Advances in the Applications of Graphene-Based Nanocomposites in Clean Energy Materials

Extensive use of fossil fuels can lead to energy depletion and serious environmental pollution. Therefore, it is necessary to solve these problems by developing clean energy. Graphene materials own the advantages of high electrocatalytic activity, high conductivity, excellent mechanical strength, strong flexibility, large specific surface area and light weight, thus giving the potential to store electric charge, ions or hydrogen. Graphene-based nanocomposites have become new research hotspots in the field of energy storage and conversion, such as in fuel cells, lithium-ion batteries, solar cells and thermoelectric conversion. Graphene as a catalyst carrier of hydrogen fuel cells has been further modified to obtain higher and more uniform metal dispersion, hence improving the electrocatalyst activity. Moreover, it can complement the network of electroactive materials to buffer the change of electrode volume and prevent the breakage and aggregation of electrode materials, and graphene oxide is also used as a cheap and sustainable proton exchange membrane. In lithium-ion batteries, substituting heteroatoms for carbon atoms in graphene composite electrodes can produce defects on the graphitized surface which have a good reversible specific capacity and increased energy and power densities. In solar cells, the performance of the interface and junction is enhanced by using a few layers of graphene-based composites and more electron-hole pairs are collected; therefore, the conversion efficiency is increased. Graphene has a high Seebeck coefficient, and therefore, it is a potential thermoelectric material. In this paper, we review the latest progress in the synthesis, characterization, evaluation and properties of graphene-based composites and their practical applications in fuel cells, lithium-ion batteries, solar cells and thermoelectric conversion.


Introduction
Green energy is also called clean energy, including solar energy, wind energy, water energy, tidal energy, geothermal energy and biomass power, whereas fossil fuels are nonrenewable energy sources, including coal, crude oil, natural gas, oil shale, oil sand and combustible ice under the sea [1,2]. The burning of fossil fuels produces a huge amount of toxic gases, polluting the environment and the food chain and causing the greenhouse effect [3]. With the rapid expansion of the world economy and the rising population, human beings are increasingly dependent on fossil fuels [4]. From 1987 to 2012, the average growth rate of energy consumed in the past 25 years has risen by more than 50%, and the demand will double in the next decade; therefore, it is necessary to seek clean energy sources [5][6][7]. Currently, clean energy material faces the challenges of high cost, low There are many reports on the preparation of graphene-based nanocomposites-for instance, ZnO nanoparticles can be dispersed with distilled water and tert-butanol to form a ZnO paste, and carbon nanomaterials were added to prepare graphene/ZnO nanocomposites films [32]. To prepare the WS3−X/rGO hybrids, GO was dispersed in dimethylformamide for ultrasonic treatment, and then, an aliquot of the dispersed droplet was immobilized on the bare, glassy carbon electrode; the suspension was dried under a lamp [33]. Zhang et al. [34] demonstrated how TiO2-graphene aerogel hybrids were prepared. At first, anhydrous ethanol was added to the GO solution, which was prepared by a modified Hummer's method and ultrasonically processed. Then, tetrabutyl titanate was added, and the reactant was sonicated. Next, the mixture was transferred to a stainless steel autoclave lined with Teflon, and TiO2-graphene hydrogel was subjected to hydrothermal treatment. Han et al. [35] fabricated SnO2/rGO nanocomposites using SnCl2 and GO precursors though a simple sonochemical method. In the process, Sn 2+ was first uniformly dispersed on the surface of GO, and then, SnO2 nanoparticles were oxidized in situ on both sides of graphene sheets with reducing GO under the action of continuous ultrasound.
In the electrochemical method, appropriate electrolyte and electrolysis conditions are the key to preparing graphene [21]. The electrochemical exfoliation of charcoal and graphite with different electrolytes was studied. Copper foil was used as a cathode and graphene as an anode. Both electrodes were immersed in electrolytes. Direct current voltage was applied to the electrochemical process to exfoliate graphite. In the same procedure, after using carbon as an anode to peel off charcoal, graphene sheets were obtained, and it was found that graphite rods were more effective than charcoal at forming suitable graphene sheets in the peeling process.
In the mechanical exfoliation method, the van der Waals force between the graphite molecular layers was overcome by physical force, and the graphite sheets were separated to obtain graphene. Ma et al. [36] reported that graphite powder was added into honey to obtain monolayer and multilayer graphene by mechanical exfoliation and stirring.
The top-down and bottom-up methods are commonly used to prepare graphene quantum dots [37]. In the top-down method, large-sized materials are cut into carbon sources containing a large amount of graphene structure and divided into quantum-sized graphene quantum dots through a series of physical, chemical or electrochemical methods. In the bottom-up method, graphene quantum dots are synthesized by the chemical In the electrochemical method, appropriate electrolyte and electrolysis conditions are the key to preparing graphene [21]. The electrochemical exfoliation of charcoal and graphite with different electrolytes was studied. Copper foil was used as a cathode and graphene as an anode. Both electrodes were immersed in electrolytes. Direct current voltage was applied to the electrochemical process to exfoliate graphite. In the same procedure, after using carbon as an anode to peel off charcoal, graphene sheets were obtained, and it was found that graphite rods were more effective than charcoal at forming suitable graphene sheets in the peeling process.
In the mechanical exfoliation method, the van der Waals force between the graphite molecular layers was overcome by physical force, and the graphite sheets were separated to obtain graphene. Ma et al. [36] reported that graphite powder was added into honey to obtain monolayer and multilayer graphene by mechanical exfoliation and stirring.
The top-down and bottom-up methods are commonly used to prepare graphene quantum dots [37]. In the top-down method, large-sized materials are cut into carbon sources containing a large amount of graphene structure and divided into quantum-sized graphene quantum dots through a series of physical, chemical or electrochemical methods. In the bottom-up method, graphene quantum dots are synthesized by the chemical method from small molecular carbon sources. Graphene quantum dots can be controlled in the process of preparation and post-treatment, so it is necessary to carefully optimize the preparation method. The surface passivation, functionalization, heteroatom doping or recombination of graphene quantum dots meets the needs of different applications. In the SiC epitaxy growth method, SiC single crystal is heated at a high temperature so that the Si atoms on the surface of SiC are evaporated and separated from the surface and the remaining carbon atoms are reconstructed by self-assembly to obtain graphene based on silicon carbide substrate [38].
In the CVD method, graphene is synthesized by decomposing methane and other carbonaceous compounds at a high temperature [39]. The growth mechanism can be divided into the following two categories: (a) The mechanism of carburizing and carbon deposition is that for metal substrates with a high carbon capacity such as nickel. The carbon atoms generated by the cracking of the carbon source penetrate the metal substrate at a high temperature, and then, the nucleation is precipitated from the inside when the temperature is lowered. (b) The surface growth mechanism, where the carbon atoms generated by the pyrolysis of a gas carbon source are adsorbed on the metal surface by copper and other low carbon metal matrix nucleates into graphene with growth at high temperatures. In rapid thermal treatment (RTT), high-quality graphene is synthesized from amorphous silicon carbide (a-SiC) by RTT. The SiO 2 /Si substrate is first covered with a-SiC film, and Cu and Ni films are deposited sequentially [40]. Then, the sample is thermally annealed by RTT and high-quality graphene is prepared in 3 min.
Ball milling is a common technology in the powder industry, which is a good method to produce shear force [24]. Both wet and dry ball milling methods are widely used to prepare graphene. The spheroidal grinding method is characterized by shear stress and high energy, which can prepare functional graphene.

Characterization of Graphene-Based Nanocomposites
In order to obtain the physicochemical properties of graphene-based nanocomposites (structure, shape, particle size, size distribution, valence state and specific surface area), these materials can be characterized by various techniques (Table 1). X-ray photoelectron spectroscopy (XPS) is mainly used to study the elements on the surface of materials and their different valence states. Peng et al. [41] synthesized high-performance graphenesupported electrocatalysts by laser-assisted solution. The valence state characterization of Ru by XPS showed that Ru mainly existed in the Ru 4+ state, while Pt mainly existed in the Pt 0 form, and a small part was Pt 2+ . A multi-collector inductively coupled plasma mass spectrometer (MC-ICPMS) is an elemental analysis instrument with a plasma ion source which can be used for the determination of element content and isotope analyses of samples. Vio et al. [42] revealed four lanthanide samples (Nd, Sm, Eu and Gd) that were isotopically characterized by MC-ICPMS. The isotope composition of heavy metals in pollutants can be used to determine the source; then, effective methods can be adopted to remove the pollutants. For example, Fan et al. [43] used rGO-supported nanoscale zero-valent iron composites to remove cadmium in aqueous solutions.
XRD was used for phase analysis of crystalline materials based on diffraction peak intensity, angular positions, relative intensity order and the shape of the diffraction peaks [44]. Iqbal et al. [45] fabricated a high-sensitivity shortwave infrared hybrid photodetector by integrating an organic charge transfer compound on a graphene transistor. XRD characterization confirmed that the prepared tetrathiafulvalene-chloranil film was polycrystalline, in which tetrathiafulvalene and chloranil molecules were stacked alternately along the a-axis.
Thermogravimetric analysis (TGA) is a method to measure the relationship between mass and temperature or time under programmed temperature. The composition, thermal stability and possible intermediate products could be obtained by analyzing the thermogravimetric curve [44]. Feng et al. [46] studied three graphene materials with different structures as the cathode of lithium-sulfur batteries. Pure sulfur has the characteristic of one-step weight loss in the temperature range of 190-300 • C by TGA. In the same temperature range, the three carbon/binary composites reproduced the weight loss curve of pure sulfur, and this the reason for the loss of sulfur. Scanning electron microscopy (SEM) is a surface imaging method in which the sample surface is scanned by an incident electron beam [47]. This interacts on the sample surface to generate signals reflecting the atomic composition and morphology of the sample surface, as well as the surface and cross-section morphology. Li et al. [48] studied the intercalation modulation method of graphene assembly and analyzed the fracture morphology of the material by SEM. It was found that the fracture of the plasticized graphene oxide paper was H-type. This indicated that the interlayer slip was activated by solvent intercalation, which determined the plastic transition of the paper. The structural information of molecular vibrational and rotational energy levels can be obtained by Raman spectroscopy analysis, so as to identify Crystals 2021, 11, 47 6 of 26 chemical bonds and functional groups of materials [44]. Tavakkoli et al. [49] reported graphene-based nanocomposites that were characterized by Raman spectroscopy, and it was found that the I D /I G = 0.06 of the material without N doping, and the I D /I G = 0.1 of the N-doped sample. The results demonstrated that the graphitization degree of N-doped carbon materials was high. EELS is the inelastic scattering of an incident electron beam in a sample, which is analyzed to obtain the element composition, chemical bond and electronic structure of samples [50]. Wang et al. [51] showed that the chemical bond configuration of BeO in a graphene reaction cell had obvious in-plane and inter-plane anisotropy through EELS characterization.
FESEM can be used to observe the surface morphology of nanomaterials. Zhang et al. [52] utilized FESEM to observe N-doped GO aerogels and graphene foam skeleton composites. It was found that this material maintained a porous and interconnected network structure and GO plates possessed obvious defects, which made N doping possible. Nuclear magnetic resonance (NMR) is a non-destructive testing method which does not interfere with the measurement system. Information such as molecular structure and dynamics can be obtained by detecting nuclear magnetic resonance signals. Among the NMR methods, 1 H-NMR and 13 C-NMR are the most widely used to determine molecular structures. Calovi et al. [53] described that solid-state NMR could confirm C-related dangling bonds and groups linked to hydrogen such as carboxyl, alcohol, epoxy and hydroxyl. X-ray powder diffraction (PXRD) is usually used to analyze the phase, purity, grain size, crystallinity and lattice parameters of materials. Xiong et al. [54] adopted PXRD to characterize the prepared COF ETBC-TAPT -graphene heterostructure ultra-sensitive photodetector and concluded that COF ETBC-TAPT was highly crystalline due to strong diffraction peaks at 3.08 • and 4.31 • .
Cyclic voltammetry (CV) is commonly used to detect electrode reaction parameters and determine its control steps and reaction mechanism. Zhang et al. [55] obtained the potassium storage mechanism of rGO by cyclic voltammetry and other methods, which showed that rGO significantly increased the specific capacity, cycle performance and rate performance of potassium-ion battery. Nitrogen adsorption can be used to analyze the specific surface area, pore size distribution, porosity, surface properties and other parameters of materials [56,57]. Cheng et al. [58] performed nitrogen adsorption/desorption (77.4 K) characterization on three-dimensional graphene and N-doped three-dimensional graphene (3D-PNG) materials. The results showed that the isotherms of 3D-PNG belong to type IV, when the relative pressure P/P 0 is 0.4 to 1, which is different from that of 3D-rGO. Generally, small-angle X-ray diffraction (SAXRD) is an important method to determine the ultra-large crystal plane spacing or film thickness as well as the micro periodic structure and periodic pore distribution of thin films. Hou et al. [29] characterized graphene composites by SAXRD and concluded that the Mn-doped Fe/rGO nanocomposites were partially ordered mesoporous materials.
HRTEM is a method to observe the internal structure, atomic arrangement and many fine structures (such as dislocations, twins, etc.) in crystals. Gao et al. [59] observed that graphene nanosheets were surrounded by graphite structures formed by polyacrylonitrile carbonization through HRTEM. Combined with other characterization techniques, it was disclosed that graphene could improve the mechanical properties of carbon fibers. An atomic force microscope (AFM) detects an object through the probe, obtains the threedimensional surface image of the sample surface and calculates the roughness, thickness and analysis of the three-dimensional morphology image. Yuan et al. [60] characterized graphene films grown on Cu substrates using an AFM. The images showed that there were clear atomic steps on the surface of the material, and wrinkles appeared along or perpendicular to the Cu atomic steps, indicating the existence of graphene films.
Fourier-transform infrared spectroscopy (FTIR) is applied to observe the bonding structure of a sample and interaction in a mixed film; the stretching and bending vibration absorption spectra of chemical bonds in the sample can be obtained [61]. Ali et al. [62] revealed via an FTIR study that the surface defects of N-doped graphene quantum dots were caused by strong surface oxidation or surface functional groups. Scanning electrochemical microscopy (SECM) is an electrochemical method based on a microscope which provides the transverse resolution characterization of an electrode surface, and the tip is located near the sample surface as a local sensor for detecting reaction products generated on the catalyst surface [55]. Besides, the spatial distribution of the surface selective region can be obtained by the submicron resolution of SECM [62]. Tavakkoli et al. [49] used SECM to analyze graphene-based nanocomposites simultaneously doped with single atoms of N, Co and Mo. Because more active centers are formed on the Ni substrate, the oxygen evolution reaction occurs near the catalyst/Ni interface. Transmission electron microscopy (TEM) can be employed to observe various defects inside materials and observe atomic structures directly by using accelerated and aggregated electron beams. For instance, Wang et al. [63] characterized graphene-carbon nanotube composites using TEM and found that the carbon nanotubes are distributed between graphene sheets, which increases the space between graphene and effectively reduces the accumulation of graphene. Spherical aberration-corrected scanning transmission electron microscopy and energy-dispersive X-ray spectroscopy (Cs-corrected STEM) could clearly reveal the size and interface structures among nanodots and oxide layers. Zhang et al. [64] used Cs-corrected STEM to study the synthesis of nitrogen-doped graphene-based Fe single-atom catalysts (Fe/NG) and found that nitrogen-doped graphene has Fe single-atom groups confined by nitrogen atoms. The valence state, stereochemistry and electronic configuration of the central atom can be obtained by X-ray absorption near-edge structure spectroscopy (XANES) [65]. Schlebrowski et al. [66] reported that near-edge X-ray absorption fine structure (NEXAFS) was used to determine the binding state of surface carbon atoms.
Synchrotron radiation source is a new type of light source with high intensity and high brightness. There are many experimental stations in the synchrotron radiation laboratory, e.g., XPS, SAXRD, XANES, X-ray absorption fine structure (XAFS), X-ray lithography, high spatial resolution X-ray imaging and 3D printing [67,68]. Orderliness XRPD, HRTEM, Nitrogen adsorption [59] Pore sizes/volumes Nitrogen adsorption [58] Specific surface area Nitrogen adsorption [58]

The Applications of Graphene-Based Nanocomposites
Graphene is a two-dimensional carbon nanomaterial with a hexagonal honeycomb lattice structure composed of carbon atoms and sp 2 hybrid orbitals which is widely used in many fields. In the following parts of this section, we will introduce the applications of graphene-based nanocomposites in HFCs, LIBs, solar cells and thermoelectric conversion.

. Hydrogen Fuel Cells
Water splitting is used as one of the basic hydrogen production units for hydrogen fuel cells, which is based on the following two semi-reactions [62]. Overall: HFCs are a power-generation device that converts the chemical energy of hydrogen and oxygen directly into electricity, which can reach more than 50% power generation efficiency [69]. The working principle of hydrogen fuel cells is the reverse reaction of electrolyzing water (Equations (1)- (3)). After hydrogen and oxygen are supplied to an anode and a cathode, respectively, hydrogen at the anode generates hydrogen ions through a catalyst, and the lost electrons reach the cathode through an external circuit to generate electric energy; meanwhile, hydrogen ions on the anode side reach the cathode through the PEM and react with oxygen to generate water. HFCs have the advantages of long driving distance, high power, no noise and no pollutant emission, and HFC vehicles are expected to become an environment-friendly vehicle in the future [70].
Two key components of HFCs are membrane electrode assembly and the bipolar plate; the membrane electrode consists of a PEM, a catalytic layer and a gas diffusion layer. The main effect of the PEM is to transport protons, separate reaction gases and insulate electrons [71]. The catalyst layer mainly carries the catalyst, which can promote the redox process of hydrogen and oxygen on the electrode and electric current is generated. The gas diffusion layer is composed of a base layer and a microporous layer, which requires high electrical conductivity, thermal conductivity and hydrophobicity. These key materials determine the life and performance of HFCs.

Classification of Fuel Cells
Fuel cells are electrochemical power generation devices that do not need to go through the Carnot cycle and have high energy conversion efficiency, which directly converts fuel (e.g., hydrogen, methanol, ethanol, natural gas and hydrocarbons) and an oxidant (e.g., air or pure oxygen) into electric energy and water in the presence of a catalyst [72][73][74]. Fuel cells have a wide range of sources of fuel and can help solve both energy and environmental problems. With the continuous breakthrough of research, fuel cells have also been applied in portable equipment, power stations, etc. [75].
Fuel cells can be divided into six categories according to electrolyte, fuel and temperature, as shown in Figure 2 [71,76]. Alkaline fuel cells (AFC) use KOH solution as an electrolyte, which can operate at lower temperatures around 80 • C; PEMFCs and direct methanol fuel cells (DMFCs) use polymer membrane as an electrolyte to transport protons. Among them, PEMFCs use PEM and Pt as catalysts and need to work under acidic conditions [73]. The phosphoric acid fuel cell (PAFC) uses pure phosphoric acid as the electrolyte and works at 200 • C. Molten carbonate fuel cells (MCFCs) use molten carbonate as an electrolyte and the operating temperature can reach 650 • C. The solid oxide fuel cell (SOFC) uses ceramic (solid oxide) materials as the electrolyte to conduct oxygen anions from the cathode to the anode at a working temperature of 1000 • C [77]. In addition, microbial fuel cells are a new type of bioelectrochemical device which utilize a biocatalyst to convert the chemical energy stored in organic wastewater into power supply and remove pollutants while generating electricity [78]. However, microbial fuel cells have high production cost and low power density and energy conversion efficiency. The principle, electrode, catalyst, advantages and disadvantages of the fuel cells are shown in Table 2.
anions from the cathode to the anode at a working temperature of 1000 °C [77]. In addition, microbial fuel cells are a new type of bioelectrochemical device which utilize a biocatalyst to convert the chemical energy stored in organic wastewater into power supply and remove pollutants while generating electricity [78]. However, microbial fuel cells have high production cost and low power density and energy conversion efficiency. The principle, electrode, catalyst, advantages and disadvantages of the fuel cells are shown in Table 2.

Applications of Graphene-Based Nanocomposites in Hydrogen Fuel Cells
The incorporation of GO into the PEM of HFCs can improve the proton conductivity and stability of the membrane and reduce the fuel permeability of the membrane [79]. Choi et al. [80] fabricated a GO/Nafion composite membrane and found that the methanol permeability of the Nafion membrane was greatly reduced by adding GO as a filler. Because GO is hydrophilic, it can interact with the non-polar main chain and polar ion cluster of the Nafion PEM to modify the microstructure of the hydrophilic and hydrophobic regions of PEM. Thus, the phase separation of the membrane is reduced in the presence of water molecules and the permeability of the proton exchange membrane is improved. Hu et al. [81] reported breakthroughs in the study of the proton transport characteristics of graphene-based membrane materials and found that two-dimensional

Applications of Graphene-Based Nanocomposites in Hydrogen Fuel Cells
The incorporation of GO into the PEM of HFCs can improve the proton conductivity and stability of the membrane and reduce the fuel permeability of the membrane [79]. Choi et al. [80] fabricated a GO/Nafion composite membrane and found that the methanol permeability of the Nafion membrane was greatly reduced by adding GO as a filler. Because GO is hydrophilic, it can interact with the non-polar main chain and polar ion cluster of the Nafion PEM to modify the microstructure of the hydrophilic and hydrophobic regions of PEM. Thus, the phase separation of the membrane is reduced in the presence of water molecules and the permeability of the proton exchange membrane is improved. Hu et al. [81] reported breakthroughs in the study of the proton transport characteristics of graphene-based membrane materials and found that two-dimensional nanocomposites with a single atomic layer thickness such as graphene and boron nitride could be used as good proton-conducting membranes.
HFCs mostly use precious metal as catalysts, such as Ru, Rh, Pd, Ag, Ir, Pt and Au, which are expensive. Precious metal catalytic activity can be improved and the cost can be reduced by improving the carrier type, structural constitution and surface characteristics. For example, when single-atom Rh is anchored on nitrogen-doped carbon, the strong interaction between carbon and Rh can effectively prevent the migration and aggregation of Rh metal particles and improve the stability of the catalyst [82]. Carbon-supported trimethoxy nanocomposites as fuel cell catalysts can elevate the activity of electrocatalysts and reduce the cost [83]. Graphene can complement the network of electroactive materials to buffer the change in electrode volume and prevent the breakage and aggregation of electrode materials [84][85][86][87]. Density functional theory is the powerful tool electronic structures based on first principles and has been widely used in fuel cells [88]. Mineva et al. [89] used density functional theory to study the active sites of Fe-N-C cathode catalysts for fuel cells.

Selection of Catalysts for Hydrogen Fuel Cells
The choice of carrier material is key to the activity, selectivity, durability and costeffectiveness of a catalyst [79]. One of the ways to improve the performance of fuel cells is to increase the power density by reducing the membrane thickness and membrane resistance of the fuel cell [80]. Electrocatalysis is the main driving factor of the reactions between the anode and cathode sides of fuel cells, and the performance of each electrode reaction is dependent on the specific surface area of the catalyst. The ideal catalyst carrier should have the following conditions [90][91][92][93][94]: (i) It has conductivity, or a catalyst carrier with good conductivity is used to obtain high conductivity; (ii) It has certain chemical stability, i.e., under the conditions of achieving the target reaction, the surface of the electrocatalyst will not be inactivated prematurely due to electrochemical reactions; (iii) The larger the surface area, the more active sites and the better the performance of the material; (iv) Appropriate porosity is conducive to electron and proton conduction, material transmission and water discharge; (v) The catalyst is easy to recycle, thus reducing costs and avoiding environmental pollution.  LIBs belong to the category of a secondary battery (rechargeable battery), which is essentially a lithium-ion concentration difference battery [95,96]. The positive and negative electrodes are composed of two different lithium-ion intercalation compounds. When charging, lithium ions are extracted from the positive electrode compound and inserted into the lattice, and the positive electrode is in a lithium-poor state; during discharge, lithium ions are extracted from the negative electrode and inserted into the positive electrode, and the positive electrode is in a lithium-rich state. LIBs have been widely used in cell phones, hybrid electric vehicles and pure electric vehicles, unmanned aerial vehicles and satellites due to their advantages of high specific energy, high working voltage, long cycle life, no memory effect and low pollution [92][93][94][95][96][97][98][99][100].
During the charge-discharge cycle of LIBs, the deposition and exfoliation of lithium metal is usually uneven, which can lead to the growth of lithium dendrites [101]. The growth of lithium dendrites can cause instability of the electrode and electrolyte interface of the LIBs during the cycle, destroying the solid electrolyte interface film formed, and the lithium dendrites will continue to consume the electrolyte during the growth process and produce metal lithium. During the growth process of lithium dendrites, the electrolyte will be consumed continuously and lead to irreversible deposition of lithium metal, which causes a low coulomb efficiency. The formation of lithium dendrites can even pierce the diaphragm and result in internal short-circuit of a lithium-ion battery, which leads to thermal runaway of the battery and brings about combustion and explosion.
The electrode material of LIBs is mostly composed of metal ions such as nickel manganese cobalt or nickel cobalt aluminum; however, cobalt metal is scarce and expensive. With the increasingly prominent issue of energy shortage and environmental pollution, people urgently need low-cost cobalt-free batteries to get rid of the dependence of electrode materials on cobalt.
Graphene is a potential conductive agent for LIBs which has high electrical conductivity and flexible, two-dimensional, ultra-thin structural characteristics [98]. Through the "surface-point" contact mode between graphene and active materials, the low conductivity threshold of graphene allows a small amount of graphene to effectively increase the electronic conductivity of the electrode and greatly reduce the use of conductive agents as inactive materials, and this can effectively increase the volume energy density of the battery. Graphene has many applications in LIBs, as shown in Figure 3.

Classification of Lithium-Ion Batteries
According to cathode materials, the commonly used LIBs can be divided into the following six categories: lithium nickel cobalt manganese oxide, lithium nickel cobalt aluminate, lithium iron phosphate, lithium cobalt oxide, lithium carbonate battery and lithium manganese oxide, among which nickel cobalt manganese oxide and nickel cobalt aluminate are ternary cathode materials. The electrochemical performance of lithium-ion batteries mainly depends on the structure and performance of the electrode materials and the electrolyte materials [103]. Cathode materials are not only used as electrode materials to participate in electrochemical reaction but also the source of lithium ion.

Applications of Graphene-Based Nanocomposites in Anodes for Lithium-Ion Batteries
As one of the main functional components of LIBs, anode materials play a vital role in the performance of LIBs. However, the theoretical capacity of industrial graphite anode is relatively low, which leads to a reduction in the energy density of LIBs; hence, using graphite as the anode of LIBs faces a clear challenge. Silicon, phosphide, metal oxide, sulfide, etc., can significantly increase the energy density of LIBs and their low cost has been extensively studied, as well as large volume change and poor cycle stability [104,105]. On the basis of these materials, composite graphene can solve the problem, e.g., cobalt sulfide's own capacity decay, low conductivity and cyclicity, which are related to the volume expansion of the electrode, resulting in specific and capacity decays and the formation of polyazo anions [105]. Graphene blocks dissolute in electrolytes by absorbing polyazo anions on amorphous surfaces and enhance ionic conduction and improve the cycle life.
Hou et al. [106] fabricated G⊥FP@C-NA nanocomposites as the anode materials of LIBs (Figure 4). The structure of the materials was carbon-coated FeP nanorod array (FP@C-NA)grown vertically on the conductive reduced graphene oxide (G) network. Due to the unique structure of G⊥FP@C-NA, it exhibits obvious conductivity and structural stability in lithium storage. For example, this material provides high lithium storage capacity (1106 mA h g −1 at 50 mA g −1 ), excellent rate capability (565 mA h g −1 at 5000 mA g −1 ) and long-term cycle stability (1009 mA h g −1 at 500 mA g −1 after 500 cycles, The solid-liquid composite electrolyte is a gel electrolyte composed of a solid polymer and a liquid electrolyte, which has high ionic conductivity, low electronic conductivity and good thermal and mechanical stability [102]. Inorganic gels can reduce the interaction between anions and cations, thereby improving the dissociation of lithium salts and enhancing the migration of lithium ions. Ceramic-graphene nanocomposites are currently tough solid electrolytes with high electrical conductivity, which can greatly solve the above problems [101].

Classification of Lithium-Ion Batteries
According to cathode materials, the commonly used LIBs can be divided into the following six categories: lithium nickel cobalt manganese oxide, lithium nickel cobalt aluminate, lithium iron phosphate, lithium cobalt oxide, lithium carbonate battery and lithium manganese oxide, among which nickel cobalt manganese oxide and nickel cobalt aluminate are ternary cathode materials. The electrochemical performance of lithium-ion batteries mainly depends on the structure and performance of the electrode materials and the electrolyte materials [103]. Cathode materials are not only used as electrode materials to participate in electrochemical reaction but also the source of lithium ion.

Applications of Graphene-Based Nanocomposites in Anodes for Lithium-Ion Batteries
As one of the main functional components of LIBs, anode materials play a vital role in the performance of LIBs. However, the theoretical capacity of industrial graphite anode is relatively low, which leads to a reduction in the energy density of LIBs; hence, using graphite as the anode of LIBs faces a clear challenge. Silicon, phosphide, metal oxide, sulfide, etc., can significantly increase the energy density of LIBs and their low cost has been extensively studied, as well as large volume change and poor cycle stability [104,105]. On the basis of these materials, composite graphene can solve the problem, e.g., cobalt sulfide's own capacity decay, low conductivity and cyclicity, which are related to the volume expansion of the electrode, resulting in specific and capacity decays and the formation of polyazo anions [105]. Graphene blocks dissolute in electrolytes by absorbing polyazo anions on amorphous surfaces and enhance ionic conduction and improve the cycle life.
Hou et al. [106] fabricated G⊥FP@C-NA nanocomposites as the anode materials of LIBs ( Figure 4). The structure of the materials was carbon-coated FeP nanorod array (FP@C-NA)grown vertically on the conductive reduced graphene oxide (G) network. Due to the unique structure of G⊥FP@C-NA, it exhibits obvious conductivity and structural stability in lithium storage. For example, this material provides high lithium storage capacity (1106 mA h g −1 at 50 mA g −1 ), excellent rate capability (565 mA h g −1 at 5000 mA g −1 ) and long-term cycle stability (1009 mA h g −1 at 500 mA g −1 after 500 cycles, 1310 mA h g −1 at 2000 mA g −1 after 2000 cycles). This attractive array structure design can also be extended to other different energy storage systems, which will promote the development of next-generation electrochemical energy storage devices with high power and high energy density.  The stacking of graphene sheets in the process of electrode preparation leads to slow ion transport and decreased surface activity, which hinders the practical application. Based on this, compared with in-plane graphene, porous graphene has the dual advantages of graphene and porous materials. Porous graphene nanosheets and interconnected nanocages have layered porous channels, fast ion transmission, complete conductive networks and good structural durability to improve the cycle stability [107,108] Zhu et al. [109] pointed out that these porous graphene nanosheets and interconnected nanocages could effectively prevent the aggregation of graphene and maintain its layered structure, which is very important for high-power battery electrodes.
Silicon has been considered as a good alternative material which can significantly increase the energy density when used as an anode material. Therefore, the energy density of a battery constructed with a silicon anode was about 25% higher than that of a commercial battery using a traditional graphite anode [105].
Han et al. [35] revealed that when using a SnO2/graphene nanocomposite as the LIB anode, the initial discharge specific capacity was as high as about 1610 mA h g −1 , which could significantly improve the specific capacity of the electrode. After 100 cycles, the specific capacity remained at 87%, showing a good cycle stability. The main reason was that graphene could provide mechanical support to reduce the volume change of the SnO2 anode and prevent nanoparticle agglomeration, and it could also act as a conductive network to promote charge transfer and Li + diffusion. The reaction mechanism was The stacking of graphene sheets in the process of electrode preparation leads to slow ion transport and decreased surface activity, which hinders the practical application. Based on this, compared with in-plane graphene, porous graphene has the dual advantages of graphene and porous materials. Porous graphene nanosheets and interconnected nanocages have layered porous channels, fast ion transmission, complete conductive networks and good structural durability to improve the cycle stability [107,108] Zhu et al. [109] pointed out that these porous graphene nanosheets and interconnected nanocages could effectively prevent the aggregation of graphene and maintain its layered structure, which is very important for high-power battery electrodes.
Silicon has been considered as a good alternative material which can significantly increase the energy density when used as an anode material. Therefore, the energy density of a battery constructed with a silicon anode was about 25% higher than that of a commercial battery using a traditional graphite anode [105].
Han et al. [35] revealed that when using a SnO 2 /graphene nanocomposite as the LIB anode, the initial discharge specific capacity was as high as about 1610 mA h g −1 , which could significantly improve the specific capacity of the electrode. After 100 cycles, the specific capacity remained at 87%, showing a good cycle stability. The main reason was that graphene could provide mechanical support to reduce the volume change of the SnO 2 anode and prevent nanoparticle agglomeration, and it could also act as a conductive network to promote charge transfer and Li + diffusion. The reaction mechanism was proposed according to the following equation: Sn(OH)Cl + GO → SnO 2 /graphene + HCl (5) In the reaction, SnCl 2 is firstly hydrolyzed to Sn(OH)Cl, and subsequently, Sn(OH)Cl is oxidized to SnO 2 , reducing GO to graphene under continuous ultrasonic force. It is established that two-step reactions are involved in SnO 2 -based electrodes:

Applications of Graphene-Based Nanocomposites in Cathodes for Lithium-Ion Batteries
Among the various components involved in LIBs, cathodes currently limit the energy density and dominate the batteries' cost [110]. Lithium transition metal phosphates or oxides are used as cathode materials for LIBs, e.g., LiFePO 4 or LiCO 2 [106,110,111]. Venkateswara et al. [112] showed the results of using a lithium transition metal oxide (LiNi 1/3 Co 1/3 Mn 1/3 O 2 ) as a cathode for the second generation of LIBs, which faced capacity fading. Graphene is incorporated into these materials to obtain a higher charge-discharge rate and capacity retention rate [103]. Graphite materials instead of lithium-containing transition metal oxides as cathodes of LIBs can reduce the overall cost and potential environmental pollution [113].
Wang et al. [107] used density functional theory to study a high-carbon acyl C 6 O 6 molecular functionalized graphene electrode as an LIB cathode. Binding energy calculations revealed that C 6 O 6 molecules maintain good electronic conductivity after being adsorbed on graphene by physical adsorption. The reduced voltage of the C 6 O 6 @graphene cathode was between 1.5 and 2.6 V, and the energy density was about 155 mA h g −1 , which is a promising renewable electrode material. Lee et al. [114] demonstrated that Na 2 C 6 O 6 , as the cathode of the battery, had an energy density of 726 Wh kg −1 , an energy efficiency above 87% and a good cycle retention, and the storage of tetrasodium in the Na 2 C 6 O 6 electrode was realized.
Vanadium pentoxide has limitations such as insufficient electronic conductivity, sluggish ion diffusion and volume expansion. In order to solve these problems, Nakhanivej et al. [115] demonstrated spray-frozen assembly into hierarchically structured open porous vanadium pentoxide/reduced GO composite microballs for high-performance LIB cathodes. When the current density increases from 100 to 1000 mA h g −1 , the capacity retention of vanadium pentoxide/reduced GO composite microballs is about 51.3%, which is much greater than that of vanadium pentoxide particles (36.4%). The capacity retention of 80.4% with a coulombic efficiency of 97.1% over 200 cycles is two times greater than that of V 2 O 5 particles, indicating improved cycle stability.

Solar Cells
Solar cells are devices that convert light energy into electrical energy directly through photoelectric or photochemical effects, which are usually called photovoltaic cells [116]. Solar cells have become the focus of attention due to high efficiency, long-term stability and low cost.

Classification of Solar Cells
According to different materials, solar cells are divided into crystalline silicon solar cell, multiple compound thin-film solar cells, polymer multilayer modified electrode solar cells, nanocrystalline solar cells, organic solar cells and plastics solar cells. Among them, crystalline silicon solar cells are the first generation of solar cells and are currently the most mature [117]. These are divided into monocrystalline silicon solar cells and polycrystalline silicon solar cells. The structure of monocrystalline silicon solar cells is mainly p-n junction. Second-generation solar cells mainly refer to thin-film solar cells, represented by copper indium gallium selenium films and cadmium telluride (CdTe) films ( Figure 5) [118,119]. Emerging solar cells include dye-sensitized cells, organic solar cells, perovskite solar cells and quantum dot (QD) solar cells [120,121]. Additionally, plastics solar cells have simple production processes, low energy consumption and low cost and the materials can be degraded, which is green and environmentally friendly products. plastics solar cells have simple production processes, low energy consumption and low cost and the materials can be degraded, which is green and environmentally friendly products.

The Applications of Graphene-Based Nanocomposites in Solar Cells
As a new type of material, graphene has been used for transparent electrodes of solar cells and electron acceptor materials in solar cells. The application of graphene in solar cells is analyzed to provide theoretical support for the continuous research on new materials and technologies and will effectively promote the overall development of the new energy industry.
Dye-sensitized solar cells as third-generation photovoltaic cells are a new generation of non-silicon-based solar cells, which mainly use nano-TiO 2 and photosensitive dye as raw materials to simulate the photosynthesis of plants in nature and convert solar energy into electricity [104,[120][121][122][123]. Cheng et al. [124] reported that the conversion efficiency of TiO 2 @rGO hybrids was higher than that of TiO 2 (4.78%). Kavan et al. [122] revealed that the energy conversion efficiency of dye-sensitized solar cells with three-dimensional honeycomb graphene pair electrode has reached up to 7.8%.
Few layers of graphene/n-Si Schottky junction solar cells can further improve the energy conversion efficiency [124,125]. However, this material is highly difficult to prepare and thus its practical application is clearly limited [126]. Sun et al. [127] reported that N-layer graphene can be obtained without any organic impurities between the layers. After the Cu (or Ni) foil is etched, the two-layer graphene film is directly transferred to the third layer of graphene on the Cu (or Ni) foil, thereby forming a three-layer graphene film. Finally, the multilayer graphene can be transferred to other substrates and the polymethylmethacrylate (PMMA) acetone on the top can be removed, as shown in Figure 6. terials and technologies and will effectively promote the overall development of the new energy industry. Dye-sensitized solar cells as third-generation photovoltaic cells are a new generation of non-silicon-based solar cells, which mainly use nano-TiO2 and photosensitive dye as raw materials to simulate the photosynthesis of plants in nature and convert solar energy into electricity [104,[120][121][122][123]. Cheng et al. [124] reported that the conversion efficiency of TiO2@rGO hybrids was higher than that of TiO2 (4.78%). Kavan et al. [122] revealed that the energy conversion efficiency of dye-sensitized solar cells with three-dimensional honeycomb graphene pair electrode has reached up to 7.8%.
Few layers of graphene/n-Si Schottky junction solar cells can further improve the energy conversion efficiency [124,125]. However, this material is highly difficult to prepare and thus its practical application is clearly limited [126]. Sun et al. [127] reported that N-layer graphene can be obtained without any organic impurities between the layers. After the Cu (or Ni) foil is etched, the two-layer graphene film is directly transferred to the third layer of graphene on the Cu (or Ni) foil, thereby forming a three-layer graphene film. Finally, the multilayer graphene can be transferred to other substrates and the polymethylmethacrylate (PMMA) acetone on the top can be removed, as shown in Figure  6. In 2018, Li et al. [128] introduced the self-doping process of the graphene-silicon Schottky junction induced by steam oxidation of HNO3. The p-type self-doping of graphene oxide with HNO3 enhanced the intrinsic potential and conductivity, thus increasing the charge transfer and inhibiting the carrier recombination at the graphene-silicon Schottky junction. After HNO3 oxidation treatment, the open circuit voltage increased from 0.36 to 0.47 V, the short-circuit current increased from 0.80 to 7.71 μA, and the ideal coefficient was optimized from 4.4 to 1.0. The enhanced performance of graphene-silicon solar cells was due to the oxidation-induced p-type self-doping of the graphene-silicon junction.
Organic photovoltaic (OPV) cells use photosensitive organics as semiconductor materials and generate voltage to form current through the photovoltaic effect [129]. So- In 2018, Li et al. [128] introduced the self-doping process of the graphene-silicon Schottky junction induced by steam oxidation of HNO 3 . The p-type self-doping of graphene oxide with HNO 3 enhanced the intrinsic potential and conductivity, thus increasing the charge transfer and inhibiting the carrier recombination at the graphene-silicon Schottky junction. After HNO 3 oxidation treatment, the open circuit voltage increased from 0.36 to 0.47 V, the short-circuit current increased from 0.80 to 7.71 µA, and the ideal coefficient was optimized from 4.4 to 1.0. The enhanced performance of graphene-silicon solar cells was due to the oxidation-induced p-type self-doping of the graphene-silicon junction.
Organic photovoltaic (OPV) cells use photosensitive organics as semiconductor materials and generate voltage to form current through the photovoltaic effect [129]. Solar cells have the advantages of low manufacturing cost, light weight, strong versatility and ease of use. Graphene and its derivatives can be used as interface materials in organic solar cells, e.g., transparent electrode, active layer, interface layer and counter electrode [130,131]. Lee et al. [132] functionalized GO with (3-glycidoxypropyl) trimethoxysilane (GPTMS) and triethoxymethylsilane (MTES) to obtain GPTMS-sGO and MTES-sGO. The power conversion efficiency (PCE) values of OPV batteries of GPTMS-sGO and MTES-sGO are 3.00% and 3.08%, respectively. The PCE value is a measure of the ability of solar cells to convert light energy into electrical energy and is determined by the short-circuit current density and filling coefficient.
Perovskite solar cells (PSCs) are solar cells using perovskite-type organic metal halide semiconductors as light absorbing materials, which are considered as one of the most promising power-generating pieces of equipment in the future [133]. Two-dimensional-type PSCs have obvious long-term stability and are more suitable for application. Graphene and its derivatives have environmental stability and compactness, and their multi-functional groups can modify perovskite films and the interface layers of devices. Liu et al. [134] prepared 2D PSCs with dimethylformamide (DMF) and γ-butyrolactone (GBL) as precursor solvents. The results show that the GBL-treated 2D PSC shows better quality, less pinholes and higher crystallinity, and the PCE is 11.17%, which is significantly higher than that of the DMF-treated devices. Meanwhile, the open-circuit voltage and short-circuit current density are improved, with stable current output and negligible hysteresis. Redondo Obispo et al. [135] studied graphene doping in the poly(3,4-ethylene dioxythiophene)/poly (styrene sulfonate) PEDOT/PSS as the hole extraction layer (HTL) of inverted perovskite solar cells. The results showed that the conductivity of the PEDOT/PSS layer was improved by adding graphene, and the PCE of the solar cell was increased by 33%. This was likely to be caused by the increased conductivity of graphene-doped PEDOT/PSS, which could improve the charge collection efficiency.
Graphene plays a role in the front and back buffer layers in copper (In, Ga) (S, Se) materials for thin-film solar cells; for instance, Kur transmittance and higher thermal stability [136]. The nano layer has some functions, e.g., reducing the consumption of aluminum and the cost of deposition, which offer high conductivity and transmittance. The conductive metal contact material can be controlled by changing the number of graphene layers. In addition, a highly conductive three-dimensional graphene network is introduced at the contact between the CdS/CdTe interface and the back of the CdS/CdTe thin-film solar cells and the conversion efficiency is improved.

Thermoelectric Conversion
Thermoelectric conversion realizes the direct output of heat energy to electric energy [137]. A thermoelectric power generation device is a basic thermoelectric power generation unit that realizes electric energy output. Thermoelectric devices have the advantages of long service life, no noise and environmental friendliness, and they can be integrated in various components and have a certain flexibility. Thermoelectric conversion is a type of green energy which is mostly used in waste heat recovery in aerospace, aviation and civil industries [138].
Thermoelectric materials mainly include inorganic and organic polymer materials. Inorganic thermoelectric materials comprise Bi 2 Te 3 -, SiGe-, Al 2 O 3 -and CoSb 3 -based materials, and organic polymer thermoelectric materials primarily comprise polyaniline, PEDOT, polypyrrole and polythiophene [103,139]. Organic polymer thermoelectric materials are basically composed of conductive polymers and their inorganic filler composites. Graphene is also considered as a promising thermoelectric material because of its intrinsic ultra-high conductivity and moderate Seebeck coefficient [103]. The thermal conductivity can be effectively improved by combining carbon materials with organic polymer materials with low thermal conductivity [140].

Classification of Thermoelectric Materials
According to their operating temperature, thermoelectric materials can be divided into three categories [103]. (a) Bismuth telluride and its alloy: this material is widely used in thermoelectric coolers, and its optimum operating temperature is lower than 450°C; (b) lead telluride and its alloys: they are widely used in thermoelectric generators, and their optimum operating temperature is about 1000°C; (c) Si-Ge alloy: this kind of material is also used in thermoelectric generators, and its optimum operating temperature is about 1300 • C.

The Applications of Graphene-Based Nanocomposites in Thermoelectric Materials
The excellent thermoelectric materials should have higher electric conductivity and lower thermal conductivity, while the intrinsic graphene electron mobility can improve the conductivity of the materials. Although the thermal conductivity of graphene is excellent, graphene can reduce the lattice thermal conductivity of thermoelectric materials by embedding itself into the matrix of traditional thermoelectric materials, thereby obtaining new composite thermoelectric materials with high energy conversion efficiency [141].
Chen et al. [142] reported the radiative heat and photothermoelectric conversion characteristics of three-dimensional microporous graphene excited by a terahertz wave. Due to the rich chemical modification and band gap regulation properties of three-dimensional graphene, oxygen-containing functional group evolution and carrier type transition could occur through high-temperature treatment. Meanwhile, three-dimensional microporous graphene has a pore size equivalent to the terahertz wavelength, so the incident terahertz wave is completely absorbed, which can increase the temperature gradient of the device and improve the sensitivity of terahertz detectors. Feng et al. [143] reported a graphene/CoSb 3 composite thermoelectric material that was prepared by the chemical method and sintering process. When the temperature was 800 K, the ZT max of the graphene-based nanocomposite was 0.6, which was 1.3 times higher than that of CoSb 3 .

Assessment Methods for Four Clean Energy Sources
Performance evaluation methods of graphene-based nanocomposites in the applications of clean energy materials have gradually become a research focus. This section uses mathematical models to assess the performance of hydrogen fuel cells, lithium-ion batteries, solar cells and thermoelectric materials.

Assessment of Hydrogen Fuel Cells
The Scherer equation (Equation (8)) is used to calculate the average crystallite size of graphene-based nanocomposites by using planes from XRD data.
where d is the average particle size (nm); B2θ is the full width of the diffraction peak at half maximum in radians; λ denotes the wavelength and θ stands for the angle at the position of the peak maximum [144]. The ion-exchange capacity (IEC) of a film was determined using a titration method. At first, the dried membrane was put into 1.0 M HCl for 24 h, washed with distilled water several times and dried in a vacuum oven at 50 • C for 24 h. Then, the above-mentioned dried film was immersed in 1.0 M NaCl solution for 24 h [25]. Finally, phenolphthalein was added as a pH indicator and titrated with 0.01 M NaOH standard solution, and the meq/g value of IEC was obtained by Equation (9): Volume o f NaOH Consumed × Normality o f NaOH Dry weight o f the membrane samples (9) Water uptake of a membrane is calculated using Equation (10): where W 1 and W 2 are the mass of the wet membrane after immersion in demineralized water for 24 h and the mass of the dry membrane after drying in an oven at 100 • C for 24 h, respectively. In order to obtain the resistance of a film, the proton conductivity σ (s/cm) of the membrane is calculated using the following Equation (11): where L is the thickness of the membrane (cm); R is the resistance of the membrane (Ω) and A denotes the surface area of the membrane (cm 2 ). The membrane resistance is measured by electrochemical impedance spectroscopy [145]. The equivalent circuit is fitted to a diagram using software, and the resistance of the film is obtained. Methanol permeability is calculated from the slope of the graph between diffusion time and methanol concentration by the following equation (12): where C A and C B are the concentrations of methanol on side A and side B, respectively; A is the area of the membrane; L denotes the thickness of the membrane; V B is volume of the cells in side B and D and K are the methanol diffusion and partition coefficient, respectively [146]. The product of DK is membrane methanol permeability (P), the methanol concentration on side B increases linearly with the increase of permeation time [147]. P can be calculated by the Equation (13): where m is the slope of C B (T) linear graph. The membrane characteristics required for high performance of fuel cells are high proton conductivity and low methanol permeability [43]. The membrane performance is calculated by the Equation (14): where Φ is a selectivity parameter, σ is the ionic conductivity and P is the methanol permeability.

Assessment of Lithium-Ion Batteries
In the molecular dynamics simulation of LIB electrolytes, the cluster and solvent molecules were identified and classified by using graph theory tools (Table 3) [148,149]. Table 3. Calculation for lithium-ion batteries' electrolytes.

Calculation Type Formula Description Reference
Diffusivity where r i (t) is the position of molecule i at time t; t 0 represents the set average value at the starting time. [148] Einstein relation where D is the self-diffusion coefficient and i represents a species. [148] Arrhenius where k is the rate constant, R is the molar gas constant, T denotes the thermodynamic temperature, Ea is the apparent activation energy and A stands for the pre-exponential factor.

Calculation Type Formula Description Reference
Ionic conductivity where λ (t) is the value of λ for computing the time t, which means the set average value at the starting time T 0 ; e denotes the charge of the electron; V is the volume of the simulation box; K B is the Boltzmann constant; T represents the temperature; Z i is the charge of the ion i in the unit of electron; r i (t) is the position of the ion i in the time t and accounts for the sum of all N ions in the simulation box. [148] Electrical conductivity σ = nqµ where σ describes electrical conductivity and µ represents the electrical mobility. [149] Nernst-Einstein µ = qD kT where D denotes the diffusion coefficient. [149]

Assessment of Solar Cells
There are three steps in the photovoltaic conversion process of most solar cells [150,151]. (a) Light is absorbed and generated electron-hole pairs in the semiconductor layer; (b) electrons and holes is separated by the built-in electric field in the junction area; (c) a small amount of carriers are extracted to the external electrode [152,153]. The difference between the two types of junctions is the size of the built-in electric field and the height of the barrier, which are closely related to the rate of hole/electron separation and recombination. A steady progress has been made in hybrid graphene/silicon Schottky solar cells based on unconventional Schottky Junctions, and the calculation of power conversion efficiency of hybrid graphene/silicon Schottky solar cells is shown in Table 4 [154]. Table 4. Rules held for the Schottky solar cells.

Calculation Type Formula Description Reference
Transport mechanism where J S is the reversed saturated current density, A* is the effective Richardson constant, n denotes the diode ideality factor, e is the elementary charge, K B represents the Boltzmann's constant, T is the absolute temperature, Φ SBH is Schottky barrier height and V denotes junction bias. [150,151] Extracting series resistance (R S ) J = J S exp eV−IR S nK B T the R S of graphene/silicon device can be extracted by plotting dV/dlnJ as a function of J. [152] Filling factor (FF) FF = V P I P V OC I SC where V P and I P are the voltage and current at the optimal working point, respectively; V OC is open-circuit voltage and I SC is short-circuit current. [153] Power conversion efficiency (PCE) η = V OC I SC FF P in where P in stands for the power of incident light. [154] Ideality factor n n = q/κTdV/d ln J where q is the electron charge, T denotes the temperature, κ is the Boltzmann constant, V is the forward voltage, and J stands for the current density. [142] In crystalline silicon solar cells, increasing minority carrier lifetime is the key to improve the efficiency of solar cells [145]. The minority carrier lifetime is related to the Crystals 2021, 11, 47 20 of 26 diffusion length of crystalline silicon solar cells, typically the diffusion length is 100-300 µm. The minority carrier lifetime is calculated with the Equation (15): where L is the diffusions length in meters; D is the diffusivity in m 2 /s and τ denotes the lifetime in seconds. Effect of light transmittance on thickness of graphene electrode, T and Rs of graphene transparent electrode is a key factor for the performance of PSCs, which strongly depended on the thickness of graphene thin film or the number of graphene layers (Table 5) [155,156]. Table 5. Calculation of the thickness of graphene electrodes.

Calculation type Formula Description Reference
Graphene where N is the number of layers. [155] Graphene thin film R S = σ 2D N −1 R s is linked to the bi-dimensional conductivity σ 2D . [156] Free-space impedance ε 0 denotes the free-space electric constant, c is the speeds of light and Go/σ 2D constants of graphene thin films represent taken from the literature. [156]

Assessment of Thermoelectric Materials
The conversion efficiency of thermoelectric module is determined by Carnot efficiency, which is calculated by Equation (16): where T hot is the hot-side temperature; T cold denotes the cold-side temperature; T m represents the average temperature [137]; ZT is the thermoelectric figure-of-merit. ZT is calculated by the Equation (17): where S, σ, k L , k e , and T are the seebeck coefficient, electrical conductivity, lattice thermal conductivity, electronic thermal conductivity and absolute temperature, respectively [76]. Temperature difference (T hot -T cold ) and the improvements in thermoelectric conversion efficiency entirely relied upon the ZT enhancement.

Summary and Future Perspectives
This paper reviews the applications of graphene-based nanocomposites in hydrogen fuel cells, lithium-ion batteries, solar cells and thermoelectric conversion, as well as their evaluation methods. Moreover, the preparation and characterization of graphene-based composites are introduced.
There are some research interests in graphene-based nanocomposites used in energy devices, which have been proven to be suitable for the development of basic materials for alternative energy sources; for example, (a) in hydrogen fuel cells, graphene is used as an electrode material to enhance electrocatalytic activity; (b) in lithium-ion batteries, graphene-based nanocomposites show better performance as they have high power density and energy density and a fast charging speed; (c) in solar cells, graphene-based composites are used in photovoltaic devices because of their unique characteristics of high carrier mobility and low resistivity, transmittance and filling two-dimensional networks, which improves the power conversion efficiency; (d) in thermoelectric materials, graphene-based nanocomposites with a polycrystalline structure can achieve high-efficiency thermoelectric energy conversion. Currently, in Ningxia, China, commercial production of graphene version 3.0 has started, with an annual output of 2000 tons of graphene-modified conductive paste (20 tons of graphene). In terms of power distribution, AI will be a core part of future smart grids; this technology will continue to collect and integrate data from millions of smart sensors so that decisions can be made in a timely manner to allocate energy resources in the best way to realize complementary, stable, safe, economic and efficient clean energy systems of water, wind, light and storage. In order to gain further knowledge to accelerate the development of this field, we hope that this review can provide a valuable reference for future research.

Conflicts of Interest:
The authors declare no conflict of interest.