
To understand the structural stability of the Ag-1D a-LiFeO 2 nanorods, the morphology of Ag-1D a-LiFeO 2 after 50 cycles was investigated by SEM analysis (Fig. S7). It was observed that the Ag nanoparticles still adhere tightly to the 1D a-LiFeO 2 nanorods even after 50 cycles at 0.5 C. This indicates enhanced structural stability achieved for the deposition of Ag nanoparticles on the 1 D a-LiFeO 2 nanorods via LBL method.

Conclusion
In this work, we report a novel electrostatic layer-bylayer self-assembly method using PAH and PSS charged polyelectrolytes to grow in situ Ag nanoparticles on the surface of a-LiFeO 2 nanorods. When tested as a cathode material for LIBs, the Agmodified 1D a-LiFeO 2 nanorods delivered a highly reversible cycling capacity of 194.5 mAh g -1 at 0.1 C after 50 cycles and show an excellent capacity at higher rates (144.8 mAh g -1 at 0.5 C after 50 cycles) with a capacity retention of ca. 94% and 90%, respectively. By contrast, the unmodified 1D a-LiFeO 2 sample showed a capacity retention of ca. 82% and 74% under identical conditions. Based on our data, the enhanced rate capability and improved cycling performance of Ag-1D a-LiFeO 2 can be ascribed to: (1) synergistic effects by combining Ag NPs with 1D microstructures, where the nanorod microstructure provides facilitated lithium and electron pathways given their large specific surface areas, and the evenly coated Ag NPs on the surface of these improve the electronic conductivity (8.2 9 10 -9 S cm -1 ) of the sample given their high electrical conductivity; (2) the large Li ? diffusion coefficient of 8.24 9 10 -12 and high exchange current density of 2.791 9 10 -4 mA cm -2 observed; and (3) the excellent structural stability upon long cycling without undergoing severe agglomeration.  Table 2 Representative fitting data of the EIS data corresponding to Fig. 8a R/sample 1D a-LiFeO 2 Ag-1D a-LiFeO 2 (R S ? R SEI ? R CT ) 0 (X) 157.2 100.5 (R S ? R SEI ? R CT ) 50 (X) 411.7 221.6

Compliance with ethical standards
Conflict of interest The authors declared that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licen ses/by/4.0/.