www.advmat.de www.MaterialsViews.com
REVIEW
Three-Dimensional Self-Supported Metal Oxides for Advanced Energy Storage Brian L. Ellis, Philippe Knauth, and Thierry Djenizian*
batteries have received considerable attention. Besides their application in electric and hybrid electric vehicles, Li-ion batteries have become the energy storage system for portable electronic devices which have revolutionized society during the last decade. Smartphones and tablets are the most spectacular examples; they are now fully part of our lives. The advent of modern microelectronic devices such as backup power for computer memories, microelectromechanical systems (MEMS), medical implants, hearing aids, “smart” cards, radio-frequency identification (RFID) tags and remote sensors has necessitated the development of highperformance power sources at the microscale. These autonomous devices operate independently and thus on-chip electricity is indispensable and rechargeable micropower sources are specially designed to be integrated into microelectronic circuit boards in order to simplify fabrication. The market for these microelectronic devices is rapidly expanding worldwide and various technological challenges must be overcome in order to improve the performance and further reduce the size of these products. The energy needs of these devices are very low: depending on the application, the power needed can range from tens of nanowatts to tens of milliwatts.[2] With the high power density and energy density available in large-scale supercapacitors and Li-ion batteries respectively, microscale energy storage devices comprised of the same materials have been extensively studied to achieve the voltage and energy density requirements. Oxide nanomaterials have been extensively researched as high-capacity electrode materials in both battery and supercapacitor applications. Electrodes fabricated by self-supported techniques offer several advantages over slurry-cast electrodes in terms of surface area and improved electronic transport. As a result, the field of 3D self-supported oxide electrode materials for energy storage has become a burgeoning research field. This review will outline the role of 3D self-supported nanostructured oxides for energy storage in Li-ion and Na-ion batteries and electrochemical supercapacitors. Several synthetic strategies used to assemble these self-supported oxide materials will be described including solution deposition and electrodeposition, template-free electrochemical anodization, hard and soft template-directed synthesis and lithographic techniques. The morphologies and electrochemical performance of the oxides prepared by the various techniques will be compared. The performance of several self-supported 3D oxide nanostructured electrodes for energy storage applications shows promise
The miniaturization of power sources aimed at integration into micro- and nano-electronic devices is a big challenge. To ensure the future development of fully autonomous on-board systems, electrodes based on self-supported 3D nanostructured metal oxides have become increasingly important, and their impact is particularly significant when considering the miniaturization of energy storage systems. This review describes recent advances in the development of self-supported 3D nanostructured metal oxides as electrodes for innovative power sources, particularly Li-ion batteries and electrochemical supercapacitors. Current strategies for the design and morphology control of self-supported electrodes fabricated using template, lithography, anodization and self-organized solution techniques are outlined along with different efforts to improve the storage capacity, rate capability, and cyclability.
1. Introduction Nowadays, industrial and transportation activities consume a large amount of fossil fuels. The increasing energy demand accelerates depletion of stocks towards a worrying supply shortage predicted for the next century. Furthermore, emission of greenhouse gases causes a serious issue of global warming that will impact the whole society. To anticipate these concerns, sustainable energy storage using non-depleted natural resources must be seriously developed. A comparison of the characteristics of common energy storage devices is shown in Figure 1; fuel cells and batteries are considered to be high energy density devices, while supercapacitors deliver high power density with a short response time.[1] Due to their superior electrochemical performance in association with safety characteristics and ecological features, rechargeable Li-ion Dr. B. L. Ellis, Prof. T. Djenizian Aix-Marseille University CNRS, LP3 Laboratory UMR 7341, 13288, Marseille, France E-mail: [email protected] Prof. P. Knauth, Prof. T. Djenizian Aix-Marseille University CNRS, MADIREL Laboratory UMR 7246, 13397, Marseille, France Prof. P. Knauth, Prof. T. Djenizian FR CNRS 3104, ALISTORE-ERI, France Prof. P. Knauth, Prof. T. Djenizian FR CNRS 3459 Réseau sur le Stockage Electrochimique de l’Energie (RS2E), Paris, France
DOI: 10.1002/adma.201306126
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
www.advmat.de
REVIEW
www.MaterialsViews.com
Brian Ellis received his Ph.D. in chemistry from the University of Waterloo (Canada) in 2013 after his research on polyanion positive electrode materials for lithium-ion and sodium-ion batteries in the laboratory of Prof. Linda Nazar. He is currently a postdoctoral fellow in the laboratory of Prof. Thierry Djenizian at l’université d’Aix-Marseille in France, where his research centers on the synthesis of composite oxide nanomaterials. Figure 1. A comparison of specific power versus specific energy for various energy storage devices. Adapted with permission.[1] Copyright 2004, American Chemical Society.
for the fabrication of high energy density storage devices on the microscale.
1.1. Lithium-Ion Thin Film and Microbatteries For over four decades, microbatteries have evolved enormously with varying compositions, form factors and performance characteristics. The most common microbattery design is the all solid-state thin film battery; a schematic representation is shown in Figure 2a. Conventionally, these microbatteries are composed of thin film materials as anode, cathode and electrolyte. They are fabricated by sequential layer-by-layer deposition of the cell components using physical vapor deposition (PVD) techniques such as direct current or radio frequency magnetron sputtering or thermal evaporation, and have quite small dimensions (∼1000 mm3) with the total thickness of the stack of films typically in the range of 10–15 μm. Electrode and electrolyte materials studied in thin film batteries have generally followed those developed for larger-scale batteries. The first practical thin film lithium battery, reported 20 years ago by Kanehori et al., was comprised of a TiS2 positive electrode, a glassy lithium phosphosilicate solid electrolyte layer and a metallic lithium negative electrode.[3] In early studies, metallic Li was frequently used as the negative electrode, although this leads to the obvious challenge of protection of the lithium metal and lithium compounds from reactions with air. As research on positive and negative electrode materials progressed for large batteries, many new materials were used in thin film batteries, including positive electrode materials such as V2O5,[4,5] layered oxides such as LiCoO2,[6–11] spinel oxides based on LiMn2O4,[9–13] and olivine phosphates such as LiFePO4[14–16] plus negative electrode materials such as carbons,[17–19] Si,[20–22] and intercalation oxides such as TiO2[23,24] and Li4Ti5O12.[25] Solid electrolyte research has also produced new materials which exhibit good Li-ion conductivity at room temperature and electrochemical stability at high potential such as polymers,[26–29] ion-conductive glasses of the Li2O-B2O5P2O5[30,31] and Li2S-P2S5 systems,[8,32,33] plus nitrogen-doped lithium phosphosphate (LiPON),[34–37] lithium phosphosilicate
2
wileyonlinelibrary.com
Philippe Knauth is a Full Professor of Materials Chemistry at Aix-Marseille University. He was invited professor at the National Institute of Materials Science (Japan), at the University of Rome (Italy) and he was a visiting scholar at the Massachusetts Institute of Technology. His research activities are focused on the study of ionic conduction at interfaces and nanostructured materials for energy storage and conversion. Thierry Djenizian is a Full Professor of Materials Chemistry at Aix-Marseille University. He received his Ph.D. degree in 2002 from the Swiss Federal Institute of Technology in Lausanne and the University of ErlangenNuremberg. His research activities are mainly focussed on the electrochemical nanostructuring of materials for applications in the field of energy storage and conversion. He is one of the Conference Chairs of Porous Semiconductors Science and Technology international conferences.
(LiSiPON)[38,39] and lithium borophosphate (LiPONB).[40,41] Compared to conventional lithium and Li-ion batteries, these all-solid-state microbatteries have the intrinsic advantage of being free of flammable organic electrolyte solvents which eases concerns over safety and electrolyte degradation.[42,43] The main disadvantage of solid electrolytes is that the ionic conductivity is usually significantly lower than that of liquid-based electrolytes usually found in conventional cells.
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
www.advmat.de www.MaterialsViews.com
REVIEW
The growth of the electroactive material directly on a conductive substrate (“selfsupported electrodes”) eliminates the need for electrode additives and the added step of slurry casting during electrode fabrication. The synthesis of nanomaterials using this self-supported strategy has given rise to a new and feverishly researched class of electrode materials, 3D self-supported nanostructures. 3D self-supported nanostructures may be defined as non-dense, nanoarchitectured layers consisting of one or more phases deposited onto or grown directly from a conductive substrate, without the incorporation of binders or conductive additives. A schematic representation of an idealized 3D self-supported electrode is shown in Figure 3. In the case of both Li-ion batteries and supercapacitors, 3D self-supported nanostructures hold many inherent advantages for small-scale energy storage over electrodes comprised of nanoparticles of electroactive material dispersed in a matrix of carbon and polymer binder: Figure 2. Schematic representation of a) 2D thin film microbattery and b) 3D microbattery. i) High surface area: the lack of dense packing increases the relative surface area of the electrode material, which is extremely important for both Li-ion batteries and supercapacitors. Li-ion Although some thin-film batteries are commercially availbattery electrodes function on the basis of the reversible uptake able, a shift to research on fully 3D microbatteries is underway. of lithium from the electrolyte, thus increasing the electrode A schematic diagram of a 3D microbattery is depicted in surface area improves the accessibility for lithium ions and Figure 2b. The key advantage of a 3D microbattery is the the electrochemical kinetics. Supercapacitors store charge at higher surface area as compared to a planar film that allows the surface, thus more surface area is a key for improving the for increased contact area with the electrolyte and more surenergy density and storage capacity. face sites available for reversible reactions with lithium ions. ii) Tunable free volume of the electrode: The incomplete This in turn leads to improved battery kinetics and increased packing of the electrode surface is particularly important power density while maintaining a small areal footprint. The for oxide electrodes in Li-ion batteries which experience a key to the fabrication of a fully 3D microbattery is the preparalarge volume expansion upon lithium insertion, in particular, tion of 3D nanoarchitectured electrodes with high capacity at those oxides which are transformed into lithium-metal alloys, fast charge and discharge rates and excellent long-term cycling allowing the expansion of the electrode material into the vacant characteristics. As outlined in this review, many self-supported spaces between nanowires or inside of nanotubes. However, oxides have been found to satisfy these criteria. an excessive porosity reduces the volumetric storage capacity of the electrode and an optimum value of the packing density must be found. 1.2. The Advantages of 3D Self-Supported Nanostructures for iii) Nanostructured active electrode: The nanostructure of Energy Storage electrodes implies a small transport length for Li-ion diffusion either through the lattice (in the case of intercalation materials) Nanomaterials for energy storage applications have been feror through the regions of lithium oxide (in the case of convervently researched for several years, owing to several propersion or alloying reactions). As the nanostructure is directly ties beneficial to the performance of these materials in energy bonded to the surface of the conductive substrate, there is no storage devices including high electrode and electrolyte contact need for binders to help with adhesion to a current collector, area, short electron and ion transport distances and in certain nor is there a need for electronically conductive additives, such cases, electrochemical reactions which are not otherwise posas carbon. This increases the gravimetric energy density in sible with bulk materials.[44] While the synthesis of an approboth batteries and supercapacitors. However, this implies that priate nanomaterial is a key aspect in the overall performance the electronic and ionic conductivity of the 3D electrode is sufof a given electrode, the design and fabrication of the electrode ficiently high. itself should be optimized. In typical large-scale electrodes, iv) Improved electronic and ionic conductivity: The ionic slurries comprised of nanoparticles of electroactive material, and electronic conductivity of nanorods, nanowires or conductive additives such as carbon as well as polymer binders nanotubes may be enhanced by the 1D growth and by the are cast onto metal foils.
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
3
www.advmat.de
REVIEW
www.MaterialsViews.com
While 3D self-supported nanostructures of several classes of electrode materials including phosphides,[51] antimonides,[52] silicon,[53–55] tin[56–58] and metal alloys[59,60] have been previously reported, the high functionality of oxides in the field of energy storage has made self-supported oxide nanostructures a fast-growing field.
2. The Utilization of Oxides in Energy Storage Devices 2.1. Lithium-Ion Batteries Lithium-ion batteries are based on reversible redox reactions between lithium ions and positive and negative electrode materials, which allow for the storage and release of chemical energy. Up to now, the negative Figure 3. Schematic representation of an ideal 3D self-supported nanostructure. electrode material used in most commercial Li-ion batteries has been graphite. Graphite has limited Li-ion storage capability with a low gravimetric low-temperature synthetic methods leading to out-of-equicapacity of 372 mA h g−1, thus different materials such as librium lattice defects.[45] In oxides, an oxygen deficiency at ambient temperature and pressure due to the formation of oxides have been investigated in order to reach higher capaciextrinsic oxide ion vacancies may be represented with Krögerties for lithium storage. While some oxides with layered archiVink nomenclature:[46] tecture or the spinel structure are suitable as positive electrode materials, most oxides react with lithium at a low potential and Oo Vo•• + 2e ′ + 1 / 2O2( g ) are therefore useful as negative electrodes. Many binary oxide (1) materials have high theoretical capacities above 600 mA h g−1. •• As such, oxide materials show great promise as negative elecPositively charged oxide ion vacancies (VO ), generated by the trodes for Li-ion batteries with high capacity and long cycle removal of oxygen from the lattice, are compensated by the injeclife. The possible mechanisms for the reaction of a metal oxide tion of electrons (e′) into the conduction band. This reaction is (MOx) with lithium have been reviewed in detail[61] and are outparticularly easy at surfaces leading typically to a positive surface potential. Oxides with a large relative surface area are thus anticlined briefly below. ipated to present a larger oxygen deficiency and consequently, a larger electronic conductivity, due to excess electrons.[47] 2.1.1. Displacement (Conversion) Reaction Furthermore, during synthesis, nanostructured oxides may be doped with other ions present in the reaction medium, The displacement mechanism involves the transformation of which may either be adsorbed at the surface leading to surface the metal oxide into the fully reduced metal and lithium oxide states and/or space charge regions[48] or incorporated into the by the process shown below: oxide bulk leading to further ionic defects.[49] A large quantity of adsorbed or dissolved dopant ions will therefore enhance the MOx + 2x Li + + 2xe − ↔ x Li 2O + M (2) electronic and/or ionic conductivity of oxides in nanotubular or nanowire form.[50] v) Adherence of the active material to the substrate: With the The reversibility of this reaction is dependent upon the forgrowth of the nanostructured active material directly onto the mation of metallic nanoparticles dispersed intimately within conductive substrate, there is no need for binders or conductive the lithium oxide matrix to maintain electronic conductivity. additives for electrode fabrication and as such, self-supported elecThis reaction was first demonstrated for CoO, MnO, FeO, trodes may be used exactly as prepared. This is in contrast with and NiO[62] and has since been found to occur in many low the fabrication of electrodes for modern large-scale energy storage valent (Mx+, x < 4) 3d metal oxides of Cr, Mn, Fe, Co, Ni, and devices in which performance is optimized by mixing powdered Cu.[61] active materials with conductive additives such as carbon and polymer binders before being coated onto a metal foil. vi) Assortment of possible nanostructures: Typical nano2.1.2. Formation of Lithium-Metal Alloys architectures for these materials are nanotubes or nanowires, although other morphologies such as vertically aligned platelets An alloy is formed by reaction between a metal oxide and may be obtained depending on the synthesis method and conlithium for oxides of metals such as Zn, Ge, Sn, Sb, In and ditions, as described in later sections of this review. Pb[61] and proceeds in two distinct steps as shown below:
4
wileyonlinelibrary.com
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
www.advmat.de www.MaterialsViews.com
(3)
M + y Li + + ye − ↔ Li y M
(4)
Firstly, the metal oxide is reduced to the elemental metal by a displacement reaction, which, depending on the metal oxide, may be irreversible or poorly reversible owing to the poor conductivity of the metal oxide. As a result, the practical capacity of this group of oxides is based on the alloying reaction between the metal and lithium, which is the second step of the mechanism. The alloying and dealloying of lithium with one of the metals listed above is typically accompanied by a large change in volume of the electrode. Repeated expansion and contraction has been shown to cause cracking and eventual disintegration of the electrodes on extended cycling, which results in poor long-term performance.
2.1.3. Intercalation For oxides with a layered or open-framework structure and highly charged metal ions, lithium ions may be reversibly intercalated and deintercalated from the structure. Lithium ions diffuse to vacant lattice sites, accompanied by a change in the oxidation state of the transition metal. The mechanism is shown below: MOx + y Li + + ye − ↔ Li y MOx
(5)
The volume expansion in these reactions is small compared to that for lithium-metal alloying and as a result, the lattice of the intercalated material closely resembles that upon deintercalation. The structural integrity of the host lattice is also conserved in these reactions, which often allows for high capacity retention on repeated cycling and long cycle life. Some negative electrode oxide materials that undergo Li+ intercalation include TiO2, MoO2, and MoO3; positive electrode materials include layered oxides based on LiCoO2 and spinels based on LiMn2O4.[61] Recently, concerns over global lithium supplies have spurred increased interest in the field of Na-ion batteries for low-cost, environmentally benign energy storage. Several suitable positive and negative electrode materials have been reported for this application.[63] TiO2 is an example of an oxide negative electrode material which intercalates sodium ions.[64] Because of the small active footprint induced by most thinfilm and 3D self-supported electrodes, the electrochemical performance is commonly expressed in terms of capacity, energy or power per area, instead of per mass as commonly quoted for large-scale Li-ion batteries. The rate at which a battery is charged or discharged is referred to as the C-rate: a battery charged over a period of n hours is said to charge at a 1/n C rate.
2.2. Electrochemical Supercapacitors Electrochemical supercapacitors take advantage of near-surface charge storage mechanisms to store energy. As a result, there is
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
REVIEW
MOx + 2x Li + + 2xe − → x Li 2O + M
no significant mechanical stress (due to reorganization, expansion or contraction) on the electrode structure, which allows supercapacitors to display better durability compared with batteries, as cycle life can be in the range of 105–106 cycles. Surface charge storage also allows supercapacitors to store and release energy in a short timeframe of a few seconds as opposed to the several minutes to hours required for battery charge and discharge. In addition, supercapacitors can achieve 1–2 orders of magnitude greater power density than batteries, albeit at reduced energy density. The specific capacitance (in F g−1) of an electrochemical supercapacitor can be calculated with the following equation: C=
ε rε 0 A d
(6)
where εr is the relative electrolyte dielectric constant, ε0 is vacuum permittivity, A is the specific area of the electrode (often 500–2000 m2 g−1), and d is a distance of the order of the Helmholtz double layer width (charge separation distance on the nanoscale). There are two categories of electrochemical supercapacitors. The first type includes electrostatic double-layer supercapacitors (EDLS) in which a potential-dependent accumulation of charges occurs at the electrode/electrolyte interface, without any chemical reaction. As they display high conductivity, porosity and electrochemical stability, the most common EDLS electrode materials studied are carbon nanomaterials[65] of various forms including aerogels,[66–68] carbon nanotubes (CNTs),[69,70] graphene,[71–73] carbide-derived carbons[74,75] and onion-like carbons.[76–78] Metal oxides fall under the second category, Faradaic supercapacitors (FS) or “pseudocapacitors”. The oxide material at the electrode undergoes a reversible redox reaction and traps electrolyte ions near the oxide surface which results in a Faradaic current passing through the cell. Unlike EDLS which only store charge at the surface of the electrode, the electrochemical reaction in FS occurs both at the electrode surface and a few nanometers below the surface which results in significantly larger energy density values and higher capacitances, 10–100 times those found in EDLS. Oxides such as MnO2, NiO, Co3O4 and Fe3O4 have been extensively studied, as discussed throughout this article. The deposition of oxide nanomaterials onto nanostructured current collectors to produce 3D self-supported electrodes is a key design strategy to improve the energy and power density in EDLS.[79]
2.3. Characterization of Structure/Property Relationships in Nanoarchitectured Electrodes The physical, chemical, and electrochemical properties of nanostructured metal oxide electrodes are intimately correlated to their architecture. Besides the use of conventional characterization techniques employed in materials science, the complexity of the materials requires the combination of different analyses to optimize the fabrication strategies and clearly understand the relationships existing between the structure and properties. Compared to their bulk counterparts, 3D self-supported
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
5
www.advmat.de
REVIEW
www.MaterialsViews.com
nanostructures reveal a high surface-to-volume ratio, the existence of different interfaces, and the presence of voids. Thus, microscopy techniques including field emission scanning electron microscopy (FESEM) and high resolution transmission electron microscopy (HRTEM) are systematically employed to examine nanostructure morphology and determine the nature of the interfaces. In general, physisorption analyses like the Brunauer–Emmett–Teller (BET) technique give valuable information about surface area, pore volume and pore size distribution. The electronic and ionic conductivity can be assessed by non-destructive methods such as electrochemical impedance spectroscopy (EIS). Combined with equivalent circuit modelling, EIS data are used to analyze the dynamics of basic electrochemical processes although interpretation is often complicated due to the coexistence of different interfaces. It is also crucial to characterize the chemical composition of the surface using spectroscopic techniques such as Raman, FTIR and solid-state NMR. X-ray photoelectron spectroscopy (XPS) provides important information on the nature of chemical bonds close to the particle surface (
REVIEW
Three-Dimensional Self-Supported Metal Oxides for Advanced Energy Storage Brian L. Ellis, Philippe Knauth, and Thierry Djenizian*
batteries have received considerable attention. Besides their application in electric and hybrid electric vehicles, Li-ion batteries have become the energy storage system for portable electronic devices which have revolutionized society during the last decade. Smartphones and tablets are the most spectacular examples; they are now fully part of our lives. The advent of modern microelectronic devices such as backup power for computer memories, microelectromechanical systems (MEMS), medical implants, hearing aids, “smart” cards, radio-frequency identification (RFID) tags and remote sensors has necessitated the development of highperformance power sources at the microscale. These autonomous devices operate independently and thus on-chip electricity is indispensable and rechargeable micropower sources are specially designed to be integrated into microelectronic circuit boards in order to simplify fabrication. The market for these microelectronic devices is rapidly expanding worldwide and various technological challenges must be overcome in order to improve the performance and further reduce the size of these products. The energy needs of these devices are very low: depending on the application, the power needed can range from tens of nanowatts to tens of milliwatts.[2] With the high power density and energy density available in large-scale supercapacitors and Li-ion batteries respectively, microscale energy storage devices comprised of the same materials have been extensively studied to achieve the voltage and energy density requirements. Oxide nanomaterials have been extensively researched as high-capacity electrode materials in both battery and supercapacitor applications. Electrodes fabricated by self-supported techniques offer several advantages over slurry-cast electrodes in terms of surface area and improved electronic transport. As a result, the field of 3D self-supported oxide electrode materials for energy storage has become a burgeoning research field. This review will outline the role of 3D self-supported nanostructured oxides for energy storage in Li-ion and Na-ion batteries and electrochemical supercapacitors. Several synthetic strategies used to assemble these self-supported oxide materials will be described including solution deposition and electrodeposition, template-free electrochemical anodization, hard and soft template-directed synthesis and lithographic techniques. The morphologies and electrochemical performance of the oxides prepared by the various techniques will be compared. The performance of several self-supported 3D oxide nanostructured electrodes for energy storage applications shows promise
The miniaturization of power sources aimed at integration into micro- and nano-electronic devices is a big challenge. To ensure the future development of fully autonomous on-board systems, electrodes based on self-supported 3D nanostructured metal oxides have become increasingly important, and their impact is particularly significant when considering the miniaturization of energy storage systems. This review describes recent advances in the development of self-supported 3D nanostructured metal oxides as electrodes for innovative power sources, particularly Li-ion batteries and electrochemical supercapacitors. Current strategies for the design and morphology control of self-supported electrodes fabricated using template, lithography, anodization and self-organized solution techniques are outlined along with different efforts to improve the storage capacity, rate capability, and cyclability.
1. Introduction Nowadays, industrial and transportation activities consume a large amount of fossil fuels. The increasing energy demand accelerates depletion of stocks towards a worrying supply shortage predicted for the next century. Furthermore, emission of greenhouse gases causes a serious issue of global warming that will impact the whole society. To anticipate these concerns, sustainable energy storage using non-depleted natural resources must be seriously developed. A comparison of the characteristics of common energy storage devices is shown in Figure 1; fuel cells and batteries are considered to be high energy density devices, while supercapacitors deliver high power density with a short response time.[1] Due to their superior electrochemical performance in association with safety characteristics and ecological features, rechargeable Li-ion Dr. B. L. Ellis, Prof. T. Djenizian Aix-Marseille University CNRS, LP3 Laboratory UMR 7341, 13288, Marseille, France E-mail: [email protected] Prof. P. Knauth, Prof. T. Djenizian Aix-Marseille University CNRS, MADIREL Laboratory UMR 7246, 13397, Marseille, France Prof. P. Knauth, Prof. T. Djenizian FR CNRS 3104, ALISTORE-ERI, France Prof. P. Knauth, Prof. T. Djenizian FR CNRS 3459 Réseau sur le Stockage Electrochimique de l’Energie (RS2E), Paris, France
DOI: 10.1002/adma.201306126
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1
www.advmat.de
REVIEW
www.MaterialsViews.com
Brian Ellis received his Ph.D. in chemistry from the University of Waterloo (Canada) in 2013 after his research on polyanion positive electrode materials for lithium-ion and sodium-ion batteries in the laboratory of Prof. Linda Nazar. He is currently a postdoctoral fellow in the laboratory of Prof. Thierry Djenizian at l’université d’Aix-Marseille in France, where his research centers on the synthesis of composite oxide nanomaterials. Figure 1. A comparison of specific power versus specific energy for various energy storage devices. Adapted with permission.[1] Copyright 2004, American Chemical Society.
for the fabrication of high energy density storage devices on the microscale.
1.1. Lithium-Ion Thin Film and Microbatteries For over four decades, microbatteries have evolved enormously with varying compositions, form factors and performance characteristics. The most common microbattery design is the all solid-state thin film battery; a schematic representation is shown in Figure 2a. Conventionally, these microbatteries are composed of thin film materials as anode, cathode and electrolyte. They are fabricated by sequential layer-by-layer deposition of the cell components using physical vapor deposition (PVD) techniques such as direct current or radio frequency magnetron sputtering or thermal evaporation, and have quite small dimensions (∼1000 mm3) with the total thickness of the stack of films typically in the range of 10–15 μm. Electrode and electrolyte materials studied in thin film batteries have generally followed those developed for larger-scale batteries. The first practical thin film lithium battery, reported 20 years ago by Kanehori et al., was comprised of a TiS2 positive electrode, a glassy lithium phosphosilicate solid electrolyte layer and a metallic lithium negative electrode.[3] In early studies, metallic Li was frequently used as the negative electrode, although this leads to the obvious challenge of protection of the lithium metal and lithium compounds from reactions with air. As research on positive and negative electrode materials progressed for large batteries, many new materials were used in thin film batteries, including positive electrode materials such as V2O5,[4,5] layered oxides such as LiCoO2,[6–11] spinel oxides based on LiMn2O4,[9–13] and olivine phosphates such as LiFePO4[14–16] plus negative electrode materials such as carbons,[17–19] Si,[20–22] and intercalation oxides such as TiO2[23,24] and Li4Ti5O12.[25] Solid electrolyte research has also produced new materials which exhibit good Li-ion conductivity at room temperature and electrochemical stability at high potential such as polymers,[26–29] ion-conductive glasses of the Li2O-B2O5P2O5[30,31] and Li2S-P2S5 systems,[8,32,33] plus nitrogen-doped lithium phosphosphate (LiPON),[34–37] lithium phosphosilicate
2
wileyonlinelibrary.com
Philippe Knauth is a Full Professor of Materials Chemistry at Aix-Marseille University. He was invited professor at the National Institute of Materials Science (Japan), at the University of Rome (Italy) and he was a visiting scholar at the Massachusetts Institute of Technology. His research activities are focused on the study of ionic conduction at interfaces and nanostructured materials for energy storage and conversion. Thierry Djenizian is a Full Professor of Materials Chemistry at Aix-Marseille University. He received his Ph.D. degree in 2002 from the Swiss Federal Institute of Technology in Lausanne and the University of ErlangenNuremberg. His research activities are mainly focussed on the electrochemical nanostructuring of materials for applications in the field of energy storage and conversion. He is one of the Conference Chairs of Porous Semiconductors Science and Technology international conferences.
(LiSiPON)[38,39] and lithium borophosphate (LiPONB).[40,41] Compared to conventional lithium and Li-ion batteries, these all-solid-state microbatteries have the intrinsic advantage of being free of flammable organic electrolyte solvents which eases concerns over safety and electrolyte degradation.[42,43] The main disadvantage of solid electrolytes is that the ionic conductivity is usually significantly lower than that of liquid-based electrolytes usually found in conventional cells.
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
www.advmat.de www.MaterialsViews.com
REVIEW
The growth of the electroactive material directly on a conductive substrate (“selfsupported electrodes”) eliminates the need for electrode additives and the added step of slurry casting during electrode fabrication. The synthesis of nanomaterials using this self-supported strategy has given rise to a new and feverishly researched class of electrode materials, 3D self-supported nanostructures. 3D self-supported nanostructures may be defined as non-dense, nanoarchitectured layers consisting of one or more phases deposited onto or grown directly from a conductive substrate, without the incorporation of binders or conductive additives. A schematic representation of an idealized 3D self-supported electrode is shown in Figure 3. In the case of both Li-ion batteries and supercapacitors, 3D self-supported nanostructures hold many inherent advantages for small-scale energy storage over electrodes comprised of nanoparticles of electroactive material dispersed in a matrix of carbon and polymer binder: Figure 2. Schematic representation of a) 2D thin film microbattery and b) 3D microbattery. i) High surface area: the lack of dense packing increases the relative surface area of the electrode material, which is extremely important for both Li-ion batteries and supercapacitors. Li-ion Although some thin-film batteries are commercially availbattery electrodes function on the basis of the reversible uptake able, a shift to research on fully 3D microbatteries is underway. of lithium from the electrolyte, thus increasing the electrode A schematic diagram of a 3D microbattery is depicted in surface area improves the accessibility for lithium ions and Figure 2b. The key advantage of a 3D microbattery is the the electrochemical kinetics. Supercapacitors store charge at higher surface area as compared to a planar film that allows the surface, thus more surface area is a key for improving the for increased contact area with the electrolyte and more surenergy density and storage capacity. face sites available for reversible reactions with lithium ions. ii) Tunable free volume of the electrode: The incomplete This in turn leads to improved battery kinetics and increased packing of the electrode surface is particularly important power density while maintaining a small areal footprint. The for oxide electrodes in Li-ion batteries which experience a key to the fabrication of a fully 3D microbattery is the preparalarge volume expansion upon lithium insertion, in particular, tion of 3D nanoarchitectured electrodes with high capacity at those oxides which are transformed into lithium-metal alloys, fast charge and discharge rates and excellent long-term cycling allowing the expansion of the electrode material into the vacant characteristics. As outlined in this review, many self-supported spaces between nanowires or inside of nanotubes. However, oxides have been found to satisfy these criteria. an excessive porosity reduces the volumetric storage capacity of the electrode and an optimum value of the packing density must be found. 1.2. The Advantages of 3D Self-Supported Nanostructures for iii) Nanostructured active electrode: The nanostructure of Energy Storage electrodes implies a small transport length for Li-ion diffusion either through the lattice (in the case of intercalation materials) Nanomaterials for energy storage applications have been feror through the regions of lithium oxide (in the case of convervently researched for several years, owing to several propersion or alloying reactions). As the nanostructure is directly ties beneficial to the performance of these materials in energy bonded to the surface of the conductive substrate, there is no storage devices including high electrode and electrolyte contact need for binders to help with adhesion to a current collector, area, short electron and ion transport distances and in certain nor is there a need for electronically conductive additives, such cases, electrochemical reactions which are not otherwise posas carbon. This increases the gravimetric energy density in sible with bulk materials.[44] While the synthesis of an approboth batteries and supercapacitors. However, this implies that priate nanomaterial is a key aspect in the overall performance the electronic and ionic conductivity of the 3D electrode is sufof a given electrode, the design and fabrication of the electrode ficiently high. itself should be optimized. In typical large-scale electrodes, iv) Improved electronic and ionic conductivity: The ionic slurries comprised of nanoparticles of electroactive material, and electronic conductivity of nanorods, nanowires or conductive additives such as carbon as well as polymer binders nanotubes may be enhanced by the 1D growth and by the are cast onto metal foils.
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
3
www.advmat.de
REVIEW
www.MaterialsViews.com
While 3D self-supported nanostructures of several classes of electrode materials including phosphides,[51] antimonides,[52] silicon,[53–55] tin[56–58] and metal alloys[59,60] have been previously reported, the high functionality of oxides in the field of energy storage has made self-supported oxide nanostructures a fast-growing field.
2. The Utilization of Oxides in Energy Storage Devices 2.1. Lithium-Ion Batteries Lithium-ion batteries are based on reversible redox reactions between lithium ions and positive and negative electrode materials, which allow for the storage and release of chemical energy. Up to now, the negative Figure 3. Schematic representation of an ideal 3D self-supported nanostructure. electrode material used in most commercial Li-ion batteries has been graphite. Graphite has limited Li-ion storage capability with a low gravimetric low-temperature synthetic methods leading to out-of-equicapacity of 372 mA h g−1, thus different materials such as librium lattice defects.[45] In oxides, an oxygen deficiency at ambient temperature and pressure due to the formation of oxides have been investigated in order to reach higher capaciextrinsic oxide ion vacancies may be represented with Krögerties for lithium storage. While some oxides with layered archiVink nomenclature:[46] tecture or the spinel structure are suitable as positive electrode materials, most oxides react with lithium at a low potential and Oo Vo•• + 2e ′ + 1 / 2O2( g ) are therefore useful as negative electrodes. Many binary oxide (1) materials have high theoretical capacities above 600 mA h g−1. •• As such, oxide materials show great promise as negative elecPositively charged oxide ion vacancies (VO ), generated by the trodes for Li-ion batteries with high capacity and long cycle removal of oxygen from the lattice, are compensated by the injeclife. The possible mechanisms for the reaction of a metal oxide tion of electrons (e′) into the conduction band. This reaction is (MOx) with lithium have been reviewed in detail[61] and are outparticularly easy at surfaces leading typically to a positive surface potential. Oxides with a large relative surface area are thus anticlined briefly below. ipated to present a larger oxygen deficiency and consequently, a larger electronic conductivity, due to excess electrons.[47] 2.1.1. Displacement (Conversion) Reaction Furthermore, during synthesis, nanostructured oxides may be doped with other ions present in the reaction medium, The displacement mechanism involves the transformation of which may either be adsorbed at the surface leading to surface the metal oxide into the fully reduced metal and lithium oxide states and/or space charge regions[48] or incorporated into the by the process shown below: oxide bulk leading to further ionic defects.[49] A large quantity of adsorbed or dissolved dopant ions will therefore enhance the MOx + 2x Li + + 2xe − ↔ x Li 2O + M (2) electronic and/or ionic conductivity of oxides in nanotubular or nanowire form.[50] v) Adherence of the active material to the substrate: With the The reversibility of this reaction is dependent upon the forgrowth of the nanostructured active material directly onto the mation of metallic nanoparticles dispersed intimately within conductive substrate, there is no need for binders or conductive the lithium oxide matrix to maintain electronic conductivity. additives for electrode fabrication and as such, self-supported elecThis reaction was first demonstrated for CoO, MnO, FeO, trodes may be used exactly as prepared. This is in contrast with and NiO[62] and has since been found to occur in many low the fabrication of electrodes for modern large-scale energy storage valent (Mx+, x < 4) 3d metal oxides of Cr, Mn, Fe, Co, Ni, and devices in which performance is optimized by mixing powdered Cu.[61] active materials with conductive additives such as carbon and polymer binders before being coated onto a metal foil. vi) Assortment of possible nanostructures: Typical nano2.1.2. Formation of Lithium-Metal Alloys architectures for these materials are nanotubes or nanowires, although other morphologies such as vertically aligned platelets An alloy is formed by reaction between a metal oxide and may be obtained depending on the synthesis method and conlithium for oxides of metals such as Zn, Ge, Sn, Sb, In and ditions, as described in later sections of this review. Pb[61] and proceeds in two distinct steps as shown below:
4
wileyonlinelibrary.com
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
www.advmat.de www.MaterialsViews.com
(3)
M + y Li + + ye − ↔ Li y M
(4)
Firstly, the metal oxide is reduced to the elemental metal by a displacement reaction, which, depending on the metal oxide, may be irreversible or poorly reversible owing to the poor conductivity of the metal oxide. As a result, the practical capacity of this group of oxides is based on the alloying reaction between the metal and lithium, which is the second step of the mechanism. The alloying and dealloying of lithium with one of the metals listed above is typically accompanied by a large change in volume of the electrode. Repeated expansion and contraction has been shown to cause cracking and eventual disintegration of the electrodes on extended cycling, which results in poor long-term performance.
2.1.3. Intercalation For oxides with a layered or open-framework structure and highly charged metal ions, lithium ions may be reversibly intercalated and deintercalated from the structure. Lithium ions diffuse to vacant lattice sites, accompanied by a change in the oxidation state of the transition metal. The mechanism is shown below: MOx + y Li + + ye − ↔ Li y MOx
(5)
The volume expansion in these reactions is small compared to that for lithium-metal alloying and as a result, the lattice of the intercalated material closely resembles that upon deintercalation. The structural integrity of the host lattice is also conserved in these reactions, which often allows for high capacity retention on repeated cycling and long cycle life. Some negative electrode oxide materials that undergo Li+ intercalation include TiO2, MoO2, and MoO3; positive electrode materials include layered oxides based on LiCoO2 and spinels based on LiMn2O4.[61] Recently, concerns over global lithium supplies have spurred increased interest in the field of Na-ion batteries for low-cost, environmentally benign energy storage. Several suitable positive and negative electrode materials have been reported for this application.[63] TiO2 is an example of an oxide negative electrode material which intercalates sodium ions.[64] Because of the small active footprint induced by most thinfilm and 3D self-supported electrodes, the electrochemical performance is commonly expressed in terms of capacity, energy or power per area, instead of per mass as commonly quoted for large-scale Li-ion batteries. The rate at which a battery is charged or discharged is referred to as the C-rate: a battery charged over a period of n hours is said to charge at a 1/n C rate.
2.2. Electrochemical Supercapacitors Electrochemical supercapacitors take advantage of near-surface charge storage mechanisms to store energy. As a result, there is
Adv. Mater. 2014, DOI: 10.1002/adma.201306126
REVIEW
MOx + 2x Li + + 2xe − → x Li 2O + M
no significant mechanical stress (due to reorganization, expansion or contraction) on the electrode structure, which allows supercapacitors to display better durability compared with batteries, as cycle life can be in the range of 105–106 cycles. Surface charge storage also allows supercapacitors to store and release energy in a short timeframe of a few seconds as opposed to the several minutes to hours required for battery charge and discharge. In addition, supercapacitors can achieve 1–2 orders of magnitude greater power density than batteries, albeit at reduced energy density. The specific capacitance (in F g−1) of an electrochemical supercapacitor can be calculated with the following equation: C=
ε rε 0 A d
(6)
where εr is the relative electrolyte dielectric constant, ε0 is vacuum permittivity, A is the specific area of the electrode (often 500–2000 m2 g−1), and d is a distance of the order of the Helmholtz double layer width (charge separation distance on the nanoscale). There are two categories of electrochemical supercapacitors. The first type includes electrostatic double-layer supercapacitors (EDLS) in which a potential-dependent accumulation of charges occurs at the electrode/electrolyte interface, without any chemical reaction. As they display high conductivity, porosity and electrochemical stability, the most common EDLS electrode materials studied are carbon nanomaterials[65] of various forms including aerogels,[66–68] carbon nanotubes (CNTs),[69,70] graphene,[71–73] carbide-derived carbons[74,75] and onion-like carbons.[76–78] Metal oxides fall under the second category, Faradaic supercapacitors (FS) or “pseudocapacitors”. The oxide material at the electrode undergoes a reversible redox reaction and traps electrolyte ions near the oxide surface which results in a Faradaic current passing through the cell. Unlike EDLS which only store charge at the surface of the electrode, the electrochemical reaction in FS occurs both at the electrode surface and a few nanometers below the surface which results in significantly larger energy density values and higher capacitances, 10–100 times those found in EDLS. Oxides such as MnO2, NiO, Co3O4 and Fe3O4 have been extensively studied, as discussed throughout this article. The deposition of oxide nanomaterials onto nanostructured current collectors to produce 3D self-supported electrodes is a key design strategy to improve the energy and power density in EDLS.[79]
2.3. Characterization of Structure/Property Relationships in Nanoarchitectured Electrodes The physical, chemical, and electrochemical properties of nanostructured metal oxide electrodes are intimately correlated to their architecture. Besides the use of conventional characterization techniques employed in materials science, the complexity of the materials requires the combination of different analyses to optimize the fabrication strategies and clearly understand the relationships existing between the structure and properties. Compared to their bulk counterparts, 3D self-supported
© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
5
www.advmat.de
REVIEW
www.MaterialsViews.com
nanostructures reveal a high surface-to-volume ratio, the existence of different interfaces, and the presence of voids. Thus, microscopy techniques including field emission scanning electron microscopy (FESEM) and high resolution transmission electron microscopy (HRTEM) are systematically employed to examine nanostructure morphology and determine the nature of the interfaces. In general, physisorption analyses like the Brunauer–Emmett–Teller (BET) technique give valuable information about surface area, pore volume and pore size distribution. The electronic and ionic conductivity can be assessed by non-destructive methods such as electrochemical impedance spectroscopy (EIS). Combined with equivalent circuit modelling, EIS data are used to analyze the dynamics of basic electrochemical processes although interpretation is often complicated due to the coexistence of different interfaces. It is also crucial to characterize the chemical composition of the surface using spectroscopic techniques such as Raman, FTIR and solid-state NMR. X-ray photoelectron spectroscopy (XPS) provides important information on the nature of chemical bonds close to the particle surface (
