Development, design and applications of structural capacitors

Applied Energy 231 (2018) 89–101

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Development, design and applications of structural capacitors

T

D.D.L. Chung Composite Materials Research Laboratory, Department of Mechanical and Aerospace Engineering, University at Buffalo, The State University of New York, Buffalo, NY 14260-4400, USA

H I GH L IG H T S

capacitors are multifunctional structural materials. • Structural provide the capacitor function for the purpose of electrical energy storage. • They paper reviews the scientific development of structural capacitors. • This paper also enunciates the design and applications of structural capacitors. • This Structural capacitors will provide an untapped form of energy storage. •

A R T I C LE I N FO

A B S T R A C T

Keywords: Structural capacitor Energy storage Composite materials Capacitance Carbon fiber Polymer

Structural capacitors are multifunctional structural materials that provide the capacitor function for the purpose of electrical energy storage. This paper reviews the development of structural capacitors and enunciates their design and applications. A structural capacitor is commonly a polymer-matrix structural composite with a dielectric film between the electrodes, which are an electronic conductor, commonly the continuous carbon fiber laminae that serve to reinforce the composite. The dielectric film is preferably small in thickness and serves to avoid short circuiting of the two electrodes. In order to maximize the capacitance by having the structural capacitor constitute capacitors in parallel, the dielectric film is preferably positioned at every interlaminar interface of the composite, such that alternating electrodes in the stack are connected to opposite polarities of the AC electric field source. A structural supercapacitor requires the matrix to be a solid electrolyte. From the viewpoints of structural performance, safety, service life and high frequency capability, structural dielectric capacitors are closer to commercialization readiness than structural supercapacitors. Structural capacitors have not yet been commercialized, but they are expected to provide an untapped, extensive, save and distributed means of energy storage, and allow aircraft, satellites, automobile, ships, wind turbines, buildings, solar panels, display panels, outdoor lighting, computers, cell phones, etc., to store energy in their structures.

1. Introduction

used to power devices such as sensors. The global energy storage market is growing significantly to an annual installation size of 6 GW in 2017 and over 40 GW by 2022 [1]. This growth is from an initial base of only 0.34 GW installed in 2012 and 2013 [1]. In particular, the California Public Utilities Commission has approved a target that requires the three largest investor-owned utilities, aggregators, and other energy service providers in California to procure 1.3 GW of energy storage by 2020 [1]. Current technologies for energy storage [2–6] include batteries (electrochemical devices in which energy is stored in the form of chemical energy, e.g., lithium-ion, sodium-sulfur and lead-acid batteries) [7–11], flow batteries (rechargeable batteries that involve two chemical components dissolved in liquids that are contained within the system

Due to the intermittent nature of the generation of electrical energy by photovoltaics, wind and other renewable energy sources, the storage of the generated electrical energy is needed. By storing the energy, energy can be available at times when no energy is generated by the renewable energy source. In addition, energy storage is needed to help manage the energy utilization of the electrical grid. The demand for electricity varies by hour, day and season. The energy management involves monitoring the electricity demand, supply, reserve margins and the mix of electricity generating technologies. For such energy management, large-scale energy storage is needed. Large-scale energy storage is to be distinguished from small-scale energy storage that is

E-mail address: ddlchung@buffalo.edu. URL: http://alum.mit.edu/www/ddlchung. https://doi.org/10.1016/j.apenergy.2018.09.132 Received 25 April 2018; Received in revised form 10 September 2018; Accepted 11 September 2018 0306-2619/ © 2018 Elsevier Ltd. All rights reserved.

Applied Energy 231 (2018) 89–101

D.D.L. Chung

two electrodes being different in composition, so that a redox reaction as in a battery contributes to the capacitance). In relation to structural supercapacitors, both types suffer from the problems related to service life and safety. The rendering of the capacitor function to a structural material makes the material multifunctional. The multifunctionality is attractive for smart structures. Non-structural functions that have been reported include strain/stress sensing [34], structural health monitoring [35], electric power generation [36,37], energy storage [38–47], heat dissipation [48], deicing [49] and vibration damping [50]. The attainment of multifunctionality without the embedment or attachment of devices (e.g., strain gages, commercial capacitors, commercial batteries, etc.) is particularly attractive. Compared to the use of devices, it gives lower cost, higher durability and absence of mechanical property loss. Because of the desire for saving weight, a low-density structural material is commonly attractive. Thus, continuous fiber polymer-matrix composites [51], which are well-known for their combination of low density, high elastic modulus and high strength, are attractive for serving as the base material for modification to render the capacitor function. This review is thus focused on structural capacitors in the form of polymer-matrix composites. The objective of this paper is to review the development of structural capacitors, with particular attention on the engineering design and applications. Engineering design is a practically important area that has not been enunciated coherently in prior work. The applications are critical to the implementation of the technology, but they have not been given the needed attention.

and most commonly separated by a membrane, such that each liquid circulates in its own respective space, e.g., vanadium redox and zincbromine flow batteries) [12–18], supercapacitors (electrochemical devices in which energy is stored in the form of electrical energy) [9,11,19,20], flywheels (device involving accelerating a rotor or flywheel to a very high speed and maintaining the energy in the system as rotational energy) [3,21,22], compressed air energy storage (device based on the principle that the compression of air creates heat and the expansion of air removes heat) [23–27], thermal energy storage (using stored excess heat to obtain energy when it is needed) [21,22,28–32] and pumped hydropower (energy derived from falling water or fast running water) [26,33]. For large-scale energy storage, scale-up is necessary. Compressed air energy storage and pumped hydropower are relatively suitable for scale-up. However, both underground compressed air energy storage and pumped hydropower energy storage are limited by geological and environmental constraints. Intense research is ongoing concerning energy storage using batteries, supercapacitors, flywheels and thermal means. In order for batteries and capacitors to be able to provide large-scale energy storage, the size of these devices must be increased greatly. Such an increase can be enabled by the use of large structures as the form of these devices. This means that the structure serves not only the structural function, but also the energy storage function. Such multifunctional structures are known as structural batteries and structural capacitors. A further advantage of such multifunctional structures is that weight and volume are not issues, as the devices effectively vanish in the structures. Furthermore, the multifunctional structures enable energy storage to occur in a spatially distributed fashion (as in community energy storage that is grid-connected and utility owned and operated) rather than being limited to centralized locations, thereby increasing the total capacity of energy storage. The operation of a capacitor does not involve energy conversion, as only electrical energy is involved, but the operation of a battery involves the conversion between chemical energy and electrical energy. Due to this energy conversion, a disadvantage of batteries is that the electrochemically active species can be depleted after a period of use and hence the battery ceases operation. This disadvantage is particularly serious for structural batteries, as a structure may be required to be in service for decades. Capacitors excel in their high power capability, whereas batteries excel in their high energy capability. This difference in behavior is commonly expressed in terms of the Ragone plot. High power is needed for electric vehicles, hybrid vehicles, electric drive-trains, aircraft powertrains and shipboard power systems that need high power for acceleration. It is also needed for electric and hybrid propulsion systems, for regenerative braking (which helps improve the fuel efficiency of a vehicle that is operated under a stop-and-go urban driving condition), and for equipment that needs high pulse power, such as rail guns, electromagnetic armor and airborne lasers. Energy storage is also needed for self-powered structures and for structures that need emergency power. There are two main types of capacitors, namely dielectric capacitors (also known as electrostatic or electrolytic capacitors) and supercapacitors (also known as ultracapacitors). The former is purely electrical in nature, whereas the latter is electrochemical in nature. Due to its electrochemical nature, a supercapacitor requires an electrolyte, which complicates multifunctional structural material design, limits the service life and poses safety concerns. In contrast, a dielectric capacitor does not involve any electrolyte. Since service life and safety are essential for structural capacitors, dielectric structural capacitors are more promising than structural supercapacitors, in spite of the fact that the capacity for small-scale energy storage tends to be greater for a supercapacitor than a dielectric capacitor. Among supercapacitors, there are two types, which are the electric double layer capacitor (symmetrical, with the two electrodes being the same in composition) and the pseudocapacitor (asymmetrical, with the

2. Polymer-matrix structural composites The dominance of polymer-matrix composites among composites with various matrices (polymer, carbon, ceramic, metal, cement, etc.) stems from the relative ease (low cost) of fabrication and the relatively good bonding ability of polymers. Applications include aircraft, unmanned aerial vehicles, satellites, automobile, sporting goods, wind turbines, structural repair, etc. The continuous fibers that are most commonly used for structural composites are carbon fiber, glass fiber and Kevlar (polyaramid) fiber. Carbon fiber is advantageous is its high tensile modulus and the low magnitude of the coefficient of thermal expansion, as well as its high temperature resistance, chemical resistance, electrical conductivity and thermal conductivity. In particular, the electrical conductivity is valuable for structural capacitors, as the carbon fibers can be used as the electrodes in the capacitor. Continuous fiber polymer-matrix composites are structural composites that exhibit excellent mechanical performance, as widely used in advanced composites, such as those for airframes. Discontinuous fibers are not as effective as continuous fibers as a reinforcement, due to the imperfect bond between the fibers and the matrix. Nanofibers and nanotubes are not as effective as continuous fibers as a reinforcement, due to their discontinuity, limited degree of alignment, and limited volume fraction in a composite. In contrast, due to their continuity, high degree of alignment, and high volume fraction in a composite, continuous fibers are highly effective as a reinforcement. On the other hand, nanofibers and nanotubes can be used as a secondary reinforcement in a composite that contains continuous fibers as the primary reinforcement. Depending on the combination of the primary and secondary reinforcements, the secondary reinforcement serves to improve properties such as the vibration damping ability, electrical conductivity, thermal conductivity and dimensionless thermoelectric figure of merit [36,37,50,52,53]. The combined use of continuous fibers and nanofillers in the same composite provides hierarchical (multi-scale) composites. Continuous fiber polymer-matrix composites that exhibit high mechanical performance are primarily of one of two forms. These forms include (i) multidirectional fiber laminates (made by the stacking and 90

Applied Energy 231 (2018) 89–101

D.D.L. Chung

Fig. 2. An optical microscope image of the cross-sectional view of a part of a continuous carbon fiber epoxy-matrix composite. Only an interlaminar interface and its two adjacent laminae are shown. The fibers in the top lamina are in the plane of the image; those in the bottom lamina are perpendicular to the plane of the image.

Fig. 1. An optical microscope image of the cross-sectional view of a part of a continuous carbon fiber epoxy-matrix composite. Only ten laminae are shown, with the fiber direction being the same in each lamina and the fiber direction of a lamina being one of four directions (successively 0°, 90°, +45°, −45°, followed by repeating this layer-up order). By having the four directions, the composite behavior is roughly isotropic in the plane of the laminate. This fiber lay-up configuration is said to be quasi-isotropic.

laminae are used as the electrodes in a dielectric structural capacitor, the through-thickness distance between the laminae should be as small as possible in order to maximize the capacitance (Eq. (1)). The presence of additional constituents at the interlaminar interface in order to render the capacitor function to the composite tends to increase this distance, thereby decreasing the capacitance. An example of an additional constituent is a dielectric film positioned at the interlaminar interface in order to avoid short-circuiting between the continuous fiber laminae, which serve as electrodes. In case of a structural supercapacitor, an electrolyte is needed. For this purpose, an electrolyte in the form of a solid polymer is commonly used. The electrolyte polymer tends to be inadequate in the mechanical properties compared to the typical polymers used as the matrix of the composites. The higher is the ionic conductivity, the lower is the modulus of elasticity [54]. Thus, the presence of the electrolyte polymer as at least a part of the matrix of the composite tends to reduce the mechanical performance of the composite.

consolidation of fiber laminae, with the fibers in each lamina being either unidirectional or woven and the number of fibers stacked along the thickness of each lamina typically ranging from 25 to 50), and (ii) multidirectionally wound fiber tubings (made by filament winding). For high-performance structural composite panels, non-woven fibers are used and the fiber orientation is commonly the same for the fibers in the same lamina, but differ for the fibers in different laminae (Fig. 1). In any of these forms, the composite is highly anisotropic, with the strength, modulus, electrical conductivity and thermal conductivity being all much higher in the fiber direction of the composite than the other directions. The interlaminar interface (i.e., the interface between the adjacent laminae) is the mechanically weak link in the composite. Delamination is the most common form of damage in the composite. With the fiber diameter typically of the order of 10 µm, the interlaminar interface region has a typical thickness of the order of 10 µm (Fig. 2), although this thickness has a substantial range, due to the variable thickness of the resin layer on the surface of the prepreg used in fabricating the composite. Thus, the interlaminar interface region is a polymer-rich region (Fig. 2), which is the region that is commonly used for positioning additional constituents for the purpose of modifying the composite. Ideally, the mechanical performance of a structural material is not reduced by the modification needed to render the capacitor function. In case of a structural material in the form of a continuous fiber polymermatrix composite, the mechanical strength and elastic modulus are governed by the fiber volume fraction. In order to maximize the strength and modulus, the fiber volume fraction is essentially maximized (typically up to about 60%, as limited by the packing geometry). To maintain high strength and modulus, the modification used to render the capacitor function to a structural composite should ideally not decrease the fiber volume fraction. As the modification involves the addition of one or more constituents to the composite and the added constituents occupy non-zero volumes, the maintaining of the fiber volume fraction is challenging. In case that the continuous fibers are electronically conductive (whether the conduction is due to electrons or holes) and the fiber

3. Design principles of capacitors The capacitance C of a parallel-plate capacitor is given by (1)

C = ε0 κA/ l −12

where ε0 is the permittivity of free space (8.85 × 10 F/m), κ is the relative permittivity, A is the area of the capacitor, and l is the thickness of the dielectric layer in the capacitor. Due to Eq. (1), a large value of A is attractive for attaining a high capacitance C. Because of the large area of a structure, particularly one in the form of a panel, a structural capacitor is attractively associated with a large value of A. The electrical energy E stored in a capacitor subjected to an AC voltage V is given by

E = (1/2) CV 2,

(2)

where C is the capacitance of the capacitor. The maximum energy stored corresponds to the energy E for the case of V being the amplitude of the AC voltage. Eq. (2) means that the capacitance C governs the energy E. In addition, Eq. (2) indicates that the ability to sustain a high voltage, i.e., a high electric field, as enabled by a high dielectric strength, contributes to providing a high value of the energy E. Glass fiber fabric is an attractive dielectric material for providing a relatively high value of the dielectric strength [55,56]. It is well-known that capacitors can be connected in series or in 91

Applied Energy 231 (2018) 89–101

D.D.L. Chung

copper yttrium-iron-garnet hybrid polyphenylene-sulfide-matrix composites [74] and other materials. Optical, electromagnetic and magnetic applications are relevant. Eqs. (3) and (4) also apply to the case of C1 and C2 being opposite in sign (i.e., κ1 and κ2 being opposite in sign). In accordance with Eq. (3), C is infinity when C1 and C2 are opposite in sign but equal in magnitude. Indeed, by studying a positive-permittivity carbon fiber polymer-matrix composite and a negative-permittivity carbon fiber composite that are stacked (hence in series) (Fig. 4(c)), it has been shown that, when C1 and C2 are opposite in sign but similar (not exactly equal) in magnitude, the series capacitance per unit geometric area is as high as 370 μF/m2 [75]. As a consequence of Eq. (3), under the condition that v1/κ1 and v2/κ2 are equal in magnitude but opposite in sign, 1/κ is zero, which means that κ is infinity. With v1/κ1 and v2/κ2 being opposite in sign but similar (not exactly equal) in magnitude for the positive-permittivity composite and negative-permittivity composite, the resulting series κ is as high as 78,000 [75]. The series capacitance or permittivity thus obtained are exceptional high in magnitude. The permittivity of polymer-matrix composites such as continuous carbon fiber polymer-matrix composites has been studied by numerous workers at radio wave and microwave frequencies, due to the relevance to electromagnetic interference shielding and radar radiation absorption [76–88]. However, this frequency range is considerably higher than that which is relevant to typical capacitors. Furthermore, the electromagnetic wave propagation involves the mechanism known as radiation (involving a transmitter and a receiver, as in broadcasting) at radio wave and microwave frequencies, but involves the mechanism known as conduction (involving a physical connection between the adjacent devices) at lower frequencies. A dielectric capacitor has the conventional parallel-plate capacitor geometry, in which a dielectric material is sandwiched by conducting plates (electrodes). The polarization of the dielectric material is mainly responsible for the capacitance. The higher the relative permittivity (also known as the relative dielectric constant) of the dielectric material, the greater is the polarization and hence the higher is the capacitance. Barium titanate (BaTiO3) and related materials are well-known for the high values of their relative permittivity. In order for the capacitor to be low in the energy loss, which is commonly expressed in terms of the loss tangent (tan δ), the dielectric material should be a good electrical insulator (i.e., high in the electrical resistivity) and should exhibit low dipole friction (i.e., easy flipping of the direction of the polarization in response to the AC electric field applied between the two electrodes). In general, the dipole friction increases with increasing frequency of the AC electric field, so that the relative permittivity and capacitance decrease with increasing frequency. A supercapacitor is an electrochemical device that involves two electrodes (electronic conductors that are connected to the outer circuit) that are in contact with an electrolyte (an ionic conductor but an electronic insulator), with the two electrodes separated by a sheet known as a separator (a membrane that electrically separates the two electrodes, so that the electrodes do not short circuit, in addition to allowing certain ions to go through it so that it does not cause an open circuit to the system). The separator is preferably thin, for the purpose of increasing the energy density of the supercapacitor. The two electrodes are usually the same in composition. For example, both electrodes are carbon or graphite. The electrolyte contains ions. For a conventional non-structural supercapacitor, the electrolyte is in liquid or gel form. However, for a structural supercapacitor, the electrolyte is in solid form. At a given instant during the AC electric field application, one electrode is electrically positive while the other electrode is electrically negative. As the polarity of the AC field switches, the electrodes also switch sign. The positive ions are attracted to the electrode that is electrically negative, while the negative ions are attracted to the electrode that is electrically positive. Thus, there is a layer of positive ions on the surface of the negative electrode, thus providing two layers of opposite charges on the electrode surface. These two layers (known as

parallel. In the case of the series connection, the capacitance C of the series combination is given by the equation

1/ C = 1/ C1 + 1/ C2 + ⋯+1/ Cn

(3)

where C1, C2, … Cn are the capacitances of the n capacitors involved. In terms of the relative permittivity, Eq. (3) becomes

1/ κ = v1/ κ1 + v2/ κ2 + ...+vn/ κn

(4)

where κ, κ1, …, κn are the values of the relative permittivity of the series combination, capacitor 1, …, and capacitor n, respectively, and v1 and v2, …, νn are the volume fractions of capacitors 1 2, … , and n, respectively [57]. A parallel-plate capacitor in the form of two electrodes sandwiching a dielectric layer actually consists of three capacitors in series, namely the capacitance Cv of the volume of the dielectric material and the capacitance Ci of each of the two interfaces between the dielectric layer and the two electrodes. Hence, the capacitance C of the overall capacitor is given by

1/ C = 1/ Cv + 2/ Ci

(5)

The lower is Ci, the higher is 2/Ci, and the greater is the influence of Ci on C. Hence, a high value of Ci indicates of a good interface. In general, the interface is not perfect and hence Ci is not equal to infinity, so that the last term in Eq. (5) cannot be neglected. Unfortunately, the last term of Eq. (5) is often assumed to be zero. The last term being not zero means that

C < Cv ,

(6)

so that the capacitance of the capacitor is less than what one expects based on calculation using Eq. (1). The decoupling of Cv and Ci can be achieved by measuring C for three or more thicknesses of the dielectric layer and plotting 1/C against the thickness l of the dielectric layer, as illustrated in Fig. 3 [58–68]. The plot is a straight line. The intercept of the line with the 1/ C axis at l = 0 is equal to 2/Ci. In accordance with Eq. (1),

1/ Cv = l/(ε0 κA)

(7)

Thus, the slope of the line in Fig. 3 is equal to 1/(ε0 κA). Hence, κ (a property of the dielectric material) can be obtained from this slope. The decoupling of the volumetric and interfacial contributions to the capacitance is important for understanding the science of the capacitor behavior. The relative permittivity is normally positive (Fig. 4(a)), but negative permittivity (Fig. 4(b)) has been reported in special cases. For example, negative permittivity has been reported in graphene with magnetic nanoparticles [69], multiwalled carbon nanotube polyanilinematrix composites [70], Fe3O4 polyaniline-matrix composites [71], nickel-alumina meta-composites [72], perovskite La1-xSrxMnO3 [73],

Fig. 3. Schematic plot of 1/C vs. l, for the determination of Ci and κ based on Eqs. (4) and (6). The slope equals 1/(ε0κA), where κ is the relative permittivity of the dielectric material, ε0 is the permittivity of free space, l is the thickness of the dielectric layer and A is the area of the dielectric layer. The intercept on the vertical axis at l = 0 equals 2/Ci. 92

Applied Energy 231 (2018) 89–101

D.D.L. Chung

+

+

+

-

+

-

+

-

-

-

(a)

(b)

+ +

Fig. 4. Illustration of (a) positive permittivity, (b) negative permittivity, and (c) positive and negative permittivity components in series electrically. An electric dipole is indicated by + and − , each inside a circle, that are connected by a vertical line. The + and – signs without encasing circles indicate the applied voltage polarity.

(c) loss in mechanical properties, the fiber volume fraction must remain high. The activation of the fibers, as conducted to increase the fiber surface area [42,43], tends to degrade the mechanical properties of the fibers.

the electric double layer) are separate from each other by atomic distances, i.e., of the order of Angstroms. Since the capacitance C is inversely related to the distance between the charges of opposite sign (Eq. (1)), the small distance means a high capacitance. Similarly, a layer of negative ions are on the surface of the positive electrode, thus providing another capacitor on this surface. Hence, the overall capacitor consists of two capacitors in series (Eq. (3)). Further increase of the capacitance of a supercapacitor is obtained by increasing the surface area of each electrode (e.g., by surface reaction, nanofiber growth, etc.) and/or adding an electronic conductor in the form of particles (e.g., activated carbon, carbon black, etc.) or discontinuous fibers (e.g., short carbon fibers, carbon nanofibers, carbon nanotubes, etc.) of large specific surface area to the electrolyte, so that the electronic conductor units touch one another. The collection of these units located in the electrolyte on the side of the separator near the positive electrode touches this electrode, thus becoming positive. The collection of these units located in the electrolyte on the side of the separator near the negative electrode touches this electrode, thus becoming negative. Hence, the presence of the conductive units in the electrolyte effectively increases the area of each electrode greatly. However, these additional conductive units are not always used in a structural supercapacitor, due to the need to maintain the continuous fibers (which bear the load) at a high volume fraction. As the AC electric field changes polarity, the ions in the electrolyte move in the opposite direction. When the AC frequency is high, the ion movement may not respond fast enough, thus resulting in a decrease in the capacitance. Thus, supercapacitors are limited to operation at relatively low frequencies, such as frequencies at or below the kHz range. In contrast, dielectric capacitors can operate at much higher frequencies. High frequencies are necessary for high-speed computers and other electronics. A dielectric capacitor requires a dielectric material sandwiched by electrodes that are electronic conductors. No electrolyte is involved, so safety and long-term charge-discharge cycling are not issues. Compared to supercapacitors and batteries, which require electrodes and electrolytes, dielectric capacitors are attractive for their safety, long-term cyclability, structural simplicity and high frequency capability. The cyclability limit makes structural supercapacitors not able to function for the full service life of the associated structure. Batteries also suffer from their limited lifetime, due to the eventual depletion of the electrochemically active species. This lifetime limit makes structural batteries [89–96] not able to function for the full service life of the associated structure. In contrast, structural dielectric capacitors are expected to be able to function for the full service life of the associated structure. The abovementioned structural simplicity of dielectric capacitors results in relatively low tendency for the capacitor structure to degrade the mechanical properties of the structural composite. In order to avoid

4. Design principles of structural capacitors 4.1. Structural capacitors in the form of carbon fiber polymer-matrix composites Due to the electronic nature of the electrical conductivity of carbon fibers, continuous carbon fiber polymer-matrix composites can be rendered the dielectric capacitor function by utilizing the fibers as the electrode. The carbon fibers in a fiber laminate always have a degree of contact in the through-thickness direction. The contact partly stems from the slight waviness (known as marcelling) of the fibers in the composite and occurs both within a lamina and across the interlaminar interface [51]. It also partly stems from the flow of the polymer matrix precursor during composite fabrication, and the resulting occurrence of parts of a fiber that are not covered by the precursor. This flow is particularly significant when the precursor is low in viscosity, as in the case of epoxy resin as the precursor. A fiber is anisotropic in both mechanical and functional properties, due to the preferred orientation of the structural units of the fiber along the fiber axis. For example, in case of a carbon fiber, the carbon layers in the fiber are preferentially oriented along the fiber axis [97]. The preferred orientation is known as texture. This type of preferred orientation is known as the fiber texture. Due to the anisotropy of a fiber, the waviness is not attractive for the mechanical performance of the composite. As a result of the fiber-fiber contact in the through-thickness direction (the direction perpendicular to the plane of the laminae), the electrical conductivity in this direction of a composite laminate is never zero [98,99], even in the absence of delamination or other defects. This occurs in spite of the insulating nature of the polymer matrix. Therefore, in order to render the dielectric capacitor function to a carbon fiber polymer-matrix composite, a dielectric film needs to be positioned between two adjacent laminae in the laminate during the composite fabrication. 4.2. Structural capacitor configurations There are four configurations of a structural capacitor, labeled configurations A, B, C and D (Fig. 5). All four configurations are in the form of a composite consisting of electrodes and dielectric layers. An electrode can be a conductive structural material, such as a carbon fiber polymer-matrix composite lamina. A dielectric layer can be a dielectric structural material, such as a glass fiber polymer-matrix composite lamina. Alternatively, the dielectric layer can be a polymer film. 93

Applied Energy 231 (2018) 89–101

D.D.L. Chung

conductive, such as glass fiber composites. For this configuration, the electrode material can be a metal foil, a metal-coated polymer film, a carbon fiber fabric, etc. Configuration A (Fig. 5(a)) has a dielectric film positioned between two electrode layers, such that the top and bottom surfaces of the composite are connected to the opposite polarities of the AC electric field source. The presence of the dielectric film causes the part of the laminate above the film to be an electrode and the part of the laminate below the film to be the other electrode, so that the composite is a single capacitor. Configuration B (Fig. 5(B)) has a dielectric film positioned between every two adjacent electrode layers, such that the top and bottom surfaces of the composite are connected to the opposite polarities of the AC electric field source. The composite is the combination of a number of capacitors, with each capacitor associated with one dielectric film. The combination is capacitors in series. According to Eq. (3), the series configuration of multiple capacitances causes 1/C to be higher than the case of a single capacitor, so that C is lower. Thus, the use of capacitors in series is not attractive for attaining a high capacitance, unless the two capacitors have capacitances that are opposite in sign, i.e., antiferroelectric coupling between the two capacitors (Fig. 4(c)) [75]. Configuration C (Fig. 5(c)) has a dielectric film positioned between every two adjacent electrode layers, such that alternate electrodes in the stack are connected to opposite polarities of the AC electric field source [100]. Thus, the combination of capacitors is capacitors in parallel. The capacitance C of a parallel combination of capacitors C1, C2, …, and Cn is given by

+ -

(a)

+

-

(b)

C = C1 + C2 + …+Cn

According to Eq. (8), configuration C (Fig. 5(c)) provides a high capacitance that is the sum of the capacitances of the various capacitors in the combination. In case that all the capacitors in the parallel combination are equal in the capacitance, which is referred to as Cj, Eq. (8) becomes

+ -

C = nCj,

+

-

(c)

(8)

(9)

In contrast, for configuration A,

C = Cj

(10)

For configuration B in the case that all the capacitors in the series combination are equal in the capacitance, which is referred to as Cj, Eq. (3) becomes,

+

C = Cj / n.

(11)

Hence, configuration C (Fig. 5(c)) is more attractive than both configuration B (Fig. 5(b)) and configuration A (Fig. 5(a)) for providing a high capacitance. Configuration C is also attractive for its suitability for realization in a lamellar structural composite material. A configuration not shown in Fig. 5 involves the stack of alternating electrode and dielectric layers being rolled together to form a cylinder. Preferably, the rolling involves winding on a mandrel, as in the composite fabrication process known as filament winding. However, the winding needs to be designed so that the electrical connection to the electrodes is feasible. The cylindrical shape is suitable for a composite structure that is in the form of a cylinder or pipe, such as a truss. However, the number of layers in the cylinder tends to be more limited than that in configuration C. Configuration D (Fig. 5(d)) has two coplanar electrodes on the surface of a dielectric layer. The two electrodes are separated by a gap. The smaller the gap, the higher is the capacitance, which is in the inplane direction. In contrast, the capacitance is in the through-thickness direction for configurations A, B and C. The area of the in-plane capacitor of configuration D relates to the product of the dimension of the electrode in the direction perpendicular to the page in Fig. 5(c) and the thickness of the dielectric layer. However, in case that the current does not penetrate the entire thickness of the dielectric layer, the area of the

(d) Fig. 5. Four configurations of a structural capacitor that is in the form of a composite consisting of electrode and dielectric layers. An electrode can be a conductive structural material, such as a carbon fiber polymer-matrix composite lamina. A dielectric layer can be a dielectric structural material, such as a glass fiber polymer-matrix composite lamina. (a) Configuration A with a dielectric film positioned between two electrode layers, such that the top and bottom surfaces of the composite are connected to the opposite polarities of the AC electric field source. (b) Configuratino B with a dielectric film positioned between every two adjacement electrode layers, such that the top and bottom surfaces of the composite are connected to the opposite polarities of the AC electric field source. (c) Configuration C with a dielectric film positioned between every two adjacent electrode layers, such that alternate electrodes in the stack are connected to opposite polarities of the AC electric field source. (d) Configuration D with two coplanar electrodes on the surface of a dielectric layer.

Configurations A, B and C are suitable for composite materials that are conductive, such as carbon fiber composites. For these configurations, the dielectric layer can be a polymer film, a glass fiber fabric, etc. Configuration D is suitable for composite materials that are non94

Applied Energy 231 (2018) 89–101

D.D.L. Chung

in-plane capacitor is smaller that that indicated by the abovementioned product. Thus, the area of the in-plane capacitor tends to be small compared to that of the through-thickness capacitor of configuration A, B or C. As a result of the small area and large gap distance in the inplane capacitor of configuration D, the capacitance is not high for configuration D. Configuration C requires that an electrical contact is applied to each electrode layer in the stack. This requirement complicates the implementation, as each electrode layer needs to protrude from the stack. In contrast, electrical contacts need to be applied to only the top and bottom surfaces in configurations A and B. It is more convenient to apply contacts to an exterior surface than applying contacts to an interior electrode. The capacitance per unit area of the composite panel is a suitable quantity used to describe the extent of energy storage. This quantity refers to the capacitance per unit area, regardless of the number of laminae in the composite. The capacitance per unit area reported in this paper corresponds that for configuration A. The capacitance per unit area of the composite panel is to be distinguished from the capacitance per unit area (known as the specific capacitance or electrochemical capacitance) of an electrode material that is intended for use in a supercapacitor, as evaluated by cyclic voltammetry. According to Eqs. (9) and (10), the capacitance per unit area for configuration C is n times that for configuration A, if all the n capacitors in the combination in configuration C are of the same capacitance. The capacitance per unit mass is another common quantity used to describe the extent of energy storage. However, for a composite panel, the capacitance per unit area is a more suitable quantity. It is not necessary for the film to be positioned between every two adjacent laminae. The presence of the dielectric at every interlaminar interface would result in a number of capacitors in series. Due to Eq. (1), the thickness of the dielectric film should be minimized for the purpose of obtaining a high capacitance. Ideally, the thickness of the dielectric film is small enough that the film resides in the interlaminar interface region without increasing the thickness of this region. An increase of the thickness of the interlaminar interface region would cause a decrease in the volume fraction of the continuous fibers, thereby reducing the modulus and strength of the composite. The dielectric film is preferably porous, so that the precursor (e.g., the resin) of the polymer matrix of the composite can penetrate the pores in the film. The penetration causes mechanical interlocking between the film and the matrix, thus enhancing the bond between the film and the matrix and reducing the chance of delamination at the part of the interlaminar interface that contains the dielectric film. The pore size of the dielectric film should be large enough for the polymer matrix precursor to be able to penetrate. The smaller the pores, the higher is the pressure needed to force the precursor to enter the pores. A higher pressure means a higher cost of fabrication. On the other hand, the pores must be small enough to avoid the fibers belonging to different electrodes from getting through the pores. The ease of penetration of the pores of the dielectric film by the polymer matrix precursor also depends on the viscosity of the precursor. The higher the viscosity, the more difficult is the penetration. In case of epoxy as the polymer matrix, the precursor is the epoxy resin, which is relatively low in viscosity. In case of a thermoplastic polymer (e.g., a polyamide such as Nylon) as the matrix, the temperature needs to be above the melting temperature of the polymer in order for the viscosity to be sufficiently low. Although the thermoplastic polymer flows above the glass transition temperature (which is below the melting temperature), the viscosity is usually not low enough at temperatures between the glass transition temperature and the melting temperature [101–103]. Surface treatment of the dielectric film can be used to enhance the bond between the film and the polymer matrix of the composite. An example of a surface treatment involves the use of plasma, as performed on a dielectric film in the form of polyethylene terephthalate (PET)

Fig. 6. A two-dimensional array of capacitors in parallel, as obtained by using two fiber laminae with orthogonal fiber directions. The intersection (overlap) of a fiber group of one lamina and a fiber group of the other lamina gives a capacitor in the array. The fibers within a group are electrically shorted together at each of their two ends. The orthogonal fiber groups serve as an x-y grid of electrical interconnections for applying an AC electric field to a chosen capacitor in the array.

[39]. 4.3. Two-dimensional array of capacitors By using two adjacent laminae that have fibers in perpendicular directions, one obtains a two-dimensional array of interlaminar interfaces [104,105], and hence a two-dimensional array of capacitors (Fig. 6). Each capacitor is formed by the intersection (overlap) of a fiber group of one lamina and a fiber group of the other lamina. The fibers within each group are electrically shorted together at each of their two ends. The orthogonal array of fiber groups also serves as an x-y grid of electrical interconnections for applying an AC electric field to the capacitor that is at the intersection of the particular set of orthogonal fiber groups. By applying an AC electric field to multiple capacitors in the array using multiple sets of intersecting orthogonal fiber groups, one obtains capacitors in parallel (Eq. (4)). Hence, by choosing the number of capacitors to receive the electric field, C can be tuned. The larger is the number of capacitors chosen, the higher is the resulting parallel capacitance. In addition, the spatial distribution of the capacitors can be tuned by the choice of capacitors in the array. Although a two-dimensional array of interlaminar interfaces has been used for two-dimensionally spatially resolved temperature sensing [104,105], it has not yet been used for providing the abovementioned capacitor array. 4.4. Electrical contacts The design must include consideration of the configuration, positions, dimensions and material of the electrical contacts in relation to the size and shape of the structure. Examples of configurations are shown in Fig. 5. Even in the case that an electrically conductive constituent (e.g., carbon fiber) in the structural composite serves as electrodes for the capacitor, electrical contacts to the structural composite are still needed in order to apply an AC electric field (i.e., voltage) to the capacitor and to measure the capacitance. In this connection, the contribution of the electrical contacts to the measured capacitance should be considered. The electrical contact contributes due to the interface between the contact material and the surface of the structure. 95

Applied Energy 231 (2018) 89–101

D.D.L. Chung

composite, if these pores are not completely infiltrated by the polymer matrix. Glass fibers are higher in density and lower in modulus than carbon fibers, but they are lower in cost. Glass fiber polymer-matrix composites can be rendered the capacitor function by sandwiching the composite with carbon fiber polymer-matrix composite electrodes or metallized polymer film electrodes. The use of a glass fiber fabric (13 µm) as the dielectric film between carbon fiber epoxy-matrix composite electrodes, as fabricated by consolidation of the film with the carbon fiber epoxy prepreg, has been reported as a structural capacitor [114]. However, the capacitance of the resulting capacitor has not been reported. The thickness of a glass fiber fabric is typically large compared to that of a polymer film. For example, it is 108 µm in a structural capacitor using glass fiber fabric as the dielectric material and metallized polymer film as the electrodes [115]. With a glass fiber epoxy-matrix composite (thickness 108 µm) and each electrode in the form of a polyimide film (12.7 µm thick) coated with aluminum (100 nm thick) on one side, the capacitance obtained is 482 nF/m2 at 100 Hz [100,115,116]. Although barium titanate exhibits a value of the relative permittivity, the addition of barium titanate particles to the writing paper to form a composite dielectric film between carbon fiber epoxy-matrix laminae results in capacitance decreasing from 1.23 to 1.06 µF/m2. This negative effect of barium titanate is because of the increase in thickness of the dielectric film from 40 to 130 µm upon addition of the barium titanate [38]. The modification of a continuous carbon fiber by coating with barium strontium titanate and further coating the barium strontium titanate layer with gold results in a coaxial capacitor in the form of a fiber [117]. In this fiber, the barium strontium titanate layer serves as the dielectric material, while the carbon fiber and gold layer serve as the electrodes. The high porosity of the barium strontium titanate layer is expected to hinder the practical use of this fiber in a structure.

Since the electrical contacts are electrically in series with the capacitor, in accordance with Eq. (5), the capacitance of this interface should be high, so that this interface does not contribute much to the measured capacitance. In addition, the electrical resistance of the electrical contacts should be small, as this resistance is associated with a voltage, thus decreasing the available voltage across the capacitor itself. The larger is the contact area, the smaller is the contact resistance [106]. Therefore, the size of the contact should be large enough. Conductive adhesives may be used to form electrical contacts with polymer-matrix structural composites [106,107]. If the thickness of the polymer matrix covering the carbon fibers on the surface of the composite is not large (e.g., less than about 20 µm), the surface matrix does not need to be removed prior to the electrical contact application [107]. This is because the carbon fibers in the composite are not perfectly straight [51], so that parts of some fibers reside in the surface matrix. Otherwise, the surface matrix may be removed locally by mechanical means, such as sanding, prior to the electrical contact application at the local regions. In case that the capacitor involves metal electrodes (metal foil or metal coating on a substrate) that are accessible from outside the capacitor, soldering [108] may be used to form electrical contacts with the metal electrodes. Pressure contacts [109] may be used in the absence of solder or conductive adhesive, whether the contact is applied to a metal or a polymer-matrix composite, but they are associated with relatively high electrical resistance and the pressure application complicates the practical implementation. Yet another method of making electrical contact to a polymer-matrix composite is to place the contact material (such as a metal foil) on the surface of the composite during the composite fabrication, so that the contact material becomes adhered to the composite during the composite fabrication [110]. 5. Structural dielectric capacitors This section reviews the development of structural dielectric capacitors. By using cellulosic writing paper (40 µm thick) as the dielectric film at an interlaminar interface in a continuous carbon fiber epoxymatrix composite, Luo and Chung [38] reported for the first time a structural capacitor. This capacitor is a dielectric capacitor in the through-thickness direction with a capacitance of 1.23 and 1.17 µF/m2 at 2 kHz and 2 MHz, respectively. The decrease in capacitance from 2 kHz to 2 MHz is quite small. The curvature of the surface of the carbon fibers in each electrode increases the surface area of each electrode beyond the geometric area, thus contributing to providing a relatively high capacitance [38]. This first report of a structural capacitor [38] was a decade later confirmed by Carlson et al. [39,40], who reported a capacitance of 450 nF/m2 at 0.1 Hz, as obtained using PET of thickness 50 µm as the dielectric film. Other than PET, dielectric polymers used include polyamide and polycarbonate [111]. The mechanical integrity of the structural capacitors has been shown to be adequate [111]. In particular, the capacitance is maintained even in the presence of considerable intralaminar cracking in the carbon fiber composite [112]. In 2018, Chan et al. [113] reported a capacitance of 17 µF/m2 (frequency unspecified) when a graphene oxide layer of thickness 4 µm is used as the dielectric layer, compared to a capacitance of 2.2 µF/m2 (frequency unspecified) when a printing paper of thickness 98 µm is used as the dielectric layer. The high capacitance achieved by using graphene oxide paper as the dielectric layer is attributed to the small thickness of this layer [113]. However, the electrical conductivity of the graphene oxide paper increases with increasing frequency when the frequency exceeds 1 kHz [113]. As a consequence, the capacitor using graphene oxide as the dielectric layer is not effective at frequencies above 1 kHz. In contrast, the capacitor using writing paper as the dielectric layer is effective even at a high frequency of 2 MHz [38]. Furthermore, graphene oxide is sensitive to moisture, which strongly affects the electrical behavior [64]. In addition, the high porosity associated with the oriented pores between the layers in the graphene oxide paper is expected to reduce the strength and modulus of the

6. Structural supercapacitors This section reviews the development of structural supercapacitors. Structural supercapacitors have been made using continuous carbon fibers as the electrodes [41–43,118,119]. A structural supercapacitor consists of continuous carbon fiber laminae as the electrodes, a separator (e.g., a microporous polymer membrane, glass fiber fabric, cellulosic paper, etc.) [120] between the electrodes, and a polymer matrix that is formed from a mixture of the base resin (e.g., vinyl ester resin, epoxy-amine resin, etc.) and the electrolyte (e.g., polyethylene glycol (PEG), lithium salt electrolyte solutions such as 1-M lithium trifluromethanesulfonimide in anhydrous propylene carbonate, and imidazolium-based ionic liquids) [118,121,122]. Because of the difference in electrolyte composition among various supercapacitors reported in the literature, quantitative comparison of the supercapacitor performance in terms of the electrode composition is not feasible. The specific gravimetric capacitance (capacitance per unit mass) is to be distinguished from the specific areal capacitance (capacitance per unit area), which is scientifically more meaningful. Using woven carbon fiber fabric as the electrodes, the obtained specific areal capacitance is 670 nF/m2 for a composite thickness of 90 µm in the absence of a conductive filler [118]. The surface modification of the carbon fibers by oxidation, activation or CuO nanowire growth increases the specific surface area, thereby increasing the capacitance, energy density and power density of the supercapacitor [55,123–125]. Copper oxide is a transition metal oxide that can participate in the oxidation-reduction reaction during capacitor operation. The CuO nanowire growth is particularly effective, giving specific surface area 133 m2/g (compared to 0.589 m2/g for the unmodified carbon fiber fabric), specific gravimetric capacitance 2480 mF/g (compared to 150 mF/g for the unmodified carbon fiber fabric) and power density 4.14 W/kg (compared to 1.03 W/kg for the unmodified carbon fiber fabric) [55,124]. 96

Applied Energy 231 (2018) 89–101

D.D.L. Chung

The coating of the carbon fibers with a conductive material with a high surface area promotes the electrode performance. Examples of such conductive materials are carbon aerogel [55,126] and reduced graphene oxide in combination with a metal organic framework [127]. For example, the carbon aerogel coating increases the specific surface area from 0.209 to 163 m2/g and increases the specific gravimetric capacitance from 0.06 to 14.3F/g [55,126]. A metal organic framework is a hybrid inorganic–organic material that exhibits very high surface area (up to 10,000 m2/g), well-defined pore size (up to 9.8 nm), large pore volume and low density [128]. In the presence of a conductive filler (e.g., carbon nanotubes [129]) in the matrix, the capacitance of a structural supercapacitor is greatly increased. In the absence of a conductive filler, by impregnating the carbon fiber fabric with carbon aerogel, the capacitance is increased by orders of magnitude [55,126,130]. These effects are attributed to the improved electrical connection among the carbon fibers and the improved electrical connection between the fibers and the matrix. In addition, the presence of the carbon aerogel enhances the in-plane shear strength and modulus of the composite. The growth of CuO nanowires, ZnO nanotubes, SnO2 nanorods or carbon nanotubes on the carbon fibers also enhances the surface area, thereby increasing the capacitance [119,131–133]. Discontinuous filamentous carbons, such as carbon nanofibers [134], in place of continuous carbon fibers, result in inferior mechanical properties. Cellulose paper is mechanically not adequate for structural supercapacitors, although it has been reinforced with chopped carbon fibers to alleviate this problem [135]. Conducting polymers such as polyaniline are attractive electrode materials due to their high kinetics of the electrochemical charge-discharge processes. However, due to the volume change during the doping and undoping, conducting polymer electrodes suffer from poor cycle life. The support of the conducting polymer by carbon fibers alleviates this problem [136–138]. The activation of the carbon fibers [139] enhances the specific surface area of the electrode, thereby increasing the capacitance of the resulting structural supercapacitor [56,140–143]. However, the mechanical properties are inferior for activated carbon fibers compared to the corresponding fibers that have not been activated. Unexpectedly, the mechanical properties of the structural supercapacitor (with glass fiber fabric as the separator) are improved by the activation of the carbon fibers. This improvement is attributed to the enhanced bond between the fibers and the polymer matrix due to the activation of the fibers [56]. In order to increase the capacitance through the contribution of the pseudocapacitance, metal oxides are added to the electrodes such that the two electrodes are different in composition. Such metal oxides include NiCo2O4 [144–146], Na2Ti5O11 [147], MnO2 [139,148–150], NiO [151,152], Ni(OH)2 [145,146,153], V2O5 [150] and Co(OH)2 [154].

of the large size (both area and thickness) of a structure enabling a high capacitance, in spite of a low capacitance per unit volume. In relation to the size, a large area helps due to Eq. (1), and a large thickness helps due to Fig. 5(c). The energy storage function should be maintained throughout the service life of the structure. For some structures, such as buildings, the service life can be as long as a century. This demand on the longevity of the energy storage function makes structural dielectric capacitors advantageous over structural supercapacitors. Concerning the structure development, the functional and mechanical aspects of the design of the structure need to be merged [116,155], along with the design of the electrical contacts. In particular, in a structural capacitor, the continuous fiber volume fraction should remain high, as required for a structural composite and as recommended for best combined functional and mechanical performance [116]. A challenge is to achieve a high fiber volume fraction in the presence of a dielectric layer between the fiber laminae. For this purpose, the thickness of the dielectric layer needs to be small. Furthermore, the adhesion of the dielectric layer with the matrix (e.g., epoxy) of the composite needs to be adequate. For example, it can be enhanced by plasma treatment of the dielectric layer [39]. Structure development includes the development of the method of fabrication of the structural capacitor, since the fabrication method affects the structure of the resulting composite. For example, a higher curing temperature during composite fabrication decreases the thickness of the interlaminar interface, thereby increasing the fiber volume fraction [51]. According to Eq. (1), a small decrease in the interlaminar interface thickness can significantly increase the capacitance. However, a small decrease in the interlaminar interface thickness increases the fiber volume fraction slightly, if any, so it increases the mechanical properties only slightly, if any. Another structural aspect pertains to the degree of fiber alignment, i.e., the degree of waviness of the fibers. The degree of fiber alignment is governed by the method of composite fabrication. An increase in the degree of fiber alignment decreases the chance of fiber-fiber contact across the interlaminar interface, thereby reducing the need for a relatively thick dielectric layer at the interface. Due to the strong effect of the composite fabrication method on the structural capacitor performance, the structure development should be performed with the involvement of composite engineers. Structural development should be conducted with inclusion of the electrical contacts in the overall design. The positions, dimensions and quality of the electrical contacts matter and should be considered prior to composite fabrication. For example, the shape and dimensions of the composite structure may be altered in order to accommodate the electrical contacts. Concerning the application development, the structural capacitor needs to be designed and tested for each application, as different applications can differ in the requirements for the energy storage capacity, mechanical performance, size, shape, weight and service life. For example, the service life is shorter and the size is smaller for a computer (for which the case is a structural capacitor) than a car (for which parts of the body are structural capacitors). Application development should be performed with the involvement of the industry associated with the particular application. The relevant industries include the aerospace, automobile, wind turbine, photovoltaic, display and computer industries, as discussed in Section 8.

7. Development challenges The technical challenges relate to the capacitor development, structure development, and application development. Concerning the dielectric capacitor development, the dielectric material can be improved in terms of increasing the permittivity, decreasing the energy loss, and increasing the dielectric strength. Concerning the supercapacitor development, the electrolyte can be improved in terms of increasing the ionic conductivity and improving the mechanical properties, and the supercapacitor can be improved in terms of enhancing the charge-discharge cyclability, safety, and mechanical properties. For both dielectric capacitors and supercapacitors, the energy density needs improvement, although this issue is less serious for supercapacitors than dielectric capacitors. In general, capacitors have high power density but low energy density, whereas batteries have high energy density, but low power density. However, in the context of structural capacitors, the energy density is not a great concern, because

8. Applications The technological needs for structural capacitors in relation to energy storage have been discussed in the Introduction. In particular, energy storage is critical to the viability of renewable energy utilization [156]. Large-scale centralized energy storage, smaller scale grid storage and distributed energy storage are all relevant [157]. In spite of these needs, the commercialization of structural capacitors has not yet occurred. The tardiness of the commercialization is partly because the 97

Applied Energy 231 (2018) 89–101

D.D.L. Chung

than glass fibers, so the use of carbon fiber composites instead of glass fiber composites can reduce the weight of the wind turbine blades. The weight reduction is particularly needed for large blades, which are increasingly common. The support structure of solar panels may be used as structural capacitors. Similarly, the support of outdoor lighting may be used as structural capacitors for storing the energy provided by solar panels. Displays such as light-emitting diode (LED) displays also require electric power and their support structures may be in the form of structural capacitors. Due to the need for light weight (as particularly needed for roof-top solar panels and large displays), the use of fiber composites for solar and LED support structures is additionally attractive. In 2016, about 40% (or about 39 quadrillion British thermal units) of the total U.S. energy consumption was consumed by residential and commercial buildings [165]. There is increasing push for zero-energy buildings, which require the use of renewable energy and the incorporation of energy storage. Currently, the energy storage relies on batteries and/or a central energy storage facility. Structural capacitors in the form of architectural/building components will enhance the energy storage capacity of a building and reduce the need for batteries or central energy storage. For building and infrastructural applications, structural capacitors in the form of cement-matrix composites are relevant. Cement-based structural supercapacitor consisting of two graphene electrodes sandwiching a hardened cement paste sheet that has been infused with a liquid electrolyte in the form of a 1-M KOH solution has been reported [166]. Although the specific capacitance reaches 10 F/g [166], the durability of the graphene electrodes and the service life of the supercapacitor are questionable. The permittivity of cement-based materials depends on the admixtures [66,167,168], defects [169] and aggregate proportions [170]. The capacitance of cement-based materials in the absence of any particular admixture changes with stress, due to piezoelectricity [171,172]. In measuring the capacitance of a cement-based material, the size of the sandwiching electrodes relative to that of the cementbased specimen matters, as relatively small electrodes result in fringing electric field that causes the apparent permittivity to be higher than the true permittivity [173]. Structural vibration control is needed for aircraft, satellites, automobile, turbine blades, bridges, high-rise buildings, high-speed rail, civil structures, and many other structures, since the vibration can reduce the structural performance and safety, and in some cases (as in rail), increase noise. Active vibration control involves the use of sensors and actuators, which need to be powered. Structures in the form of structural capacitors can power such devices, as particularly needed for structures that are mobile or are located in remote areas that are away from the electrical grid.

interdisciplinary subject is a stretch for both structural composite material manufacturers and capacitor manufacturers. Structural capacitors are expected to provide an untapped, extensive, save and distributed means of energy storage. The distributed nature refers to the fact that the energy is not stored in a central stationary location, but is stored in various structures that are located differently and that may be mobile, as in the case of a car. From the viewpoints of structural performance, safety, energy storage service life and high frequency capability, structural dielectric capacitors are closer to commercialization readiness than structural supercapacitors. The safety concern is particularly high for supercapacitors that involve liquid electrolytes [158–160] and remains for supercapacitors that involve an electrolyte in the form of an ionic liquid together with a solid electrolyte [161]. The service life of supercapacitors is limited by the charge-discharge cyclability [162,163]. The service life of a rechargeable structural battery is also limited by the charge-discharge cyclability. For a structural battery that is not rechargable, the service life tends to be even shorter, as it ends when one of the electrochemically active constituents has been consumed. Although structural batteries have been reported [45,46,89–95,164], they are much farther from commercialization readiness than structural dielectric capacitors. Commercialization readiness requires the battery or capacitor function of a structure to be able to last for the service life of the structure (years or decades, depending on the structure), in addition to being safe and being adequate in the mechanical performance. Numerous types of structure can store energy by serving as structural capacitors. They include mobile and immobile structures with a wide size range, which span from buildings to laptop computers. Mobile structures cannot totally rely on energy that is stored in a central location or energy from the electrical grid, so their suitability for the structural capacitor approach is particularly clear. Aircraft, unmanned aerial vehicles, satellites, automobile, ships, laptop computers and cell phones are examples of mobile structures. The structural capacitor approach would reduce the need for batteries, which can also store energy, but suffer from safety concerns (particularly fire safety concerns in case of the lithium-ion battery), limited service life (even for rechargeable batteries) and substantial volume requirement. For example, batteries currently used for electric or hybrid vehicles require such a large volume that the trunk of a car may need to be used largely for the batteries. In contrast, the use of the structure of a car for energy storage will impose essentially no volume requirement. Laptop computers and cell phones currently use batteries for energy, but their use of cases in the form of structural capacitors can enhance safety and reduce the weight compared to the use of batteries. The carbon fiber polymer-matrix composite cases that serve as structural capacitors can serve the additional function of electromagnetic interference (EMI) shielding [76–88], which is needed for both computers (which need protection from EMI) and cell phones (which need to avoid the emission of EMI). Wind turbines, buildings, solar panels, display panels and outdoor lighting are not mobile, but they may be situated in a location where access to the electrical grid or a central energy storage is not feasible or not convenient. Wind turbines and solar panels generate energy intermittently, for which storage is necessary. The wind turbine blades are made of fiber composites that may be used as structural capacitors. Wind turbine blades are commonly made of glass fiber composites rather than carbon fiber composites, due to the low cost of glass fibers compared to carbon fibers. However, carbon fiber prices are decreasing. Moreover, due to the electrical conductivity of carbon fibers, the use of carbon fiber composites for wind turbine blades will enable de-icing and anti-icing by resistance heating, in addition to the use of the turbine blade as a structural capacitor. The removal of ice from wind turbine blades is seriously needed, as the ice on the blades can interfere with the function of the blades. In addition, carbon fibers have lower density

9. Conclusions Structural capacitors are structural materials (commonly polymermatrix structural composites) that have been modified in order to render the capacitor function for the purpose of electrical energy storage. They are a type of multifunctional structural material. This paper reviews the development of structural capacitors, including structural dielectric capacitors and structural supercapacitors, and provides the first enunciation of their engineering design and applications. The modification of a structure to render the capacitor function involves the positioning of a dielectric film between the electrodes, which are an electronic conductor, commonly the continuous carbon fiber laminae that serve to reinforce the composite. The electrode preferably has a high specific surface area, which can be increased by grafting nanofibers or nanotubes on the surface of the electrode material, impregnating the electrode material with a nanomaterial such as carbon aerogel, or activating the electrode material (such as forming activated carbon fibers). 98

Applied Energy 231 (2018) 89–101

D.D.L. Chung

The dielectric film is preferably small in thickness, so that the energy density of the structural capacitor is relatively high. In case of a structural dielectric capacitor, a thin dielectric film is also attractive for providing a high capacitance. For mechanical performance of the composite, the dielectric film should be well bonded to the polymer matrix of the composite. Porosity in the film (which allows mechanical interlocking between the film and the polymer matrix) and surface modification of the film can be used to enhance this bond. The dielectric film serves to avoid short circuiting of the two electrodes that sandwich the film. It is commonly a polymeric film. For a structural dielectric capacitor, the dielectric film should be an electrical insulator that exhibits high electrical resistivity in relation to electronic conduction. For a structural supercapacitor, the dielectric film (known as a separator) should be an ionic conductor and an electronic insulator. The design of a structural capacitor should include consideration of the capacitance of the interface between the dielectric film and electrode. The design should also address the electrical contacts on the structural capacitor. The capacitance of a structural capacitor can be increased significantly by having the capacitor consist of capacitors in parallel. For this purpose, the dielectric film is preferably positioned at every interlaminar interface of the composite, such that alternating electrodes in the stack are connected to opposite polarities of the AC electric field source (configuration C, Fig. 5(c)). In addition to a separator, a structural supercapacitor requires the matrix to be a solid electrolyte, which is typically in the form of a mixture of a conventional polymer (e.g., epoxy) and an ionic species. The electrolyte requirement makes the attainment of high levels of strength, modulus and safety relatively difficult. Capacitors that are in parallel can alternatively be obtained by using a two-dimensional array of capacitors. By using a two-dimensional array of interlaminar interfaces between laminae with fibers in different directions and using the fiber groups in each lamina as electrical interconnections, a two-dimensional array of capacitors in parallel can be obtained, with selected capacitors in the array activated by applying an AC electric field to these capacitors using the electrical interconnections associated with these capacitors.

materials. J Vacuum Sci Technol B 2017;35(4):040801/1–040801/27. [15] Ding Y, Yu G. The promise of environmentally benign redox low batteries by molecular engineering. Angew Chem Int Ed 2017;56(30):8614. [16] Weber AZ, Van Nguyen T. Redox flow batteries-reversible fuel cells. J Electrochem Soc 2016;163(1):Y1. [17] Soloveichik GL. Flow batteries: current status and trends. Chem Rev (Washington, DC, United States) 2015;115(20):11533–58. [18] Li Z, Pan MS, Su L, Tsai P, Badel AF, Valle JM, et al. Air-breathing aqueous sulfur flow battery for ultralow-cost long-duration electrical storage. Joule 2017;1(2):306–27. [19] Ramachandran R, Chen S, Kumar GG. An overview of electrochemical energy storage devices of various electrodes and morphological studies of supercapacitors. Int J Electrochem Sci 2015;10(12):10355–88. [20] Vatamanu J, Bedrov D. Capacitive energy storage: current and future challenges. J Phys Chem Lett 2015;6(18):3594–609. [21] Farhadi M, Mohammed O. Energy storage technologies for high-power applications. IEEE Trans Ind Appl 2016;52(3):1953–61. [22] Werfel FN, Floegel-Delor U, Rothfeld R, Riedel T, Goebel B, Wippich D, Schirrmeister P. Superconductor bearings, flywheels and transportation. Superconductor Sci Technol 2012;25(1):014007/1–014007/16. [23] Najjar YSH, Abubaker AM. Using novel compressed-air energy storage systems as a green strategy in sustainable power generation-a review. Int J Energy Res 2016;40(12):1595–610. [24] Crotogino F, Schneider G, Evans DJ. Renewable energy storage in geological formations. Proc Inst Mech Eng, Part A: J Power Energy 2018;232(1):100–14. [25] Liu W, Ramirez A. State of the art review of the environmental assessment and risks of underground geo-energy resources exploitation. Renew Sustain Energy Rev 2017;76:628–44. [26] Vo TTQ, Xia A, Rogan F, Wall DM, Murphy JD. Sustainability assessment of largescale storage technologies for surplus electricity using group multi-criteria decision analysis. Clean Technol Environ Policy 2017;19(3):689–703. [27] Manfrida G, Secchi R. Performance prediction of a small-size adiabatic compressed-air energy storage system. Int J Thermodyn 2015;18(2):111–9. [28] Reay DA. Thermal energy storage: the role of the heat pipe in performance enhancement. Int J Low-Carbon Technol 2015;10(2):99–109. [29] Lefebvre D, Tezel FH. A review of energy storage technologies with a focus on adsorption thermal energy storage processes for heating applications. Renew Sustain Energy Rev 2017;67:116–25. [30] Pereira da Cunha J, Eames P. Thermal energy storage for low and medium temperature applications using phase change materials – a review. Appl Energy 2016;177:227–38. [31] Pfleger N, Bauer T, Martin C, Eck M, Woerner A. Thermal energy storage – overview and specific insight into nitrate salts for sensible and latent heat storage. Beilstein J Nanotechnol 2015;6:1487–97. [32] Laughlin Robert B. Pumped thermal grid storage with heat exchange. J Renew Sustain Energy 2017;9(4):044103/1–044103/23. [33] Strang KD. Innovative closed-loop underground pumped hydroelectricity geothermal storage system. Int J Environ Stud 2017;74(3):363–85. [34] Wang D, Chung DDL. Through-thickness piezoresistivity in a carbon fiber polymermatrix structural composite for electrical-resistance-based through-thickness strain sensing. Carbon 2013;60(1):129–38. [35] Chung DDL. Damage detection using self-sensing concepts. J Aerospace Eng (Proc Inst Mech Eng, Part G) 2007;221(G4):509–20. [36] Han S, Chung DDL. Through-thickness thermoelectric power of a carbon fiber/ epoxy composite and decoupled contributions from a lamina and an interlaminar interface. Carbon 2013;52:30–9. [37] Han S, Chung DDL. Carbon fiber polymer-matrix structural composites exhibiting greatly enhanced through-thickness thermoelectric figure of merit. Composites Part A 2013;48:162–70. [38] Luo X, Chung DDL. Carbon fiber polymer-matrix composites as capacitors. Compos Sci Technol 2001;61:885–8. [39] Carlson T, Asp LE. Structural carbon fibre composite/PET capacitors – effects of dielectric separator thickness. Composites Part B 2013;49:16–21. [40] Carlson T, Ordeus D, Wysocki M, Asp LE. Structural capacitor materials made from carbon fibre epoxy composites. Compos Sci Technol 2010;70(7):1135–40. [41] Jiang Q, Yang R, Fu GG, Xie DY, Huang B, He ZW, et al. Preparation of the carbon nanotube/carbon fiber composite and application as the electrode material of the electrochemical super capacitor. Mater Sci Forum 2011;687:158–62. [42] Salinas-Torres D, Sieben JM, Lozano-Castello D, Cazorla-Amoros D, Morallon E. Asymmetric hybrid capacitors based on activated carbon and activated carbon fibre-PANI electrodes. Electrochim Acta 2013;89:326–33. [43] Jin Z, Tian Y, Su LJ, Qin CL, Zhao DY, Li RQ, et al. Hybrid supercapacitors based on polyaniline/activated carbon fiber composite electrode materials. Adv Mater Res (Durnten-Zurich, Switzerland) 2013;800:505–8. [44] Asp LE. Multifunctional composite materials for energy storage in structural load paths. Plastics, Rubber Compos 2013;42:144–9. [45] Yu Y, Zhang B, Feng M, Qi G, Tian F, Feng Q, Yang J, Wang S. Multifunctional structural lithium ion batteries based on carbon fiber reinforced plastic composites. Compos Sci Technol 2017;147:62–70. [46] Park MY, Kim J, Kim DK, Kim CG. Perspective on carbon fiber woven fabric electrodes for structural batteries. Fibers Polym 2018;19(3):599–606. [47] Pereira T, Guo Z, Nieh S, Arias J, Hahn HT. Energy storage structural composites: a review. J. Compos. Mater. 2009;43:549–60. [48] Han S, Chung DDL. Increasing the through-thickness thermal conductivity of carbon fiber polymer-matrix composite by curing pressure increase and filler incorporation. Compos Sci Technol 2011;71:1944–52. [49] Chung DDL. Self-heating structural materials. Smart Mater Struct 2004;13(3):562–5. [50] Han S, Chung DDL. Mechanical energy dissipation using carbon fiber polymermatrix structural composites with filler incorporation. J Mater Sci

References [1] http://energystorage.org/energy-storage/facts-figures (as viewed on Sept. 8, 2018). [2] Gallo AB, Simoes-Moreira JR, Costa HKM, Santos MM, Moutinho dos Santos E. Energy storage in the energy transition context: a technology review. Renew Sustain Energy Rev 2016;65:800–22. [3] Ould Amrouche S, Rekioua D, Rekioua T, Bacha S. Overview of energy storage in renewable energy systems. Int J Hydrogen Energy 2016;41(45):20914–27. [4] Zhou Z, Benbouzid M, Charpentier JF, Scuiller F, Tang T. A review of energy storage technologies for marine current energy systems. Renew Sustain Energy Rev 2013;18:390–400. [5] Shakeri M, Soltanzadeh M, Berson RE, Sharp MK. Efficiency of solar electricity production with long-term storage. J. Solar Energy Eng. 2015;137(1):011007/1–9. [6] Barnhart CJ, Benson SM. On the importance of reducing the energetic and material demands of electrical energy storage. Energy Environ Sci 2013;6(4):1083–92. [7] Zhang C, Wei Y, Cao P, Lin M. Energy storage system: current studies on batteries and power condition system. Renew Sustain Energy Rev 2018;82(Part_3):3091–106. [8] Wang Y, Zhang B, Lai W, Xu Y, Chou S, Liu H, et al. Room-temperature sodiumsulfur batteries: a comprehensive review on research progress and cell chemistry. Adv Energy Mater 2017;7(24):1602829. [9] Lin X, Salari M, Arava LMR, Ajayan PM, Grinstaff MW. High temperature electrical energy storage: advances, challenges, and frontiers. Chem Soc Rev 2016;45(21):5848–87. [10] Abraham A, Housel LM, Lininger CN, Bock DC, Jou J, Wang F, et al. Investigating the complex chemistry of functional energy storage systems: the need for an integrative, multiscale (molecular to mesoscale) perspective. ACS Central Sci 2016;2(6):380–7. [11] Chen K, Xue D. Materials chemistry toward electrochemical energy storage. J Mater Chem A 2016;4(20):7522–37. [12] Dinesh A, Olivera S, Venkatesh K, Santosh K, Priya MS, Geetha Murugesan, et al. Iron-based flow batteries to store renewable energies. Environ Chem Lett 2018;16(3):683–94. [13] Kamat PV, Schanze KS, Buriak JM. Redox flow batteries. ACS Energy Lett 2017;2(6):1368–9. [14] Qi Z, Koenig GM. Review article: flow battery systems with solid electroactive

99

Applied Energy 231 (2018) 89–101

D.D.L. Chung

2012;47:2434–53. [51] Chung DDL. Processing-structure-property relationships of continuous carbon fiber polymer-matrix composites. Mater Sci Eng R 2017;113:1–29. [52] Han S, Chung DDL. Strengthening and stiffening carbon fiber epoxy composites by halloysite nanotubes, carbon nanotubes and silicon carbide whiskers. Appl Clay Sci 2013;83–84:375–82. [53] Han S, Chung DDL. Increasing the through-thickness thermal conductivity of carbon fiber polymer-matrix composite by curing pressure increase and filler incorporation. Compos Sci Technol 2011;71(16):1944–52. [54] Asp LE, Greenhalgh ES. Structural power composites. Compos Sci Technol 2014;101:41–61. [55] Chan K, Jia B, Lin H, Hameed N, Lee J, Lau K. A critical review on multifunctional composites as structural capacitors for energy storage. Compos Struct 2018;188:126–42. [56] Shirshova N, Qian H, Shaffer MSP, Steinke JHG, Greenhalgh ES, Curtis PT, et al. Structural composite supercapacitors. Composites Part A 2013;46:96–107. [57] Chung DDL. Functional materials. World Scientific; 2010. Ch. 2. [58] Wang A, Chung DDL. Dielectric and electrical conduction behavior of carbon paste electrochemical electrodes, with decoupling of carbon, electrolyte and interface contributions. Carbon 2014;72:135–51. [59] Hong X, Chung DDL. Exfoliated graphite with relative dielectric constant reaching 360, obtained by exfoliation of acid-intercalated graphite flakes without subsequent removal of the residual acidity. Carbon 2015;91:1–10. [60] Moalleminejad M, Chung DDL. Dielectric constant and electrical conductivity of carbon black as an electrically conductive additive in a manganese-dioxide electrochemical electrode, and their dependence on electrolyte permeation. Carbon 2015;91:76–87. [61] Delixiati A, Chung DDL. Bentonite-derived materials preferably with nanocarbon incorporation exhibiting exceptionally high dielectric loss at relatively low electrical conductivity. J. Mater. Sci. 2016;51:969–78. [62] Hong X, Yu W, Wang A, Chung DDL. Graphite oxide paper as a polarizable electrical conductor in the through-thickness direction. Carbon 2016;109(2016):874–82. [63] Hong X, Yu W, Chung DDL. Electric permittivity of reduced graphite oxide. Carbon 2017;111:182–90. [64] Hong X, Yu W, Chung DDL. Significant effect of sorbed water on the electrical and dielectric behavior of graphite oxide. Carbon 2017;119:403–18. [65] Wang M, Chung DDL. Understanding the increase of the electric permittivity of cement caused by latex addition. Composites Part B 2018;134:177–85. [66] Wang M, Chung DDL. High electric permittivity of polymer-modified cement due to the capacitance of the interface between polymer and cement. J Mater Sci 2018;53(10):7199–213. [67] Chakraborty P, Gundrati NB, Zhou C, Chung DDL. Effect of stress on the capacitance and electric permittivity of three-dimensionally printed polymer, with relevance to capacitance-based stress monitoring. Sens Actuators A 2017;263C:380–5. [68] Gundrati NB, Chakraborty P, Zhou C, Chung DDL. Effects of printing conditions on the molecular alignment of three-dimensionally printed polymer. Composites Part B 2018;134:164–8. [69] Zhu J, Luo Z, Wu S, Haldolaarachchige N, Young DP, Wei S, et al. Magnetic graphene nanocomposites: electron conduction, giant magnetoresistance and tunable negative permittivity. J Mater Chem 2012;22(3):835–44. [70] Gu H, Guo J, He Q, Jiang Y, Huang Y, Haldolaarachige N, et al. Magnetoresistive polyaniline/multi-walled carbon nanotube nanocomposites with negative permittivity. Nanoscale 2014;6(1):181–9. [71] Kavas H, Guenay M, Baykal A, Toprak MS, Sozeri H, Aktas B. Negative permittivity of polyaniline-Fe3O4 nanocomposite. J Inorg Organometallic Polym Mater 2013;23(2):306–14. [72] Shi Z, Chen S, Sun K, Wang X, Fan R, Wang X. Tunable radio-frequency negative permittivity in nickel-alumina “natural” meta-composites. Appl Phys Lett 2014;104(25):252908/1–5. [73] Yan K, Fan R, Shi Z, Chen M, Qian L, Wei Y, et al. Negative permittivity behavior and magnetic performance of perovskite La1-xSrxMnO3 at high-frequency. J Mater Chem C 2014;2(2014):1028–33. [74] Tsutaoka T, Fukuyama K, Kinoshita H, Kasagi T, Yamamoto S, Hatakeyama K. Negative permittivity and permeability spectra of Cu/yttrium iron garnet hybrid granular composite materials in the microwave frequency range. Appl Phys Lett 2013;103(26):261906/1–5. [75] Takizawa Y, Chung DDL. Continuous carbon fiber polymer-matrix composites in unprecedented antiferroelectric coupling providing exceptionally high throughthickness electric permittivity. J Mater Sci 2016;51(14):6913–32. [76] Luo X, Chung DDL. Electromagnetic interference shielding using continuous carbon fiber carbon-matrix and polymer-matrix composites. Composites Part B 1999;30:227–31. [77] Chung DDL. Electromagnetic interference shielding effectiveness of carbon materials. Carbon 2001;39:279–85. [78] Wu J, Chung DDL. Increasing the electromagnetic interference shielding effectiveness of carbon fiber polymer-matrix composite by using activated carbon fibers. Carbon 2002;40(ER3):445–7. [79] Chung DDL. Carbon materials for structural self-sensing, electromagnetic shielding and thermal interfacing. Carbon 2012;50:3342–53. [80] Wang R, Yang H, Wang J, Ma Z, Li F. Preparation and characterization of conductive filler used for electromagnetic shielding materials. Appl Mech Mater 2014;492:268–72. [81] Liu L, He P, Zhou K, Chen T. Microwave absorption properties of carbon fibers with carbon coils of different morphologies (double microcoils and single nanocoils) grown on them. J Mater Sci 2014;49(12):4379–86. [82] Yuan Q, Su C, Huang J, Gan W, Huang Y. Process and analysis of electromagnetic shielding in composite fiberboard laminated with electroless nickel-plated carbon fiber. BioResources 2013;8:4633–46.

[83] Singh BP, Choudhary Saini VP, Mathur RB. Designing of epoxy composites reinforced with carbon nanotubes grown carbon fiber fabric for improved electromagnetic interference shielding. AIP Adv 2012;2:022151. [84] Singh AP, Garg P, Alam F, Singh K, Mathur RB, Tandon RP, et al. Phenolic resinbased composite sheets filled with mixtures of reduced graphene oxide, γ-Fe2O3 and carbon fibers for excellent electromagnetic interference shielding in the Xband. Carbon 2012;50:3868–75. [85] Wang R, He F, Wan Y, Qi Y. Preparation and characterization of a kind of magnetic carbon fibers used as electromagnetic shielding materials. J Alloys Compounds 2012;514:35–9. [86] Mall S, Rodriguez J, Alexander MD. Electromagnetic interference and electrical conductivity behavior of carbon/polycyanate composite with nickel nanostrands™ under fatigue. Polym Compos 2011;32:483–90. [87] Hua Y, Yamanaka A, Ni Q. Electromagnetic shielding properties of super fiberreinforced composites. Adv Mater Res (Zuerich, Switzerland) 2010;123–125:65–8. [88] Micheli D, Laurenzi S, Mariani PV, Moglie F, Gradoni G, Marchetti M. Electromagnetic shielding of oriented carbon fiber composite materials, In: European Space Agency [Special Publication] SP-702:a3/1; 2012. [89] Zhang Y, Ma J, Singh AK, Cao L, Seo J, Rahn CD, et al. Multifunctional structural lithium-ion battery for electric vehicles. J Intelligent Mater Syst Struct 2017;28(12):1603–13. [90] Feng M, Wang S, Yu Y, Feng Q, Yang J, Zhang B. Carboxyl functionalized carbon fibers with preserved tensile strength and electrochemical performance used as anodes of structural lithium-ion batteries. Appl Surf Sci 2017;392:27–35. [91] Ihrner N, Johannisson W, Sieland F, Zenkert D, Johansson M. Structural lithium ion battery electrolytes via reaction induced phase-separation. J Mater Chem A 2017;5(48):25652–9. [92] Singh AK, Cao L, Ma J, Seo J, Bakis CE, Zhang Y, et al. Design, manufacture and test of a novel structural battery based on sandwich construction. J Sandwich Struct Mater 2015;17(6):666–90. [93] Ihrner N, Johansson M. Improved performance of solid polymer electrolytes for structural batteries utilizing plasticizing co-solvents. J Appl Polym Sci 2017;134(23). [94] Hagberg J, Leijonmarck S, Lindbergh G. High precision coulometry of commercial PAN-based carbon fibers as electrodes in structural batteries. J Electrochem Soc 2016;163(8):A1790–7. [95] Kim HC, Sastry AM. Effects of carbon fiber electrode deformation in multifunctional structural lithium ion batteries. J Intell Mater Syst Struct 2012;23(16):1787–97. [96] Asp LE, Greenhalgh ES. Structural power composites. Compos Sci Technol 2014;101:41–61. [97] Chung DDL. Carbon composites. 2nd ed. Elsevier; 2016. [98] Wang X, Chung DDL. Sensing delamination in a carbon fiber polymer-matrix composite during fatigue by electrical resistance measurement. Polym Compos 1997;18(6):692–700. [99] Wang S, Chung DDL. Interlaminar interface in carbon fiber polymer-matrix composites, studied by contact electrical resistivity measurement. Compos Interfaces 1999;6(6):497–506. [100] Baechle DM, O'Brien DJ, Wetzel ED. Structural dielectrics for multifunctional capacitors. In: Proc. SPIE 6929, Behavior and Mechanics of Multifunctional and Composite Materials 2008, 69292L; 23 April 2008. [101] Mei Z, Chung DDL. Kinetics of autohesion of thermoplastic carbon-fiber prepregs. Int J Adh Adh 2000;20:173–5. [102] Mei Z, Chung DDL. Thermal stress induced thermoplastic composite debonding, studied by contact electrical resistance measurement. Int J Adh Adh 2000;20:135–9. [103] Mei Z, Chung DDL. Glass transition and melting behavior of carbon fiber reinforced thermoplastic composite, studied by electrical resistance measurement. Polym Compos 2000;21(5):711–5. [104] Wang S, Kowalik DP, Chung DDL. Self-sensing attained in carbon fiber polymermatrix structural composites by using the interlaminar interface as a sensor. Smart Mater Struct 2004;13(3):570–92. [105] Wang S, Chung DDL. Temperature/light sensing using carbon fiber polymer-matrix composite. Composites Part B 1999;30(6):591–601. [106] Kim KD, Chung DDL. Electrically conductive adhesive and soldered joints under compression. J Adh Sci Tech 2005;19(11):1003–23. [107] Wang S, Pang DS, Chung DDL. Hygrothermal stability of electrical contacts made from silver and graphite electrically conductive pastes. J Electron Mater 2007;36(1):65–74. [108] Luo X, Chung DDL. A comparative study of silver-epoxy and tin-lead solder in their joints with copper, through mechanical and electrical measurements during debonding. J. Mater. Sci 1999;34:273–6. [109] Leong C, Chung DDL. Pressure electrical contact improved by carbon black paste. J Electron Mater 2004;33(3):203–6. [110] Todoroki A. Effect of degree of cure on sandwich structural capacitor using ionconductive polymer with carbon fabric skins. Open J Compos Mater 2016;6(4):112–20. [111] Carlson T, Ordeus D, Wysocki M, Asp LE. CFRP structural capacitor materials for automobile applications. Plastics, Rubber Compos 2011;40(6/7):311–6. [112] Carlson T, Asp LE. An experimental study into the effect of damage on the capacitance of structural composite capacitors. J Multifunctional Compos 2013:91–7. [113] Chan K, Jia B, Lin H, Zhu B, Lau K. Design of a structural power composite using graphene oxide as a dielectric material layer. Mater Lett 2018;216:162–5. [114] Todoroki A, Shiomi H, Mizutani Y, Suzuki Y. Electrical shorting between the carbon-fiber cloth electrodes of structural capacitors with a glass-fiber cloth separator. Open J Compos Mater 2014;4:140–7. [115] O’Brien DJ, Baechle DM, Yurchak OB, Wetzel ED. Effect of processing conditions and electrode characteristics on the electrical properties of structural composite capacitors. Composites Part A 2015;68:47–55. [116] O’Brien DJ, Baechle DM, Wetzel ED. Design and performance of multifunctional

100

Applied Energy 231 (2018) 89–101

D.D.L. Chung

2010;160:655–63. [144] Zhang J, Liu P, Jin C, Jin L, Bian S, Zhu Q, et al. Flexible three-dimensional carbon cloth/carbon fibers/NiCo2O4 composite electrode materials for high-performance all-solid-state electrochemical capacitors. Electrochim Acta 2017;256:90–9. [145] Li W, Xin L, Xu X, Liu Q, Zhang M, Ding S, et al. Facile synthesis of three-dimensional structured carbon fiber-NiCo2O4-Ni(OH)2 high-performance electrode for pseudocapacitors. Sci Reports 2015;5(9277):1–6. [146] Huang L, Chen D, Ding Y, Wang ZL, Zeng Z, Liu M. Hybrid composite Ni(OH)2@ NiCo2O4 grown on carbon fiber paper for high-performance supercapacitors. ACS Appl Mater Interfaces 2013;5(21):11159–62. [147] Zhang C, Xi Y, Wang C, Hu C, Liu Z, Javed MS, et al. High-performance flexible supercapacitors based on C/Na2Ti5O11 nanocomposite electrode materials. J Mater Sci 2017;52:13897–908. [148] Rafique A, Massa A, Fontana M, Bianco S, Chiodoni A, Pirri CF, et al. Highly uniform anodically deposited film of MnO2 nanoflakes on carbon fibers for flexible and wearable fiber-shaped supercapacitors. ACS Appl Mater Interfaces 2017;9:28386–93. [149] Chi HZ, Yin S, Cen D, Chen K, Hu Y, Qin H, et al. The capacitive behaviours of MnO2/carbon fiber composite electrode prepared in the presence of sodium tetraborate. J Alloys Compounds 2016;678:42–50. [150] Liu W, Liu N, Shi Y, Chen Y, Yang C, Tao J, et al. A wire-shaped flexible asymmetric supercapacitor based on carbon fiber coated with a metal oxide and a polymer. J Mater Chem A 2015;3:13461–7. [151] Xiang D, Liu X, Dong X. A facile synthetic method and electrochemical performances of nickel oxide/carbon fibers composites. J Mater Sci 2017;52:7709–18. [152] Cheng S, Yang L, Liu Y, Lin W, Huang L, Chen D, et al. Carbon fiber paper supported hybrid nanonet/nanoflower nickel oxide electrodes for high-performance pseudo-capacitors. J Mater Chem A 2013;1:7709–16. [153] Ma Y, Chen W, Zhang P, Teng F, Zhou J, Pan X, et al. Ni(OH)2 nanosheets grown on a 3D graphene framework as an excellent cathode for flexible supercapacitors. RSC Adv 2014;4:47609–14. [154] Tai Z, Lang J, Yan X, Xue Q. Mutually enhanced capacitances in carbon nanofiber/ cobalt hydroxide composite paper for supercapacitor. J Electrochem Soc 2012;159(4):A485–91. [155] Gallagher T, LaMaster D, Ciocanel C, Browder C. Electro-mechanical characterization of structural supercapacitors. In: Proc. SPIE 8342(Behavior and Mechanics of Multifunctional Materials and Composites); 2012. p. 83420S/1–83420S/8. [156] Chen YJ. Crucial role of energy storage in the use of renewable energy. Asian J Chem 2014;26(11):3351–4. [157] Cook TR, Dogutan DK, Reece SY, Surendranath Y, Teets TS, Nocera DG, et al. energy supply and storage for the legacy and nonlegacy worlds. Chem Rev (Washington, DC, United States) 2010;110(11):6474–502. [158] Pereira NM, Trigueiro JPC, Monteiro IF, Montoro LA, Silva GG. Graphene oxide – ionic liquid composite electrolytes for safe and high-performance supercapacitors. Electrochim Acta 2018;259:783–92. [159] Abdallah T, Lemordant D, Claude-Montigny B. Are room temperature ionic liquids able to improve the safety of supercapacitors organic electrolytes without degrading the performances? J Power Sources 2012;201:353–9. [160] Dou Q, Lei S, Wang D, Zhang Q, Xiao D, Guo H, et al. Safe and high-rate supercapacitors based on an “acetonitrile/water in salt” hybrid electrolyte. Energy Environ Sci 2018. Pages Ahead of Print. [161] Gonzalez F, Gregorio V, Rubio A, Garrido L, Garcia N, Tiemblo P. Ionic liquidbased thermoplastic solid electrolytes processed by solvent-free procedures. Polymers (Basel, Switzerland) 2018;10(2):124/1–124/17. [162] Yan S, Xu L, Jiang J, Xiao H, Li X. An extra-long-life supercapacitor based on Co3O4/NiCo2O4/NiO/C&S composite by decomposition of Co/Ni-based coordination complex. J Alloys Compounds 2018;764:684–90. [163] Zou X, Zhou Y, Wang Z, Chen S, Li W, Xiang B, et al. layered graphene monoliths for long-life supercapacitor. Chem Eng J (Amsterdam, Netherlands) 2018;350:386–94. [164] Deshmukh KU, Jamdar NS. Large scale stationary electrical storage batteries and research area. Int J Pure Appl Res Eng Technol 2016;4(8):599–607. [165] https://www.eia.gov/tools/faqs/faq.php?id=86&t=1 (as viewed on April 23, 2018). [166] Zhang J, Xu J, Zhang D. A structural supercapacitor based on graphene and hardened cement paste. J Electrochem Soc 2016;163(3):E83–7. [167] Wen S, Chung DDL. Effect of admixtures on the dielectric constant of cement paste. Cem Concr Res 2001;31(4):673–7. [168] Haddad AS, Chung DDL. Decreasing the electric permittivity of cement by graphite particle incorporation. Carbon 2017;122C:702–9. [169] Wang Y, Chung DDL. Capacitance-based defect detection and defect location determination for cement-based material. Mater Struct 2017;50:237. [170] Wang Y, Chung DDL. Capacitance-based nondestructive detection of aggregate proportion variation in a cement-based slab. Composites Part B 2018;134:18–27. [171] Chung DDL, Wang Y. Capacitance-based stress self-sensing in cement without any admixture requirement. submitted for publication. [172] Shi K, Chung DDL. Piezoelectricity-based self-sensing of compressive and flexural stress in cement-based materials without any admixture requirement and without poling. Smart Mater Struct 2018;27:20. 105011. [173] Wang Y, Chung DDL. Effect of the fringing electric field on the apparent electric permittivity of cement-based materials. Composites Part B 2017;126:192–201.

structural composite capacitors. J Compos Mater 2011;45(26):2797–809. [117] Lin Y, Zhou Z, Sodano HA. Barium titanate and barium strontium titanate coated carbon fibers for multifunctional structural capacitors. J Compos Mater 2013;47(12):1527–33. [118] O'Brien DJ, Gienger EB, Snyder JF, Baechle DM, Xu K, Wetzel ED. Structural power for U.S. Army applications. In: Int. SAMPE Tech. Conf. 42; 2010. p. sa12/ 1–a12/12. [119] Javaid A, Ho KKC, Bismarck A, Steinke JHG, Shaffer MSP, Greenhalgh ES. Carbon fibre-reinforced poly(ethylene glycol) diglycidylether based multifunctional structural supercapacitor composites for electrical energy storage applications. J Compos Mater 2016;50(16):2155–63. [120] Gallagher T, LaMaster D, Ciocanel C, Browder C. Electro-mechanical characterization of structural supercapacitors. In: Proc. SPIE 8342, Behavior and Mechanics of Multifunctional Materials and Composites 2012, 83420S (28 March 2012). [121] Cho B, Choi J, Kim K. Preparation and properties of solid polymer electrolyte based on imidazolium-based ionic liquids for structural capacitors. Fibers Polym 2017;18(8):1452–8. [122] Hudak NS, Schlichting AD, Eisenbeiser K. Structural supercapacitors with enhanced performance using carbon nanotubes and polyaniline. J Electrochem Soc 2017;164(4):A691–700. [123] Li S, Zhao Y, Zhang Z, Tang H. Preparation and characterization of epoxy/carbon fiber composite capacitors. Polym Compos 2015;36(8):1447–53. [124] Deka BK, Hazarika A, Kim J, Park Y, Park HW. Multifunctional CuO nanowire embodied structural supercapacitor based on woven carbon fiber/ionic liquid–polyester resin. Composites Part A 2016;87:256–62. [125] Javaid A, Ho KKC, Bismarck A, Steinke JHG, Shaffer MSP, Greenhalgh ES. Improving the multifunctional behaviour of structural supercapacitors by incorporating chemically activated carbon fibres and mesoporous silica particles as reinforcement. J Compos Mater 2018. [126] Qian H, Kucernak AR, Greenhalgh ES, Bismarck A, Shaffer MSP. Multifunctional structural supercapacitor composites based on carbon aerogel modified high performance carbon fiber fabric. ACS Appl Mater Interfaces 2013;5:6113–22. [127] Srimuk P, Luanwuthi S, Krittayavathananon A, Sawangphruk M. Solid-type supercapacitor of reduced graphene oxide-metal organic framework composite coated on carbon fiber paper. Electrochim Acta 2015;157:69–77. [128] Hou C, Xu Q. Metal-organic frameworks for energy, Adv. Energy Mater. 2018. Pages Ahead of Print. [129] Ciocanel C, Browder C, Simpson C, Colburn R. The challenges of achieving good electrical and mechanical properties when making structural supercapacitors. In: Proc. SPIE 8689, Behavior and Mechanics of Multifunctional Materials and Composites; 2013, 868917 (3 April 2013). [130] Shirshova N, Qian H, Houll M, Steinke JHG, Kucernak ARJ, Fontana QPV, et al. Multifunctional structural energy storage composite supercapacitors. Faraday Discuss 2014;172:81–103. [131] Deka BK, Hazarika A, Kwon O, Kim D, Park Y, Park HW. Multifunctional enhancement of woven carbon fiber/ZnO nanotube-based structural supercapacitor and polyester resin-domain solid-polymer electrolytes. Chem Eng J 2017;325:672–80. [132] Jiang Q, Yang R, Fu GG, Xie DY, Huang B, He ZW, et al. Preparation of the carbon nanotube/carbon fiber composite and application as the electrode material of the electrochemical super capacitor. Mater Sci Forum 687(Funct Electron Mater) 2011:158–62. [133] Kwon OB, Deka BK, Kim J, Park HW. Electrochemical performance evaluation of tin oxide nanorod-embedded woven carbon fiber composite supercapacitor. Int J Energy Res 2018;42:490–8. [134] Kim SY, Yang K, Kim B. Enhanced electrical capacitance of heteroatom-decorated nanoporous carbon nanofiber composites containing graphene. Electrochim Acta 2014;137:781–8. [135] Lyu Shaoyi, Chang Huanjun, Feng Fu, La Hu, Huang Jingda, Wang Siqun. Cellulose-coupled graphene/polypyrrole composite electrodes containing conducting networks built by carbon fibers as wearable supercapacitors with excellent foldability and tailorability. J Power Sources 2016;327:438–46. [136] Almeida DAL, Fonseca CP, Baldan MR, Ferreira NG. CF/PANI/MWNT composites material, a novel electrode to supercapacitor. ECS Trans 2012;41(21):13–9. [137] Liu M, He S, Fan W, Miao Y, Liu T. Filter paper-derived carbon fiber/polyaniline composite paper for high energy storage applications. Compos Sci Technol 2014;101:152–8. [138] Jin Z, Tian Y, Su LJ, Qin CL, Zhao DY, Li RQ. Zhao J. Hybrid supercapacitors based on polyaniline/activated carbon fiber composite electrode materials. In: Advanced Materials Research (Durnten-Zurich, Switzerland) 800(Recent Development on Material Science and Environmental Material); 2013. p. 505–8. [139] Paul P, Gon DP, Banerjee D. Activated carbon fiber from PAN precursor and its application. Chem Fibers Int 2011;61(4):208–9. [140] Yang Q, Dong L, Xu C, Kang F. High-performance supercapacitors based on graphene/MnO2/activated carbon fiber felt composite electrodes in different neutral Electrolytes. RSC Adv 2016;6:12525–9. [141] Song K, Ni H, Fan L. Flexible graphene-based composite films for supercapacitors with tunable areal capacitance. Electrochim Acta 2017;235:233–41. [142] Javaid A, Ho KKC, Bismarck A, Shaffer M. Multifunctional structural supercapacitors for electrical energy storage. J Compos Mater 2014;48(12):1409–16. [143] Wilson GJ, Looney MG, Pandolfo AG. Enhanced capacitance textile fibres for supercapacitors via an interfacial molecular templating process. Synth Met

101