Enhancement of LiFePO4 (LFP) electrochemical performance through the insertion of coconut shell-derived rGO-like carbon as cathode of Li-ion battery

An old coconut shell as a green biomass was known as a potential carbon materials for rGO and cost effectiveness. The objective of this study is synthesizing an rGO-like carbon (C) compound from coconut shells and inserting into LiFePO4 (LFP), as Li-ion battery cathode. Thus, an LFP/rGO nanocomposite was successfully fabricated using an unconventional approach which is the combination of the sol–gel technique and mechanical ultracentrifugation. LiFePO4 precursors were prepared from commercial starting materials, using the sol–gel technique, and the composites’ carbon weight content was varied between 15 and 30%. This process was subsequently followed by evaluating the microstructural characteristics and electrochemical properties as cathode for the Li-ion batteries. The results showed a high tendency of achieving maximum efficiency with merged LFP and rGO, although LFP molecules appear scattered but are firmly attached to each rGO structure, acting as a "bridge" between the surrounding particles. This reduced graphene oxide (rGO) link is relatively effective in limiting LFP grain growth as well as expanding the surface area, leading to a declined Li-ion diffusion rate. Consequently, the bridge presence also demonstrated a significant effect by enhancing the conductivity, electrical capacity, and performance of the LFP/rGO cycle than pure LFP. Furthermore, the percentage ratio of the synthesized LFP/rGO cathode (85:15) attained higher cycle capacity, compared to 70:30 on the level of 0.1 C, with specific discharging average of 128.03 mAhg−1 and retention capacity of 97.75% after 50 cycles, at room temperature and a rate of 0.1 C.


Introduction
Lithium iron phosphate (LFP) was firstly reported in 1997, by Goodenough et al., became a promising cathode material for rechargeable Li-ion batteries, and has long been studied intensively afterward [1,2]. This is due to the easy to obtain, non-toxic, environmentally friendly, and cheaply manufactured constituent materials [2][3][4]. In addition, LFP with olivine structure (LiFePO 4 ) shows a high theoretical capacity (170 mAhg -1 ), operating voltage suitable (3.5V vs Li ? /Li), long life, and high thermal stability [3,4]. However, limitation of lithium ions such as low intrinsic electronic conductivity and slow diffusion rate for the application in Li-ion batteries are particularly at low temperatures and high current densities [3,5]. Therefore, a combination of carbon bridging and particle size reduction is one of the most effective strategies to overcome this problem [2,5,6]. This carbon bridging effort aims to inhibit crystal growth [7,8], blocking particle agglomeration [4,9], strengthen contact inter-particles [10], and shorten the lithium ion diffusion pathways [11,12], while the presence of a carbon layer prevents Fe 2? oxidization to Fe 3? ; thus, the material's purity is guaranteed [12].
Coconut shells have been studied as a potential carbon materials for rGO [13,14] with advantages not only in the green energy issue but also in the cheaper cost compared to the commercial ones [15]. Carbon material is one of the optimal selection of the electrode material [16]. Recent studies showed that carbon-based integrations, including carbon nanotubes (CNT) or graphene, accelerate the movement of photo-generated electrons and generate easier charge separations [17] while enhancing the thermoelectric characteristics [18]. Graphene exhibits exceptional electron transport properties and very substantial carrier-mobility [19]. In terms of electrical conductivity, the substance demonstrates high capacity to increase power conversion efficiency [20] and its stability [21]. Moreover, graphene as a carbon-based nanocomposite advantages are mostly on the economic concern and the environmental impact [22]. Reduced graphene oxide (rGO), as part of Graphenebased nanomaterials, is having particular interest among the various activated carbon materials available due to its high electronic conductivity, wide surface area, structural flexibility, chemical stability, wide operating temperature range, and unique characteristics of a single atomic thick layer [23][24][25].
Furthermore, rGO is a major functional graphene derivative formed from the graphene oxide reduction process that involves the addition of electrons to an atom or atomic network [26]. Based on the unique structure and excellent characteristics, the nanocomposite also occurs among the most promising materials for possible application in lithium ion battery electrodes [23,27,28]. Recently, graphene and its derivatives, including rGO-based samples, have been analyzed extensively after demonstrating significantly enhanced electrochemical performance and stability [29,30].
The preparation of LFP/rGO nanocomposites does not only tend to improve the cathode material's conductivity but also shortens the lithium ions' diffusion path and increasing the lithium ion diffusion rate and then improving the material's electrochemical performance [31]. During the last decade, research on bridging LFP cathode material with graphene and its derivatives, such as rGO in composite form, were conducted [3,23]. Shang et al. synthesized a LFP/graphene composite by the liquid phase method that showed an initial specific discharge capacity of 160 mAhg -1 at 0.1 C [32]. Meanwhile, Yuan et al. synthesized a LiFePO 4 /rGO composite through a simple solid phase combined with a carbothermal reduction method that resulted a first special discharge capacity of 151.5 mAhg -1 at 0.1 C, and 149.2 mAhg -1 after 50 cycles, with a coulombic efficiency of 98.5% [33]. Furthermore, Dhindsa et al. synthesized LiFePO 4 /graphene nanocomposites by the sol-gel method and the constant current test results showed that LiFePO 4 /G composite had a higher initial discharge capacity at a 0.1 C current density, compared to pure LiFePO 4 [34]. Similarly, Yuan et al. successfully prepared rGO composites based on LiFePO 4 using the hydrothermal method [35], and almost all the result indicated better electrochemical performance and stability in the composites and LFP/graphene composites have also been successfully prepared through in situ solvothermal methods [36,37], solid-state reaction [38,39], and co-precipitation method [40]. Generally, the composites produced from these various synthesis methods are significantly improving the LiFePO 4 electrode's electrochemical performance [35]. In terms of simplicity, manufacturing of LiFePO 4 nanocomposites in order to control LiFePO 4 's morphology is more preferable since nanocomposites are made by simply mixing components and conductivity controllable by carefully selecting its components [29,39,41]. However, certain mechanisms employ expensive and environmentally unfriendly commercial chemicals in LFP/graphene nanocomposites, with a complex manufacturing process. Therefore, further research that utilizes low cost and sustainable (green) materials through a new and simpler approach appears very necessary [26,[42][43][44][45].
Previously [41], preparation method of LFP/rGO nanocomposites is using polymerized Polypyrrole in situ and [7] is using limited precipitation polymerization in situ at room temperature. Nano-structured LiFePO4 particles and rGO potentials are believed to be suitable candidates for further developing of cathode materials for lithium-ion batteries. In this study, an old coconut shell waste (biomass) was used to prepare an rGO-like carbon, as a ''green'' (sustainable) compound, acting as material for technological applications in the form of high-performance cathode materials for lithium ion batteries. The aim of this study is to provide an overview on old coconut shell (biomass) utilization to generate rGO as a sustainable compound in battery technology. As a consequence, a novel approach was introduced to evaluate the microstructural characteristics and electrochemical properties of rGO-based LFP cathodes. Specifically, the research scope covers the following: (i) obtaining a close relationship between the LFP and rGO substances needed for maximum efficiency; (ii) providing vital ideas related to natural resource utilization, including coconut shells (biomass), used to prepare the rGO-like carbon as a ''green'' (sustainable) compound in technological applications in form of high-performance cathode materials for lithium ion batteries; and (iii) developing an efficient method to synthesize functional LFP/ rGO nanocomposites as superior lithium ion battery cathode materials, compared to previous studies [7,41]. A combination of sol-gel and mechanical ultracentrifugation techniques with emphasis on super high centrifugation speed is required.

Method
The material preparation is particular focused on LiFePO 4 and rGO composites. During LiFePO 4 nanoparticles preparation, firstly, the LiFePO 4 nanoparticles were prepared by sol-gel approach using Li 2 CO 3 , FeCl 2 Á4H 2 O, and NH 4 H 2 PO 4 powders as precursors. Secondly, the materials were mixed and homogenized using a magnetic stirrer into a solution. Thirdly, the resulted sol solution was slowly dripped with NH 4 OH until a pH of 7 was obtained, while continuously stirred and heated at 100°C for 3 h until a gel was formed. Finally, the precursor gel was dried at 120°C, and further heating at 700°C and 750°C in an argon gas environment for 10 h at 5°C/s resulted in xerogel and calcinated substances, respectively. Meanwhile, in rGO preparation, the first step is rGO-like carbon compound was synthesized from old coconut shells, subjected to be burning, pulverizing as well as sieving. The second step is the obtained charcoal powder was carbonized in a furnace at 400°C for 5 h. At third step, a mechanical and chemical exfoliating process using HCl was carried out for 6 h to unbond rGO layers and resulting thinner particle size.
Furthermore, the LFP/rGO nanocomposite cathode was prepared by mechanical ultracentrifugation approach. This was performed by adding the rGO solution into LiFePO 4 nanoparticles and mixed with normal butanol (N-butanol). The mixture was dispensed and stirred with a magnetic stirrer at 120°C for 5 h. The resulting sample was then ultrasonised until the rGO layers were estimated to be covering the LPF particles and subjected to be polarized. Subsequently, the newly formed LFP/rGO slurry was centrifuged at 6000 rpm for 40 min to generate a LFP/rGO nanocomposite structure, prior to drying at 80°C for 2 h. This process is believed to serve as an advancement of the previous method [46].
The crystal structure, phase composition, and lattice parameters of LFP nanoparticles as well as the rGO composites were characterized using the Philips X'Pert MPD X-Ray diffraction instrument (multipurpose diffractometer) and Cu Ja (k = 1.54060 Å ) radiation. Also, the rGO occurrence was determined by Raman spectroscopy (Raman spectrometer-Renishaw InVia), while the transmission electron microscope (TEM, Tecnai G2-200 keV, Philips) was used to observe the morphology and microstructure. In addition, the morphology and crystal size were examined by scanning electron microscopy (SEM-EDX, ZEISS).
The LiFePO 4 -based nanocomposite cathode's electrochemical properties were measured using a CR2016 coin cell, with a lithium metal disc as the electrode counter. Meanwhile, the working electrodes were assembled by pressing a mixture of the synthesized LiFePO 4 /rGO, polyvinylidene fluoride (PVDF) binder, and super-p addictive material, in a 0.5:0.033:0.022 weight ratio. The electrolyte was created from 1M LiPF 6 dissolved in a mixture of ethylene carbonate (EC) and dimethyl carbonate (DMC) (1:1:1, v/v/v) as well as Li-foil and Celgard 2300 microfilm used as a separator and counter electrode. Furthermore, the coin cells were assembled in a glove box filled with high-purity Ar gas, while the LiFePO 4 /rGO electrical conductivity test was conducted with a four-point probe. The galvanostatic charge/discharge is tested in the voltage range of 0.5-4.0 V vs Li/Li ? at various current levels, while electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV) were measured above the CHI-660D electrochemical workstation, at 0.1 mVs -1 in 0.5-2.1 V vs Zn°/Zn 2? . EIS measurements were performed over the frequency range of 0.01 Hz-100 kHz at the discharge stage, with 5 mV amplitude. Figure 1 shows the X-ray diffraction (XRD) patterns revealed by the two LFP sample structure shapes and phase purities. Based on phase identification from the XRD pattern results, the two LFP sample's crystalinity were analyzed. Both patterns show similar LiFePO 4 main diffraction peaks and match the orthorhombic structure as well as Pnma space group. The sharp, narrow diffraction peaks without detectable phase impurities indicate that the LFP sample has high crystalinity. For the sample calcined at 750°C, an olivine phase without detectable impurity phases was observed, while the 700°C sample showed a 97% pure olivine phase and a phase impurity in the form of FeO 11 P 4 . The LFP sample calcined at 750°C had a sharper diffraction peak with narrower full width at half maximum (FWHM) and a higher intensity, compared to the 700°C counterpart. According to the peak FWHM, the crystal size of the samples prepared at 700°C and 750°C calculated by the Scherrer equation resulted in 99 nm and 100 nm, respectively. Figure 2 shows the XRD spectrum and Raman shift curve of coconut shell-based rGO after furnace carbonization for 5 hours at 400°C. Based on the illustration, the rGO phase was observed by the presence of two broad peaks at positions 23.5°and 42.5°along the 002 and 100 planes, respectively. Similarly, two main crests with wavenumbers 1370 cm -1 (band D) and 1620 cm -1 (band G) were also recorded. The defect band (D) was caused by crystal defects during synthesis, while the graphite band (G) represented the graphite bonds shared by the entire carbon allotropes, including graphite, graphene, activated carbon, and carbon nanotubes (CNT). In pure graphene, Fig. 1 XRD patterns of LFP preparation at temperatures of 700°C and 750°C Fig. 2 The XRD spectrum and Raman shift curve of rGO based on coconut shell the crystal structure is dominated by single-dimensional hexagonal carbon bonds, indicating a perfect state and low-intensity D band. This condition is validated by calculating the I D /I G ratio that represents the proportional value of the intensities of the defect (D) to graphene (G). Based on the relationship, greater I D /I G estimates tend to increase the defects, resulting in a weak hexagonal bonding in graphene sheet that possibly confirms an rGO phase. Meanwhile, the broad peak in 2D band shows the number of layers that are subsequently formed. Figure 3 represents the transmission electron microscopy (TEM) images of rGO samples. The TEM results revealed the occurrence of several aggregated layers that are individually stacked, although with slight wrinkles and transparency, due to the reduction process.

Results and discussion
The LFP/rGO nanocomposite particles' morphological and microstructure are observable with SEM-EDX (Fig. 4). Based on the images, LFP particles have spherical lumps and are firmly attached to rGO, resembling a thin flake ( Fig. 4a and b). LFP's spherical lumps indicate agglomeration, where several particles are joined together to look larger, with a non-homogenous and unevenly distributed grain size in certain places. This agglomeration occurs since the LFP powder, rGO solution, and N-butanol are not evenly mixed in the slurry-making process, while the black lumps indicate that the cathode sheet has shafts. Figure 4b shows that large LFP particles are scattered and firmly attached to each rGO layer side, thus acting as a ''bridge'' among the particles surroundings. Lithium ion has a greater diffusion distance between the crystal grain's boundary and center due to the larger crystal size. Furthermore, the rGO bridge is quite effective in limiting the LFP's grain growth and widening the surface area, shortening the lithium ion's diffusion length. Therefore, the close relationship between LFP and rGO nanoparticles encourages maximum efficiency while implementing the framework. The presence of Fe, P, O, and C were confirmed in EDX elemental analysis for the LFP/rGO nanocomposite at Fig. 4c. The graph showed various colors of the element mapping, including Fe (pink) on the energy position 0.65 keV and 6.40 keV, P (blue) at 2.0 keV, and O (green) at 0.46 keV that formed the violet color and was assumed as LFP nanoparticle. Meanwhile, C element represented rGO with red color on the energy position of 0.24 keV and K element (light blue) was also founded in the sample. Furthermore, the raw atomic ratios were C, O, Fe, P, and K with 41.93%, 35.41%, 13.98%, 8.46%, and 0.22%, respectively (Table 1). This EDX element mapping approach also can be found in [8,47]. Figure 5 shows the LFP/rGO nanocomposites' electronic conductivity and stored energy that are much higher, compared to pure LFP. Mixing LiFePO 4 particles with rGO by mechanical ultracentrifugation plays a significant role in enhancing the electronic conductivity and stored energy. These properties increase with the raise in rGO percentage. The optimum result of 7.84 9 10 -4 S/cm was obtained for samples with a mass ratio of 70% LFP and 30% rGO, while the counterparts for pure LiFePO 4 (theoretical) and LFP synthesized by the sol-gel method are * 10 -9 S/cm and 2.09 9 10 -7 S/cm, respectively.
Meanwhile, energy stored from LFP/rGO nanocomposites show a unique phenomenon and the optimum value of 6.50 9 10 -3 J was obtained by samples with a mass ratio of 85% LFP and 15% rGO. This explains that the addition of rGO bridging LFP particles must be carried out within certain limits and is reinforced by the LFP/rGO nanocomposite cathode material's electrochemical properties. A lithium battery's energy capacity depends on how many lithium ions are stored in the electrode structure and how much the ions are able to move during charging and discharging due to the amount of electron current stored and distributed is proportional to the amount of lithium ions diffused. Further discussion is to refer to the two samples, for a comparative study. Figure 6 shows the LFP/rGO nanocomposite's cyclic voltammogram curve with the addition of 15% and 30% rGO percentages in the first charge-discharge cycle [48]. Both curves show sharp and symmetric anodic/cathodic peaks, indicating good electrochemical performance due to single electron transfer reaction during the cycle. In addition, both electrodes showed a Fe 2? /Fe 3? redox peak at a   In the meantime, for LFP/rGO nanocomposites with a 85-15% mass ratio, the peak anodic was located at 3.46 V and corresponds to the oxidation of Fe 2? to Fe 3? though the cathodic peak at 3.30 V corresponds to the reduction of Fe 3? to Fe 2? with a 0.16 V potential interval between the two redox peaks. The anodic and cathodic peaks' voltage difference within the same cycle correlates to the redox reaction's polarization and inverse indicating the battery material's reversibility. A reduction in voltage difference lowers the polarization and increases the battery material's reversibility, resulting in greater cycle stability, so consistent with the charge-discharge graph and accelerates the Li ? ions' diffusion rate. The observed separation between the oxidation and reduction peaks is often used to distinguish electrochemical reversibility from electrode materials with larger separations which indicates lower reversibility. This narrow separation from the redox peak implies that the LFP/rGO nanocomposite has excellent electrochemical kinetics. Figure 7 shows the Nyquist plot derived from the electrochemical impedance spectroscopies (EIS) of both LFP/rGO nanocomposite samples which are measured to further prove that the rGO thin-layers in LFP/rGO nanocomposites increase the material's electronic conductivity. According to the image, the Nyquist plots obtained are in the form of semicircles and slopes. The semicircle pattern in the high to medium frequency region shows the lithium ion charge-transfer process on the LFP/rGO and electrolyte surfaces. Meanwhile, the straight line (slopes) pattern in the lower frequency region represents the lithium ion diffusion process into the electrode bulk material, commonly known as Warburg diffusion. This pattern shows that the electrodes are capable of storing lithium ions, therefore suitable for use in lithium-ion batteries. The charge-transfer resistance width also determines the battery's electrical conductivity of the battery. An increase in pattern narrowness leads to an increase of its electrical conductivity. A comparison of the EIS profile semicircle diameters shows that the LFP/rGO nanocomposite cathode with a weight ratio of 70-30% (* 350 ohm) has a much smaller charge-transfer resistance, compared to the counterpart with a ratio of 85-15% (* 430 ohm). This much smaller solidelectrolyte interface resistance ought to be due to the rGO layers' presence in the LFP/rGO nanocomposite, with a much better electronic conductivity.
The galvanostatic charge-discharge (Fig. 8) was performed to evaluate LFP/rGO nanocomposite's electrochemical properties. The results obtained were as a graph of a relationship between the voltage and the charge-discharge capacity, provided with a constant current and a cut-off voltage of 2.5-4.2 volts during the analysis. Furthermore, the cutoff voltage is the initial voltage before treatment (charging and discharging). Lithium ions move from anode to cathode until a maximum voltage of 4.2 volts is reached during charging. Subsequently, the ions move from cathode to anode until a minimum voltage of 2.5 volts is reached (discharging). There are flat plateaus of LFP/rGO nanocomposite from 3.4 V to 3.5 V that is exclusively for the transition of single phase of lithium iron phosphate to become ironphosphate. According to the image, the first cycle's charge profile shows a stable voltage at 3.5 V (versus Li ? /Li) and this should be corresponding to the redox pair Fe 2? /Fe 3? . The discharge profile from the first cycle shows a stable voltage at 3.3 V (versus Li ? / Li) with a very small separation (0.2 V) being stable at the charge profile. The sample of LFP/rGO nanocomposite with ratio 85-15% shows the highest discharge capacity up to 128 mAhg -1 and the lowest is showed by ratio 70-30% with discharge capacity 127 mAhg -1 . This excellent electrochemical performance is attributed to the large specific surface area and a wide bridging layer, ensuring electrons pass through each LiFePO 4 particle, shortening electronic transport pathways, and then reducing interface resistance. The lithium ions easily intercalate into the LiFePO 4 framework through the rGO layer and in turn avoids particle structure collapse during the charge shedding. Therefore, the LFP/rGO nanocomposite structure endures high current density charge/discharge. Figure 9 describes the performance of the LFP/rGO structure, with percentage ratios of 85:15 and 70:30, respectively, at rate 0.1 C. The graph displays that specific discharging average of both samples are 127.73 mAhg -1 and 128.03 mAhg -1 after 50 cycles. It shows a very well cyclic performance by retention capacities 97.73% and 97.75%. However, the LFP/rGO nanocomposite with ratio 85:15 obtained superior properties, compared to 70:30 on the level of 0.1 C. Furthermore, the sufficient stability of both samples without decreasing capacities denoted strong electrodes and stable solid electrolyte interfaces (SEI) as well as exhibited reversible kinetic reactions during cyclic reiteration.

Conclusion
Based on the overall results and discussions, an appropriate approach in synthesizing and characterizing the microstructure and electrochemical properties of rGO-based LFP cathode materials from old coconut shells has been proposed. LFP/rGO nanocomposites were successfully synthesized by combining the sol-gel route and mechanical ultracentrifugation techniques as an alternative to lithium ion battery cathode materials. The content of rGO prepared from biomass (old coconut shell) was varied between 15% and 30%. The results showed higher electrical conductivity and capacity as well as better cycle performance in LFP/rGO nanocomposites compared to pure LiFePO4 particles. Specific findings confirmed that the percentage ratio of synthesized LFP/rGO nanocomposite cathode (85:15) attained higher cycle capacity, compared to 70:30 on the level of 0.1 C, with specific discharging average of Fig. 8 Initial curve of charge and discharge of cathode material nanocomposite LFP/rGO sample at a ratio of 70-30% and 85-15% Fig. 9 Cycle life performance of the LFP/rGO at ratio of 70%:30% and the LFP/rGO at ratio of 18%:15% and was operated between 4.2 and 2.5 V at the current rate of 0.1 C 128.03 mAhg-1 and a retention capacity of 97.75% after 50 cycles, at room temperature and a rate of 0.1 C. The improved electrical conductivity and electrochemical performance of the prepared samples are believed to be done due to a three-dimensional conduction network provided by rGO-like carbon sheets as observed by electron microscopy. The findings in combination of the LFP and rGO nanoparticles drive maximum efficiency in their implementation. Moreover, the LFP particles are scattered and firmly attached to each rGO layer side, thus acting as a ''bridge'' among the LFP particles surroundings. The lithium ion has a greater diffusion distance between the crystal grain's boundary and its center due to larger crystal size. The rGO bridge is quite effective in limiting LFP grain growth as well as widening the surface area and shortening the lithium ions diffusion rate. In short, this work provides ideas in the utilization of biomass as a source of carbon compounds, including an rGO-like carbon, for technological applications of cathode materials for lithium ion batteries.
In the future with the same approach, it will be continued to synthesize another local green biomass and not only for lithium ion batteries but also will be developed for an alternative super-capacitor.