Tin Oxide Based Nanomaterials and Their Application as Anodes in Lithium‐Ion Batteries and Beyond

Abstract Herein, recent progress in the field of tin oxide (SnO2)‐based nanosized and nanostructured materials as conversion and alloying/dealloying‐type anodes in lithium‐ion batteries and beyond (sodium‐ and potassium‐ion batteries) is briefly discussed. The first section addresses the importance of the initial SnO2 micro‐ and nanostructure on the conversion and alloying/dealloying reaction upon lithiation and its impact on the microstructure and cyclability of the anodes. A further section is dedicated to recent advances in the fabrication of diverse 0D to 3D nanostructures to overcome stability issues induced by large volume changes during cycling. Additionally, the role of doping on conductivity and synergistic effects of redox‐active and ‐inactive dopants on the reversible lithium‐storage capacity and rate capability are discussed. Furthermore, the synthesis and electrochemical properties of nanostructured SnO2/C composites are reviewed. The broad research spectrum of SnO2 anode materials is finally reflected in a brief overview of recent work published on Na‐ and K‐ion batteries.


Introduction
Lithium-ion batteries (LIBs) represent the most advanced electrochemical energy-storage technology for powering mobile and consumer applications,w ith energya nd powerd ensities greatly exceeding those of other battery systems. Although enormous progress in the performance of LIBs has been achieved in recent decades, making even large-scale energy storage applications, such as electric vehicles feasible, the constantly growing demandf or electricale nergy storaged evices necessitates the development of novel battery chemistries to furtheri ncrease the energy density on the cell level. [1] By using materials with different energy-storage mechanisms, such as alloyingo rc onversion,i nsteado ft he state-of-the-art insertion anode material, graphite, is ap romising way to significantlyi ncrease the charge-storage capacity.
Tin-based conversion and alloyinga node materials gained considerable attention in recent years due to their high theoreticalc apacity.M etallic tin, tin alloys, stannates, or tin chalcogenides such as tin (di)sulfide and tin (di)oxide were intensively investigated as battery anode materials. [2] Among the listed materials classes, metallic tin features the highest theoretical capacity, buts uffers from severes tabilityi ssues upon cycling. Althoughn anostructuring or alloyingw ere shown to be promising concepts to improvel ong-term stability,t he use of metallic tin as an anode remains very challenging. [2] Tind ioxide (SnO 2 )a nd layered sulfides (SnS or SnS 2 )e xhibit comparable theoretical capacities. However,t in oxidess how faster lithiation/delithiation kinetics and ag reatly enhanced cyclability,w hereas the Li insertion and conversionr eaction with SnS 2 is only partly reversible. [2] Therefore, SnO 2 is believed to be ap otentialc andidate as an activea node material for nextgeneration LIBs.
It was more than 20 years ago that tin oxide materials were reported, for the first time, by Idota et al. from the Fuji Photo Film Celltec Co. (Japan) company as highly promising anode materials. [3] Since that time, tin oxide containing materials have gainedt remendousa ttentiond ue to the high theoretical and volumetric capacity,b iological compatibility,e nvironmental friendliness, and rather lowc ost. Moreover, the low discharge potentialo fS nO 2 makesi te ven more attractive as an anode materialinL IBs. [4] The lithium reactionw ith SnO 2 has been long believed to proceedt hrough two major steps, namely,aconversion reaction followed by as ubsequent alloying/dealloying process;this was substantiated by various in situ studies. [5] However,m ore recent theoretical calculations [7] and in situ scanningt ransmission electron microscopy on nanowires [7b] suggested the occurrence of Li + insertion into the SnO 2 lattice preceding the abovementioned steps ( Figure 1).
Herein, recent progress in the field of tin oxide (SnO 2 )-based nanosized and nanostructured materials as conversion and alloying/dealloying-type anodes in lithium-ion batteries and beyond( sodium-a nd potassium-ionb atteries) is briefly discussed. The first section addresses the importance of the initial SnO 2 micro-and nanostructure on the conversion and alloying/dealloying reaction upon lithiation andi ts impact on the microstructure and cyclability of the anodes. Af urther section is dedicated to recenta dvances in the fabrication of diverse 0D to 3D nanostructures to overcome stability issues induced by large volume changes during cycling. Additionally,t he role of doping on conductivitya nd synergistice ffectso fr edoxactive and -inactive dopants on the reversible lithium-storage capacity and rate capability are discussed. Furthermore, the synthesis and electrochemical properties of nanostructured SnO 2 /C composites are reviewed. The broad research spectrum of SnO 2 anode materials is finally reflected in ab rief overview of recent work published on Na-and K-ion batteries.
Based on the latest findings,t he total process of the lithium reactionwith SnO 2 can be presented as Equations (1)- (3).
insertion ðintermediate phaseÞ : conversion : Florian Zoller is aP h.D. student in the Fattakhova-Rohlfing group at the Uni-versitätD uisburg-Essen (UDE). He received his BSc. degree in chemistry and biochemistry and his M.Sc. degree in chemistry from Ludwig-Maximilians-UniversitätMünchen (LMU). His current research interests include novel nanostructured lithium-ion battery anode and cathode materials as well as electrocatalysis.
Daniel Bçhm is aP h.D. student in the Fattakhova-Rohlfing group at the Ludwig-Maximilians-UniversitätM ünchen (LMU). He received his B.Sc. degree in chemistry and biochemistry and his M.Sc. degree in chemistry from the LMU. His current research interests include novel nanostructured lithiumion battery anode materials and electrocatalysis for water-splitting applications.
Thomas Bein is Chair of Physical Chemistry at the Ludwig-Maximilians-UniversitätM ünchen (LMU). He leads a research team dedicated to the discovery and translation of novel functional nanostructures related to renewable energy conversion technologies and biomedical applications.
Dina Fattakhova-Rohlfing is Head of the Department of Electrochemical Storage at the Institute of Energy and Climate research (IEK-1) at Forschungszentrum Jülich (FZJ) and Professor at the UniversitätD uisburg-Essen (UDE). Her research is focused on the development of materials for electrochemical applications, including electrocatalysis and electrochemical energy storage. Sn þ x Li þ þ x e À ! Li x Sn ð0 x 4:4Þ Ab initio calculations for the first lithiation cycle predicted Li 2 SnO 3 and Li 8 SnO 6 as compositions of intermediate phases. [7a] Recently,F erraresi et al. found strong experimental evidence for the existence of these phases by combining electrochemistry,p ostmortem X-ray photoelectron spectroscopy (XPS), and SEM imaging together with DFT calculations. [6] The few available reports in the literature indicatet hat the compositiona nd spatiald istribution of intermediate LiÀSnÀOp hases and the reversibility of subsequent reactions steps are strongly affected by the composition and morphologyo fp arent SnO 2 electrodes. The crystallinity andc omposition( exact stoichiometry,d efects, surface termination, impurities), as important parameters of SnO 2 materials, all influenced by the choice of precursors and the fabrication method, are known to affect their electrochemicalp erformance and stability.S tudies on af lat amorphous SnO 2 film as amodel electrode demonstrate that the reversibility of the reactions teps strongly depends on the reactions during the first lithiation cycle, as proposed by calculations on the Li x Sn phase diagram. [6,7] The typical cyclic voltammogram ( Figure 1) furthermore shows redox features of side reactions at the interface that are associated with solid-electrolyte interface (SEI) formation and electrolyte reduction, which contributet oi rreversible capacity loss (ICL) of SnO 2based anodes in the first cycles. [6] In as ubsequentc onversion reaction[ Eq. (2)],t he intermediate Li x SnO 2 compounds are reduced to metallic Sn, which crystallizes in aL i 2 Om atrix. [7] Thec onversion reactiono fS nO 2 to metallict in is reported to be irreversible for bulk SnO 2 ,b ut it can be (partially) reversible for nanosized SnO 2 ;t his greatly depends on the particles ize and morphology. [4b, 5a, 8] Upon further Li-ionu ptake, the surrounding matrix with metallicS np articles is lithiated to form Li x Sn alloys [Eq. (3)].I t has been shown that, starting from the b-Sn phase, am ixture of cubic a-a nd the tetragonal b-Sn (Figure 2b,c) is formed; the a-phaseisstabilized for small nanostructures. [7a, 9] The alloying/dealloying process between Sn and Li + is considered to be reversible. [8c,d] According to experimentally determined and ab initio calculated Li x Sn phase diagrams,t he followingL i ÀSn alloys are proposed to form during the lithiation/delithiationc ycles with in-creasingL ic ontent:L iSn, Li 13 Sn 5 ,L i 7 Sn 2 ,u pt oL i 17 Sn 4 (Figure 3a-d). [7a] The specific capacity of the SnO 2 anodes is greatlyd ependent on the reversibility of different reaction steps.T he theoretical capacity of the complete reaction, including both conversion anda lloying is as high as 1494 mAh g À1 ,b ut it reduces to 783 mAh g À1 if only the alloying/dealloying reactioni sr eversible. It should be noted, however,t hat, even if only partial reversibility of the alloying/dealloying step is possible, the specific capacity still significantly exceeds that of graphite (372 mAh g À1 ). [10] Apart from the quasi-irreversibility of the conversion reaction and subsequents evere capacity loss during the first cycles, SnO 2 -based anodess uffer from large volume changes of up to 250 %d uring the alloying and dealloying process. [5c] This causes internal stress that leads to pulverization of the electrode. Moreover,i ns itu XRD and TEM measurements also reveal that the formed tin particles can agglomeratei nto tin clusters that are less electrochemically active. Last, but not least, volumec hanges impede the formation of as table SEI, which prevents further electrolyte decomposition. These factors are responsible for fast capacity fading andd ecreased cycling performance upon repeated cycling, [5a, 8a,c,d, 9, 10] and are the main reasonst hat SnO 2 -based anodesh ave not yet been commercialized.
The shortcomings of SnO 2 -based anodesa re mainly addressedb yu sing two main strategies. One is to tailor bulk SnO 2 down to the nanosize and/ort on anostructure the SnO 2 compounds towards nanoparticles, [9,11] 1D nanorods, [12] nanowires, [13] nanotubes, [14] 2D nanosheets, [15] and 3D porous [16] or hollow [8d, 17] structures.N anosized materials are known to accommodate large volume changes and to shortend iffusion paths for electrons and lithium ions. Porous or hollow structured (nanosized)S nO 2 can provide additional free space to reduce the problemso fp ulverization and large volume changes. [1b] Another effective approach is the fabrication of composites of SnO 2 andc arbonaceous materials. The carbonaceouss upports increase the overall conductivity of the composites and can also buffer large volumec hanges of SnO 2 during alloying and dealloying. There are many reports on carbon coating of SnO 2 , [18] as well as composites consistingo fS nO 2 and carbonaceous materials, including carbon nanotubes (CNTs), [19] fibers, [20] aerogels, [21] hollow spheres, [22] and graphene. [23] Herein, we introduce recent developments regarding different tin oxide based anode materials systems, with af ocus on the properties of the materials that affect their applicationi n future energy-storage devices.B ased on the analysis of key electrochemical properties, the phasesi dentified during electrochemical transformationsa nd the consequences arising for the reversibility of their transformations, the general goal of this Minireviewi st oi ndicate solutionst om aximize the initial storagec apacity and to overcome ICL, which is mainly associated with the conversion reaction. The most promising strategies to improve the performance of SnO 2 -based anodes, such as nanostructuring, doping, and composite formation, to obtain high-rate and high-capacity anodes for future LIBs, and potentially also for sodium-(NIBs) andp otassium-ion batteries (KIBs), are discussed in separate sections.

Nanostructured Phase-Pure SnO 2 LIB Anodes
Large volume changes,t ogether with repeated cycling of bulk SnO 2 ,l eads to pulverization of the anode and to decreased electricalc ontact, which causes ad rastic loss in capacity within af ew cycles.O ther seriousd rawbacks of pure SnO 2 are its low electronic and ionic conductivity.Avery lowroom-temperature conductivity of SnO 2 of 1.82 10 À8 Scm À1 [24] drastically limits its storagea nd rate capability as an anode material. The measured apparent lithium-ion diffusion coefficient is also low; the reportedv alues range from 10 À16 -10 À14 cm 2 s À1 for as puttered metallicS nf ilm (3 mmt hick) to 10 À15 -10 À13 cm 2 s À1 for amorphous SnO 2 tin oxide films ( % 1.5 mm). [25] Similar to other electrode materials with comparable properties (Si can be mentioned asa ni mportant example), nanostructuringi sc onsidered to be ap romising strategy to mitigate the intrinsic drawbacks of the materials. Nanocrystalline SnO 2 ,w ith various nanomorphologies, can accommodate volumeexpansiont hrough built-in porosity and reduce the agglomeration of Sn clusters by ah omogeneous dispersion within an Li 2 Om atrix. It can furthermore decrease the required Li + diffusion pathway by as ignificantly increased electrodeelectrolyte interface, and thereby enable ah igherf lux of ions, resultingi nh igh rate-capablea nodes. [1a, 26] In addition, nanostructured SnO 2 may displaya lteredp roperties, depending on the synthetic conditions, such as as ignificantly increasede lectrical conductivity of 0.1-0.9 Scm À1 measured on single nanorods [27] or the preservation of nanocrystallinity indicated by the presenceo fa na-Sn phase upon repeated cycling. [7a, 9] The presence of an anocrystalline a-Sn phase is thereby correlated to the reversibility of the alloying reaction; however,i ti sn ot clear whether it is actually the phase that influencesr eversibility.The a-phase, which is more stable on an anoscale, might indicate the intactness of the initial nanomorphology and, partic-ularly,t he fine distributiono fS nw ithin the Li 2 Om atrix, which is important for reversibility.
Ac omprehensive review,w ith af ocus on synthetic routes and electrochemical performance of phase-pure SnO 2 -based anodes, was published by Chen and Lou in 2013. [1a] Hence, we aim to providea nu pdateo nr ecent developments of SnO 2based nanostructures applicable as anodesi nL IBs and to link the properties of materials andi nitial SnO 2 morphologiesd efined by the synthetic parameters with electrochemical performance and stability of the resulting electrodes.

Nanoparticles
Diffraction studies on SnO 2 anodes revealed an ICL due to the formation of the amorphous Li 2 Om atrix and afterwards the loss in reversible capacity upon cycling. The reversibility of the reactionu pon cycling wasc orrelated to the initial SnO 2 crystallite size. [28] Generally,i tc an be said that only if the active Sn materialr esultingf rom the conversion of nanosized SnO 2 crystals is well dispersed in the amorphous Li 2 Om atrix is ar eversible alloying reactionw ithoutd rastic capacity fading possible ( Figure 4a). Larger Sn particles that are not homogeneously dispersed in the amorphous Li 2 Om atrix aggregate to form even larger clusters upon cycling, which leads to mechanical and electronic disintegration of the electrode (Figure 4b). [4b, 28] In 2004, Ahn et al. reported SnO 2 particlesa bout1 1nmi n size, which were prepared through ac olloidal method, to be an optimum size for lithium storage and reversibility with respect to the alloying reaction. [8a] In contrast, even smaller SnO 2 nanoparticles (e.g.,2nm) have shown ah igh ICL. As ap ossible reason, increased SEI formation on very small nanoparticles, due to their larger electrochemical surfacea rea, as well as a decreased formation of the surrounding Li 2 Op hase, were proposed;t his may lead to increaseda ggregation, and thus, capacity fading. [8a] Conclusions about the optimum particle sizes are, however,n ot corroborated by other reports and seem to be strongly influenced by the synthetic route. Thus, Kim et al. reportedt hat hydrothermally synthesized particleso fa bout 3nmi ns ize showed an optimum initial ( % 740 mAh g À1 at 60 mA g À1 for the first cycle discharge current) and reversible capacitya nd cycling stability (negligible fading over 60 cycles at 300 mA g À1 discharge current). [9] It can be suggested that the optimum size of SnO 2 nanocrystals, with respect to reversible capacity and capacity retention, is strongly dependento n the exact nature and amount of amorphous Li 2 Om atrix surrounding Sn formed during the initial conversion reaction, which is, in turn, affected by the SnO 2 nanoparticle synthetic route and initial cycle lithiation parameters (see also the discussion about the reactionmechanism in the Introduction).
Ar ecent study by Hu et al. suggested that the capacity decay of SnO 2 -basede lectrodes with larger nanoparticles was not directly induced by mechanical disintegration of the electrode due to large volume changes, but associated with a graduald egradation of the reversible conversion reactiona t potentials below 1.0 Vv ersus Li/Li + . [10] Thermal and stressdriven Sn coarsening that could be correlated to the average crystallite size has been identifieda samain factor responsible for the reversibility of the conversion reaction, andthus, the reversible capacity of SnO 2 -based electrodes. Furthermore, a quantitative relationb etween Sn-grain coarsening and the initial SnO 2 crystallite size was found, with ac ritical size of 11 nm for af ully reversible conversion reaction. Smaller crystallites with high-density Sn/Li 2 Oi nterfaces are reported to possess fast enough interdiffusion kinetics that enableafully reversible conversion reaction.
Through their synthetic approachbased on magnetron-sputtered pure SnO 2 thin films, Hu et al. obtained an initial capacity of 1066 mAh g À1 ,w ith ar eversible capacity of about 915 mAh g À1 at ar ate of 0.2 Ag À1 after 20 cycles,w hich remained stable for over 100 cycles followed by as low decay. [10] Af urther recente xample of SnO 2 nanoparticlesi ncludes the fast and scalable microwave-assisted hydrothermals ynthesis of fine particles of about 14 nm in size. An initial discharge capacity of about 1197 mAh g À1 ,w ith ar eversible capacity of 520 mAh g À1 (2nd cycle), and ac apacity retention of about 53 %( 273 mAh g À1 )a fter 50 cycles at 100 mA g À1 were reported for this material by Yine tal. [11a] Jiang et al. demonstrated al arge-scale hydrothermals ynthesis of SnO 2 nanoparticles about 6nmi ns ize. [11b] Fabricated anodest hat werec ycled between 0.01 and 3.0 Vv ersus Li/Li + showed an initial discharge capacity of 2223 mAh g À1 at ar ate of 0.1 Ag À1 with af ast capacity fading to about 800 mAh g À1 within the first 20 cycles and as low decay to 760 mAh g À1 after 40 cycles. [11b] The reported capacity outperforms the valuesp ublished for other morphologies, such as nanosheets, -tubes, -rods,o r-spheres, and is in the range of tin oxide based carbon and transition-metal oxide composites.
To enhance the rate capability and lithium-storage capacity of SnO 2 -based anodes, Hameed et al. used ah ydrothermals ynthetic method with the micelle-forming surfactant Tween-80 to obtainm esoporous powders of connected SnO 2 nanoparticles ( Figure 5) or -rods. [29] The resulting electrodes showed an initial dischargec apacity of 1877.8 mAh g À1 ,w ith fast capacity fading within the first 20 cycles to stabilizew ith prolonged cycling at 641.1 mAh g À1 at ah igh discharge rate of 200 mA g À1 (doubled in comparison to the majority of examples reportedi nt he literature). The rate capability of the porous nanoparticle electrodes is thereby outstanding,w ith values of 629, 490, and 340 mAh g À1 at current densities of 300, 500, and 1000 mA g À1 , respectively;t his is attributed to their open and accessible morphology. [29] Apart from 0D structures,i nt he form of nanoparticles, considerable effort was made, in recent years, to fabricate anodes with diverse 1D to 3D morphologies. The goal is to form an optimized electrode-electrolyte interface that enables fast lithium diffusion kinetics from the electrolyte, but also am aximized utilization of active material by offering short diffusion pathways in nanostructures. The second aim is to fabricate "breathable" structures that can accommodate volume changes induced by the alloying/dealloying reaction during cycling, and therefore, preventm echanical and electricald isintegration of the active material.

Nanorods
The synthesis of high aspectr atio SnO 2 structures was initially demonstrated by Liu et al. in 2001 for an inverse microemulsion system (ImE). [30] The reactionc onditions, including the choice of precursors and ah igh calcinationt emperature ( % 800 8C), are thereby similart ot hose used in the molten salt synthetic method widely applied for the formation of nanostructured ceramic powders.
Since then, severalg roups have adapted the concept of ImEbased synthesis;f irst with ah igh or moderate temperature and/or salt-assisted calcination and later also by using as olvo-/ hydrothermala pproach at temperatures as low as 150 8C. [1a, 12, 31] In 2010, Xi and Yi synthesized nanorods with diameters down to 1-1.5 nm that exhibited as trong quantum confinement effect,i ncreasing E g by about 0.9 eV relative to that of bulk SnO 2 . [32] Am ain focus of the work, however,w as to investigate the nanorod growth mechanism through time-dependent diffraction and high-resolution (HR) TEM measurements.
According to Equation (4), the formation of sphere-like SnO 2 nanoparticles is driven by am ild hydrolysis reaction (aqueous urea solution at 90 8C): Larger cubelike SnO 2 nanoparticles with defined crystal facets evolve from ac lassical crystallization process known as Ostwaldr ipening.T he 1D nanorod morphology is then obtained without templating agentso rl ong-chain organic solvents througha ne nergetically drivena ssembly of particles on their (001) facets to reduce the surface energy,u ltimately leading to ag rowth alongt he [001] orientation.T hese 1D aggregates of SnO 2 nanoparticles recrystallize to finally form singlecrystalline SnO 2 nanorods. [32] Examples of the nanorod morphologye mployed in SnO 2based anodes in recent years include the synthesis of SBA-15templated active materialb yJ iao et al. in 2014. [33] In this work, as olution of SnCl 2 is used for the infiltration of am esoporous silica (SBA-15) hard template, which is removed after drying and calcination of the SnO 2 nanorods inside its aligned pores ( Figure 6).
The resulting anode material showed an initial discharge capacity of 1119mAh g À1 and ar eversiblec apacity of about 700 mAh g À1 (2ndc ycle) that declined to about 300 mAh g À1 within 50 cycles at ar ate of 100 mAh g À1 ,w hichc orresponded to acapacity retention of about 43 %. [33] In 2015, Han et al. synthesized larger,h ighly alignedS nO 2 nanorodsi nt he size range of about5 0 100-150nmo naselfproduced Na 2 Sn(OH) 6 substrate through ao ne-step, templatefree hydrothermals ynthetic method. [34] Single-crystalline rods grew along the [001] orientationo nt he substrate and exposed (110) facets. An initial discharge capacity of 1930 mAh g À1 was determined for this material, with ah igh reversible capacity of around1 000mAh g À1 that was retained at about 60 %a tarate of 100 mA g À1 after 20 cycles ( % 600 mAh g À1 ). In 2017, Sennu et al. used am odified precipitation route, with ar elated mild hydrothermalt reatment and calcination, to obtain bundles of SnO 2 nanorods with dimensions of 2-3.5 and 0.2-0.3 mminl ength and diameter, respectively. [35] The material morphologyr esembling marinea lgae is polycrystalline in nature and built up from individual SnO 2 particles of around 10-20 nm (Figure 7).

Nanowires and -tubes
SnO 2 conversion and alloying anodes with 1D nanowire morphologyw ere fabricated by various synthesis approaches in recent years.
Wu et al. synthesized nanowireso fa bout 200 nm in diameter and several micrometers in length through ac arbon-assisted thermal evaporation technique under ambient conditions in as inglez one tube furnace. [36] Ap romising initial reversible capacityo fa bout 1350 mAh g À1 ,w ithac apacity retention of about 46 %( % 620 mAh g À1 )a fter 50 cycles, was achieved at 100 mA g À1 .
Lee and Kim synthesized SnO 2 nanowirea rrays by meanso f chemicalv apor deposition (CVD) with distinct patternsb y using ap hotolithographic process. [13b] Theb est performing samples of this type showeda ni nitial discharge capacity of about 1600 mAh g À1 and ar eversible capacity of about 700 mAh g À1 that faded to about 500 mAh g À1 within 50 cycles ( % 71 %c apacity retention), and down to 400 mAh g À1 within 100 cycles.
In 2017, Lee et al. [13a] were able to synthesize hierarchically branched SnO 2 nanowires through at wo-step CVD method, which showedaslightly increased performance compared with that of the work of Lee and Kim. [13b] The material also showedi nitial discharge and reversible capacities of about 1600 and 800 mAh g À1 ,r espectively,w ith 69 %c apacity reten-tion ( % 550 mAh g À1 )a fter 50 cycles, and about 400 mAh g À1 after 100 cycles at ar ate of 0.1 C(1C= 400 mA g À1 ). [13a] Related nanotube SnO 2 morphologies were recently investigated by Han et al. in an oxalate-assisted "redox etching and precipitating"' route involving MnOOH nanowires and Sn 2 + ions. SnO 2 nanotubes with ad iametero f2 00-250nma nd several micrometers in length were synthesized (Figure 8). [37] Electrode measurements showed an initial discharge capacity of about 2000mAh g À1 with ah igh reversible capacity of 1400 mAh g À1 that faded to 700 mAh g À1 within 50 cycles (50 % capacityr etention). Extended cycling showedarather high stability of the electrode material, with ad ischargec apacity of 500 mAh g À1 after 100 cycles at an elevated rate of 500 mA g À1 . [37] 2.4. Nanosheets The 2D SnO 2 nanosheet morphologya nd its application as a LIB anode material wast horoughly discussed in ar eview by Chen and Lou in 2012. [38] The electrochemical performance of nanosheet-based anodeswas found to be greatly influenced by the morphology, crystallinity,a nd phase purity of SnO 2 ,w ith as trong effect of the precursors used on the resulting product. Thus, anisotropic growth of SnO 2 with the formation of nanosheets was successfully achieved through hydrothermal synthesis with SnCl 2 as the precursor. [38] However,t he presenceo ff luoride ions, either by using SnF 2 as the tin oxide precursor or by using an additional fluoride source,s uch as NH 4 F, with the actual tin oxide precursor (e.g.,S nCl 2 )w as shown to lead to the formation of an SnO/SnO 2 mixture (for SnF 2 as the precursor) or phase-pure SnO 2 nanosheets (for NH 4 Fa sa na dditive), respectively,u nder hydrothermalc onditions. [39] Ar ecent example for the fabrication of SnO 2 nanosheets is given by the work of Narsimulu et al.,w ho described the surfactant-andt emplate-free hydrothermal and microwave-assisted synthesis of agglomerated SnO 2 nanosheets (Figure 9). [15a]  The respective electrodes showed am oderate initial discharge capacityo f1 350 mAh g À1 ,w ith ar eversible capacity of 873 mAh g À1 that faded to 258 mAh g À1 within 50 cycles at a rate of 100 mA g À1 . [15a]

3D hollow nanostructures
Beyondt he 0D, 1D, and 2D SnO 2 materials introduced above, porous3 Dm orphologies were fabricated in recent years. Amongt hem, hollow and porousn ano-and microspheres, [16a, 17] as well as 3D ordered macroporous structures, [40] were synthesized and proposed to possesss tructural flexibility to counteract fast pulverization of the anode by volume changes induced upon cycling.
Ap romising synthetic route was presented by Li et al.,w ho used negatively charged carbonaceousm icrospheres (CMSs) prepared through ah ydrothermalm ethod that electrostatically boundS n 4 + ions on their surface. [16a] After calcination in air with simultaneous templater emoval,h ollow dumbbell-shaped microspheres of several micrometers were obtained ( Figure 10).
Electrochemical measurements revealav ery high and reversible lithium-ion storage capability of the material of about 1000 mAh g À1 in the second discharge cycle ( % 1750 mAh g À1 in the first discharge cycle) that is maintained after 100 cycles, with about 600 mAh g À1 at ar ate of 500 mA g À1 and still above 500 mAh g À1 with an appliedr ate of 1Ag À1 .T he capability of the hollow spherical structures to effectively buffer volume changes and to allow high rate applications is reflectedb yg alvanostatic charge/discharge measurements, with rates of up to 1600 mA g À1 and ar esulting capacity of over 500 mAh g À1 , which returnsto695 mAh g À1 if the rate is decreased to the initial value of 100 mA g À1 . [16a] Another way to obtain large hollow SnO 2 structures with rodlike shapes was developed by Wang et al. [41] In their syn-  thetic approach, ag enetically modified microbial Escherichia coli template binds aS n 2 + precursor on its surface through electrostatic interactions. Subsequent calcination results in the formation of about 400 600 nm rodlikeh ierarchical SnO 2 structures composed of smaller nanosheets andparticles. However,t he electrochemical performance of the prepared anodes is moderate, with an initial discharge capacity of about 975 mAh g À1 and ac apacity retention fading to 194 mAh g À1 ( % 20 %o ft he initial value) after 50 cycles at ar ate of 200 mA g À1 .

Doped SnO 2 LIB Anodes
Element doping is ak nown approacht oo ptimize the electrochemicalp erformance of SnO 2 -based electrodes. Dopingc an lead to ag reatly increased electronic conductivity,w hich is beneficial for the electrode performance.
Pure SnO 2 is aw ide band gap semiconductor,w itha no ptical band gap of 3.6 eV at room temperature. It exhibits an intrinsic n-type conductivity due to the presence of shallow donor levels located at 0.46 eV below the conduction band, which is attributed to ionizedd efects (e.g.,u nintentional hydrogen atom doping), according to computational studies by Singh et al. and more recently by Villamagua et al. [42] Fluorine doping is reported to increase the conductivity up to about 5 10 3 Scm À1 by substituting O 2À in the crystal structure, and thereby creating shallow donors that enhance the n-type conductivity significantly. [43] Due to better synthetic control than that with fluorine doping, p-type doping with Group III atoms (In, Ga, and Al) [42b] or n-type doping with Group Va toms (Sb doping), [44] which createss hallow levels,w as thoroughly exploredi nr ecent years. In addition to increasing conductivity,t ransition-metal doping is reported to decrease large volumec hanges upon lithiation/delithiation. [45] In recent years, av ariety of transitionmetal dopantsf or SnO 2 were proposed in the literature;t hese can be divided into two groups:r edox-inactivea nd -active elements that can undergo conversion/alloying reactions with lithium ions in the potential range applicable for SnO 2 -based anodes. [46] Niobium, [47] titanium, [48] zirconium, [46] palladium, [49] and tungsten [50] can be assigned to the first group. Doping with these transition-metal ions does not result in an observable gain in capacity because the lithiation/delithiation curves of SnO 2 anode materials remainu nchanged, without additional redox features from the doping elements in the respective potential window.H owever,d oped tin oxidess how as ignificantly increased cyclability and rate capability. [46] The beneficial effect on the cycling performance provided by both redox-active and -inactivet ransition-metal doping in conversion-typea nodes (ZnO, SnO 2 )w as initially attributed to the decreased crystal size observed upon doping;t husl imiting the aggregation of primary nanoparticles and enabling ar eversible lithium alloy formation. [51] Recent investigations suggest that the improved performance of doped tin oxides results from an increasei n the conductivity of the active materialc aused by an additional charge percolation pathway provided by the transition-metal (dopant)i on network in the SnO 2 structure, as well as through an increasei nt he intrinsic conductivity through newly generated surfaceo xygen vacancies. [49][50][51] Thed egree of conversion reactionv ersus side reactions, such as particle aggregation,i s thereby correlated with the reaction kinetics,w hich depend strongly on the electron-transfer properties and local current density. [49,50] Apart from increased conductivity,acatalytic effect of transition-metal ions on decomposition of the Li 2 O phase is discussed;t his furtherp romotes ar eversible conversion reaction. [49] In the context of widely appliedS nO 2 /graphene composites, transition-metal doping( W-doped SnO 2 ) has also been shown to reduce the charge-transfer resistance betweena ctive materialp articles and graphene through an increased interaction at the interface. [50] Redox-actived opantsi ncludem anganese, [46,52] iron, [46,52,53] antimony, [44b, 54] cobalt, [4a, 45a,c,d, 46, 52, 55] nickel, [46] copper, [46] zinc, [45d, 46, 56] and molybdenum. [45b] In addition to the effect of redox-inactive dopants discussed above, their corresponding metal oxides can, in principle, undergo ac onversion reaction with lithium over the applied potentialr ange of the anode, resultingi nat heoretical gain in capacity (see also Section 4.4). [46] However, the increased capacity does not necessarily translate into an increased energy density of af ull-cell assembly because dopants( e.g.,C u) can causeavoltage hysteresis;t hus lowering the total storable energy. [46] Moreover,o ther dopants or multidoping strategies have been reported, for example, Mg, [45d] Al, [57] In, [58] F, [45c, 59] N, [60] P, [61] S/F, [62] Co/F, [45c] and Co/N. [45a] Among others, cobalt is an interesting redox-actived opant because Co-doped SnO 2 shows av olume buffering effect that is attributed to ar educed and maintained small SnO 2 primary particles ize upon cycling. Furthermore, Co-doped SnO 2 demonstrates am easurable gain in capacityv ersusu ndoped SnO 2 , with ad ecreased voltage hysteresis and increased coulombic efficiency. [45a,d] Nithyadharseni et al. compared Co-, Mg-, and Zn-doped SnO 2 nanoparticles. [45d] The compoundsw ere prepared through sol-gels ynthesis with ethylene glycol, dimethyl ether,a nd citric acid. They found that cobalt doping led to a superior electrochemical performance. The Co-dopede lectrodes deliver as pecific capacity of 573 mAh g À1 ,c ompared with 330 mAh g À1 for the undoped sample, after 50 cycles at 60 mA g À1 .T hey attributedt his to structurals tability and CoÀ Sn intermetallic interactions. Lübke et al. reporteds imilar results and confirmed that, in their case, Co doping was also superior to that of Nb-, Ti-, Zr-, Fe-, Cu-, Zn-, Mn-,a nd Ni-doped materials. [46] Not only does the nature of the dopant, buta lso the doping ratio, strongly influence the electrochemical performance, as studied by Ma et al.,w ho compared pure SnO 2 with Co-doped SnO 2 with cobalt concentrationso f5 ,1 0, and 15 %. [4a] They found that the particle size decreased with increasing dopant concentration. Ad opant ratio of 10 %( Sn 0.9. Co 0.10 O 2 )p rovided the best cycling stabilityo ff our investigated materials. After 50 cycles at 0.1 Ag À1 ,aspecific capacity of 493 mAh g À1 was obtained for the Sn 0.9 Co 0.10 O 2 sample, compared with 242, 464, and 476 mAh g À1 for SnO 2 ,S n 0. 85  furthere nhanced by carbon coating. [4a] The influence of carbon and its derivatives on the electrochemical performance of SnO 2 /C composites is reviewed in more detail in Section 4. Very promisingr esultsregardingt he incorporation of transition metals into SnO 2 were also reported by Wang et al. [53b] Thea uthors compared the electrochemical performance of an Fedoped SnO 2 /reduced graphene oxide (rGO) nanocomposite with undoped SnO 2 /rGO and pristine SnO 2 nanoparticles;a ll of them synthesized through aw et chemical approach. TEM measurements showed that the 6-8 nm small SnO 2 and FeÀ SnO 2 nanoparticles wereh ighlyd ispersed ( Figure 12) over the rGO sheets; this is beneficial for buffering volume changes upon cycling (see Section 4.3), and hence,i nfluences the cycling performance:t he bare SnO 2 electrode reached only 172 mAh g À1 after 60 cycles at 0.1 Ag À1 compared with 905 mAh g À1 for the rGO composite after 100 cycles ( Figure 12). The FeÀSnO 2 /rGO nanocomposite even retained a capacityof1353 mAh g À1 after 100 cycles.The performance improvement is attributed to iron doping because it leads to better electrical conductivity and encourages the conversion reaction. Consequently,t he rate performance of the FeÀSnO 2 / rGO nanocomposite is also superior to that of the undopeda nalogue. [53b]

SnO 2 -Based Composite LIB Anodes
The use of SnO 2 togetherw ith ac arbonaceous material has positive effects on the electrochemical performance. [4a] The carbonaceouss upport can bufferv olumec hanges that occur during the alloying/dealloying processes, suppress pulverization and agglomeration of the electrode material, and enhance the overall electrical conductivity in the material. [18c,f] SnO 2 / carbon composites are synthesized either from SnO 2 active material together with am olecular organicc arbon precursor or from preformed carbon allotrope based precursors. Beyond the use of carbon,v arious metal-based components, especially transition-metal chalcogenides, were investigated for the fabrication of composite anodesw ith SnO 2 for superior electrochemicalperformance.

Amorphous carbon (SnO 2 /C composites)
There are different synthetic routes to obtain an amorphous carbon layer coated on SnO 2 as an active electrode material. One approach is to use both SnO 2 and carbonaceous precursors to form SnO 2 and the carbon layer in situ. [18a, 22b, 63] Af urther synthetic route utilizes preformed 3D carbon structures present duringS nO 2 synthesis. [64] At hird possible strategy is to synthesize SnO 2 first and subsequently treat it with ac arbon precursor. [18b-f, 19b, 65] This is especially helpful for retaining the morphology of SnO 2 compounds with exceptional structures.
Zhou et al.,f or example, used the last approach to preserve the "sub-microbox" structure of SnO 2 . [18c] They used N-doped carbon,i nstead of pure carbon, which was supposed to further enhancet he conductivity and electrochemical performance. The sub-microboxes were prepared by means of am ultistep synthetic strategy in which Fe 2 O 3 sub-microcubes served as templates to be covered with SnO 2 particles in an in situ hydrothermal process. The resulting core-shell structure was then covered with as mooth layer of polydopamine, whichwas converted into N-doped carbon by annealing at 500 8Cu nder nitrogen. Finally,t he Fe 2 O 3 core was removed by etchingw ith oxalic acid. The resulting SnO 2 /N-doped carbon (SnO 2 /NC) submicroboxes have an average size of 400 nm constructed from  nanoparticles with sizes of 4-5 nm. Zhouetal. could show that SnO 2 /NC displayed abetter cycling performance and rate capability than that of uncoated SnO 2 sub-microboxes. After 100 cycles at 0.5 Ag À1 ,c apacities of 491 and 75 mAh g À1 were obtained for the NC-coated and "pure"S nO 2 sub-microboxes, respectively ( Figure 13). The authors attributed the superior electrochemical performance of the SnO 2 /NC sub-microboxes to the large specific surface area andp ore volume, small particle size, and increased conductivity supplied by the NC.

CNTs(SnO 2 /C composites)
CNTsa re an important example of 1D nanostructured carbon support materials. The use of CNTst ogether with SnO 2 can add attractive features. The CNTsc an improve the electrical conductivity,b uffer volume changes during alloying/dealloying with Li ions, and enablef ast electron-transfer pathways. [19c-e] The first step in the synthesis of SnO 2 /CNT composites is often ah arsh treatment of pristine CNTsw iths trong acids or strong oxidizinga gents. This creates functional groups on the CNTs that can be used to anchorS nO 2 particles. [19e, 66] Such treatment leads, however,t os tructural damage and decreased electrical conductivity. [19e] Ma et al. reported as ynthesis without the oxidation of CNTs. [19e] They used glucose as am ediating agent during hydrothermals ynthesis to assist in the in situ formation of 7nmS nO 2 particles and serve as ac arbon source. The glucose-assisted SnO 2 /CNT composites exhibited as uperior cycling performance. After 150 cycles at 1Ag À1 ,aspecific capacity of around9 00 mAh g À1 was retained, compared with around 450 mAh g À1 for the unmediated SnO 2 /CNT composite. Pure SnO 2 exhibits even lower values. The glucose-assisted SnO 2 / CNT composites alsos howed as uperiorc ycling performance at different Cr ates;t his was also attributed to the unique struc-ture and, consequently,e nhanced electrical conductivity. [19e] Cheng et al. reported that the SnÀCb ond content played a crucial role. [19d] They synthesized SnO 2 /CNT composites through ah ydrothermala pproach by using commercial functionalized multiwalled CNTsf ollowed by an annealing step at different temperatures. The SnÀCf raction strongly depends on this step. The composite annealed at 500 8Ce xhibited the best cycling and rate performance, compared with those of composites heated at 400 and 600 8C. The first compound demonstratesacapacity of around 600 mAh g À1 after 400 cycles at 0.2 Ag À1 ,w hereas the other two have capacities of only 323 and 211mAh g À1 ,r espectively,a fter 200 cycles. The authors attributed the promising electrochemical performance to thei nterplay of the particlesize;conductivity;and, most importantly, favorableSn ÀCbondingi nt he SnO 2 /CNT composite. [19d]

Graphene (SnO 2 /C composites)
Graphene is an important 2D carbonaceous support material with exceptionalp roperties, such as very good electrical conductivity, large surfacea rea, high theoretical capacity of 744 mAh g À1 ,a nd excellent mechanical properties. The last of these,f or example, can help to avoid aggregation of SnO 2 particles and buffer volumec hanges duringa lloying/dealloying with Li ions;t hus leading to better cycling stability (Figure 14). [23a, 67] SnO 2 /graphene composites can be obtained by simply mixingS nO 2 with graphene or graphene oxide (GO) or through an in situ method, whichi sm ore common. [23c, 68] For the latter, graphene or GO is treated with at in precursor (e.g.,S nCl 4 or SnSO 4 )t of orm SnO 2 particles attached to the graphene or GO surface. In particular, functional groups such as epoxide, carbonyl, or hydroxyl, which can be found on the GO surface, are attractive anchor points for the tin precursors. [23b, 67a, 69] If GO has not been reduced to graphene during the synthesis, there are two popularo ptions:t he use of as trong reducing agent (e.g.,h ydrazine) or heating the sample under ar educing or inert atmosphere,f or example, H 2 or N 2 .T he obtained graphene/rGOh as as uperior conductivity to that of GO. [   They used ap H-dependent, one-pot hydrothermal methodt o grow SnO 2 nanoparticles (2-5 nm) in situ onto the surface of graphene sheets.T he SnO 2 /rGOn anocomposite delivers as pecific capacity of 942 mAh g À1 after 80 cycles at 100 mA g À1 , comparedw ith 827 and 142 mAh g À1 for SnO 2 /GO and pristine SnO 2 particles, respectively.T he SnO 2 /rGOn anocomposite also exhibits asuperior rate capability. [23b] However,S nO 2 particles can aggregate on graphene sheets during cyclic lithiation/delithiation reactions, whichc ould lead to al oss in capacity. [69] Carbon coating of SnO 2 particles can avoid the formation of such agglomerates, as discussed previously herein. Hence,t he use of both carbon coating andg raphene as as upport is reported to be advantageous for the electrochemical performance. For example, Zhang et al. presented ac arbon-coated SnO 2 graphene (rGO/PC/SnO 2 )n anocomposite with an improved rate performance and cycling stability to that of an uncoated reference composite. [69] The SnO 2 nanoparticles are formed in situ on the GO sheets through a solvothermala pproach, with as ize of around 4nm. The additionally added glucose served both as as oft template anda sa carbon-coating source.T he rGO/PC/SnO 2 nanocomposite exhibits ac apacity of 1468mAh g À1 after 150 cycles at 0.1 C, relative to 200 mAh g À1 for the uncoated sample. The rate performance of the coated nanocomposite is also superior.T he authors argued that this excellent performance was caused by the small particle size, good conductivity,l arge electrolyteactive material interface, and mechanical stabilization of the nanocomposite.
Importantly,n ot only SnO 2 ,b ut also graphene sheets,c an suffer from some kind of aggregation.G raphene sheets tend to restack due to p-p interactions, whichi mplies an inferior compensation of the volumec hanges of SnO 2 and, as ac onsequence, ar educed electrochemical performance. [1b, 23a] Fabrication of 3D structures and/or the introductiono fabuffering layer are reported to preventt he restacking of individual graphene sheets,w hich has positive effects on the electrochemical performance. [23a, 71] The 3D graphene structures, such as graphene foams, aerogels, or skeletons, can have an increased surfacea rea and more voids to host and/or encapsulate SnO 2 particles. The latter can be beneficial to alleviatev olume changes;h ence increasing the structurals tability and electrochemicalp erformance of SnO 2 /graphene composites. [71,72] Liu et al.,f or example, used as pray-dryinga pproach to preparea SnO 2 /skeleton-structured 3D network of graphene sheets. [71] Their composite exhibits as pecific capacity of 1140 mAh g À1 after 120 cycles, relative to 121 mAh g À1 after 50 cycles for pristine SnO 2 (at 100 mA g À1 ). They attributed the improved electrochemical performance to the skeleton-like 3D structure, which could buffer the volume changes of SnO 2 and was beneficial for electrolyte transport and the diffusion of lithium ions.
Another strategy to improvet he performance of SnO 2 -based anode materials is to use doped SnO 2 nanoparticlesa nd graphene as as upport material. [44b, 48, 50, 53b, 54b, 55, 56, 59b-d] Zoller et al. demonstrated that the electrochemical performance of Sbdoped SnO 2 (ATO)/rGOc omposite was superior to that of SnO 2 /rGO and unsupported ATOp articles. The composites and pure ATOw ere synthesized through am icrowave-assisted sol-vothermal approach, whichl ed to SnO 2 and ATOp articles of around3 -4 nm in size. The superior electrochemical performance of the ATO/rGO composite, relative to those of SnO 2 / rGO andp ure ATO, was especially demonstrated in performance tests at high Crates of up to 60 C ( Figure 15). [44b] Additionally,g raphene can be functionalized and doped with nitrogen [8c, 73] and/ors ulfur, [74] whichc an further enhance the electrochemical rate performance of the SnO 2 /graphene composites, as demonstrated in the recent work by Wu et al. [74a] The authors showedt hat SnO 2 quantum dots anchoredo ns ulfur-doped rGO( S-rGO) outperformed the analogous undoped rGO composite in terms of rate capability and cycling stability;t his was attributed to sulfur doping of graphener esulting in an improved structural stabilitya nd better chargea nd ion conduction at the electrode interface.
In the case of SnO 2 /metal sulfide (M x S y ;M= Sn, Mo) composites, the individual compounds have different band gap energies that enablet he formation of heterojunctions. [77, 78b, 95] As mentioned in Section3,S nO 2 is aw ide band gap (3.8 eV)ntype semiconductor,w hereas SnS is an arrow-band-gap (1.3 eV) p-type semiconductor,f or example. [76] Ap -n heterojunction forms at the interface between SnO 2 and the metal sulfide. This entails holes diffusing from the metal sulfide to SnO 2 and electrons diffusing in the opposing direction;t hus leadingt ot he formation of ad epletion region and the formation of an internal electric field. This enhances charge-transfer kinetics through increased carrier mobilitiesa nd thereby eventually results in ahigherconductivity. [76,96] In this context, Ye et al. demonstrated that SnO 2 /SnS NC composite showed as uperior electrochemical performance to those of pure SnS, SnO 2 ,a nd SnO 2 /NC, reaching values of 550, 300, 200, and 50 mAh g À1 ,r espectively,a fter 100 cycles at 0.1 Ag À1 (Figure 16). [76a] The authors also demonstrated an improvedr ate performance for the SnO 2 /SnS/NCn anocomposite; thus underlining the beneficial effect of the formation of the SnO 2 /SnS heterojunction on the conductivity of the activem aterial.
However,t he improved electrochemical performance of SnO 2 /metalo xide (M x O y ;M = Co, Cu, Fe, Mn, Mo, Ni, W, etc.) hybridsc ompared with that of SnO 2 is associated with sequential lithiation at different potentials of SnO 2 and M x O y . [81b, 82a,d, 85, 86b, 87b] Consequently,i ft he SnO 2 nanoparticles are reduced, at the same time, the M x O y particlesare practically electrochemically inactive and can buffer volume changes and preventn ewly formed Sn particles from aggregating. [82d] Additionally,i tw as reportedt hat, upon cycling, in situ generated metal nanoparticles from the M x O y phase catalytically decomposed the formed Li 2 Om atrix, which increased the overall capacity and cycling stability. [79, 80, 81b, 82a, 84b, 86b, 87b] Notably,t itanium oxides in SnO 2 /M x O y composites are "zero" or low-strain materials that displayn egligible volumec hanges upon lithiation/delithiation,w ith the downside of al ow specific capacity.T itanium oxidesc an therefore be used to preserve the nanostructure of SnO 2 by physicalc onfinementa nd anchoring. [88c,d, 97] The class of 2D metal carbides and nitrides knowna sM Xene has gained considerable attention for composite formation in recenty ears. [98] The synergistic effect in SnO 2 /MXenea nodes is based, on one hand, on thev ery good electronic conductivity and enhanced lithium-ion transport ability of the layered MXene structures, together with their mechanical flexibility, which is important for buffering the volume changes of SnO 2 . On the other hand, SnO 2 prevents the MXene sheets from restacking, and thus, improves the cyclability remarkably. [94] This was, for example, successfully demonstrated by Liu et al. [94a] They compared the cycling performance of aS nO 2 nanowire/Ti 3 C 2 (MXene) nanosheetc omposite, SnO 2 nanowires, and Ti 3 C 2 (MXene) nanosheets (Figure 17), and obtained values of 530, 31, and 139 mAh g À1 ,r espectively,a fter 500 cycles at 1Ag À1 .T he rate performance measurements also confirmed the improved electrochemical performance of the SnO 2 nanowire/MXene composite.
Af urther example of aS nO 2 /non-carbonaceous composite was presentedb yI dota et al.,w ho embedded redox-activeS n II centers into an amorphous glass-forming matrix of À(MÀO)À elements composed of B III ,P V ,a nd Al III ,r esulting in an amorphous SnM x O y composite. [3] Ar eversible capacity of > 600 mAh g À1 was reported at ac harge/discharge current of 48 mA g À1 ,w ith ac apacity retention of > 90 %a fter 100 cycles in af ull-cell configurationw ith aLiCoO 2 (LCO) cathode.

Full LIB Cell Performance with SnO 2 -Based Anodes
Because SnO 2 -based materials exhibit very promising results in half-cells (meaning with Na or Li metal foil as the cathode), there is growing interesti nt esting these materials in full cells to evaluatet heir possible application in LIBs. Mismatching charge/discharge potentials and kinetics of corresponding anode-cathode materials may result in low performance and/ or fast degradation of the active material. [99] Ta ble 1p resents an overview of performance data for full-cell assemblies employing SnO 2 -based anodesand the most commonly used lithium cobalt oxide based cathode materials.
Wu et al. reported ac omposite consisting of hollow SnO 2 nanospheres, NC, and rGO sheets. [101] This unique structure enabled an encouraging electrochemical performance, also on the full-cell level, with commercial LCO as the cathode material ( Figure 18). The full cells were investigated over ap otential range of 1.2-4.2 V. After 90 cycles at 0.1Ag À1 ,ad ischarge capacity of 346 mAh g À1 (based on the weighto ft he anode) was reported;t his equaled ac apacity retention of approximately 67 %.

SnO 2 -based NIB anodes
Since the first successful demonstration of SnO 2 as ap romising anode material in LIBs, there has been growing interesti nt he  (5) and (6), resulting in at otal theoretical specific capacity of 1398 mAh g À1 : [2] conversion : alloying=dealloying : Sn þ x Na þ þ x e À ! Na x Sn ð0 x 3:75Þ ð6Þ The larger ionic diameters of Na + and K + (K + > Na + > Li + ; 1.38 > 1.02 > 0.76 ,r espectively), however,a ggravate problemsc aused by volumec hanges upon charge/discharge, and result in ad ecreased cycling performance compared with that of Li + . [101,103] To tackle these problems,s trategies successfully employed for SnO 2 -based anodesi nL IBs, such as nanosizing, 3D structuring, or the introduction of carbonaceous support materials, were also suggested to improve the electrochemicalperformance in KIBs and NIBs. [2,103,104] The use of SnO 2 together with rGO is an example of this development.Joetal. synthesized aSnO 2 /rGO composite that exhibiteda ni mproved electrochemical performance to that of  In their approach, SnO 2 particles were first solvothermally prepared and then attached to the rGO sheets through al ayer-by-layer self-assembly process ( Figure 19). Cycling tests at 0.1Ag À1 revealed capacities of 492 mAh g À1 (capacity retention:8 0.2 %r elative to that of the first charging cycle) for the composite and 194 mAh g À1 (42.5 % retention of the initial charge capacity) for SnO 2 after 100 cycles. The rate performance of the SnO 2 /rGO composite could also be significantly increased from about 250 to 425 mAh g À1 at 2.4 Ag À1 compared with that of bare SnO 2 .
[104b] For the construction of ahigh energy density sodium ion full cell, they further paired the SnO 2 -nanoparticle/rGO anode with aC -NaCrO 2 cathode. The resulting NaCrO 2 //SnO 2 /rGO composite full cells showeda ne xcellentc ycling stabilitya tar ate of 0.5 C (55 mA g À1 ), with ac apacity retention of 84 %a fter 300 cycles and high rate capabilityt ested up to 10 C( 87 mAh g À1 based on the cathode mass at 1.1 Ag À1 ). [104b] Af urthere xample of as odium-ion full-cell assembly was reported by Lee et al. [102] In their work, aS nO 2 /3D graphene composite prepared through ah ydrothermala pproachw as paired with self-producedN a 3 V 2 (PO 4 ) 3 (NVP) serving as ac athode. The anode material was preactivated before the first cycle to avoid alkalinei on consumption during SEI formation.I nt he case of the SnO 2 /3D graphene-NVPf ull cells, as pecific capacity of 71 mAh g À1 (based on the weighto ft he anode) was reached after 100 cycles at ar ate of 0.05 C.
Af urtheri ncrease in performance was achieved by Wang et al.,w ho used al ayer-by-layer assembly technique with a porphyrin derivative as an interfacial linker to homogeneously attach SnO 2 crystalsa bout5nm in size onto Na nd Sc odoped graphene. [100] By combining it with aN VP/C cathode, ar emarkable full-cell capacity of 108.2 mAh g À1 was measured after 100 cycles at ar ate of 0.1 Ag À1 .
Ta ble 2g ives ab rief overview of recently published sodiumion full-cell battery performance data.

SnO 2 -based KIB anodes
Inspired by as tudy on KÀSn alloying and intercalation by Sultana et al., [105] Wang et al. published an in situ TEM and diffraction study on the potassiation of Sn nanoparticles in KIBs. [106] They observedahighv olumee xpansion of about 197 %a fter   an uptake of only one equivalent of K, with the formation of a KSn phase identified by electron diffraction, accompanied by the reversible formation of nanopores and finally pulverization of the active material. [106] However,i nafollow-up study by Ji et al.,o nd ual-ion batteries, with Sn foil as the anode, ah igher potassium uptake could be observed by means of ex situ XRD measurements, with the formation of aK 2 Sn phase as afinal alloying product. [107] Large volumec hanges induced by the potassiation of metallic Sn and accompanying capacity fading caused by electrode pulverization constitute significant challenges for its application as an anode materialinKIBs. However,ithas been demonstrated that the use of SnO 2 -based electrodes, instead of Sn, can significantly mitigate these effects.S imilar to lithiation processes, the K 2 Om atrix formed in the conversion reactions and surrounding the newly formed Sn (nano) particlesc an buffer volumec hanges upon alloyinga nd suppress aggregation. [108] The positive influence of the K 2 Om atrix formed around Sn nanoparticles on the structurali ntegrity of the tin oxide based anodesf or KIBs, in contrast to metallic Sn-based electrodes, was demonstrated, for example, by Shimizu et al. (Figure 20). [109] They precipitated SnCl 2 precursor,w ith subsequent thermalo xidation, to obtain a1 0mms ized flowerlike morphology composed of SnO 2 sheets of about 100 nm as primary building blocks. The resulting electrodes exhibit ar ather limited potassium storage capability of about 25 mAh g À1 at a rate of 0.025 Ag À1 ,b ut demonstrate stabilityo ver 50 cycles. [109] Huang et al. recently investigated the potassium-storage capabilityo fS nO 2 -carbon nanofibers synthesized by meanso f electrospinning of ap recursor solution containing SnCl 2 /polyacrylonitrile (PAN)/polymethylmethacrylate (PMMA), with a subsequentp yrolysis step, to obtain fibers with ad iametero f about 490 nm and several micrometers in length. [104a] The focus of their work was on enhancing electrode conductivity by the addition of graphene to the electrospinning process and as ynergistic effect on the K + storage behavior among the SnO 2 ,r GO, and carbon constituents. As ar esult, the capacity could be increased from about 170 (SnO 2 /C) to 250 mAh g À1 (SnO 2 /rGO/C)u pon cycling at ar ate of 0.1 Ag À1 . [104a] In af ollow-up paper by Huang et al.,Pdoping of SnO 2 /rGO/ Cb yp hosphorica cid was reported, with the aim of furtheri ncreasingt he electrochemical performance. [103] The electrospinning processo faGO/(H 3 PO 4 )/SnCl 4 /PVP-containing precursor solution yieldedn anofibers of about1 50 (non-P-doped) and 120 nm in diameter (P-doped) and micrometers in length. The cycling performance at ar ate of 0.1 Ag À1 could be increased from about 206 (undoped material) to 285 mAh g À1 (P-doped), both determined for the 60th cycle. The authors hypothesize that modification with H 3 PO 4 had several beneficial effects on the K + diffusion kinetics. These include the formationo fabeneficialm esoporous structure,a ni ncrease in conductivity,a nd a widening of the interlayer spacing of rGO, which is reflected in ar eversible capacity of 200 mAh g À1 at ah igh rate of 1Ag À1 . [103] The bestp erforming SnO 2 anode for KIBs so far,t ot he best of our knowledge,w as recently published by Suo et al.,w ho prepared ab inder-free SnO 2 -nanosheet/stainless-steelm esh (SSM) anode through solvothermal synthesis with aS nCl 2 precursor in the presence of the mesh ( Figure 21). [108] An initial discharge capacity of 603 mAh g À1 was determined for this material, which stabilized within 5c ycles at ar eversible capacity of about 450 mAh g À1 .W ithin 100 cycles, am oderate decrease in capacity to 339 mAh g À1 was observed. The prepared anode materiala lso showedagood rate capability of 125 mAh g À1 at 1Ag À1 .
Ta ble 3p resentsa no verview of SnO 2 -baseda node materials for application in KIBs testedinh alf-cell configurations.

Summary and Outlook
The alloying of alkali ions with tin results in ah igh theoretical volumetric and gravimetricc harge capacity,w hich is accompanied by volumec hanges of up to 200 [106] -250 % [5c] (for K + and Li + ,r espectively). Large volume changes pose am ajor challenge for the mechanical and structurali ntegrity of the electrode upon cycling. [5c, 106] To address this problem, much effort was dedicated to fabricate diverse 0D-3DS nO 2 nanostructures. Based on an analysiso ft he most recent developments, herein, we aimed to elucidate the relationship between the nanostructure, synthetic route employed (resulting phase), and the elec-  trochemical performance of phase-pure SnO 2 .I tc an be concluded that the optimum size of SnO 2 nanocrystals, with respect to reversible capacity and cyclability,s trongly depends on the exact nature( crystallinity and dominating crystal facets determined by the synthetic conditions) and spatial distribution of nanosized Sn and its surrounding amorphous Li 2 O matrix formed during the initial conversion reaction. From the performance data of recently published articles with differing SnO 2 nanomorphologies andc rystallite sizes, we conclude that particles with as ize smaller than 10 nm may yield anodesw ith ah ighi on-storage capacity and reversibility, [11b] which, however,c annot effectively be enhanced by nanostructuring.
As another meanst oi mprove the electrochemical performance of SnO 2 anodes, doping with either redox-active or -inactivea toms was explored by many research groups.W e concludet hat the increasei ne lectrochemical performance (capacity and rate) observedi sassociated with an increaseinconductivity (known for Sb) [44a] inducedb yamodificationo ft he band structure of the wide band semiconductor SnO 2 .A dditionally,a mong av ariety of investigated transition metals, cobalt is very promisingb ecause Co-doped SnO 2 was also reported to show av olume buffering effect, which might additionally increase its cyclability. [4a] On the electrode level,c arbon composite formation in the form of SnO 2 /(doped)graphene,S nO 2 /CNT,S nO 2 /amorphous carbon,a nd/or their combinationw as discussed as av ery efficient strategy to improvet he anode performance, in terms of storagec apacity and cyclability.G raphene-type carbon( undoped rGO [44b] or doped with N, [8c, 73] S, [74] or P), [103] with ah igh surfacea rea and high conductivity,i so ften used as as upport for the homogeneous attachment of nanosized SnO 2 -based active materials. To gether with at hin layer of amorphous carbon obtained through the pyrolysis of organic molecules in the precursor mixture, this results in ah ighly conductive, flexible, and porousm atrix. [23b,c, 44b] Theb est performing composite anodesw ith transition-metal-doped nanostructured SnO 2 showedaremarkabler eversible capacity of over 1200 mAh g À1 (after 100 cycles at 0.1 Ag À1 ), [94a] which greatly outperformed that of standard graphite anodes (e.g., % 226 mAh g À1 cycled at 0.5 Cf or 100 cycles with al oading of 10.1 mg cm À2 ) [110] in classicalL IBs by more than af actor of five. However,r egarding the future commercialization of SnO 2 -based anodes, two objectives need to be addressed. First, high-capacity and rate-capable anodes, with mass loadings in the range of 10 mg cm À2 , [111] need to be realized. Second, and most important,f or practical applicationsi st he combinationw ithasuitable high-rate-capa-ble, high-voltage cathode materialt oo btain full cells with equal or increased energy density to that of classical LIBs employing only carbonaceous anodes.
Future work could include the combination of SnO 2 -based anodesw ith high-voltage cathodes, exceeding the stability window of conventionalc arbonate electrolytes (EC, ethyl methyl carbonate, diethyl carbonate, etc.), whichwould require the use of respective additives or ionic-liquid-based electrolytes. [112] From the perspective of increased operational safety,w hich is already increased at the anode side by the replacement of graphitic carbon with SnO 2 ,asolid electrolyte that allowsf or a high-voltage window (e.g.,N aSICON-type or LiGe 2 (PO 4 ) 3type) [112] would be beneficial. The high rate capability and increasedg ravimetric capacity,r elative to that of graphite electrodes, paired with increased operational safety renders SnO 2based anodesinteresting for applicationsi nf uture energy-storage devices in the industrial and automotive sector.
NIBs with SnO 2 -based anodesh ave gained considerable attention in recent years, with the first published examples of full cells. Knowledge transfer from the design of LIBs resulted in the fabrication of full cells with reversible capacities of up to about 108 mAh g À1 after 100 cycles at 0.1 C. [100] It can be expectedt hat research into SnO 2 -baseda nodesf or NIBs will intensify due to the general attractiveness of NIBs, such as low cost, high abundance of sodium, low toxicity, and increased safety due to al ack of dendrite formation.
Although research into KIBs with SnO 2 -based anodes is very new,r apid progress has been made due to knowledge transfer (synthesis of active materials, anode architecture, andm ethodology)f rom LIBs and NIBs with SnO 2 -based anodes. However, the processes taking place during reversible potassiation/depotassiation of tin and occurring intermediate phases [2] still have to be clarified, although the first publications have identified possible KÀSn alloys. [106,109] Fabricated KIB half-cells have shownacapacityo fu pt o3 51 mAh g À1 for ap ure, binder-free SnO 2 nanosheet anode, [108] and resultsf or the first full cells are expected in the near future.