Providing and storing clean energy to an exponentially growing population is becoming an ever-increasing challenge. With fossil fuels being a finite resource, new techniques must be developed to store and supply energy.
Energy storage is one of the many challenges faced when approaching the issue. Ideally a device used for energy storage should have two key attributes: High energy density and high-power density. Energy density is defined by the capacitance of the battery whereas power density is defined by loading capability (It’s ability to deliver power)1
The two most common energy storage devices currently being used are and not exclusive to, lithium-ion batteries (LiBs)and supercapacitors (SC’s). 2 The difference between conventional LiBs and SC’s will be discussed in detail later in this dissertation. The development of more sophisticated battery technology over recent years has significantly improved the energy density of said batteries by increasing the small size high capacitance ratio batteries for small hand held electronic devices.
Some of main issues LiBs face are: charge time, lifetime, power density and excess heat.3
Figure 1 Ragone plot showing energy density vs. power density for various devices. 4
One of the most promising alternative energy storage solutions are SC’s. SC’s (also known as ultracapacitors) are storage devices that offer high capacitance via electrostatic charge.5
The high capacitance alongside their long cyclic life of 10,000 cycles of modern SC’s 6 make them ideal for the use in many applications that require energy to fill a short term power need 7 such as loading cranes, forklifts, electric vehicles and even intermittent generators (wind farms), and smart grids as a result of their fast charge propagation 8.
Their properties of low resistance present the advantage of a significantly lower amount of excess heat energy produced avoiding the risk of overheating that conventional LiB’s face. These charge cycle life properties give SC’s a significant advantage compared to the cycle life a conventional LiB (approx. 300+ cycles) 9
Having less energy density than a conventional LiB, SC’s are still not viable replacements for the use as primary storage for many electronic devices.
If the possibility of creating a supercapacitor that exhibits high energy density as well as power density, the applications for such a device could extend further than uses in specialist equipment or the aerospace industry. The upcoming ban made by the UK government of the production of new diesel and petrol cars in the UK by 2040 10 galvanises the need to research more into more efficient and clean alternative energy storage solutions.
For example, the use of electric cars could become more widespread if the charge time and duration of charge is significantly improved; all of which a supercapacitor could achieve once overcoming the obstacle that is energy density.
In an age where demand for energy Is rising exponentially, consumers want to be able to charge their devices quickly whilst retaining the ability of large energy storage and overcoming any safety hazards that LiBs possesses (prone to overheating and causing electrical fires). The research into the latest generation supercapacitors paves the way for a greener and more energy efficient world.
This dissertation will analyse the recent findings in supercapacitor technology, more specifically the advances in metal ceramic (metal nitrides and oxides) based supercapacitors compared to the ones commonly used at present, as well as an outlook into the future of this technology
2. The Supercapacitor
A capacitor is defined as a “a passive terminal electrical used to store energy electrostatically in an electric field separated by a dielectric (i.e. insulator)”. 11 One of the main features that separates a SC from a conventional capacitor is its utilization of electrode materials with large surface areas alongside use of thin electrolytic dielectrics; thus having a significantly large capacitance than that of a regular capacitor 12. The electrolyte in SC’s is the conductive connection between two electrodes2 distinguishing itself from the common electrolytic capactitor. Due to no change in phase or chemical reaction occurring the discharge-charge cycle can theoretically be repeated over and over 13 Capacitance is an objects ability to store an electrical charge 14. Supercapacitors could essentially bridge the gap between batteries and capacitors and perhaps in the near future completely replace batteries for conventional energy storage.
Figure 2 comparison of the performance of a conventional supercapacitor vs a Lithium -ion (general)11
The information presented in figure 2 clearly shows that the charge time and cycle life of a SC is significantly greater than a general Lithium-ion storage device. However. Its energy density is significantly lower alongside the greater cost which questions the economic viability of using supercapacitors as storage. The self-discharge rates of commercially available SC’s are extremely high (100 to 50 percent in 30 to 40 days)7 whereas LiB are said to self-discharge about 2 percent per month.15 This will be discussed further in detail later in this review.
Supercapacitors are currently separated into 3 distinct types: EDLCs (electrochemical double-layer capacitors), Pseudocapacitors and hybrid capacitors. Their mechanisms for storing charge are what separates them from each other. Energy is either stored by a non-Faradaic mechanism (EDLCs). A Faradaic mechanism (Pseudocapacitors) and Hybrid is essentially is the two mechanisms combined12
Capacitance can be expressed by the following equation :
Where E is energy, C is capacitance and V2 is the square of the applied voltage.
A basic EDLC has two electrodes that are either coated or immersed with an electrolyte; applying an electric potential to they system causes electrons to flow onto the negative electrode which in turn attract positively charged ions present within the electrolyte forming to the negative electrode and the accumulation of electrons on the positive electrode thus forming an ionic layer at both electrodes allowing the charges to separated and energy to be stored statically. 16
Figure 3: The taxonomy of SC classes
As shown in Figure 3 the two electrodes are kept apart by a separator which allows the movement of charged ions through the membrane but prevents the
electronic contact between the electrodes. The material used for the separator is dependent on the electrolyte used is solid state, organic or aqueous.17
In recent years more, sophisticated techniques and materials have been researched into improving the efficiency of the electrodes used in EDLC’s. Techiques that are adopted in order to increase the energy densities of EDLCs include: encoporating nitrogen and oxygen functional groups to the surface layer, using pore sizes that are equal to the size of electrolyte ions, using redox-active species in the electrolyte, or designing ionic liquid mixtures for improving the cell voltage and temperature range.18
Graphene has gained much attention due to the properties it possesses. Graphene is a single layer of sp2 hybridised carbon atoms which are bonded together and arranged in a lattice that of a honeycomb. Being extremely lightweight and having high electrical and thermal conductivity alongside its strong mechanical strength and chemical stability makes it very appealing for it to be used in EDLC’s. 19
Figure 4: A basic EDLC diagram20 Figure 5: The “Honeycomb” lattice of Graphene21
New research into graphene based electrodes for SC’s has shown high levels of conductivity 106 S cm-16 but, are held back by their low capacitance of approximately 100–205 F g-1 6, this significantly hinders their ability for usage in primary power storage devices. A potential solution to this issue is the use of graphene in hybrid electrodes; combining the high conductivity properties of graphene with a material with better capacitance. A study carried out by the departments of chemical engineering, materials science and engineering and chemistry at Stanford University in 2011 managed to achieve this.
“By three-dimensional (3D) conductive wrapping of graphene/MnO2 nanostructures with carbon nanotubes or conducting polymer, specific capacitance of the electrodes (considering total mass of active materials) has substantially increased by ?20% and ?45%, respectively, with values as high as ?380 F/g achieved. Moreover, these ternary composite electrodes have also exhibited excellent cycling performance with >95% capacitance retention over 3000 cycles”22. It becomes apparent that these composites show a lot of promise for the future in energy storage.
Pseudocapacitors undergo electron transfer mechanisms in order to achieve charge storage rather than electrostatically compared to the likes of and EDLC.14 This transfer of charge is known as a ‘Faradic’ mechanism. There are three known systems that give rise to pseudocapacitance: Redox, Intercalation and Underpotential deposition of metal adatoms.
Pseudocapacitance can be expressed by the following equation:
Where q is the faradic charge needed for adsorption/desorption of ions, d? is the change in fractional coverage of the surface and dV is the change in potential. Due to the nature of being non linear, the capicitance is not entirely constant, thus giving rise to the term pseudocapacitance 24 Electron charges (either in excess or defiency) are accumulated on the electrodeplates with no inolvement of redox chemical changes alongside repulsion between charges. The electrons inolved in Faradaic battery- type processes are transferred between the two electrodes with different oxidation states. 25
(Should I discuss the thermodynamic and kinetic theory of pseudocapacitance?)
Equation 3 shows the general form of a redox equation:
Where Ox is the oxidized species and Red is the reduced species.
One material that is an area of interest is Ruthenium oxide (RuO2) since its first discovery in 1971 and was the first material to exhibit pseudocapacitance due to the involvement of faradaic charge-transfer reactions;18 since its discovery numerous studies managed to improve the capacitance up to 720F g-123, largely due to being able to pinpoint the importance of structural water as well as implementing a porous nanoscale architecture18. As a result hydrous ruthenium dioxide, RuO2?xH2O or RuOxHy was shown to exhibit the high capcitance values mentioned earlier.23
The equation 4 shown below is used to express the storage of protons by hydrated RuO2
RuOx(OH)y + dH+ + de- ? RuOx-d(OH)y+d (4)
The capacitance of RuOxHy has linked to several key variables variables such as: surface area, water content, electronic conductivity, and crystallinity.23
The high capacitance values of Hydrous ruthenium oxides results from a pseudocapacitive effect whereby both protons and electrons are transferred from an acidic electrolyte into the surface of hydrous ruthenium oxide, (eqn 4 describes this process).23
The pseudocapacitance of RuOx comes mainly from the surface reactions. The higher the specific surface area, the more the metal centers will be capable of providing multiple redox reactions and the higher the specific capacitance will then be. Obviously, one of the most effective ways to increase the specific capacitance is to increase the specific surface area of RuOx.182 Several approaches, such as depositing RuO2 films on substrates with a rough surface, coating a thin RuO2 film on high-surface-area materials, making nanometre-sized oxide electrodes, etc. have been explored to maximize the surface area of RuO2 by creating micropores large enough for ion diffusion.125,190–194 For example, hydrous ruthenium oxide (RuO2?xH2O) thin-film electrodes electrodeposited on titanium substrates exhibited high reversible characteristics, excellent cycle stability, and superior power characteristics. The maximum specific capacitance of the RuO2?xH2O film electrode was as high as 786 F g-1 24
The combined water in RuOx. As implied by eqn (13), the reversible redox transitions strongly depend on the processes of proton/cation exchange and electron-hopping. The quasi- metallic conductivity of RuO2 allows facile electron transfer into and through the electrode matrix, so the transfer of cations in the solid phase is critical for a RuO2 electrode. It was reported that cation diffusion in hydrated electrodes could occur via hopping of alkaline ions and H+ ions between H2O and OH- sites, suggesting that the hydrogen atoms were relatively mobile in RuO2?xH2O samples as compared to those in rigid samples.196 Thus, the combined water in RuO2 is expected to enhance the diffusion of cations inside the electrode layer.
The pseudocapacitance of RuO2?xH2Omaterials is also closely related to the degree of crystallinity. A well crystallized structure has difficulty in expanding or contracting, so it prevents protons from permeating the bulk material, leading to a diffusion limitation. As a result, the fast, continuous, and reversible faradaic reactions are compromised, and the capacitance of RuO2 with good crystallinity comes mainly from the surface reaction.205 In contrast, the redox reaction of an amorphous composite occurs not only on the surface but also in the bulk of the powder, which is why an amorphous composite exhibits far superior performance when compared with crystallized structures.24
Smaller sized particles not onlycan shorten the diffusion distance but also can facilitate proton transport in the bulk of ruthenium oxide, increase the specific surface area, and enhance the electro-active sites. Thus, the smaller the particle size, the higher the gravimetric capacitance and utilization efficiency.
Another metal oxide that has developed a significant amount of interest in the use of supecapacitors is Manganese oxide (MnO2).
Some advantages of MnO2 are its low cost, high capacitance as well as its overall electrochemical performance25.
All routes contain siginificant advantages and disadvantages scaling from cost to the yield of the desired product. 25. MnO2 is said to have a theoretical capacity of approximately 1370 F g-1 25however, in practise this value has been shown to be significantly lower. In 1999 some of the first tests on MnO2 for its suitablility as a supercapacitor materials were carried out by Lee and Goodenough recording a specific capacitance of 200 F g-126. Over recent years scientists have attempted (and succeeded) to improve this value through the synthesis of nanostructured MnO2- based electrodes. This has been achieved through many synthetic routes. In order to synthesise nanoscale MnO2 suitable for use in supercapacitor electrodes, the following routes can be used25:
· Hydrothermal synthesis
· Chemical coprecipitation
· Low-temperature solid-phase synthesis
· Microwave synthesis
These routes allow more control over the following parameters: particle size, morphology, crystallinity, high specific surface area and electrical conductivity 26
Another technique that has been used is the design of composite materials based on MnO226. As found with RuO2, the psuedocapacitance of MnO2 is also closely related to crystallinity. If the degree of crystallinity becomes too high the rate of the protonation (or deprotonation) becomes significantly hindered. Studies found that high crystallinity resulted in an increase of conductivity and a decrease in surface area simultaneously.24
Crystallised MnO2 is said to appear in several different forms, : ?-, ?-, ?-, and ?-MnO2. 24
Each form requires its own unique set of conditions in order for its formation as well as each form varying in specific capacitance 24. The morphology of plays a large apart in determining the specific capacitance of MnO2 largely due to its direct effect on the surface area 24 as well as its surface to volume ratio26. The “one dimensional nanostructured MnO2″ resulted in high specific capacitance due to its large surface areas created by the short diffusion pathways for the electrons and ions.26 A study by Wu et al found that MnO2 nanowires possessed a specific capacitance of 350 F g-1.27
The small diameter of the nanowires can provide a significant number of active sites that allow for charge transfer as well as shorten proton diffusion, resulting in higher charge/discharge rates. 26 Other morphologies of MnO2 have also been prepared such as : nanorods, nanobundles, nanobelts and hollow spheres /”hollow urchins”, 24 each resulting in different sizes of surface areas. The morphology can be controlled by changing how its prepared or by manipulating the reaction conditions
As mentioned various carbon materials along side metal oxides have been growing in popularity over recent years for the use as electrode materials in energy storage devices which are considered the key component for determining the performance of said storage devices. However one group of materials that hav started to gain a large amount of interest and research into are the metal nitirides (MN). 2
The term “Metallic Ceramics” is used to group metal nitrides alongside based upon their unique properties. Metal nitrides follow the trend of high melting points, extreme hardness and large amounts of strength; typical properties that a standard ceramic material would follow. However, a large number of MN’s are metallic conductiors exhibiting strong electronic and magnetic properties, thus increasing the interest in potentially using these MN’s in energy storage applications when compared to metal oxides such as RuO2 (which we discussed previously in this report).28
Many researchers have found that metal nitrides posses a plethora of advantages such as “outstanding electrochemical properties, high chemical stability,.2 These findings are key to improving the effeciency, and charge capacity that supercapacitors lack when compared to modern LiB’s. Metal nitrides have displayed large capacitance values alongside impressive cycling stablity 2
Numerous process methods have been used to synthesise metal nitrides some of which are as follows:
1. Heating metal to high temperature in N2 or NH3 environments
2. Ammonolysis of oxides (heating metal oxide in liquid NH3 whilst gradually increasing the temperature at a slow rate>) 29
3. Sol-gel (Allows more control over porosity of final product by adjusting the conditions of reacton accordingly) 28
4. Vapor depostion of thin films
One metal nitride in particular that my attention was is “Vanadium Nitride”. apacitance and pseudo-capacitance occurring at the nitride surface. Vanadium is an element in Group 5 of the periodic table with a half empty d-shell with variable oxidation states of +2 to +5, thus allowing it to partake in reversible Faradaic reactions yielding pseudocapacitance.
Experiments have been taking place to further improve the synthesis Experimentation on VN nanocrystals carried out by Choi et al proved that they can display high specific capacitance (ca. 400 F g-1) at a scan rate of 50 mV s-1 approximately 1000 cycles without dropping more than 10% in capacitance. This experiment has said to display the most optimal cycling performance so far reported for pseudocapacitors. 30
Unfortunately, the results came with its shortcomings, the main issue being that the use of VN as the SC electrode resulted in poor “cyclic stability and short cycle life”.
The report has shown that nitride capacitance is “directly proportional to surface area and inversely proportional to particle size”. Although the study also demonstrated that if the vanadium nitride nanoparticles were made too small it resulted in a decreases of net electronic conductivity. In addition, the capacitance behavior of ultra-fine nanoparticles of vanadium nitride has been shown to depend very much on the electrode loading. Very thin films exhibit very high gravimetric charge storage on account of the facile electron conduction and ion accessibility into pores. However, thicker films end up being limited in their charge storage capability on account of the same. The cycling stability of vanadium nitride has also been shown to depend on the particle size reflected as a possible difference in the surface oxidation state.
The Faradaic reactions are more pronounced on the ammonlysis derived VN on account of the higher surface area resulting in increased surface oxide contact with the electrolyte.