Carbon Allotropes/Epoxy Nanocomposites as Capacitive Energy Storage/Harvesting Systems

Featured Application

Lightweight, low cost, flexible systems for fast storing and retrieving electrical energy.

Abstract

The present work aims at the development and characterization of carbon/polymer matrix nanocomposites, which will be able to operate as compact materials systems for energy storage and harvesting. Series of polymer nanocomposites employing different types of carbon allotropes (carbon black nanoparticles, multi-walled carbon nanotubes, graphene nanoplatelets and nanodiamonds) were developed varying the filler type and content. The energy storage ability of the systems was examined under AC and DC conditions to evaluate the influence of temperature, DC voltage and different types of filler content upon the stored and harvested energy. Experimental data confirmed the ability of the examined systems to store energy and release it on demand via a fast charge/discharge process. The addition of carbon nanoparticles significantly enhances the energy density of the systems. The coefficient of energy efficiency (n_eff) was determined for all systems, reaching up to 80% for the nanocomposite with 5 phr (parts per hundred resin per mass) carbon black content. In order to examine the optimal operational conditions of the systems, their structural integrity and thermomechanical properties were also investigated by means of static tensile tests, Dynamic Mechanical Analysis (DMA) and Differential Scanning Calorimetry (DSC).

1. Introduction

The social challenges of resource depletion and climate change call for a cleaner, more efficient use of resources and energy. The need for energy efficiency is leading to the introduction of novel and improved materials and the design of new low-cost, low-weight and environmentally friendly device structures that can harvest and/or store and/or convert energy with higher efficiency. By these means, increased autonomy and support for diverse applications within mobile devices that range from portable (or even wearable) electronics to hybrid electric vehicles can be achieved [1]. Innovation in dielectric polymer nanocomposites, which have the intrinsic ability to store energy at the nanoscale level, could lead to structures with high performance having a profound impact on the longer-term plans for more energy-efficient technology [2]. Multifunctional polymer composites could be employed as structural energy devices, where the matrix and filler work synergistically, leading to improved functional behavior [3,4].

Polymer nanocomposites are considered to be advanced technological materials in the fields of electronics, aviation, automobile and aerospace industries [5,6] due to their low weight, low cost, easy processing and remarkable mechanical, thermal, electrical and magnetic properties that can be effectively tailored by controlling the filler type and content [7,8,9,10]. The combination of polymers that exhibit high dielectric breakdown strength with the high values of dielectric permittivity of the ceramic fillers leads to a composite system with enhanced dielectric properties [11]. In addition, dispersed ceramic nanoparticles inside the polymer matrix can act as a distributed network of nanocapacitors where energy can be stored and harvested via a fast charge/discharge process [8,12,13]. The large interfacial area between the polymer matrix and the nanoinclusions is a dominant factor that dictates the dielectric and energy performance of nanocomposite materials due to interfacial polarization [14,15], which is the primary mechanism for most capacitor applications.

Conductive elements have been employed to increase the dielectric permittivity of polymer matrices and enhance their energy storage performance [16,17]. However, increased conductivity values may increase leakage currents and diminish energy efficiency. Semiconducting ceramic nanoparticles, especially perovskite ferroelectric oxides such as BaTiO₃ and SrTiO₃ [12,13,18], have been extensively investigated for their potential energy storage performance along with piezoelectric polymers such as PVDF [3,19,20,21]. Even though there are many studies in the literature that investigate the potential energy storage capacity of polymer nanocomposites [14,22,23,24], little work has been done to determine the amount of recovered energy [8,12,13,25]. Furthermore, the disparity of the fabrication processes and/or different functionalization methods makes the direct comparison of the reported results difficult.

The main goal of this study is to determine the energy efficiency in identically prepared nanocarbon-filled epoxy nanocomposite systems by investigating the effects of filler concentration temperature, DC voltage and different types of filler content upon the stored and harvested energy. For this reason, nanocomposites comprising an epoxy resin as a matrix and different carbon allotropes as the reinforcing phase (carbon black, multi-walled carbon nanotubes, graphene nanoplatelets and nanodiamonds) were prepared and studied at various filler contents. The nanocomposites were characterized by several techniques to determine their operational window and structural integrity.

2. Materials and Methods

2.1. Materials

Four nanocomposite systems were prepared by employing carbon allotropes as fillers in an epoxy matrix. In particular, Epoxol 2004 was supplied by Neotex S.A., Athens, Greece, while fillers were obtained from Plasmachem GmbH. The type of the employed epoxy was a two-component low-viscosity resin consisting of epoxy prepolymer (diglycidyl ether of bisphenol A (DGEBA)) along with the curing agent (aromatic amine). The physical properties of the carbon allotropes, according to the manufacturer’s datasheet, are presented in

Table 1. The employed carbon allotropes as fillers.

Fillers	Size	Specific Surface
Carbon black (CB) nanoparticles	13 nm	500 m2/g
Graphene nanoplatelets (GnP)	Thickness: 1–4 nmLateral size up to 2 μm	700–800 m2/g
Multi-walled carbon nanotubes (MWCNT)	Length: 1–10 μm (3–15 walls)Outer diameter: 5–20 nmInner diameter: 2–6 nm	240 m2/g
Nanodiamonds grade G (ND)	4 nm	290–360 m2/g

.

2.2. Fabrication Method

For the preparation of the nanocomposites, the carbon allotropes were initially dispersed into the prepolymer and stirred inside a sonicator for 5 min. Then, the curing agent was added at a 2:1 (prepolymer/hardener) mixing ratio, and the whole mixture was stirred by hand in a sonicator for 10 min before it was poured into suitable silicon moulds to be cured for seven days at room temperature. After polymerization, the samples underwent a post-curing treatment at 100 °C for 4 h in order to complete the crosslinking process. Filler content for all systems was 0, 0.1, 1, 3, 5, 7 and 10 phr (parts per hundred resin per weight). A schematic representation of the fabrication process of the nanocomposites is depicted in Figure 1.

2.3. Characterization Techniques

2.3.1. Morphological Characterization

The morphology and the quality of the developed nanocomposites were examined via Scanning Electron Microscopy (SEM) using an EVO MA 10 apparatus provided by Carl Zeiss.

2.3.2. Thermomechanical Characterization

The thermomechanical properties of the nanocomposites were assessed by Differential Scanning Calorimetry (DSC) using TA Q200 and Dynamic Mechanical Analysis (DMA) using TA Q800, both provided by TA Instruments. For the DSC tests, samples were placed in an aluminum crucible, while an empty aluminum crucible served as a reference. Additionally, the mechanical properties of the nanocomposites were examined with static tensile tests at room temperature with a 5 mm/min rate, using an Instron 5582 apparatus, provided by Instron. The parameters of the DSC and DMA are shown in

Table 2. DSC parameters.

DSC
Temperature range	0–100 °C
Heating rate	5 °C/min
Module time	Data
Nitrogen flow	50 mL/min

and

Table 3. DMA parameters.

DMA
Temperature range	Ambient–100 °C
Heating rate	5 °C/min
Frequency	1 Hz
Operating mode	Three-point bending

, respectively.

2.3.3. Energy Storage

AC Measurements

The electric response of the nanocomposites under AC conditions was studied via Broadband Dielectric Spectroscopy (BDS) using an Alpha-N Frequency Response Analyzer and a Novotherm system for the temperature control. All experimental parameters for the BDS measurements are given in

Table 4. BDS parameters.

BDS
Frequency range	10−1–107 Hz
Temperature range	30–160 °C
Temperature step	5 °C/min
Stabilization time	60 s
Vrms	1 V

. The sample was placed inside the dielectric cell BDS 1200, forming a sandwich capacitor. Frequency scans were conducted under successive isothermal conditions. The experimental data were recorded using Windeta software. Devices, cell and software were obtained from Novocontrol Technologies.

DC Measurements

The energy storage/harvesting tests under DC current measurements were performed using a high-resistance DC meter (Agilent 4339B). The experimental set up included an automatic measurement process controlling the charging/discharging sequence.

Table 5. DC measurements’ parameters.

DC Measurements
Applied Voltages	50, 100, 250 V
Charging time	60 s
Discharging time	600 s
Temperature	Ambient

presents the parameters of DC measurements. The samples were put in a parallel-plate capacitor configuration. Experimental data were obtained via suitable software.

3. Results and Discussion

3.1. Morphological Characterization

The quality of the surface of nanocomposites along with the dispersion and distribution of nanoparticles in the polymer matrix is vital for the enhancement of their performance. Any cracks, voids or air trapped inside the matrix during the fabrication process would be detrimental to both the mechanical as well as the electrical properties of the nanocomposites. Scanning Electron Microscopy was employed to examine the quality of the fabricated systems. Figure 2 shows representative SEM images for the nanocomposites with 3 phr filler content. Obtained SEM images confirm the high quality of the systems since no apparent air, cracks or voids are present. Uniform distribution and dispersion of the nanoparticles inside the polymer matrix was achieved, and the formation of large agglomerates was avoided.

3.2. Thermal and Mechanical Characterization

Differential Scanning Calorimetry was employed for the investigation of the thermal properties of the nanocomposites. An endothermic step-like process was recorded in the thermograms of all examined systems, which was identified as the transition from the glassy to the rubbery state of the polymer matrix. The characteristic glass transition temperature for all systems was determined at the point of inflection using suitable software provided by TA instruments. The results are listed in

Table 6. Glass transition temperature as determined via DSC test, for all examined systems.

Specimens	CB	GnP	MWCNT	ND
	Tg (°C)	Tg (°C)	Tg (°C)	Tg (°C)
Epoxy	44.6	44.6	44.6	44.6
0.1 phr	43.7	52.0	51.7	53.2
1 phr	46.4	53.3	51.8	54.0
3 phr	55.6	53.9	53.7	53.5
5 phr	56.1	58.8	52.1	53.3
7 phr	56.3	55.3	51.7	53.7
10 phr	57.4	54.9	50.2	52.4

. The variation of the glass transition temperature with filler content can give an insight into the different interactions occurring inside the materials. Increasing T_g values indicate strong interactions between the nanoparticles and the macromolecules that obstruct the cooperative segmental movement of the polymer chains. All nanocomposites exhibit higher glass transition temperature values than the neat epoxy, which is a sign of the good adhesion of the nanofillers to the polymeric matrix. Carbon black nanocomposites exhibit a progressive increase in T_g values with filler content while, on the other hand, nanocomposites filled with nanodiamonds show a remarkable stability across the board. At high concentrations for the GnP and MWCNT-filled nanocomposites, the interactions between adjacent nanoparticles, due to their aspect ratio, become stronger, thus cancelling part of the hindrance of the chain mobility and slightly lowering the values of the glass transition temperature.

The mechanical properties of polymer nanocomposites are dictated by several parameters including the individual properties of the constituents and the distribution of fillers in the polymer matrix, as well as interfacial bonding and the fabrication method. The addition of carbonaceous nanoparticles enhances the mechanical properties of the nanocomposites as demonstrated by the static and dynamic tests (Figure 3 and Figure 4, respectively). The tensile modulus was determined as the slope of the stress–strain curve in the elastic region. Figure 3 presents the Young’s modulus and tensile strength versus filler content for all examined nanocomposites. Error bars denote the standard deviation of the representative results. The Young’s modulus increases predictably with filler concentration for all systems except the nanocomposites filled with nanodiamonds. The systematic increase in the tensile modulus values is attributed to the intrinsic stiffness and rigidness of the nanoparticles and to the interfacial bonding between the matrix and the nanoinclusions, which affects the effectiveness of the load transfer from the polymer matrix to nanofillers [26,27]. It can be seen that the addition of very small amounts of carbon nanotubes, with very large aspect ratio and stiffness, in the polymer matrix leads to the largest modulus enhancement of all examined systems. In the case of nanodiamond-filled nanocomposites, the Young’s modulus increases until the 5 phr specimen, where it obtains the maximum value and then diminishes significantly, indicating that a higher content of nanodiamonds favors their aggregation.

The macroscopic properties of heterogeneous polymer nanocomposites are governed by the local stress distribution around the inclusions [28] that form a network that is able to bear the mechanical load. The tensile strength of the nanocomposites decreases with filler content for all examined systems, exhibiting lower values of strength than the neat epoxy, as depicted in Figure 3. Nanoparticles seem to act as stress raisers, which is a common problem in particulate-filled composites [29]. The nanocomposites with MWCNT and GnP nanoparticles maintained higher values of tensile strength compared to the systems with 0D nanoreinforcements.

Figure 4 depicts the variation of the storage modulus and loss modulus (insets) as a function of temperature for all examined systems. The maximum values of the storage modulus follow the trend of the results for the elastic modulus. The increase of E′ values with filler content is ascribed to the effective load transfer from the polymer matrix to the filler, as a consequence of the fine dispersion and strong adhesion of filler in the matrix. Similarly to static tests, there is a significant increase in the storage modulus values with the addition even of a low amount of MWCNTs. This behavior reflects the intrinsic stiffness of carbon nanotubes and the formation of a load-carrying network inside the polymer matrix. There are several studies in the literature [30,31,32] reporting a sudden increase in stiffness for similar systems and suggesting the concept of a percolation threshold for the mechanical properties of polymer nanocomposites in an attempt to explain the high ranges of mechanical properties and the level of nanoparticle networks in polymer nanocomposites. However, this does not seem to be the case in the examined systems, as theoretical and computational results [30,31,32] indicate that such a threshold lies further beyond the electrical threshold, which in the examined MWCNTs system is 4.71% w/w [16]. A sharp decline in the storage modulus values suggests the transition of the polymer matrix from the glassy to the rubbery state, where the nanocomposites lose their load retention capacity. These step-like transitions are distinguished by the formation of peaks in the loss modulus spectra.

3.3. Energy Storage and Harvesting

3.3.1. AC Measurements

The ability of a nanocomposite to store electrical energy is expressed by its energy density [22]. The energy density, for all examined systems, was calculated via Equation (1):

$$\mathrm{U}=\frac{1}{2}{\mathsf{\epsilon }}_0\mathsf{\epsilon }\prime {\mathrm{E}}^2,$$

(1)

where ε₀ is the dielectric permittivity of the free space, which is equal to 8.854 × 10⁻¹² F/m, ε′ is the real part of permittivity and Ε is the field’s intensity. The calculations were made at constant electric field $\left(\mathrm{E}=1\frac{\mathrm{kV}}{\mathrm{m}}\right)$. In linear dielectric materials, polarization is proportional to the applied electric field, and the energy stored per volume increases with increasing permittivity, at constant applied field.

The variation of energy density as a function of frequency and temperature for all nanocomposites with 3 phr filler content is depicted in Figure 5. The illustrated behavior is representative of all examined systems. The energy density increases with temperature as the induced thermal energy facilitates the molecular mobility and the alignment of the dipoles with the externally applied electric field, thus increasing the achieved polarization. At high temperatures and low frequencies, energy density attains high values due to interfacial polarization. Polarization—and therefore also the energy density values—diminish rapidly with increasing frequency as permanent and induced dipoles do not have sufficient time to align themselves with the rotating external field. At intermediate frequencies and temperatures, a relaxation process arises which is related to the glass to rubber transition of the polymer matrix, also known as α-relaxation.

The relative energy density function, defined by Equation (2), was employed to highlight the contribution of the nanoreinforcements and to estimate the percentage of the stored energy in the nanocomposites compared to the energy stored in the matrix. Its normalized character eliminates the influence of geometric factors.

$${\mathrm{U}}_{\mathrm{relative}}=\frac{{\mathrm{U}}_{\mathrm{composite}}}{{\mathrm{U}}_{\mathrm{resin}}}$$

(2)

Figure 6 showcases the variation of U_relative as a function of temperature at 160 °C for all examined systems. In general, the ability of the nanocomposites to store energy increases with filler content due to the higher values of conductivity for the sp² hybridized carbon nanoparticles, such as MWCNTs, and dielectric permittivity for the sp³ hybridized nanodiamonds [16] and the induced heterogeneity of the systems. The nanocomposites with carbon black and MWCNT filler content exhibit enhanced energy storage ability that reaches up to hundreds and thousands of times, respectively, the ability of the neat epoxy matrix at the maximum point of the function at the highest filler concentrations. The enhanced performance of MWCNT and carbon black systems, especially at intermediate and high concentrations, can be attributed to the increased values of conductivity of these systems. A thorough conductivity analysis of these systems reported in [16] shows that the significant increase of conductivity corresponds to a very narrow variation of the conductive phase content. This increase can be described by means of percolation theory, which predicts that at a critical concentration (also called the percolation threshold), a conductive path is formed, allowing the migration of charge carriers throughout the material [16,33]. The abrupt increase of conductivity appears as the result of the transition from the insulating state, where limited contacts between conductive sites exist, to the state where a conductive network is formed, via physical contacts of the conductive inclusions (metallic-type conduction). The determined values of critical concentration were found to be 4.71% w/w for the MWCNT systems and 5.26% w/w for the carbon black systems. The dependence of conductivity on temperature and the Variable Range Hopping model were employed in this study [16], revealing that at concentrations higher than the percolation threshold, hopping and metallic-type conduction coexist. On the other hand, the effect of the addition of nanodiamonds on the energy storage capabilities of the systems is almost negligible. The peak formed at low frequencies is attributed to interfacial polarization due to the accumulation of free charges at the interface between the filler and the polymer matrix, while the secondary peaks at higher frequencies correspond to the different dynamics of the α-relaxation process between the neat matrix and the nanocomposites.

3.3.2. DC Measurements

The calculation of the of the stored and retrieved energy was performed by integrating the time-dependent current functions, under charging and discharging conditions, via Equation (3):

$$\mathrm{E}=\frac{1}{2}\frac{{\mathrm{Q}}^2}{\mathrm{C}}=\frac{1}{2}\frac{{\left(\int \mathrm{I}\left(\mathrm{t}\right)\mathrm{dt}\right)}^2}{\mathrm{C}},$$

(3)

where E is the stored energy at the capacitor, Q the charge and C is the capacitance of each nanocomposite as evaluated via the BDS measurements at the lowest frequency [12]. A more detailed insight into the rationale and the methodology can be found in [8,12].

Figure 7 depicts representative graphs of the stored and harvested energies for the nanocomposite systems with 3 phr filler content. The success of the energy storage/harvesting process is confirmed in the above diagrams for all examined systems. All charging curves are above the corresponding discharge curves. Systems quickly acquire high charging and discharge values (fast charge/discharge)—a characteristic feature of instantaneous power density capacitive systems. This ability to store and deliver a large amount of energy nearly instantaneously, sometimes called “pulse power”, cannot be delivered by storage technologies such as batteries and has potential applications in numerous commercial and military devices that require increased power density as well as higher charge/discharge current capabilities. The stored and recovered energies increase significantly with increasing voltage since the applied field lowers the local potential barriers, thus facilitating the migration of the charges inside the nanocomposites.

Figure 8 presents the comparative plots for the stored and retrieved energies for all examined systems at 100 V. Both energies increase with filler content due to the enhanced electrical properties of the carbonaceous nanoparticles compared to the polymer matrix as well as the interfacial effects due to the heterogeneity of the systems. It should be noted that, for the carbon black and MWCNT systems, above a critical concentration, reported as the percolation threshold in a previous study [16], the nanocomposites transition from the insulating to the conducting state. Therefore, no charge/discharge process could be established since the charge carriers do not accumulate inside the material but rather travel through it, either solely via a conducting path or in conjunction with hopping mechanisms [33]. However, the nanocomposite with 3 phr MWCNT concentration still exhibits the optimum performance.

Another critical factor for the valuation of the nanocomposites’ energy storage/harvesting performance is the calculation of the coefficient of energy efficiency (n_eff), which is defined as the ratio of the recovered to stored energy with parameters the charging voltage and time [8,12]. Figure 9 presents the results for the coefficient of energy efficiency at the same instant times of t = 5, 10 and 30 s in both charging and discharging procedures. The coefficient of energy efficiency diminishes over time for all systems, confirming once more the fast charge/discharge nature of the examined systems. n_eff increases with filler content for the nanocomposites with carbon black and MWCNT filler content. The optimum energy efficiency was noted for the nanocomposite with 5 phr carbon black content with coefficient (n_eff) values of 80 % at 100 V and 5 s. Unfortunately, once more, the nanocomposites filled with nanodiamond particles fail to record a decent performance, as their respective coefficient drops even below that of the polymer matrix.

4. Conclusions

In the present study, series of polymer nanocomposites employing different types of carbon allotropes were successfully developed varying the filler type and content. SEM images confirmed the quality of the fabricated systems and revealed fine nanodispersions of the fillers in the polymer matrix. The thermal and mechanical properties were investigated by means of static tensile tests, Dynamic Mechanical Analysis and Differential Scanning Calorimetry. All nanocomposites exhibit higher glass transition temperature values than the neat epoxy, implying strong adhesion of the nanofillers to the polymer matrix. Storage and elastic moduli attain higher values with the addition of the reinforcing nanoparticles. The energy storage ability of the systems under AC conditions was determined by means of Broadband Dielectric Spectroscopy in a wide frequency and temperature range. Energy density increases with temperature and attaines the highest values at low frequencies due to interfacial polarization. The stored and harvested energy of the nanocomposites was also evaluated under DC conditions by employing a charge/discharge procedure. The energy storage and harvesting ability of the nanocomposites increases with filler content due to the higher values of dielectric permittivity and/or the conductivity of the nanoinclusions. The nanocomposites with carbon black and MWCNT filler content exhibit enhanced energy storage ability that exceeds several times the stored energy of the polymer matrix. On the other hand, the addition of nanodiamonds fails to enhance the energy storage capabilities of the systems. Finally, the coefficient of energy efficiency (n_eff) was determined for all systems. The filler content enhances the energy efficiency of the examined systems, reaching the highest value of over 80% for the 5 phr carbon black nanocomposite.