HM Medical Clinic

Latest development of nanostructured si/c materials for lithium anode studies and applications

Contents lists available at Energy Storage Materials journal homepage: Latest development of nanostructured Si/C materials for lithium anodestudies and applications Miao Zhang Tengfei Zhang , Yanfeng Ma Yongsheng Chen a Centre for Nanoscale Science and Technology, Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), School of Materials Scienceand Engineering, Nankai University, Tianjin 300071, Chinab Key Laboratory of Functional Polymer Materials and the Institute of Polymer Chemistry, College of Chemistry, Nankai University, Tianjin 300071, China Silicon anodes for lithium-ion batteries have been extensively explored due to their high capacity, moderate Received 15 November 2015 operation potential, environmental friendliness, and high abundance. However, silicon's application as Received in revised form anodes is hindered by its poor capacity retention caused by the large volume change during lithium insertion and desertion process, its intrinsic low conductivity and the formation of unstable solid-electrolyte inter- Accepted 1 February 2016 phase (SEI) films. Recently, influential improvements have been achieved using different design methods Available online 12 February 2016 with the purpose of increasing cycle life and increasing charging rate performance. Here, we review such design methods including the rational design of nanostructured silicon, the combination of silicon with Lithium ion batteries different carbonaceous materials including traditional carbons and the utilization of nanocarbons (such as carbon nanotube, graphene and corresponding three dimensional architectures). Meanwhile, we draw the essential reason accounting for the excellent electrochemical performance of those structures. Furthermore, Carbon nanotubesGraphene we selectively depict the effects of binder, conductive additives and electrolyte composition, which also playimportant roles in silicon based battery performance.
& 2016 Elsevier B.V. All rights reserved.
n Corresponding author at: Centre for Nanoscale Science and Technology, Colla- borative Innovation Center of Chemical Science and Engineering (Tianjin), School of Energy is one of the most important topics of the 21st century.
Materials Science and Engineering, Nankai University, Tianjin 300071, China.
Ever rising demands for energy coupled with the depletion of Fax: þ86 2223 499992.
E-mail address: (Y. Chen).
finite fossil fuel and the emission of contaminative gases have 2405-8297/& 2016 Elsevier B.V. All rights reserved.
M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 encouraged scientists to develop new energy storage and conver- pulverization of the Si particles can cause lithium trapping in the sion technologies for renewable and clean energy sources. Among active material and progressively consume of active material various energy storage technologies, electrochemical storage is during cycling, all of which contribute to low Coulombic efficiency considered as one of the most promising technologies, especially and high irreversible capacity. (2) Disruption of the electron–ion for the applications of electric vehicles (EVs), plug-in hybrid transport pathways at the level of the entire electrode: The elec- electric vehicles (PHEVs) and hybrid electric vehicles (HEVs) trode is composed with Si particles as active materials, carbon Moreover, the popularization of portable electronics and com- black to enhance the conductivity and binder to facilitate the munication equipment worldwide stimulates the development of integrity of the whole electrode. During lithiation process, the Si energy storage devices, such as batteries and supercapacitors with particles expand, suppressing the surrounding materials expand- higher energy density and higher power density respectively ing at the same time, which exceed the mechanical elongation of Lithium ion batteries (LIBs) are widely used as a convenient power them, resulting in electrical detachment of the surrounding com- source for various portable electronic devices and are considered ponents with active materials. Finally, this drastic electrode mor- as potential energy source to power EVs, HEVs and PHEVs phology change combined with intrinsic low conductivity disrupt A LIB is mainly composed of an anode (negative), a cathode the electron-transport pathways and further contribute to capacity (positive), a separator, and a certain amount of electrolyte , fade. (3) Unstable solid-electrolyte interphase (SEI): When the The positive electrode materials are typically Li-containing metal potential of the anode is below ca. 1 V versus Li/Li þ, the decom- oxides and the negative electrode materials include insertion-type position of the organic electrolyte at the electrode surface is materials (such as graphite, Li4Ti5O12 (LTO), etc.), and alloying-type favorable, forming the SEI layer. If the SEI layer is dense and stable, materials. The function of the separator is to prevent short cir- it can prevent further side chemical reactions. But the SEI layer on cuiting between the cathode and anode electrodes and to provide the Si surface continuously breaks and the fresh Si unceasingly abundant channels for transportation of Li ion during charging/ exposes to electrolyte during lithium insertion and desertion, discharging. The electrolyte should be a good ionic conductor and leading to extremely thick SEI layers and excessive consumption of electronic insulator and most of them are based on the solution of lithium ions and electrolyte. As a consequence of this process, inorganic lithium salts dissolved in a mixture of organic solvents.
even if most of the Si active materials remain electrically con- Currently, most commercially available LIBs are made with gra- nected, the capacity decays because of the poor electronic con- phite as the anode material and lithium metal oxides/lithium iron ductivity and the exhaustion of the electrolyte.
phosphate as the cathode material. The theoretical capacities of To overcome the drawbacks of Si as anode materials, significant these anode and cathode materials are 372 mAh g1 and ca.
number of achievements have successfully addressed the pro- 200 mAh g1, respectively, resulting in energy density of ca.
blems. There have been several special reviews on Si as anodes for 200 Wh Kg1. However, this energy density value cannot satisfy LIBs in the recent two years (2014 and 2015) Xu et al.
the demands for higher energy density batteries used in emerging described the progress in Si-based materials utilized in LIBs in new-type electronic devices, advanced communication facilities, terms of composite systems, nanostructure designs, material and in particular, EVs or HEVs.
synthesis methods, and electrochemical performances . Les- The energy density of a battery is mainly determined by its triez et al. summarized the influence of the different parameters of output voltage and specific capacity, which are dependent on the the formulation of silicon-based composite electrode on its electrochemical properties of electrode materials . Alloy type cyclability . Metal-assisted chemical etching of silicon and anode materials with high theoretical capacity and low operation nanoscale silicon materials used as Li-ion battery anodes were voltage, such as silicon and tin, have been intensively explored to reviewed by McSweeney1 s group Terranova, Su and Zhu further increase the energy densities of LIBs for the above- et al. reviewed Si/C composites respectively ,, but they mentioned applications . Silicon (Si) is one of the most pro- did not give specific depiction of the recently reported composites mising alloy type high capacity alternatives to graphite anodes. Si of Si with nanocarbons, especially the three dimensional nano- offers a high theoretical capacity of 4200 mAh g1 because each carbons. Also, some reviews of progress of LIB materials have silicon atom can accommodate 4.4 lithium atoms corresponding to mentioned the progress in Si/C composites . However, the formation of Li22Si5 alloy. This theoretical specific capacity of the combination of nanostructured Si and nanostructured carbon silicon anode is ca. 10 times higher than that of graphite, and ca. 20 in recent years have not been systematically presented. Thus, in times higher than that of Li4Ti5O12 (LTO). The second merit of Sianodes is its moderate operation potential versus lithium (ca.
this review, we will first summarize the recent progress and 370 mV above Li/Li þ ) . Thirdly, Si is the second most abundant advances in designing nanostructured Si anode electrode materi- element on earth and it is environmentally benign. Furthermore, als. And then, more details for the advantages of combination of there are large and mature infrastructures for its processing and above mentioned nanostructured Si with various carbons, espe- there are growing approaches using cheap raw materials to fab- cially the nanocarbons, will be presented. Finally, we will move to ricate nano-Si particles For example, Cui and coworkers other additional aspects, such as the binder, electrolyte and elec- recovered Si nanoparticles directly from rice husks, an agricultural trode additives, which could also impact electrode performance.
waste, in which the silicon naturally exists in the form of silicananoparticles, which show good electrochemical properties whentested as anodes in cells 2. Rational designs of nanoscale dimensional silicon However, the use of bulk Si experiences large volume changes (undergoes up to a 300% volume expansion and contraction) as Li Nanostructured materials improve cycling stability by incor- ions enter and leave the Si lattice, which shortens the cycle life and porating pores or voids to accommodate expansion along with contributes to cell failure The mechanism that affects the short lithium diffusion distances within the electrode A electrode lifespan need to be discussed first: (1) Particle pulver- number of nanoscale morphologies have been investigated to ization: during the process that lithium ion insert into the Si lat- minimize electrode pulverization and capacity loss in silicon tice, the Si particle grow larger and larger and they will collide into anodes, including zero dimensional (0D) nanoparticles, one and squeeze each other until attaining three folds of their original dimensional (1D) nanowires and nanotubes, two dimensional (2D) volume, leading to extremely large stresses, which cause crac- thin films and three dimensional (3D) porous structures. In this king and pulverization of active particles. The cracking and section, we will discuss them one by one.

M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 Nanoscale dimension particles allow quick relaxation of stress, cycling, leading to superior capacity retention of 89% after 200 making nanoparticles more resistant to fracture than bulk parti- cycles at a rate of 1 C in practical Li-ion cells . Arrays of sealed, cles. Silicon with smaller diameters as anodes could cycle more tubular Si nanotubes, which combine the merits of nanowires and reversibly with moderately higher capacity than micron-sized hollow spheres, were expected to further accommodate the large silicon powder anodes . Some studies have been carried out volume changes, exhibit high initial Coulombic efficiencies (i.e., for the relationship between Si size and the cycle life of Si anode 485%) and stable capacity retention (480% after 50 cycles), due materials. Kim and coworkers reported 10 nm sized n-Si as the to an unusual, underlying mechanism that is dominated by free optimized sample because smaller Si particle encounter with lar- surfaces An active silicon nanotube surrounded by an ion- ger unstable SEI and larger Si particles suffer from larger strain permeable silicon oxide shell can cycle over 6000 times in half during volume changes . Wang s group reported the critical cells while retaining more than 85% of their initial capacity, 15 nm Si building block size with SiO as starting materials . As resulting from a stable SEI . However, such techniques do not for the preparation methods, Si nanoparticles (Si NPs) could be fit with standard preparation methods for manufacturing Li-ion synthesized from reduction of silica or SiCl battery electrodes because of its high cost, low syntheses effi- 4 disproportion of silicon monoxide , ball milling from bulk Si . Compared to ciency. A simple 1D free-standing carbon-coated Si nanofiber solid structures, hollow structures provide empty interior space binderless electrode, which was prepared via magnesiothermic for the volume expansion, which offers lower diffusion-induced reduction of electrospun SiO2 nanofiber paper produced by an acid stresses. The finite element modeling results show that the max- catalyzed polymerization of tetraethyl orthosilicate (TEOS) in- imum tensile stress in a hollow Si sphere is five times lower than flight, was reported with a capacity of 802 mAh g1 after 659 that in a solid sphere with an equal volume of Si . The lower cycles with a Coulombic efficiency of 99.9% 1D Si nanowires stress values mean that the hollow nanostructures will fracture are also promising in application of lithium ion batteries.
less readily. The 0D nanoparticles are most promising nanoscale Two dimensional thin silicon films have demonstrated high design for the application of Si/C materials, which we will review capacities and long cycle lives of Si nanostructures The thin in the following section. The reasons are as follows: (1) 0D film expands during lithiation along perpendicular direction, while nanoparticles are easier to combine with carbon materials, contraction during delithiation occurs both perpendicular to and including traditional carbons and nanocarbons in simple methods; in plane with the substrate, leading to cracking of the film after the (2) 0D nanoparticles are easier to fabricate with low cost in easy first discharge cycle. After the initial fracture, the active material can be cycled without additional film cracking . Thickness, Other than 0D nanoparticles, 1D silicon nanowires (NWs) have surface morphology, and the interfacial bonding degree between shown high discharge capacities and stable cycling over tens of the Si and current collector all have impact on the performance of cycles with high reversible capacities. An advanced vertical growth Si thin films as electrode The reversible capacity and of NW structure design prepared via vapor–liquid–solid method cycling life decrease with increasing films thickness, whereascapacity and cycle life increase with increasing films roughness.
not only take advantage of small NW diameter to better accom- Amorphous n-type silicon films with a thickness of 50 nm vacuum modation of the large volume changes but also electrically connect deposited onto nickel substrates exhibited an initial capacity of each Si NW to the current collector to prompt efficient charge approximately 3750 mAh g1 and without significant capacity transport (). When those nanowires were charged with C/5 decay after 200 cycles at 1 C Thicker film of 275 nm delivered rate, the capacity was stable at ca. 3500 mAh g1 for 20 cycles.
an initial reversible capacity of about 2200 mAh g1 with capacity Moreover, the capacity of the Si NWs at faster rates was also retention of 61.3% after 500 cycles. After annealing this sample for excellent with the capacity remained at 2100 mAh g1 at 1 C rate better interfacial adhesion between the Si thin film and the Cu Hence, silicon nanowire battery electrode might make a truly substrate, the capacity retention was further improved to 78.5% for promising design, which could accommodate large strains without 500 cycles . Nevertheless, Si thin film encounter with the pulverization, provide good electronic conduction, and display trouble in production.
short lithium insertion distances. Similarly, 1D Si nanotubes, which An advanced effective approach to improve the electrochemical offers lower diffusion-induced stresses for the empty space inside, performance of silicon anodes is to fabricate silicon-based 3D show that the morphology of the nanotubes did not change after composites with porous nanostructures, in which the local voidspace could partially accommodate the large volume change, thuspreventing the capacity from fading . A lotus-root-likemesoporous Si with carbon surface coating displayed a stablecapacity of ca. 1500 mAh g1 for 100 cycles at 1 C and a high ratecapability up to 15 C Macroporous silicon and carbon–siliconperiodic materials based on inverse-opal structures synthesizedvia templating with ordered colloidal spheres and subsequentsilicon deposition was demonstrated with high capacities at lowcurrents and decent capacity retentions, but their performance isseverely restricted due to the low electrical conductivity of silicon.
A capacity with carbon coated could be maintained above2100 mAh g1 for 145 cycles whereas the capacity of a siliconinverse-opal coated without amorphous carbon was completelylost by the 11th cycle Monodisperse porous silicon nano-spheres (MPSSs) were synthesized via hydrolysis process with Fig. 1. Schematic of morphological changes that occur in Si during electrochemical subsequent surface-protected magnesiothermic reduction. The Li- cycling. NWs grown directly on the current collector do not pulverize or break into ion battery (LIB) anodes based on MPSSs demonstrate a high smaller particles after cycling. Rather, facile strain relaxation in the NWs allows reversible capacity of 1500 mAh g1 after 500 cycles at C/2 .
them to increase in diameter and length without breaking. This NW anode design Furthermore, nest-like Si nanospheres exhibit superior has each NW connected with the current collector, allowing for efficient 1D elec- lithium-storage capacity, high-rate capability and long cycling tron transport down the length of every NW. The figure is reprinted from Ref. with permission.
properties as well.

M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 From previous discussion, we find that even though large carbon coated Si nanowires, nanotubes and porous structures in volume change of pure Si could be accommodated by the rational this method are usually high, ca. 2000 mAh g1 The design of nano architectures, their cycling stability still much need shortcomings of this method are bad operability, high cost and to be further improved by the protection of carbon. Because the poor scalability.
surface-to-volume ratio is high for these morphologies, the for- Similarly, Si NPs could also deposited on commercial carbon mation of stable SEI layers is less possible. At the same time, car- structures, and these composites also show high capacities after bonaceous materials normally could form stable SEI layers on their long cycle numbers. For example, Yushin s group loaded Si NPs on own surfaces. Thus, carbon coating might offer a good choice in a 3D spherical carbon-black scaffold. The final Si/C nanohybrid constructing long lifespan batteries for Si materials. Moreover, the exhibited impressive electrochemical properties, including a high incorporation of carbon can release stress of volume expansion specific capacity above 1500 mAh g1 at a rate of 1 C after 100 contributing to long cycle life. Thus, recent progress for the roles of cycles . An interesting architecture is Si nanowires internally carbon in promoting the electrochemical performance of Si anode grown from porous graphite, demonstrating a high volumetric will be discussed more thoroughly in the following section.
capacity density of 1363 mAh cm3 with 91% Coulombic efficiencyand high rate capability of 568 mAh cm3 even at a 5 C rate.
3. The composite materials of nano silicon and carbon A more cost efficient approach for preparing Si/C composite is mixing Si with carbon or blending with carbon precursors with Much effort has been not only put in preparing nanostructured following pyrolysis process. For example, directly mixing Si pow- silicon, but also come over constructing of nanohybrids with such der with graphite, mesocarbon microbeads (MCMB) or hard car- as metal and carbon for their ability to accommodate the volume bon could achieve stable electrode morphology even after exten- change and enhance electrical conductivity. Typical structures ded cycling and the capacity retention could attain ca.
include metallic nanohybrids and carbonaceous nanohybrids.
500 mAh g1 after 400 cycles (for MCMB/Si mixture) Si can There are some examples of applying Cu and Ag coatings that have also blend with carbon sources by ball-milling and then pyrolyze been shown to increase coulombic efficiency and improve rate resulting mixture at high temperature . Silicon/carbon com- capability. ,Metallic nanohybrids, due to a larger atomic posites synthesized via the above methods often display reversible density and higher cost, are practically unsuitable for the con- capacities higher than graphite . When lithium metal was struction of electron-transport pathways. As a result, a significant added in previous of ball milling, lithiated silicon–carbon com- amount of studies have focused on exploring lighter alternatives posites display ca. 0.13% capacity loss per cycle with high specific for the construction of similar electron-transport pathways in Si- capacity ( 700 mAh g1) . Ternary composites such as flake based anodes. The most common attempt is to combine the Si graphite/silicon/carbon also show good cycling retention .
structure with a conducting carbonaceous layer, which possess Various precursors can be used for encapsulation, including many merits, such as excellent flexiblity, high conductivity, light- resorcinol formaldehyde gel poly(vinylidenefluoride) (PVDF) weight, electrochemical and thermal stability, in hoping to better , polystyrene glucose , pitch , poly(vinyl chloride) retain the integrity of Si particles and restrain active materials (PVC) ,and so on. The hydrothermal method is another easy disconnected from the conductive electrode. In this section, Si/ approach to homogeneously blend Si with precursors .
carbon composites are classified into two main classifications: Spray pyrolysis, which could blend solvent type precursors toge- (1) Si/traditional carbon (Si/TC) composites, including Si particles ther and obtain powder product, was considered as an effective with deposition carbon, pyrolytic carbon and commercial bulk approach to synthesis various structures and composites with carbons; (2) Si/nanocarbon (Si/NC) composites, including Si par- good electrochemical performance . Spherical nanostructured ticles with carbon nanotubes (CNTs), graphene and corresponding Si/C composites could be prepared by spray drying technique with 3D architectures; and (3) Si/traditional carbon/nanocarbon (Si/TC/ phenol formaldehyde resin as the carbon source, which could NC) composites. The Si/TC composites will be discussed first fol- exhibit a relatively high reversible capacity and good cycle per- lowed by Si/NC composites and Si/TC/NC composites.
formance used in lithium ion batteries Furthermore, spraydrying technique could be applicable for the large-scale produc- 3.1. Si/traditional carbon composites tion of various Si/C composites .
There are some delicate designs prepared with special meaning Si/traditional carbon (Si/TC) composites using traditional car- but hard to classify. Here we will give a brief introduction of them.
bons will be discussed on the line of their preparation methods A pioneer example is a yolk-shell structure design. This structure and then some of the recent exciting and interesting design was prepared through an intermediate sacrificial silica layer with examples will be presented.
void space in between the Si particles and the carbon shell and Conventionally, coating carbon on nano Si materials commonly showed excellent capacity (2833 mAh g1 at C/10), long cycle life adopted thermal decomposition or chemical vapor deposition (1000 cycles with 74% capacity retention), and high coulombic method with precursors of acetylene gas. These methods efficiency (99.84%) Later, those hybrid nanoparticles are result in carbon layers of high uniformity, neat surface smoothness assembled into a thicker carbon layer. This microstructure could and high purity. Thus, the initial capacity and capacity retention of further lower the electrode-electrolyte contact area, resulting in Fig. 2. Schematic view for the synthesis process of Si nanowires internally grown in porous graphite. The figure is reprinted from Ref. with permission.

M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 higher Coulombic efficiency (99.87%) and volumetric capacity simultaneously electrospraying nano-Si-PAN (polyacrylonitrile) (1270 mAh cm3), and the cycling remains stable even when the clusters and electrospinning PAN fibers followed by carbonization areal capacity was increased to the level of commercial lithium-ion with uniform incorporation of Si NPs into fiber paper. The flexible batteries (3.7 mAh cm2) 3D Si/C fiber paper electrode demonstrated a very high overall In short, Si/TC composites are very promising for LIB anodes.
capacity of 1600 mAh g1 with capacity loss less than 0.079% per But the preparation of high quality carbon in easy and cost effi- cycle for 600 cycles and excellent rate capability .
cient methods are still challenging.
The carbon nanofibers can not only serve as the conductive matrix, but also encapsulate Si NPs into them. Here are two 3.2. Si/nanocarbon composites interesting examples. Hwang et al. developed an electrospinningprocess to produce core–shell fiber electrodes with Si NPs as core Nanocarbons such as fullerene carbon nanofibers (CNF) wrapped in the carbon shell. This core–shell structure exhibited –, carbon nanotubes (CNTs) and Graphene, intensively high gravimetric capacity of 1384 mAh g1, excellent cycle life of investigated for advanced energy storage with continuously gro- 300 cycles with almost no capacity loss . Later, Lee s group wing academic and technological impetus, have been incorporated demonstrated the almost full accommodation of all the volumetric with Si NPs for the preparation of LIB anode materials changes of Si by embedding Si NPs into a tunable cyclized- Thus, the latest development for the composite anode materials of polyacrylonitrile (cPAN) fiber network bonded together by a Si with these nanocarbon materials, especially CNFs/CNTs and Graphene, are discussed below.
3.2.2. Si/CNTs anode materials 3.2.1. Si/CNFs anode materials Due to their excellent conductivity, high surface area, One-dimensional (1-D) nanostructures have the advantages of mechanical flexibility, and chemical stability CNTs, inclu- high surface area and short ion diffusion length, and have been ding single wall CNTs and multiwall CNTs , are considered as viewed as components for next-generation electrochemical energy promising materials to improve the performance of silicon anode conversion and storage devices. The ductile CNF matrix can buffer materials. There are many studies combining CNT with Si in the Si volume expansion on the macro domain and maintain good mainly two forms, considering the relative position of Si particles contact with both the active materials and the electrolyte after and CNTs: Si NPs on the inner walls of CNTs and Si NPs on the lithium insertion and extraction cycles, resulting in high reversible outer walls of CNTs capacity and fairly good cyclability Sputter Si at CNF, the When Si film was deposited on and weakly bonded to the inner hybrid nanostructured Si/CNF anodes exhibited superior device surface of a nano confined and size-preserving tubular material, Si performance to materials used in previous studies, in terms of deforms upon electrochemical alloying and dealloying without both specific capacity and cycle life. The CNFs provide not only a cracking. This model was verified by Yushin s group, who utilized good strain/stress relaxation layer but also a conductive electron carbon nanotubes (CNTs) with an inner Si coating. The composite pathway . Vertically aligned carbon nanofiber (VACNF) provide samples with a Si content of 46 wt% showed a capacity of a good lithium-ion intercalation medium and a robust conductive 2100 mAh g1, very close to the theoretical maximum predicted, core to effectively connect high-capacity silicon shells for lithium- assuming Si's contribution to be 4200 mAh g1. Furthermore, a ion storage. When VACNFs coaxially coated with silicon shells, stable SEI layer on the carbon layer impermeable to solvent an excellent cycle stability, about 89% of the capacity retention molecules serves as a barrier to electrolyte decomposition and after 100 charge–discharge cycles at the C/1 rate, has been leads to a Coulombic efficiency 499.9% after the first cycle ) achieved Si/C fiber paper electrode could be synthesized by Cheng s group revealed the confinement effect of CNTs Fig. 3. Electron microscopy of the composite Si-in-C tubes: (a) scanning electron microscopy (SEM) of one of the synthesized samples, (b) SEM of the electrode attached to aCu current collector, (c) TEM and (d) its schematic of the sample after Li extraction at the 10th cycle. The figure is reprinted from Ref. with permission.

M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 when Si NPs encapsulated within the hollow cores of the CNTs.
3.2.3. Si/Graphene anode materials The volume expansion of the lithiated Si NPs is restricted by the Graphene plays important roles in electrode materials investi- walls of the CNTs () gation, due to that it could provide a conductive channel for As another combination method, Si NPs on the outer walls of electron transport, optimize electrical contact between the elec- CNTs take advantage of both the voids between the CNTs/N-CNTs trode components and also act as a buffer to volume changes and the tubular voids, thus effectively release the volume change during cycling, which are all benefiting from the graphene's large of Si Vertical aligned Si/CNTs composite can be prepared via surface area, mechanical flexibility, chemical stability and excel- a two-step CVD method: CVD growth of CNTs array on substrate lent conductivity. Scientists have developed several methods to and deposition of Si NPs on it –The covalent bonding of fabricate composites of silicon nano particles/wires and graphene CNTs to Si NPs can further enhance the electronic pathway to the (Si/Graphene) composites for LIB recently.
active material particles and helps to prevent the detachment of Si Graphene in Si/Graphene composites include the graphene reduced from graphene oxide, exfoliated graphene and CVD from CNTs upon repeated lithium insertion/extraction . The growth of graphene among which reduced graphene oxide capacity of the composite can reach 2000 mAh g1 with 0.15% (RGO) is the most common used graphene. Si/Graphene compo- decay per cycle in 25 cycles The gravimetric capacity is high, sites can be developed through a simple facile way by filtering the but volumetric (areal) capacity of Si/CNTs composite is relatively silicon/graphene oxide solution to film with a following reduction low. When the vertically aligned CNTs utilized as the conductive treatment . The reversible specific capacity of this free core and coated with pore size-graded Si film, this 4 μm thick standing Si composites can reach 1500 mAh g1 after 200 cycles.
electrode deliver a high volumetric (areal) capacity and good cycle Another interesting self-supporting binder-free silicon based stability Volumetric (areal) capacity could also be increased anode was prepared by double encapsulation of silicon nanowires through preparing a multilayer Si/CNT coaxial by a layer-by-layer (SiNWs) with two kinds of graphene (overlapping graphene assembling technique .
(G) sheaths and reduced graphene oxide (RGO) overcoats) ( Generally speaking, Si/CNTs is a good candidate for LIB anodes This resulted structure (SiNW@G@RGO) have a high rever- and the encapsulation technique is an effective way to prevent sible specific capacity of 1600 mAh g1 at 2.1 A g1, 80% capa- pulverization and stabilize SEI layer, mainly because of the sig- city retention after 100 cycles, and superior rate capability nificantly enhanced conductivity and structural durability. Major (500 mAh g1 at 8.4 A g1) .
concerns for this composite material are: (1) the complicated Though both the filtration-directed assembly approach and sim- fabrication process, which significantly increase the fabrication ple mixing method have obtained improvements on lithium storage, cost, and thus prevent its commercial application and (2) the loose they do not provide good dispersion of Si NPs between graphene binding between silicon and CNTs and the binding construction sheets and good interfacial connection between Si NPs and graphene between them is also expensive and time consuming.
sheets. An electrostatic attraction-directed self-assembly approach Fig. 4. Dynamic structural changes of a Si NP-filled CNT under electrochemical lithiation/delithiation. (a-d) Lithiation of a Si NP-filled CNT. (e-h) Delithiation of the same SiNP-filled CNT. (i) Illustration of the lithiation/delithiation of Si NP-filled CNTs. The figure is reprinted from Ref. with permission.

M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 Fig. 5. Schematic of the fabrication (upper panel) and adapting (lower panel) of SiNW@G@RGO. The fabrication process mainly includes (I) chemical vapor deposition (CVD)growth of overlapped graphene sheets on as-synthesized silicon nanowires (SiNWs) to form SiNW@G nanocables, and (II) vacuum filtration of an aqueous SiNW@G-graphene oxide (GO) dispersion followed by thermal reduction. The resulting SiNW@G@RGO can transform between an expanded state and a contracted state duringlithiation/delithiation cycles, thus enabling the stabilization of the silicon material. The figure is reprinted from Ref. with permission.
was developed with uniform dispersion of Si NPs on graphene sheetsThis approach realizes an uniform dispersion of Si NPs betweentwo layers of graphene sheets. The as-obtained composite exhibitsstable cycling performance (approximately 1205 mAh g1 after 150cycles) and excellent rate capability. The interfacial contact betweengraphene and other materials could be connected by mechanical orchemical interaction Dual chemical cross-linking and hydrogenbonding interactions between surface-modified Si NPs and grapheneoxide (GO) exhibited an outstanding capacity retention capabilityand good rate performance, delivering a reversible capacity of1000 mAh g1 after 400 cycles at a current of 420 mA g1 withalmost 100% capacity retention . Furthermore, Si wrapped ingraphene sheets can also be synthesized by a simple spray dryingprocess. This simple but effective nano/micro-assembly technology Fig. 6. Schematic fabrication process of the rolled-up Si/rGO bilayer nanomem-branes. The figure is reprinted from Ref. with permission.
can be used for large-scale production of various graphene-basedcomposite materials with high performance for electrochemical Si/RGO nanoarchitecture demonstrates long cycling life of 2000 energy storage and conversion .
cycles at 3 A g1 with a capacity degradation of only 3.3% per 100 To further improve the electrochemical properties of Si/Gra- cycles. The inner void space inside the configuration together with phene composites, works with delicate control of Si nanoparticle the mechanical feature of the amorphous Si nano membranes can size and coating graphene layers have been conducted. Si NPs buffer the strain of lithiation/delithiation against pulverization to downsized to ca. 3 nm silicon quantum dots and anchored on GNS, extend cycling life. The alternatively aligned RGO layers in this which exhibited an extraordinary rate capability due to the nanostructure can facilitate electron transport, accommodate the surface-controlled lithium storage behavior rather than conven- volume change of Si layers and prevent their aggregation. Fur- tional diffusion-controlled mechanisms . Another example is thermore, the RGO layers can protect the nanomembranes from graphene sheets of 2–10 layers directly grown on Si particles and the excessive formation of thick SEI layer to suppress the capacity the graphene layers rooted on the Si particles. This structure can accommodate the large volume change of Si particle via a sliding Si/Graphene composite is another good candidate for LIB process between adjacent graphene layers. Thus, cells using the anodes and the combination structure is effective to prevent the above composite as anode could reach a high volumetric energy mechanical pulverization, mainly because of the flexibility and densities of 972 and 700 Wh L1 at first and 200th cycle, respec- high specific area of graphene. However, the capacity is still inevitably decaying, and the proposed reasons are as follows: (1) The multilayer C/Si/C microtubes exhibited synergistic proper- the low conductive connection between graphene and Si, espe- ties and superior electrochemical performance when used as cially during the lithium insertion and desertion process; (2) not anodes for LIBs because its rolled-up layer by layer structure could yet refined structural morphology of the composite.
buffer the strain of volume changes, and delay the pulverization of Three-dimensional structures of active materials provide large the electrode materials . When the carbon layer in the roll to surface area, well defined pathways for the access of electrolyte roll nanomembranes was replaced by RGO layer, the sandwiched and mechanical stability for the integrity of electrodes .

M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 Taking advantages of such extraordinary properties of graphene carbon, encapsulating Si NPs, was anchored on the walls of RGO.
offers excellent opportunities for building attractive matrices, Pyrolytic carbon plays three important roles in Si/TC/NC compo- which not only accommodate the Si volume change, but also site: (1) provide electrical conductivity; (2) reduce direct exposure provide unobstructed electron and ion pathways.
of Si to electrolyte, thus decrease the unstable SEI growth; (3) Si on 3D graphene framework composites could be prepared serve as a glue to connect Si NPs with RGO. The RGO play key roles through constructing 3D architectures, and then Si NPs or Si pre- in: (1) provide good electrical contact and (2) provide large void cursors being loaded on the as formed framework. The first spaces to accommodate the volume change of Si. In the Si/Poly- example is a hierarchical 3D mesoporous carbon-coated Si@gra- mer/CNT composites, CNTs and PPy work together to improve phene foam nanoarchitecture, which was prepared via magne- conductivity and provide space for volume change . Si/Poly- siothermic reduction of silica on the three dimensional matrix and mer/CNT composites possess an advantage over Si/TC/RGO in that exhibited superior electrochemical performance including a high they do not need annealing at high temperature to recover the lithium storage capacity of 1200 mAh g1 at the current density of conductivity of RGO.
1 A g1 Another example is the amorphous Si NPs coated on Furthermore, Silicon could also be well-anchored onto 3D NT/ backboned-graphene nanocomposite architecture. The electrode TC frameworks. For example, Si @ 3D graphene/carbon nanotube facilitates the electron transport and lithium diffusion, resulting in (CNT)/ aerogels framework (CAs) nanohybrids could make use of remarkable first-cycle Coloumbic efficiency of 92.5% with a high both NT/TC layers and the balanced open voids, with a high specific reversible capacity of 2858 mAh g1, excellent power reversible capacity of 1011 mAh g1, and excellent capacity capability, and outstanding cycling stability ) .
retention of 96% . However, the preparation method above Compared with simple graphene sheets, which are easier to was not satisfying for the complicate fabrication of 3D supporting restack, hierarchical three-dimensional architectures could more framework and deposition of Si. Therefore, some simple methods effectively buffer the strain from the volume change of Si during for the preparation of high performance Si on 3D network anode the charging discharging process and preserve the high electrical composites were also reported. Our group fabricated a ternary Si- conductivity of the overall electrode, representing a new direction based composite Si@C/GF through simple hydrothermal reaction for fabricating robust, high-performance lithium-ion batteries and and thermal reduction, in which Si NPs were coated by a thin related energy storage applications with advanced nanostructured carbon layer by pyrolysis of phenolic resin and encapsulated in a graphene framework (GF). As a result, the double-protected Si NPshave a much improved cycle stability as well as high specific 3.3. Si/traditional carbon/nanocarbon composites capacity and good rate performance . Another simple methodof preparation of 3D framework composite is Si NPs impregnated Traditional carbon coating of nanocarbon/Si composite could assemblies of templated carbon bridged oriented graphene, which further effectively alleviate the aggregation of Si NPs by separating was prepared by a modified vacuum filtration process followed by them from each other and help nanocarbons build more efficient thermal treatment (When used as LIB anodes, the 3D fra- 3D conducting networks. In the hybrids of graphene/carbon- mework exhibited high gravimetric capacity (1390 mAh g1 at coated Si NPs, the Si NPs are wrapped between graphene sheets 2 A g1 with respect to the total electrode weight), high volu- and amorphous carbon coating layers. These two layers work together to effectively suppress the aggregation and destruction of (900 mAh g1 at 8 A g1) and excellent cyclic stability (0.025% Si NPs, keeping the overall electrode highly conductive and active decay per cycle over 200 cycles) in Li storage Another graphene/Si–C hybrid (G/Si–C) is We can summarize from , a‑Silicon @ backboned gra- reported having a high areal capacity of 3.2 mAh cm2 after 100 phene nanocomposite (a-SBG) is the most effective approach to cycles with high coulombic efficiency These findings improve the performance of Si/C materials, for it show a high demonstrate the importance of building a conductive network in initial Coulombic efficiency of 92.5% and a high capacity of the electrode level for efficient material utilization and may sug- 1103 mAh g1 at 14 A g1 after 1000 cycles with nearly 100% gest future designs of Si-based anodes capacity retention. However, the Si coating process is not com- lists some combination works of Si NPs with TC and NC, mercially viable for applications in large batteries using similar in form of preparation methods and electrochemical properties, in scalable technology. So the simple method of preparing 3D fra- which Si/TC/RGO composites are more intensively investigated mework composite with easy mixing, filtration and annealing than Si/TC/CNT. This may be resulted from that GO, precursor of process is attractive, and this composite has a relatively high initial RGO, has functional groups and negatively charged that could Coulombic efficiency of 72% and a high capacity of 1390 mAh g1 chemically bond or electrostatically connect with other carbon at 2 A g1 after 200 cycles with 95% capacity retention. Simulta- source or modified Si NPs. Most of Si/TC/RGO composite can be neously, the in-situ polymerization of the mixture of Si, Py and prepared via simple mixing of Si NPs with carbon source (carbon CNT, and coating on the current collector for cell fabrication is also sources are usually positive charged, serving as an intermedia a good method to prepare the Si anode materials.
layer to electrostatically attract with negatively charged Si NPs and In brief, Si NPs on frameworks of TC and NC is an excellent GO) and GO, and then annealing at high temperature to pyrolysis candidate for LIB anodes. The TC and NC framework could work carbon source and reduce GO at the same time. Thus, the pyrolytic together to provide both superior and robust conductivity and Fig. 7. Schematic view of a-SBG nanocomposites before and after electrochemical cycling. The figure is reprinted from Ref. with permission.
M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 Table 1Electrical properties of Si anodes with Si/TC/NC composites.
Preparation method Qr1 (mAh g1) (initial Si @PANI @RGO, pyrolysis Si&PANI&RGO, pyrolysis PAN add into GO&Si, coat on Cu foil, pyrolysis Si & PDA & GO mix – pyrolysis Si & Py &CNT – in-situ polymerization – coat on Cu 70% (binder Si wt%: mass loading of Si; Qr1: the first reversible capacity; CE: Coulombic efficiency; QdN(N): discharge capacity in Nth cycle; C.R.N.: capacity retention in Nth cycle. PANI:polyaniline; PDA: polydopamine; RF: resorcinol (R) and formaldehyde (F); PVP: polyvinylpyrrolidone; Py: Pyrrole.
Fig. 8. Electrode design and fabrication. (a) Schematic of the configuration of silicon nanoparticle-impregnated assemblies of templated carbon bridged oriented graphene(TCG-Si). (b) Schematic illustration showing the structure of TCG obtained by removing the Si template from the TCG-Si. (c) Schematic illustration of the fabrication processfor TCG-Si, where bovine serum albumin (BSA)-coated silicon nanoparticles and graphene oxide (GO) are assembled via electrostatic interactions during vacuum filtration,thus enabling the successful fabrication of TCG-Si. The figure is reprinted from Ref. with permission.
sufficient pores. Furthermore, the 3D framework could provide not only superior conductivity but also sufficient pores for lithium ionpathways and volume change buffer. Thus, the construction of The type of binder used in particulate electrodes could greatly effective 3D TC/NC architecture with Si NPs encapsulated in it influence the cycling lifetime. Poly(vinylidene) difluoride (PVDF) and simultaneously and with low cost may be a research direction in carboxymethylcellulose (CMC) are the most frequently used binder in bulk powder electrodes, but these binders cannot sustain theelongation that occurs during volumetric expansion, leading to rapidcapacity fade . Poly(acrylic) acid (PAA) possessing certain mechanical properties comparable to those of CMC but con-taining a higher concentration of carboxylic functional groups, has It is noteworthy that the binders, electrolyte composition and been shown to improve cycle life by enhancing the adhesion of conductive additives also play important roles in improving the electrode active materials to copper current collectors. Alginate, a electrochemical performance of Si based LIB anodes besides the natural polysaccharide extracted from brown algae, was introduced morphology and composition of Si materials. Here, we will as binder for Si NPs in LIBs and the capacity of Si NPs has been sig- describe the noteworthy progress in these aspects.
nificantly improved to 2000 mAh g1 at a high charge rate of 1 C M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 Table 2Preparation methods and electrical properties of Si anodes with Si disperse in 3D frameworks.
Preparation method 3D mesoporous silicon @ graphene PU as template to get GF –TEOS load SiO2 – Mg foam nanoarchitecture reduction – C2H2 to carbon coating a‑Silicon @ backboned graphene Reduction of freeze-dried GO – decomposition of nanocomposite (a-SBG) Si @ 3D graphene/carbon nanotube Polymerization of RF with CNT/G –freeze-dried –decomposition of SiH4 Si&Py&CNT – in-situ polymerization – coat on Cu foil 70% (bin- Nano Si 3D graphene-PF pyrolytic HT – add P/F – HT –annealing Si @ templated carbon-bridged Si mixed with BSA – mixed with GO – filtration – oriented graphene Si wt%: mass loading of Si; Qr1: the first reversible capacity; CE: Coulombic efficiency; QdN(N): discharge capacity in Nth cycle; C.R.N.: capacity retention in Nth cycle; RF:resorcinol (R) and formaldehyde (F).
Fig. 9. Schematic illustration of 3D porous SiNP/conductive polymer hydrogel composite electrodes. Each SiNP is encapsulated within a conductive polymer surface coatingand is further connected to the highly porous hydrogel framework. The figure is reprinted from Ref. with permission.
. Although the linear, one dimensional (1D), backbones repre- 4.2. Electrolyte and electrode additives sent a significant progress in the Si anode research, the multi-dimensional hyperbranched β-CD polymer with three dimensional The incorporation of additives into electrolytes has been con- backbones was reported to enhance Si-binder interactions as well as sidered for enhancing the stability of the passivation layer because improve mechanical stability of the electrode and therefore resolving additives can alter the composition of SEI. Lithium bis(oxalato) the chronic insufficient cycle lives of Si anode . After the concept borate (LiBOB), fluoroethylene carbonate (FEC) and vinylene car- of combining binding and conducting properties in the binder bonate (VC), undergoing reductive decomposition at the silicon surface at higher potentials than commonly-used ethylene car- comethylbenzoic acid) (PFFOMB) binder as an example, self- bonate (EC), is proposed to improve cycling stability by encoura-ging the formation of a stable SEI layer with a lowered resistance healing polymers (SHPs), which can mechanically and electrically for the diffusion of lithium ions on the surface of silicon electrodes.
heal cracks and damages, had been also demonstrated as effective This avoids not only the decomposition of electrolyte but also the binder to stabilize low-cost Si microparticle (SiMP) anodes. Com- oxidation of the Si electrode As evidenced by the dif- pared with traditional polymer binders, the self-healing chemistry is ference of SEI layer of Si film anode formed in VC-free and VC- designed to enable spontaneous repair of the mechanical damage in containing electrolytes, the SEI layer formed in VC-containing the electrode and thus increase the lifetime of the SiMP anode electrolyte possessed better properties, which was impermeable In-situ polymerization of conducting polymer hydrogel is to electrolyte and the impedance kept almost invariant upon another effective novel binder conception, which forms a continuous cycling. In addition, SEM imaging revealed that the electrode three-dimensional (3D) pathway for electronic conduction and pro- cycled with VC had a smooth surface with small cracks believed to vide sufficient voids for the expansion of Si () . By taking form during delithiation, compared to a much rougher surface advantage of the conductive polymer matrix, which provides fast with protruding crystallites and an inhomogeneous appearance electronic and ionic transfer channels, as well as free space for Si without VC Detailed investigations are needed to guide our volume changes, the electrode could be continuously deeply cycled further advancement of electrolyte additives.
up to 5000 times without significant capacity decay. The following The choice of the conductive additive and the quantities of research could concentrate on the fabrication of binders with high additive added into the electrode can have a significant impact on mechanical strength, excellent electrical conductivity and more electrochemical performance. The addition of carbon nanotubes importantly, simple preparation process with low cost.
and reduced graphene, which have larger theoretical surface area M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 compared to the active material (such as super P and acetylene black), were shown to enhance both capacity and cycling stability by forming a more efficient percolation pathway compared that when larger later active additive particles were used . It is also found that the amount of conductive additive can have a profound effect on the cycle life of the electrode, which increases with increasing conductive additive content For example, Si NPs with a mean diameter of 78 nm that were mixed with carbon black in a 1:1 weight ratio showed a reversible capacity above 1700 mAh g1 over ten cycles To obtain high energy density in practice, addition of less advanced additive and attain better electrochemical performance is attractive.
5. Conclusion and perspective The discussion in this review has illustrated that significant amount of researches are devoted into overcoming the challenges of using silicon anodes in practical lithium ion batteries. Rational designs of nanoparticles are effective to accommodate the large volume change, but they are relatively expensive and must incorporate with carbon to attend acceptable capacity retention.
Si/C composite could not only address the volume problems but also help to stabilize the SEI layers. Nano carbons features their remarkable properties and are more attractive in constructing of Si composite to further improve the cycling stability. Battery com- posites other than electrode materials, such as electrolyte addi- tives and conductive additives, are also play important roles in improving the cycle stability of Si anodes.
Even though considerable improvements has been achieved, future research is still necessary along the following directions for their practical application in commercial LIBs: (1) To the electrode material aspect: we should explore more structurally and com- positionally complex hierarchical composite nanostructures with not only internal void/pore space to buffer the large volume change but also good conductivity to attend high capacity reten- tion during long cycle life; the manufacture process of those nanomaterials should be simple and liable and the cost of raw nanomaterials should be also accessible. (2) To other aspects: new low cost, effective binders, electrode and electrolyte additives are needed to maintain efficient electronic and ionic conduction from silicon to the current collectors and stabilize the SEI layers during cycling. (3) To practical fabrication aspect: the scalability, manu- facturability, and cost of the nanomaterials are crucially important to the eventual success in the practical applications. Furthermore, the volumetric capacity, as an important parameter, should be given more attention in future research and practical manufacture.
The authors gratefully acknowledge financial support from the MOST (Grants 2012CB933401 and 2014CB643502); NSFC (Grants 51273093, 21374050, 51373078, 51422304 and 51472124); NSF of Tianjin City (Grant 13RCGFGX01121); Collaborative Innovation Center of Chemical Science and Engineering (Tianjin).
M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14 M. Zhang et al. / Energy Storage Materials 4 (2016) 1–14

Source: http://nanocenter.nankai.edu.cn/PDF/Energy%20Storage%20Materials_2016_ZM.pdf