Hybrid material design for energy applications: impact of graphene and carbon nanotubes

Introduction

Design and synthesis of new materials are critical processes in advancing the technology of energy storage and conversion [1, 2]. With problems of energy consumption and environmental deterioration facing our society, it is important to develop new materials with high performance, high stability, low cost, and low environmental impact to be integrated into electrochemical devices such as batteries, supercapacitors, and fuel cells, which can store energy produced from renewable sources for human activities [3–6]. Recently, hybrid materials based on graphitic nanocarbons (graphene sheets or carbon nanotubes) have emerged as a new class of electrode materials exhibiting more desirable electrochemical performance than conventional electrode materials that are often physical mixtures of electrically insulating active materials and conductive carbon [7, 8].

Electrochemical energy devices and materials

Electrochemical energy storage and conversion devices such as batteries, supercapacitors, and fuel cells can provide conversion between chemical energy and electrical energy with high efficiency and low environmental impact [4, 6, 9, 10]. They are important in harvesting and transforming energy for human needs from renewable sources, including sunlight, wind, water, geothermal heat, and biomass, which may lower our dependence on nonrenewable fossil fuels and reduce carbon dioxide emission. Depending on their operating mechanisms, the devices are possessed with different energy and power output characteristics suitable for certain types of practical applications.

A lithium ion battery is a category of rechargeable battery based on lithium ion moving between two electrodes (the cathode and the anode) [3, 11]. In charging, the lithium ions and electrons travel from the cathode to the anode, which reverses in discharging. High energy density and efficiency of lithium ion batteries render them the dominant power supply for the present portable electronic devices such as laptops, cell phones, digital cameras, etc. However, to propel electric vehicles, energy density, cycle life, cost, and safety of lithium ion batteries need to be further improved [5, 12].

Beyond lithium ion batteries, there are rechargeable lithium-air and lithium-sulfur batteries that can deliver much higher specific energy [12]. They are based on reversible reactions between lithium and oxygen/sulfur. However, these two battery technologies are still in their infancy [13, 14].

Supercapacitors are electrochemical capacitors with much higher capacitance than traditional capacitors [3, 4]. Unlike traditional capacitors where two metal plates are separated by a dielectric layer and charge is stored by polarization of the dielectric between the two plates, a supercapacitor device is made of two electrodes of high-surface-area or redox-active materials separated by an electrolyte layer. Supercapacitors can be divided into two categories based on charge storage mechanism. Electric double-layer capacitors storage charge from ion adsorption and desorption, while pseudocapacitors storage charge by Faradic reactions. Depending on whether the two electrodes are identical, supercapacitors can also be divided into symmetrical and asymmetrical supercapacitors. Supercapacitors can be fully charged and discharged in a very short time, ideal for power-demanding applications in both civil and military areas. As a drawback, their energy density is generally much lower than that of batteries.

Within the energy and power gap between lithium ion batteries and supercapacitors, it is desirable to develop fast rechargeable batteries with aqueous electrolytes. Such batteries can deliver a power rate that is orders of magnitude higher than traditional batteries while retaining similar specific energy.

A fuel cell is a type of non-rechargeable electrochemical device that generates electricity through the reaction of a fuel with an oxidizer [3]. They afford both high energy density and high power density, but the slow kinetics of the electrode reactions limits the efficiency.

Generally speaking, the amount of usable energy stored by a cell is determined by the cell voltage and the number of electrons transferred in the cell reaction. The theoretical cell voltage is determined by the standard electric potentials of its two electrodes, which are thermodynamic properties of the electrode materials. However, the real cell voltage is undermined by a kinetic term called “overpotential”, which is spent on overcoming activation energy barriers of electrode reactions, driving ion diffusion from electrolyte to electrode surface and within the electrode, and overcoming electrical resistance in the device. Overpotential is often larger in magnitude at higher current density. The existence of overpotential makes discharging voltage lower than charging voltage, giving voltage efficiency <100%. Another important concept is Coulombic efficiency, which reveals how many electrons are directed to the desirable reactions rather than side reactions. It is an indication of reaction reversibility and cycling stability. Combining voltage efficiency and Coulombic efficiency gives energy efficiency.

In pursuit of electrochemical energy storage and conversion devices with improved kinetics and efficiency, our research is focused on design and synthesis of new electrode materials and electrocatalysts with optimized structures and superior properties as well as understanding the chemical correlations between material structures and electrochemical performance. We utilize graphene sheets and carbon nanotubes as conducting substrates to anchor active electrode materials and catalysts by building covalent chemical bonding at the interface. The strong electrical and chemical coupling between the anchored materials and the nanocarbons not only enhances electron transfer and transport during electrochemical reactions but also provides an extra degree of freedom to control the morphology and size of the anchored materials, leading to higher electrochemical performance of the hybrid materials than conventional electrode materials.

Inorganic/nanocarbon hybrid materials for pseudocapacitors and ultrafast Ni-Fe batteries

Pseudocapacitors utilize fast and reversible redox reactions for energy storage and release, which brings an advantage of higher capacitance together with a disadvantage of slower kinetics as compared to electric double-layer capacitors [4, 15]. Rate performance of the electrode materials is limited by electron transfer and ion diffusion. Redox reactions only take place on the surface of active material particles under fast charging and discharging conditions. To improve the electrochemical performance of pseudocapacitor materials, we selectively grow nanocrystals of active materials on graphene sheets to enhance the electron transfer by the covalent bonding at the hybrid interface and the 3D conducting graphene network. Strong chemical interactions between the functionalized graphene sheets and active materials also provide an extra degree of freedom to control the morphology and size of the nanocrystals to improve utilization of active materials during electrochemical cycling.

We synthesized a Ni(OH)₂/graphene hybrid material with a two-step solution-phase method [16]. In the first step of the reaction, Ni(OAc)₂ was mixed with chemically derived graphene in a mixed solvent of N,N-dimethylformamide and water at a low temperature of 80°C to allow for sufficient interaction between the Ni ions and the oxygen functional groups on graphene surface. The interaction initiated hydrolysis of Ni(OAc)₂ and nucleation of the product on graphene surface, which eventually led to selective and uniform coating of graphene sheets with the hydrolysis product. In the second step of the reaction, the material collected from the previous step was treated under hydrothermal conditions to crystallize the coated species into Ni(OH)₂ nanocrystals and partially restore the graphitic structure of the sheets.

The Ni(OH)₂ nanocrystals grown on high-quality graphene sheets exhibited well-defined hexagonal nanoplate shapes (Fig. 1a,b) [16]. The nanoplates were a few hundred nanometers in lateral dimensions and about 10 nm in thickness, which is a reasonable depth for ion and electron penetration during fast charging and discharging. The hybrid material showed high specific capacitance and rate capability in 1 M aqueous KOH solution [17]. At a charging and discharging current density of 2.8 A/g, the Ni(OH)₂ nanoplates on graphene showed a high specific capacitance of ~1335 F/g (Fig. 1c,d). The capacitance was still as high as ~953 F/g at a high current density of 45.7 A/g (Fig. 1c,d), indicating fast charging and discharging capability in about 10 s. The hybrid was among supercapacitor electrode materials with the highest performance. The superior performance was attributed to the strong electrical and chemical coupling between the Ni(OH)₂ nanoplates and the graphene sheets as well as the desirable morphology of the Ni(OH)₂ nanoplates. Such advantages of the hybrid structure enabled the evolution of Ni(OH)₂ from a mediocre-rate battery material into a fast-rate supercapacitor material.

Fig. 1

Comparison of structural and electrochemical properties of Ni(OH)₂ nanocrystals grown on high-quality graphene (a–d) and more-oxidized graphene (e–h). (a) SEM image, (b) TEM image, (c) discharging voltage profiles and (d) specific capacitance at various current densities of Ni(OH)₂ nanoplates grown on high-quality graphene sheets. (e) SEM image, (f) TEM image, (g) discharging voltage profiles and (h) specific capacitance at various current densities of Ni(OH)₂ nanoparticles grown on more-oxidized graphene sheets. Adapted with permission from ref. [17]. Copyright © 2010 American Chemical Society.

In the system we found the oxidation degree of the graphene sheets had a strong influence on both the structural and electrochemical properties of the hybrid material (Fig. 1) [16, 17]. When sheets of more oxidized graphene were used instead of the high-quality graphene sheets, the grown nanocrystals of Ni(OH)₂ took a small-particle-like morphology due to the strong binding force of the oxygen functional groups and defects on the sheet surface (Fig. 1e,f). The hybrid material also exhibited much lower specific capacitance (Fig. 1g,h). This confirmed the interfacial coupling to be a key factor for the hybrid materials.

Combining the Ni(OH)₂/graphene hybrid electrode with a graphene (reduced graphene oxide) electrode afforded an asymmetrical supercapacitor [18]. The device showed nearly square-shaped cyclic voltammetry (CV) curves and nearly linear charging and discharging voltage profiles (Fig. 2b), indicating capacitor-like behavior. The energy and power output characteristics of the asymmetrical supercapacitor were drastically higher than those of a symmetrical supercapacitor made with two identical graphene electrodes (Fig. 2a,b). At a power density of 1 kW/kg, the Ni(OH)₂/graphene-graphene asymmetrical supercapacitor delivered an energy density of ~25 Wh/kg, ~3 times higher than that of the graphene-graphene symmetrical supercapacitor at the same power density (Fig. 2d). The improved energy density of the asymmetrical supercapacitor is attributed to the high specific capacitance of the Ni(OH)₂/graphene hybrid material.

Fig. 2

Charging and discharging voltage profiles of (a) graphene-graphene, (b) Ni(OH)₂/graphene-graphene and (c) Ni(OH)₂/graphene-RuO₂/graphene supercapacitors, and (d) comparison of energy and power output for four types of supercapacitors. The current, energy, and power density values are all based on total mass of electrode materials. In the figure, GS and RGO stand for high-quality graphene sheets and reduced graphene oxide (a type of more-oxidized graphene sheets), respectively. Adapted with permission from ref. [18]. Copyright © 2011 Tsinghua University Press and Springer-Verlag.

In the Ni(OH)₂/graphene-graphene asymmetrical supercapacitor, the specific capacitance of the Ni(OH)₂/graphene was about 5 times higher than that of the graphene [18]. The graphene electrode thus became the performance-limiting electrode material in the device. To further increase the energy density of the supercapacitor, we synthesized a RuO₂/graphene hybrid material by growing RuO₂ nanoparticles on the surface of graphene sheets. The specific capacitance of the RuO₂ nanoparticles on graphene was as high as 550 F/g. A new asymmetrical supercapacitor made with the Ni(OH)₂/graphene and RuO₂/graphene hybrid materials exhibited high specific capacitance, rate capability, and cycling stability. It showed a specific capacitance of 159 F/g (normalized to the total mass of electrode materials) at a charging and discharging current density of 1 A/g (Fig. 2c). At a high current density of 10 A/g, the specific capacitance was as still as high as 110 F/g, indicating a fast charging or discharging time of ~30 s (Fig. 2c). At a power density of 1 kW/kg, an unprecedented high energy density of ~40 Wh/kg was achieved (Fig. 2d), outperforming well-known RuO₂-based symmetrical supercapacitors. The capacitance of the device was stable over 5000 cycles.

In a later study, we discovered that the specific capacitance of Ni(OH)₂ could be further increased by using oxidized multiwall carbon nanotubes (MWNTs) instead of graphene sheets in the hybrid materials [19]. The multiwall structure of the carbon nanotube could largely avoid graphene’s dilemma of trade-off between creating oxygen functional groups and retaining electrical conductivity. The outer walls of the carbon nanotubes can be oxidized for adsorption and binding of metal ions, and the inner walls remain unaltered with superior electron transport properties. Consequently, in certain cases of electrochemical energy storage and conversion, carbon nanotubes could make better hybrid materials than graphene sheets.

The success of using Ni(OH)₂/nanocarbon hybrids as high capacitance supercapacitor electrodes led us to think about the role of redox-active materials in supercapacitors as well as the essential differences between supercapacitors and batteries. Being a battery-type electrode material, Ni(OH)₂/nanocarbon needs to be paired with a capacitor-type electrode material such as graphene or RuO₂ to render capacitor-like voltage-current characteristics to the device. However, the device does not have to behave like a capacitor to provide high power performance. In fact, batteries that can be charged and discharged at comparable rates to supercapacitors will be more desirable because of their high-energy-density characteristics. We were therefore inspired to build ultra-fast batteries using inorganic/nanocarbon hybrid materials with fast charging and discharging capabilities [19].

We used Ni(OH)₂/MWNT (Fig. 3a) and FeO_x/graphene (Fig. 3b) hybrids as the cathode and anode materials, respectively [19]. Pairing of the two electrodes in 1 M aqueous KOH solution afforded a battery-type device with high power rate. When charged at 3.7 A/g and discharged at 1.5 A/g (5 min of discharging time), the cell achieved a specific capacity of 126 mAh/g based on the total mass of active materials (Fig. 3c). The capacity was still as high as 104 mAh/g at a high discharging current density of 37 A/g, corresponding to a fast discharging time of about 10 s (Fig. 3c). At a power density of 1 kW/kg, the cell could reach a high energy density of 145 Wh/kg based on the total mass of active materials (Fig. 3d). The cell demonstrated the possibility of using strongly coupled inorganic/nanocarbon hybrid electrode materials to increase the rate performance of traditional Ni-Fe batteries by nearly 1000 times while retaining their energy density.

Fig. 3

TEM images of (a) Ni(OH)₂/MWNT and (b) FeO_x/graphene hybrid materials, and (c) discharging voltage profiles at various current densities and (d) energy and power characteristics of the Ni(OH)₂/MWNT-FeO_x/graphene battery. Adapted with permission from ref. [19]. Copyright © 2012 Nature Publishing Group.

Inorganic/nanocarbon hybrid materials for lithium ion batteries

Limited energy density is a major challenge that the present generation of lithium ion batteries face in the context of transportation electrification [11, 12]. To improve the specific energy of lithium ion batteries, it is important to explore new active electrode materials that can increase specific capacity or voltage of the batteries. Many active materials are electrically insulating. Some of them are extremely insulating, and their specific capacity and rate capability are severely limited by electronic transport in the materials. We utilize inorganic/graphene hybrid structures to tackle the problem. Strong electrical and chemical coupling between the active materials and conducting graphene sheets can substantially enhance electron transfer at the interface and electron transport to current collectors. Such intimate interactions can also be helpful in achieving desirable morphology and size of active material particles as well as their uniform and stable dispersion on the graphene surface.

Mn₃O₄ was our first material investigated [20]. It is a potential anode material with a theoretical specific capacity as high as 936 mAh/g based on a conversion reaction mechanism. Mn₃O₄ is also preferred to Co₃O₄ in consideration of environmental benignity. However, unlike Co₃O₄ for which high capacity could be readily obtained, the low electrical conductivity of Mn₃O₄ (~10^–7 to 10^–8 S/cm) had been keeping its capacity <400 mAh/g until we utilized the graphene hybrid structure to tackle the problem [20]. Our Mn₃O₄/graphene hybrid was synthesized by the same two-step solution-phase method as for the Ni(OH)₂/nanocarbon hybrids. The resulted Mn₃O₄ nanoparticles were in the size of 10–20 nm and well dispersed on the surface of graphene sheets (Fig. 4a). At a charging and discharging current density of 40 mA/g, the hybrid showed a specific capacity of ~900 mAh/g based on the mass of Mn₃O₄ only (Fig. 4b). It was the strong electrical and chemical coupling between the graphene sheets and the Mn₃O₄ nanoparticles grown on top as well as the desirable size and dispersion of the Mn₃O₄ nanoparticles that enabled the high specific capacity approaching theoretical limit. The capacity was still as high as 780 mAh/g when the current density was increased by 10 times to 400 mA/g (Fig. 4b), suggesting good rate performance of the hybrid material.

Fig. 4

(a) TEM image and (b) electrochemical charging and discharging performance of the Mn₃O₄/graphene hybrid as an anode material for lithium ion batteries. Adapted with permission from ref. [20]. Copyright © 2010 American Chemical Society.

LiMnPO₄ was our second material of target. Besides a wide range of advantages such as long cycle life, high thermal and chemical stability, environmental benignity, low cost, and constant discharging voltage, which are shared with LiFePO₄ (one of the present commercial cathode materials), LiMnPO₄ offers a discharging voltage of 4.1 V vs. Li⁺/Li, which is 0.7 V higher than that of LiFePO₄ but is still compatible with the stable voltage window of the prevalent carbonate electrolytes. Therefore, LiMnPO₄ could be a potentially promising cathode material for the next generation of lithium ion battery. However, current LiMnPO₄ materials have so far been showing considerably lower specific capacity than the theoretical value even at moderate charging and discharging rates, mainly due to its extremely low electrical conductivity.

Our strategy was to use simultaneous graphene hybridization and Fe doping to prepare a LiMnPO₄-based cathode material with decent electrochemical performance [21]. We firstly coated oxidized graphene sheets with Fe-doped Mn₃O₄ nanoparticles by a controlled hydrolysis process. The obtained material was then transformed to LiMn_0.75Fe_0.25PO₄/graphene by reacting with LiOH and H₃PO₄ under solvothermal conditions. The LiMn_0.75Fe_0.25PO₄ took a rod-like morphology with average size below 100 nm (Fig. 5a). In addition, the long axis of the nanorod was parallel with the [001] direction of the crystal and the lithium ion diffusion channel, namely, the [010] lattice direction was a short axis of the nanorod [21]. Such desirable morphology was a direct outcome of the strong chemical interaction of the LiMn_0.75Fe_0.25PO₄ nanocrystals with the graphene sheets. The LiMn_0.75Fe_0.25PO₄ nanorods grown on graphene showed a specific capacity of 155 mAh/g at C/2 charging and discharging rates (Fig. 5b). The capacity was still as high as 132 and 107 mAh/g when discharged at 20 and 50C rates, respectively (Fig. 5b). The hybrid material produced one of the highest electrochemical performances among all the reported LiMn_1–xFe_xPO₄ cathode materials.

Fig. 5

(a) TEM image and (b) specific discharging capacity of the LiMn_0.75Fe_0.25PO₄/graphene hybrid as a cathode material for lithium ion batteries. Adapted with permission from ref. [21]. Copyright © 2011 Wiley-VCH.

To further increase the battery energy density by several times may require some new technology beyond lithium ion battery [12]. With high specific capacity and low material cost, lithium-sulfur battery is a potential choice [13]. However, there are several problems such as low electrical conductivity of sulfur, solubility of polysulfide discharging intermediates and volume expansion during discharging associated with the fast capacity decay of sulfur cathode materials. Graphene hybrid structures provided us a platform to design and prepare novel-structured sulfur materials with improved electrochemical performance. We simultaneously used polyethylene glycol (PEG) coating and graphene wrapping to realize a S/PEG/graphene hybrid material (Fig. 6) [22]. The graphene wrapping increased electron conduction and helped confining polysulfide intermediates. The PEG layer served as capping agents to control the size of the S submicron particles, helped trapping polysulfides during cycling, and created a cushion between the inner S particle and the outer graphene wrapping to accommodate the volume expansion of S during discharging. The hybrid material exhibited a specific capacity of ~600 mAh/g with ~10% decay over 100 cycles at C/2 charging and discharging rates, representing substantial improvement over conventional S/C mixtures [22].

Inorganic/nanocarbon hybrid materials for electrocatalysis of oxygen reduction and evolution reactions

Oxygen reduction and evolution reactions are fundamental electrode processes for electrochemical energy devices including fuel cells, metal-air batteries, and water-splitting cells [23–26]. Sluggish kinetics of oxygen electrodes is one of the bottleneck problems limiting device performance. Developing highly active, durable, and economical electrocatalysts for oxygen electrodes is essential for advancing the technology of electrochemical energy conversion and storage [8]. Platinum and its alloy nanoparticles have thus far been the most active catalysts for oxygen reduction reactions, but they are scarce and expensive, not compatible with large-scale applications. Non-noble-metal-based materials are being highly pursued as alternative catalysts. For example, 3d-transition-metal oxides are a type of materials with bifunctional catalytic activity for oxygen reduction and evolution reactions in alkaline solutions [8]. Nevertheless, the insulating nature of the metal oxides requires the catalysts to take a different structure than those based on conducting metals. We synthesized hybrid materials by covalently anchoring 3d-transition-metal oxide nanoparticles on oxidized graphene sheets or carbon nanotubes. The hybrids exhibited comparable oxygen reduction activity and superior stability as compared to Pt/C. The hybrids were also active for oxygen evolution reactions and enabled high-performance rechargeable lithium-air and zinc-air batteries.

Fig. 6

Design and synthesis principles of the S/PEG/graphene hybrid material. Adapted with permission from ref. [22]. Copyright © 2011 American Chemical Society.

Co₃O₄ nanoparticles were grown on graphene sheets with a two-step solution-phase synthesis that was initially developed for the above-mentioned Ni(OH)₂/graphene hybrid. Co(OAc)₂ was used as the precursor, and a ethanol/water mixture was used as the solvent for both steps. Ammonia was added to the synthesis to slow down the hydrolysis reaction by coordinating to the metal ions and subsequently reduce the size of the obtained nanoparticles to about 5 nm [27]. In addition, ammonia also reacted with the oxidized graphene sheets and rendered nitrogen-containing functional groups on the graphene surface as another type of binding sites for the oxide nanoparticles. As a catalyst for oxygen reduction, the Co₃O₄/graphene hybrid material showed high activity that had not been achieved by other Co₃O₄-based materials [27]. Molecular oxygen was reduced to water on the surface of the hybrid catalyst via a 4-electron process, as revealed by both rotating disk electrode and rotating ring-disk electrode measurements. When loaded with a density of 0.24 mg/cm² on a gas diffusion electrode, the Co₃O₄/graphene catalyst delivered an oxygen reduction current density of ~30 mA/cm² at 0.8 V vs. the reversible hydrogen electrode (RHE) in 1 M aqueous KOH solution, as compared to ~50 mA/cm² for the commercial Pt/C catalyst under the same conditions. While the catalytic current of Pt/C decreased by 35% after being on stream for 7 h, no activity decay was observed for our hybrid catalyst [27]. The Co₃O₄/graphene hybrid was also an active catalyst for oxygen evolution reaction.

The oxygen reduction catalytic performance of the Co₃O₄/graphene hybrid could be enhanced by substituting Mn for one-third of the Co in the structure [28]. The resulting MnCo₂O₄/graphene hybrid catalyst showed higher catalytic current density than Pt/C at medium overpotentials when measured under the same conditions (Fig. 7). While Mn substitution enhanced the activity for oxygen reduction reaction, the oxygen evolution activity was slightly decreased [28].

Fig. 7

Oxygen reduction polarization curves recorded with CoO/MWNT hybrid, hydrothermally derived Co₃O₄/MWNT, Co₃O₄/graphene, MnCo₂O₄/graphene hybrids and Pt/C. 0.24 mg/cm² of catalyst was loaded onto Teflon-coated carbon fiber paper to prepare the gas diffusion electrode. In the figure, NCNT and N-rmGO stand for N-containing carbon nanotubes and N-containing reduced mildly oxidized graphene oxide, respectively. Adapted with permission from ref. [29]. Copyright © 2012 American Chemical Society.

Using oxidized carbon nanotubes as growth substrates for the metal oxide nanoparticles brought extra activity improvement over the original graphene hybrid catalysts, as the oxidized MWNTs provided functionalized surfaces and conducting backbones simultaneously without structural compromise [29]. In 1 M KOH, a gas diffusion electrode with 0.24 mg/cm² of the Co₃O₄/MWNT catalyst delivered an oxygen reduction current of ~50 mA/cm² at 0.8 V vs. RHE (Fig. 7), significantly outperforming the corresponding graphene hybrid.

Performance of the hybrid catalysts could be further increased by employing a solid-gas-phase annealing step in place of the solvothermal step in the synthesis [29]. The reducing ammonia gas atmosphere transformed the intermediate nanoparticles to CoO on nanocarbons. The resulting CoO/MWNT became the most active oxygen reduction catalyst in alkaline solutions. With a catalyst mass loading of 0.24 mg/cm², the gas diffusion electrode showed a catalytic current density of ~70 mA/cm² at 0.8 V vs. RHE, substantially higher than that obtained with the Pt/C catalyst (Fig. 7). Like our other hybrid catalysts, CoO/MWNT was also very stable under working conditions. The catalytic current density only decreased by 4% after being on stream for 5.5 h [29].

Recently, we have developed a highly active, durable, and cost-effective oxygen evolution catalyst by growing Ni-Fe layered double hydroxide (LDH) nanoplates on oxidized MWNTs [30]. In aqueous KOH solutions, our NiFe-LDH/MWNT hybrid catalyst exhibited both higher activity and higher stability than the precious Ir/C catalyst under the same measurement conditions (Fig. 8). The NiFe-LDH/MWNT hybrid was among the most active and durable electrocatalysts based on non-precious metals.

Fig. 8

Electrochemical performance of NiFe-LDH/MWNT hybrid oxygen evolution catalyst. Polarization curves of NiFe-LDH/MWNT and Ir/C catalysts measured on (a) rotating (1600 rpm) glassy carbon (GC) electrode (catalyst loading 0.2 mg/cm²) and (b) carbon fiber paper (CFP) electrode (catalyst loading 0.25 mg/cm²) in 0.1 and 1 M KOH solutions, and chronopotentiometry curves of NiFe-LDH/MWNT hybrid and Ir/C catalyst measured on (c) GC electrode (current density 2.5 mA/cm²) and (d) CFP electrode (current density 5 mA/cm²) in 0.1 and 1 M KOH solutions. Reprinted with permission from ref. [30]. Copyright © 2013 American Chemical Society.

The superior bifunctional catalytic activity of the spinel-oxide/nanocarbon hybrid materials suggested their promising potential as cathode catalysts for rechargeable metal-air batteries. We evaluated the MnCo₂O₄/graphene hybrid material in Li-O₂ coin cells (Fig. 9a) [31]. By comparing various cathode catalysts, we found that the discharging voltage of the cell was directly related to the oxygen reduction activity of the catalyst in both aqueous and non-aqueous solutions, but the charging performance of the cell did not seem to correlate with the oxygen evolution activity of the catalyst in aqueous solutions. At a charging and discharging current density of 100 mA/g, the cell catalyzed by the MnCo₂O₄/graphene hybrid exhibited discharging and charging voltages of 2.95 and 3.75 V, respectively (Fig. 9b), among the lowest overpotentials reported in similar electrolyte at comparable gravimetric current densities. The cell could deliver a specific capacity of 1000 mAh/g for 40 cycles with only negligible decrease in energy efficiency (Fig. 9c).

Fig. 9

Schematic structures and performance of rechargeable Li-O₂ and Zn-O₂ batteries catalyzed by inorganic/nanocarbon hybrids. (a) Schematic structure of a Li-O₂ coin cell using MnCo₂O₄/graphene as cathode catalyst. (b) Charging and discharging voltage profiles of the Li-O₂ coin cell at various current densities. (c) Charging and discharging voltages of the Li-O₂ coin cell during first 40 cycles of 1000 mAh/g. (d) Schematic structure of a tri-electrode Zn-O₂ cell using CoO/MWNT and NiFe-LDH/MWNT as cathode catalysts. (e) Charging and discharging performance of the Zn-O₂ cell catalyzed by hybrid catalysts as compared to another cell catalyzed by Pt/C and Ir/C. (a–c) Adapted with permission from ref. [31]. Copyright © 2012 Royal Society of Chemistry. (d,e) Adapted with permission from ref. [32]. Copyright © 2013 Nature Publishing Group.

Recently, we combined our most active oxygen reduction catalyst, the CoO/MWNT hybrid with our most active oxygen evolution catalyst, the NiFe-LDH/MWNT hybrid to catalyze the cathode reactions of rechargeable zinc-air batteries [32]. Charging and discharging overpotentials of the cell were substantially reduced by the two high-performance electrocatalysts, affording specific energy >700 Wh/kg and a maximum power of 265 mW/cm². With a tri-electrode configuration (Fig. 9d) to circumvent the deactivation problem of the oxygen reduction catalyst at high electrical potentials, the cell exhibited great cycling stability with negligible increase in overpotential after 200 h of operation (Fig. 9e). At a current density of 20 mA/cm², the roundtrip efficiency of our cell was ~65%.

XANES study of electrochemical interactions at the inorganic/nanocarbon interface

The electrical and chemical interactions between the inorganic nanocrystals and the nanocarbon substrates are key to the design, synthesis, and high electrochemical performance of our hybrid materials [7, 8]. XANES is an appropriate technique to study such interactions [33]. In an XANES measurement, one inner electron of an atom is ejected by an X-ray photon. Near the absorption edge, the energy of the photoelectron is relatively low, and it can be scattered multiple times by the neighbor atoms, which allows for probing of oxidation state and coordination environment of an element as well as the unoccupied band structure of a material. Another electron at a higher level will fill the vacancy generated by the X-ray photon, which will then lead to fluorescence or an Auger electron. The Auger electron signal gives surface-rich information due to the short mean free path of electrons in materials.

XANES measurement was performed for the Mn₃O₄/graphene hybrid with a scanning transmission X-ray microscope (STXM) set-up. Well-resolved C–C π* and σ* features observed in the C K-edge spectrum of the hybrid revealed that oxidized graphene in the hybrid was well reduced. Lowered π* transition of the hybrid indicated electron transfer from Mn₃O₄ to graphene. Strong C–O bond absorption of the hybrid suggested that the Mn₃O₄ nanoparticles were likely to be bound to graphene via Mn–O–C bonds. In the Mn L-edge and O K-edge spectra of the hybrid, we noticed increased absorption for Mn and decreased absorption for O, as compared to free Mn₃O₄. As the area under the absorption curve is in direct proportion to the unoccupied density of states, the data revealed slight electron relocation from Mn to O in the hybrid, induced by the interaction between Mn₃O₄ and graphene.

For the LiMn_0.75Fe_0.25PO₄/graphene hybrid, we also found the graphene in the hybrid was in a well-reduced form as suggested by the well-resolved C–C π* and σ* features (Fig. 10a). The existence of C–O groups indicated the metal-O-carbon bonding in the hybrid (Fig. 10a). There was electron transfer between the nanorods and the graphene as well (Fig. 10b–d). These results strongly suggested chemical and electrical coupling at the hybrid interface, which was responsible for the high electrochemical performance of the material. Moreover, we found possible P–O–C bonding and surface Li polyphosphate in the hybrid, which could also contribute to its high rate performance [34].

Fig. 10

XANES spectra of LiMn_0.75Fe_0.25PO₄/graphene compared to free LiMn_0.75Fe_0.25PO₄ and graphene control samples at the (a) C K-edge, (b) Mn L-edge, (c) Fe L-edge and (d) O K-edge recorded in total-electron-yield mode. In the figure LMFP and rGO stand for LiMn_0.75Fe_0.25PO₄ and reduced graphene oxide respectively. Reproduced from Ref. [34] by permission of the PCCP Owner Societies.

For the ultimate purpose of understanding chemical processes of our novel hybrid materials in working devices, we took the LiMn_0.75Fe_0.25PO₄/graphene material out of the coin cell after a full charging step and performed XANES measurement on STXM [35]. We found that Fe(II) in the original material had almost been completely oxidized to Fe(III) (Fig. 11a), but only a small portion of Mn(II) had been oxidized to Mn(III) and there was more Mn(III) detected in the bulk of the nanorods than on the surface [35]. Chemical mapping of the Fe species with different oxidation states in the materials at various states of charge (Fig. 11b) provided distribution of charged products with nanometer-scale resolution as well as delithiation reaction frontiers (boundary between the red and blue areas in Fig. 11b).

Fig. 11

(a) Fe L-edge XANES of LiMn_0.75Fe_0.25PO₄/graphene at various state-of-charge stages along with the pristine state. (b) STXM chemical maps of partially charged LiMn_0.75Fe_0.25PO₄/graphene for visualizing the Fe valance distribution. Adapted with permission from ref. [35]. Copyright © 2013 Royal Society of Chemistry.

We also used XANES to characterize the strong electrochemical coupling responsible for the synergistic effect in our inorganic/graphene hybrid catalysts. In the Co₃O₄/graphene hybrid, we found increased absorption at 288 eV on C K-edge compared to the graphene control sample, corresponding to carbon atoms attached to oxygen or nitrogen [27]. This suggested possible formation of Co–O–C and Co–N–C bonds at the oxide/graphene interface in the hybrid. We also noticed increased absorption at the Co L-edge and decreased absorption at the O K-edge in the hybrid compared to pure Co₃O₄ nanoparticles, suggesting electron dislocation from the Co site to the O site and consequently a higher degree of ionic Co–O bonding in the hybrid. Bond formation between Co₃O₄ and graphene and changes in the chemical environment of C, O, and Co atoms in the hybrid material were likely to be responsible for the synergistic oxygen reduction catalytic activity [27].

Doping Co₃O₄ with Mn led to the MnCo₂O₄/graphene hybrid, which was an even more active electrocatalyst for oxygen reduction reactions than the Co₃O₄/graphene hybrid [28]. XANES measurement revealed that Mn in the hybrid was mainly in the form of Mn³⁺ with a small amount of Mn⁴⁺ (Fig. 12d), indicating that half of the Co³⁺ in the original Co₃O₄ structure was substituted by Mn³⁺ and the amount of Co²⁺ remained almost unchanged (Fig. 12c). Since Mn³⁺, Mn⁴⁺,and Co²⁺ are generally considered to be active species for oxygen reduction, the improved catalytic activity of the MnCo₂O₄/graphene hybrid could be well explained by the XANES results (Fig. 12).

Fig. 12

XANES spectra of MnCo₂O₄/graphene hybrid compared with Co₃O₄/graphene hybrid, free MnCo₂O₄, and free graphene control samples at (a) C K-edge, (b) N K-edge, (c) Co L-edge, and (d) Mn L-edge. Reprinted with permission from ref. [28]. Copyright © 2012 American Chemical Society.

Summary and future remarks

In this article I have reviewed our recent progress in designing and synthesizing inorganic/nanocarbon hybrid materials for supercapacitors, ultrafast Ni-Fe batteries, lithium ion batteries, Li-S batteries, oxygen reduction and evolution catalysis and metal-air batteries. The key to the superior performance of the hybrid materials is building strong chemical and electrical coupling at the inorganic/nanocarbon interface. We expect more rational design and controlled synthesis could come into the developing process of new energy materials to advance the technology of electrochemical energy storage and conversion. Besides material engineering aiming for better performance, it is critically important to keep up the fundamental study of structure–property relationships in novel electrochemically functional materials. Chemical synthesis, structural characterizations, and reaction study ought to be combined more widely, frequently, and deeply in research. As the electrode materials and electrocatalysts always change their structures when exposed to electrolytes and electric potentials, endeavor should be taken in developing multiple techniques for characterizing chemical and physical structures of materials under working conditions.