Improvement in electrical characteristics by surface modification of multi-wall carbon nanotube based buckypaper for de-icing application

Abstract

Icing hazards often cause severe mobility concerns, safety risks, and even accidents in modern industries. Various ice mitigation strategies have been employed. Electro-thermal heating approach is a popular measure for ice protection, in which heat is generated electrically by internal components and then transferred to the outer surface for anti-icing and de-icing operation. Given the increasing usage of composites in aircraft structures, self-heating fibre reinforced polymer composite has been studied as an ice protection solution for the new generation aircraft. The present work focuses on improving electrical characteristics of carbon nanotube (CNT) based buckypaper (via surface functionalization) and the heating performance with an ice-phobic resin. The results indicated that surface functionalization of multi-wall carbon nanotubes (MWCNTs) was effective in obtaining MWCNT buckypaper (CNP) with improved electrical conductivity. The de-icing test confirmed the electro-thermal heating performance of the CNP based composite in a climate chamber at −20°C. The utilisation of surface modified MWCNTs in self-heating composites could be a promising strategy for maintaining lightweight and efficiency for electro-thermal systems to mitigate icing hazards.

Keywords

Multi-wall carbon nanotubes buckypaper electrical resistance surface modification de-icing

Introduction

Fibre-reinforced polymer composites are gaining an increasing attention in modern industries like aerospace, automobile, and wind energy, because of their high mechanical strength and light weight.¹ However, in a cold region or environment, icing accretion on the composite surface often cause severe mobility concerns, safety risks, and even fatal accidents. Significant attention has been paid for developing smart de-icing systems to protect from the ice hazards in many industrial applications, including aircraft,^2,3 telecommunications,⁴ and power transmission.⁵ Electro-thermal heating is one of the most popular approaches for the ice protection of polymer composites due to its capability of integration and controlling heat dissipation.⁶ Typically an electro-thermal structure consists of a metal framework and an inorganic insulator which is normally complicated and heavy. Incorporating a traditional metal foil heater in a composite structure may create a weak layer, especially under inter-laminar shear force.⁷ Thus, conductive nanomaterials into the composite structure has received increasing interest in the last decade,^8–10 which could provide excellent self-heating capabilities for ice protection without sacrificing structure integration.

Carbon nanotube (CNT) is one of the most promising materials for electro-thermal heating elements of composite structures.^8–11 Since the discovery of CNTs in 1991 by Iijima,¹² CNTs have been using in various applications because of their excellent mechanical and chemical properties, high electrical and thermal conductivity.^13–16 Various applications of carbon nanotubes have been reported, from antiballistic vests¹⁷ to electro-thermal elements.^18–20 High electrical conductivity makes CNTs more attractive for electro-thermal application. Aouraghe et al.^18,19 developed a flexible carbon nanotube film that could reach a steady-state temperature (>300°C) at low voltage, even after mechanical deformation and On/Off voltage switching cycles. Li et al.²¹ produced a transparent, flexible thin heater based on graphene and CNT sheets using a dry spinning technique. The film revealed defrosting performance at 12 V and uniform thermal distribution even after bending for 10,000 cycles. Sun et al.²² developed a CNT/thermoplastic polyurethane composite via a ball mixing and compression molding method. The obtained composite exhibited an electrical conductivity of 142.6 Sm⁻¹ and a steady temperature of 65°C under 6 V applied voltage. Further attempts have been made to improve the intrinsic electrical conduction of CNTs-based structures through surface modification of CNTs using different acid treatment. The addition of functional groups (-OH/-COOH) on the surface of CNTs can significantly increase the electrical conductivity because of the increased mobility and charge carrier concentration, thus enhancing the metallic behaviour of the nanotubes.^23–25 It was reported that the functional groups located at the contact area between the CNTs serve as conductance paths with lower electrical resistance.²⁶ Zhang et al.²⁷ showed an increase in electrical conductivity of MWCNT buckypaper from 1.3 × 10⁴ Sm⁻¹ to 3.8 × 10⁴ Sm⁻¹ after surface functionalization by acid treatment. Hong et al.²⁸ produced a transparent de-icing CNT film by spin-coating on glass substrate. Its effective de-icing capability under a freezing condition of −20°C was demonstrated with 80 V of applied voltage. Chu et al.¹⁰ fabricated single-wall carbon nanotubes (SWCNTs) based composites for electro-thermal heating and de-icing applications. The SWCNTs were used to produce a buckypaper (SW-CNP), by filtration technique in glass fibre (GF) fabrics. The SW-CNP/GF/epoxy composites exhibited excellent heating performance in the temperature range of −22°C–15°C with a wind speed of 14 m/s, and the de-icing time 4–7 min. Rashid et al.²⁹ produced a CNT heater by a roll-to-roll slot-die coating process. The coating sample kept the heated area ice-free and further ice accretion was prevented with 40 V applied voltage. Yao et al.³⁰ developed a horizontally oriented CNTs film by chemical vapour deposition and embedded it in glass fibre fabrics to prepare composites for the de-icing purpose. During the de-icing test, the CNT/GF composites removed accreted ice within 15 s under a constant voltage of 16 V. Zhao et al.³¹ developed a multi-layered ice protection coating by a combination of MWCNTs heating element with a super-hydrophobic layer. Compared to the traditional heating element, super-hydrophobic electric heating element showed up to 58% reduction in energy consumption. Recently, fibre reinforced polymer composites embedded with CNPs have been fabricated by pre-preg and resin impregnation methods, and excellent electro-thermal characteristics has been demonstrated.³² Though the CNTs based self-heating element has demonstrated much improved heating performance, the overall electrical resistivity of the CNTs-based structure can be further lowered to increase the power density substantially.

The present work investigates the effect of surface functionalization on the microstructure of CNT based buckypaper and its effect on e-heating performances when impregnated with an ice-phobic resin. The acid treated buckypapers have been integrated into a composite structure by laminating the buckypaper with GF fabric followed by the impregnation in thermosetting resin (EL2 Epoxy laminating resin and Silikopon EF). EL2 Epoxy lamination resin is a commercial product used as a reference for comparison with the icephobic Silikopon EF.³³ Microstructures and defect states of MWCNTs before and after acid treatment were systematically investigated. The self-heating element, consisting of surface-modified MWCNTs and the corresponding composites using advanced ice-phobic resin, could be a promising electrical heating strategy to maintain lightness and efficiency in ice mitigation.

Experimental details

Raw materials

Multi-wall carbon nanotubes (95% purity) were purchased from Merck KGaA (Darmstadt, Germany). The nanotubes had an average outer diameter of 6–15 nm and length 2.5–20 μm. Nitric acid (70 wt%) was supplied by Fisher Chemical UK Ltd (Loughborough, UK). Triton X-100 (C14H22O(C2H4O)_n) was provided by Merck KGaA (Darmstadt, Germany). The glass fibre fabric (GF, thickness 0.185 mm), EL2 laminating resin and its hardener AT30 SLOW were purchased from Easy Composite (Stoke-on-Trent, UK). Silikopon EF resin was provided by Evonik Industries AG (Essen, Germany), which was considered as an icephobic resin.

Acid treatment

100 mg of pristine MWCNTs were dispersed in 50 mL nitric acid. The suspension was ultrasonicated in a bath for 2 h, followed by magnetically stirring for another 2 h. At the end of the acid treatment, the suspension was filtered, washed with distilled water until pH reached 7 and then dried at 60°C overnight. The obtained product was named as NA-CNT.

Preparation of carbon nanotube buckypaper

To obtain carbon nanotube buckypaper (CNP), which was a macroscopic aggregate of CNTs, a water-based dispersion of MWCNTs (100 mg, untreated or modified) and surfactant (0.1 wt% Triton X-100) was filtered through a PTFE membrane (Omnipore Membrane Filter PTFE, 1 μm pores). The CNT layer was peeled off from the membrane after drying at 35°C overnight. The CNP layer with desired dimension (55 mm × 55 mm) was obtained. The thickness of the buckypaper was measured in nine different positions by a digital feeler gauge.

Composite fabrication

The composite samples were fabricated by laminating the buckypaper between two GF fabric layers (60 mm × 60 mm square). The GF/CNP structure was soaked in 1.5 g resin (EL2 or Silikopon EF). Two copper electrodes (5 mm × 140 mm × 0.2 mm) were placed between the first GF layer and the CNP. This basic composite configuration was used to understand the effect of surface functionalization on the electro-thermal performance of the buckypapers. An optimised device configuration has to be considered to minimise the occurrence of thermal and mechanical stresses before being implemented on the industrial scale. A quick-drying silver paint from Agar Scientific was manually applied to the copper electrodes to attach with the CNPs to minimise the electrical resistance at the contact areas. The mould was made of two PTFE sheets covered with aluminium plates fixed with clamps. A schematic representation of the mould and the lay-up configuration is represented in Figure 1. The composites were heat-treated in an oven with different curing cycles. Silikopon EF resin was used as icephobic matrix. The soaked GF/CNP was placed in the mould and cured at 110°C for 16 h, with a heating ramp of 1˚C/min. While for EL2 laminating resin, it was mixed with AT30 SLOW hardener, with a weight ratio of 100:30, and stirred at 200 rpm for 30 min. The soaked GF/CNP was placed in the mould and cured in the oven at 65°C for 12 h, with a heating ramp of 1˚C/min.

Figure 1.

Schematic diagram of the layer-up configuration in the mould.

Characterisation

Microstructural characterisation

The microstructural characterisation was performed using Scanning electron microscopy (SEM), Transmission electron microscope (TEM), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS) and confocal microscope. TEM analysis was carried out by a Jeol 2100F FEG-TEM with field emission electron gun (FEG) providing electron source at 200 kV, equipped with a GATAN Orius SC1000 camera for high angle angular dark field (HAADF). A high definition field emission scanning electron microscope (JSM - 7100F) was used for morphological characterisation of the buckypaper and the corresponding composite samples. Raman spectra were collected at room temperature using a triple monochromator (T-64,000, Jobin-Ivon/Horiba Group, Kyoto, Japan) equipped with a charge-coupled device (CCD) detector. The excitation source used a 532 nm Nd:YVO₄ diode pumped solid-state laser (SOC JUNO, Showa Optronics Co Ltd, Tokyo, Japan) operating with a nominal power of 200 mW. The lateral resolution of the Raman microprobe was in the order of 1 μm. XPS analysis was performed by a photoelectron spectrometer JPS-9010 MC (JEOL Ltd, Tokyo, Japan) with an X-ray source of monochromatic MgK_α (output 10 kV, 10 mA). The testing chamber was controlled under a vacuum level of 2 × 10⁻⁷ Pa. To neutralize the surface charge during data acquisition, a low-energy electron flood gun was used to deliver the electrons to the sample surface. A 3D laser scanning confocal microscope (Keyence VK-X150K) was used to analyse the surface rugosity of the buckypaper.

Electrical resistance

Figure 2 shows an assembled CNP heating device for testing the electrical resistance of the buckypaper. The device was composed of two Polyether Ether Ketone (PEEK) films (70 mm × 70 mm × 100 μm) placed in between two aluminium sheets (60 mm × 60 mm × 2 mm), one CNP layer in contact with two T-shape copper electrodes. The CNP films were laminated between the two PEEK sheets. Clamps were used to fix the devices and to keep the buckypaper in contact with the electrodes.

Figure 2.

CNP heating device for electrical resistance and heating performance tests: (a) schematic diagram of the layer assembly; (b) an assembled CNP heating device. CNP: carbon nanotube buckypaper.

The prepared CNP and the respective composite samples were connected to a power supply (linear DC bench power supply CSI530S, by CircuitSpecialists) that provided a constant electrical potential difference (5 V) between the two electrodes. The electrical current was recorded and the resistivity of the sample was calculated by Ohm’s law

V = I R

(1)

R = ρ \frac{L}{S}

(2)

where V is the voltage (V), I is the electrical current (A), R is the electrical resistance (Ω), ρ is the resistivity (Ω m), L is the length of the sample (mm) and S is the cross section of the sample (mm²).

De-icing evaluation

The composite samples were placed in a cold chamber at −20°C and 100 mL of distilled water were nebulized and delivered into the chamber. A 2 mm thick rim ice layer was formed on the top surface of the samples. Then an electrical potential of 5 V was applied until the ice fully disappeared from the surface. An IR thermo-camera (FLIR E4, operating in Multi-Spectral Dynamic Imaging mode) was used to measure the temperature of the sample surface every 15 s from the start of the heating.

Results and discussion

Buckypaper

To study the microstructures MWCNTs after acid treatment, Raman spectroscopy has been used and the data is presented in Figure 3 after normalization. Both samples showed similar responses from the Raman spectra. The main features in the spectra are: D band, D′ band and G band. The D band, at 1350 cm⁻¹,^34–41 is caused by disordered structure or discontinuity/defect in the carbon nanotube network. The presence of disorder in sp²-hybridized carbon systems results in resonance Raman spectra.⁴² The G band at 1580 cm⁻¹ ^43–49 is induced by the crystalline graphitic structure. The distributed impurities, or defects in the nanotube structure, can cause the G-peak to split into two peaks, G-peak (1580 cm⁻¹) and D′-peak (1610 cm⁻¹). The split is due to the interaction between the vibrational modes of the impurities and the carbon atoms of the nanotubes structure.⁴² The intensity ratio between G and D peaks remains the same in MWCNT and NA-CNT, further indicating that the amounts of amorphous and crystalline carbon remain almost unchanged. The reason behind this unchanged amorphous and crystalline carbon after acid treatment was confirmed from the TEM studies as discussed in the next paragraph.

Figure 3.

Normalized Raman spectra for multi-wall carbon nanotube (blue line) and NA-CNT (green line).

Figure 4 shows the TEM images of untreated and acid treated MWCNTs. The amorphous carbon content on the surface of NA-CNT (Figure 4(b)) is lower than the untreated one (Figure 4(a)). The oxidant treatment could have reduced the amorphous carbon.⁵⁰ An obvious structural change occurred in NA-CNT due to acid treatment, as observed in Figure 4(c) and (d). The graphene layers of CNT present different misalignments. The red circles highlight the area in which the acid may have attacked the nanotubes and the functional groups may attach onto the external wall of CNTs. Raman spectra indicates that the relative ratio between the amorphous/disordered and crystalline carbon did not change, i.e., the loss of amorphous carbon was compensated by the defect generation.³⁴

Figure 4.

Transmission electron microscope images of: (a) untreated multi-wall carbon nanotubes; (b,c) defects (red circles) on multi-wall carbon nanotubes walls after nitric acid treatment; (d) a magnified image of one acid attacked position.

The changes in chemical states and bonding nature on the surface of carbon nanotubes have been analysed from the XPS study of C 1s core-level spectra. The deconvolution of the C 1s peak (Figure 5) helps to quantify the concentrations (at%) of oxygen containing groups (C-O, C=O and -COOH) in the samples, which is given in Table 1. NA-CNP corresponds to buckypaper made of nitric acid-treated CNTs; while MWCNP refers to buckypaper fabricated from untreated CNTs. The acid treatment creates defect sites on the graphitic network. The defects can be attributed to the result of carbon atoms being replaced by functional groups such as carboxylic acid (–COOH) and hydroxyl (–OH) on the CNT surface. The binding energy of the possible oxygen-containing groups is 286.0 eV for C-O, 287.1 eV for C=O and 288.8 eV for -COOH^51–55 respectively. Any signal with eV > 289 could be classified as p-p* transition.^51,56,57 Due to the acid treatment, the concentration of the functional groups on NA-CNT is higher than that on MWCNT confirming the successful addition of the functional groups on the surface of CNTs.

Figure 5.

C1s deconvolution for NA-CNT.

Table 1.

Quantitative results from C1s scan.

Samples	C 1s
Samples	C=C(at%)	C-C(at%)	C-O(at%)	C=O(at%)	-COOH(at%)	p-p*(at%)
MWCNT	51.6	34.7	3.8	2.7	1.8	3.5
NA-CNT	61.9	7.9	13.7	12.3	2.1	2.1

The characteristics of the buckypaper like surface area, thickness, surface roughness and their standard deviations are listed in Table 2. ΔThickness is the change in thickness compared with MWCNPs.

Table 2.

Characteristics of carbon nanotube buckypaper.

Samples	Surface area(m²)	Thickness(mm)	Thickness St. dev(mm)	ΔThickness(%)	Surface roughness(μm)	Surface roughnessSt. deviation(μm)
MWCNP	3 × 10⁻³	0.120	0.010	0.0	0.98	0.18
NA-CNP	3 × 10⁻³	0.081	0.012	−32.2	0.57	0.05

SEM images of the two types of buckypaper are shown in Figure 6. It is observed that nanotube aggregates are randomly distributed on the surface of MWCNP (Figure 6(a) and (b)), with length ranging from half to one micron. However, the amount of aggregates seems to be significantly reduced on the NA-CNP surface (Figure 6(c) and (d)), which could be attributed to the oxidation effect of MWCNTs.^34,50,58 Nitric acid is a strong oxidizing agent, so that some amorphous carbon would be removed, making NA-CNP surface much flatter than MWCNP.⁵⁰ The reduction of amorphous carbon allows a more compact (around 32% less thickness) configuration in NA-CNP, without sacrificing the electrical conductivity, as noticed from the electrical resistances in Table 3. HNO₃ oxidation freed the MWCNTs from the entanglements, causing a reduction of the CNTs aggregates. The lower surface roughness of NA-CNP than MWCNP, as observed from the laser microscope measurement, also confirms the reduced amount of nanotube aggregates in NA-CNP, which allows more compactness with reduced thickness.

Figure 6.

Scanning electron microscopy images of: (a) ×10000 MWCNP; (b) ×30000 MWCNP; (c) ×10000 NA-CNP; (d) ×30000 NA-CNP.

Table 3.

Electrical properties of carbon nanotubes buckypaper.

Samples	Resistance(Ω)	ResistanceSt. dev(Ω)	ρ(10⁻³ Ω m)	Δρ(%)	σ(S/m)	Δσ(%)
MWCNP	5.5	0.5	0.66	0.0	1513.2	0.0
NA-CNP	6.0	0.8	0.49	−26.0	2045.3	35.2

The electrical properties of buckypaper are listed in Table 3, including electrical resistance and the standard deviation, electrical resistivity ρ, and electrical conductivity σ. Δρ and Δσ are the changes in electrical resistivity and electrical conductivity, respectively, compared with the reference MWCNP. A wide range of power densities of self-heating elements have been reported in the literature. The power density values of 4–12 kW/m² for metallic self-heating elements³⁰ (e.g., Boeing 787 wings), and 1–33 kW/m² for nanomaterial based self-heating elements³⁰ (e.g., carbon nanotubes and graphene) were reported. The power density values of MWCNP and NA-CNP based self-heating elements are in same range of 1.5–1.6 kW/m².^18,19,30 Metallic self-heating elements are the standard products currently in use. In terms of energy consumption, nanomaterial based self-heating elements develop more heat than the metallic one with the same applied voltage thus could develop higher power density values.

The acid treatment has effectively reduced the electrical resistivity of the CNP layer and consequently, 35% increment in electrical conductivity is achieved in NA-CNP. Although the untreated and nitric acid-treated samples show close electrical resistances, a much thinner and more compact structure is formed using the acid-treated nanotubes. As a consequence, a significantly reduced electrical resistivity is obtained. Thus, nitric acid-treated CNTs are more effective for the construction of thin-layer of electrically conducting nanostructure network.

Characterisation and performance of the composites

The composite samples have been named based on the used thermosetting resins and buckypaper: EF-MWCNP for untreated buckypaper soaked in Silikopon EF resin; EL2-MWCNP for untreated buckypaper soaked in EL2 laminating resin and the similar for NA-CNP. The thickness of the composites is listed in Table 4. To obtain the thickness of the impregnated CNP, the GF fabrics thickness has been subtracted. The resin impregnation would slightly increase the CNP thickness. The increase in GF thickness caused by the resin impregnation has been considered negligible.

Table 4.

Characteristics of the fabricated composites.

Sample	Thickness
Sample	Composite(mm)	CNP/resin(mm)	St. dev(mm)
EF-MWCNP	0.618	0.248	0.019
EF-NA-CNP	0.589	0.219	0.018
EL2-MWCNP	0.645	0.275	0.020
EL2-NA-CNP	0.564	0.194	0.016

Figure 7 show SEM images of EF-MWCNP (Figure 7(a)), EF-NA-CNP (Figure 7(b)), EL2-MWCNP (Figure 7(c)) and EL2-NA-CNP (Figure 7(d)). From the SEM images, random distribution of the nanotubes impregnated with the thermosetting resin is observed. The buckypaper has been produced by filtration of nanotube dispersion (functionalised or untreated) and then impregnated with the resin. Both processes do not prompt any particular orientation or regular distribution of the carbon nanotubes. It is hard to notice clear differences in the distribution of the nanotubes.

Figure 7.

Scanning electron microscopy images (×60000) of samples cross-sections: (a) EF-MWCNP; (b) EF-NA-CNP; (c) EL2-MWCNP; (d) EL2-NA-CNP.

Despite the similar surface morphology, the composites have different electrical characteristics. The electrical properties of the composites samples are listed in Table 5. The composites demonstrate higher electrical resistivities than the CNP (MWCNP and NA-CNP) layer themselves. The comparison of electrical resistivity between CNP and their corresponding composites is illustrated in Figure 8. The impregnation of the insulating resin decreases the contact of the neighbouring CNTs, increasing the electrical resistivity of the CNP layer. In each type of resin, the composites based on NA-CNPs exhibit a larger increase in electrical resistance. MWCNTs are randomly distributed in each conductive layer, which creates pathways for the electrical currents. The increase in electrical resistivity is due to the electrically insulating effect from the thermosetting resin that penetrates into the inner space of CNP and aggregates MWCNTs in bundles. As NA-CNP are more compact than MWCNP (as demonstrated by lower thickness), the resin would impregnate with more difficulty through the NA-CNT bundles. This may disturb the intra-bundle interactions and increase the electrical resistivity of the respective composites.^10,32,59 The electrical resistivity of the composites impregnated with the icephobic Silikopon EF resin (MWCNP or NA-CNP) is lower than those of the samples impregnated with EL2. The impregnation of EF resin reduced the formation of CNTs bundles than that of EL2 resin, leading to a smaller increase in electrical resistivity.

Table 5.

Electrical properties of the fabricated composites.

Sample	R(Ω)	St. dev(Ω)	ρ(10⁻³Ω m)	σ(S/m)
EF-MWCNP	11.8	1.2	2.9	345.8
EF-NA-CNP	36.7	2.5	8.0	125.0
EL2-MWCNP	15.8	1.8	6.1	190.2
EL2-NA-CNP	51.5	1.5	10.0	100.6

Figure 8.

Comparison of electrical resistance between buckypapers and the composite samples with different resins.

Figure 9 shows the photos of de-icing tests and the corresponding IR thermal images during the ice melting. The composites were covered by rime ice at −20°C and a 5 V electrical potential was applied to provide the heating. This test condition was used to simulate an electro-thermal ice protection system under de-icing configuration.

Figure 9.

Optical images (first row) and the corresponding IR thermal imaging (second row) of the de-icing test for an EF-MWCNP sample: (a, d) at the beginning of the heating; (b, e) during the ice melting; (c, f) at the end of the ice melting.

Figure 10 shows an increase in surface temperature of two samples (EF-MWCNP and EL2-MWCNP) during the de-icing test. The ice melting time is defined as the moment in which the temperature rises above 0°C. The ice melting time is characterised by a plateau of temperature values, i.e., from 50 s to 315 s, which is visible in Figure 10. During the ice melting, the measured temperature was slightly below 0°C (around −1°C ∼ −2°C). The IR camera measured the temperature at the centre of the sample during the ice melting (as shown in Figure 9(e)), where an ice/water mixed phase was observed. The ice melting started when melting spots appeared on the sample surface, as shown in Figure 9(b) and the procedure continued until all the ice disappeared from the sample surface, as shown in Figure 9(c). As the EF-MWCNP have an average electrical resistance lower than EL2-MWCNP, the ice melting of EF-MWCNP samples (⁓ 200 s) is faster than that of EL2-MWCNP (⁓ 315 s), which is highlighted by the shorter plateau in Figure 10 (from 50 s to 200 s). NA-CNP based composites have not been tested due to higher electrical resistance. A high electrical resistance would cause longer ice melting time.

Figure 10.

Temperature versus heating time during the de-icing test of EF-MWCNP and EL2-MWCNP samples.

The power densities during the de-icing test are listed in Table 6. The surface of the sample is approximately 0.003 m². Both power densities were lower than 1 kW/m². A higher power density value was obtained when using Silikopon EF resin (0.68 kW/m²) over EL2 epoxy (0.52 kW/m²). Power density values around 2kW/m² were often reported in the literature.^9,10,30 For a more efficient ice-protection system supplied with constant voltage, the electrical resistance of the composite should be further reduced to increase the power density and reduce the ice melting time. The formation of CNT bundles due to resin impregnation increases the electrical resistance: bigger bundles will lead to a higher electrical resistance. The electrical resistance along CNT is much lower than the junction resistance. The electrical resistance in the CNP network can be illustrated by a series of resistors, consisting of compactly connected CNT bundles. Each bundle is like a node of this continuous network. The bundles are sets of joint areas made of nanotube agglomerates. The smaller size of these agglomerates, the lower will be the electrical resistance of these nodes. During the resin impregnation, a lower viscosity may allow the resin to penetrate into the CNP and create small CNT bundles, reducing the increment in electrical resistance.

Table 6.

Power densities of heating elements in the de-icing test.

Sample	Power density(kW/m²)	St. dev(kW/m²)
EF-MWCNP	0.68	0.09
EL2-MWCNP	0.52	0.05

Conclusions

This work investigates the microstructural changes in CNT based buckypapers due to surface functionalization by acid treatment and the effect on heating performance when impregnated with an ice-phobic resin. The composites have been prepared by laminating the CNPs (both acid treated and untreated) with glass fibre fabric followed by the impregnation in thermosetting resin (EL2 Epoxy laminating resin and Silikopon EF resin).

The results indicated that surface functionalization of MWCNTs was effective in obtaining MWCNT buckypaper with improved electrical conductivity. The composites impregnated with Silikopon EF have lower electrical resistivity than the composites impregnated with EL2. The impregnation of EF resin reduced the formation of big CNTs bundles. The composites based on NA-CNPs exhibit a higher increase in electrical resistance with both resins. This was probably due to the more compact structure of the NA-CNP. It would be more difficult to impregnate the resin through the NA-CNT bundles. Diluting the thermosetting resin would allow it to penetrate into the CNP and reduce the increment in electrical resistance. The de-icing test confirmed the electro-thermal heating performance of the CNP based composite in a climate chamber at −20°C, indicating the potential to be integrated into composite structures for ice protection purposes.

Footnotes

Acknowledgements

Kyoto Institute of Technology is acknowledged for the instrument access in the placement. The placement was supported by Prof. Pezzotti Giuseppe, Prof. Marin Elia and all the colleagues of the laboratory.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work is funded by the INNOVATIVE doctoral program. The INNOVATIVE program is partially funded by the Marie Curie Initial Training Networks (ITN) action (project number 665468) and partially by the Institute for Aerospace Technology (IAT) at the University of Nottingham.

ORCID iD

Francesco Zangrossi

Appendix

References

Pozegic

Anguita

Hamerton

, et al. Multi-functional carbon fibre composites using carbon nanotubes as an alternative to polymer sizing. Sci Rep 2016; 6: 37334.

Raji

A-RO

Varadhachary

Nan

, et al. Composites of graphene nanoribbon stacks and epoxy for Joule heating and deicing of s. ACS Appl Mater Interfaces 2016; 8: 3551–3556.

Palacios

Smith

Rose

, et al. Ultrasonic de-icing of wind-tunnel impact icing. J Aircraft 2011; 48: 1020–1027.

Farzaneh

Volat

Leblond

. Anti-icing and de-icing techniques for overhead lines. In: Atmospheric icing of power networks. Berlin, Germany: Springer, 2008, pp. 229–268.

Wang

, et al. Review of de-icing methods for transmission lines. In: 2010 International Conference on Intelligent System Design and Engineering Application, Changsha, China, 13–14 October 2010, pp.310–313.

Karim

Zhang

Afroj

, et al. Graphene-based surface heater for de-icing applications. RSC Adv 2018; 8: 16815–16823.

Hung

C-C

Dillehay

Stahl

. A heater made from graphite composite material for potential deicingapplication. J Aircraft 1987; 24: 725–730.

De Volder

Tawfick

Baughman

, et al. Carbon nanotubes: present and future commercial applications. Science 2013; 339: 535–539.

Zhao

Zhang

, et al. Carbon nanotube based composites film heater for de-icing application. In: 14th European Conference on Composite Materials, Budapest, Hungary, 7–10 June 2010.

10.

Chu

Zhang

Liu

, et al. Self-heating fiber reinforced polymer composite using meso/macropore carbon nanotube paper and its application in deicing. Carbon 2014; 66: 154–163.

11.

Yao

Falzon

Hawkins

, et al. Aligned carbon nanotube webs embedded in a composite laminate: a route towards a highly tunable electro-thermal system. Carbon 2018; 129: 486–494.

12.

Iijima

. Helical microtubules of graphitic carbon. Nature 1991; 354: 56.

13.

Dai

Mau

. Controlled synthesis and modification of carbon nanotubes and C60: carbon nanostructures for advanced polymeric composite materials. Adv Mater 2001; 13: 899–913.

14.

Tehrani

Safdari

Boroujeni

, et al. Hybrid carbon fiber/carbon nanotube composites for structural damping applications. Nanotechnology 2013; 24: 155704.

15.

Khabashesku

. Covalent functionalization of carbon nanotubes: synthesis, properties and applications of fluorinated derivatives. Russ Chem Rev 2011; 80: 705–725.

16.

Shieh

Chen

Twu

, et al. The effect of pH and ionic strength on the dispersion of carbon nanotubes in poly (acrylic acid) solutions. Polym Int 2012; 61: 554–559.

17.

Mylvaganam

Zhang

. Ballistic resistance capacity of carbon nanotubes. Nanotechnology 2007; 18: 475701.

18.

Aouraghe

Mengjie

Qiu

, et al. Low-voltage activating, fast responding electro-thermal actuator based on carbon nanotube film/PDMS composites. Adv Fiber Mater 2021; 3: 38–46.

19.

Aouraghe

Liu

, et al. Flexible, quickly responsive and highly efficient E-heating carbon nanotube film. Compos Sci Technol 2019; 183: 107824.

20.

Aouraghe

Xie

, et al. Highly stretchable, fast thermal response carbon nanotube composite heater. Compos A Appl Sci Manuf 2021; 147: 106471.

21.

Hong

, et al. Transparent, flexible heater based on hybrid 2D platform of graphene and dry-spun carbon nanotubes. ACS Appl Mater Interfaces 2019; 11: 16223–16232.

22.

Sun

W-J

Jia

L-C

, et al. Highly conductive and stretchable carbon nanotube/thermoplastic polyurethane composite for wearable heater. Compos Sci Technol 2019; 181: 107695.

23.

Geng

H-Z

Kim

, et al. Effect of acid treatment on carbon nanotube-based flexible transparent conducting films. J Am Chem Soc 2007; 129: 7758–7759.

24.

Behabtu

Young

Tsentalovich

, et al. Strong, light, multifunctional fibers of carbon nanotubes with ultrahigh conductivity. Science 2013; 339: 182–186.

25.

Zhao

Wei

Vajtai

, et al. Iodine doped carbon nanotube cables exceeding specific electrical conductivity of metals. Sci Rep 2011; 1: 83.

26.

Chen

I-WP

Liang

Zhao

, et al. Highly conductive carbon nanotube buckypapers with improved doping stability via conjugational cross-linking. Nanotechnology 2011; 22: 485708.

27.

Zhang

Leonhardt

Nguyen

, et al. Roll-to-roll continuous carbon nanotube sheets with high electrical conductivity. RSC Adv 2018; 8: 12692–12700.

28.

Hong

J-w

Jung

Yong

S-m

, et al. Radio-frequency transparent carbon nanotube electrothermal film for radome de-icing application. J Mater Res Technol 2020; 9: 10854–10862.

29.

Rashid

Liang

H-L

Taimur

, et al. Roll to roll coating of carbon nanotube films for electro thermal heating. Cold Reg Sci Technol 2021; 182: 103210.

30.

Yao

Hawkins

Falzon

. An advanced anti-icing/de-icing system utilizing highly aligned carbon nanotube webs. Carbon 2018; 136: 130–138.

31.

Zhao

Chen

Liu

, et al. Development of high-efficient synthetic electric heating coating for anti-icing/de-icing. Surf Coat Technol 2018; 349: 340–346.

32.

Zangrossi

Warrior

, et al. Electro-thermal and mechanical performance of multi-wall carbon nanotubes buckypapers embedded in fibre reinforced polymer composites for ice protection applications. J Compos Mater 2020; 54: 3457–3469.

33.

Zhao

JWC

, et al. Design and durability study of environmental-friendly room-temperature processable icephobic coatings. Chem Eng J 2019; 355: 901–909.

34.

Rosca

Watari

, et al. Oxidation of multiwalled carbon nanotubes by nitric acid. Carbon 2005; 43: 3124–3131.

35.

Strano

Dyke

Usrey

, et al. Electronic structure control of single-walled carbon nanotube functionalization. Science 2003; 301: 1519–1522.

36.

Dresselhaus

Jorio

Hofmann

, et al. Perspectives on carbon nanotubes and graphene Raman spectroscopy. Nano Letters 2010; 10: 751–758.

37.

Ferrari

Basko

. Raman spectroscopy as a versatile tool for studying the properties of graphene. Nat Nanotechnol 2013; 8: 235.

38.

Dresselhaus

Saito

, et al. Raman spectroscopy of carbon nanotubes. Phys Rep 2005; 409: 47–99.

39.

Ferrari

. Raman spectroscopy of graphene and graphite: disorder, electron–phonon coupling, doping and nonadiabatic effects. Solid State Commun 2007; 143: 47–57.

40.

Lebedkin

Blum

Stürzl

, et al. A low-wavenumber-extended confocal Raman microscope with very high laser excitation line discrimination. Rev Sci Instrum 2011; 82: 013705.

41.

Thomsen

Reich

. Raman scattering in carbon nanotubes. In: Light scattering in Solid IX. Berlin, Germany: Springer, 2006, pp. 115–234.

42.

Katagiri

Ishida

Ishitani

. Raman spectra of graphite edge planes. Carbon 1988; 26: 565–571.

43.

Graupner

. Raman spectroscopy of covalently functionalized single wall carbon nanotubes. J Raman Spectrosc 2007; 38: 673–683.

44.

Blackburn

Engtrakul

McDonald

, et al. Effects of surfactant and boron doping on the BWF feature in the Raman spectrum of single-wall carbon nanotube aqueous dispersions. J Phys Chem B 2006; 110: 25551–25558.

45.

Rao

Richter

Bandow

, et al. Diameter-selective Raman scattering from vibrational modes in carbon nanotubes. Science 1997; 275: 187–191.

46.

Gupta

Hughes

Windle

, et al. Charge transfer in carbon nanotube actuators investigated using in situ Raman spectroscopy. J Appl Phys 2004; 95: 2038–2048.

47.

Ajayan

. Nanotubes from carbon. Chem Rev 1999; 99: 1787–1800.

48.

Reich

Thomsen

. Raman spectroscopy of graphite. Philos Trans A Math Phys Eng Sci 2004; 362: 2271–2288.

49.

Knoll

Marchl

Kiefer

. Raman-spectroscopy of microparticles in laser-light traps. Indian J Pure Appl Phys 1988; 26: 268–277.

50.

F-H

Lee

C-Y

C-J

, et al. Purification of multi-walled carbon nanotubes through microwave heating of nitric acid in a closed vessel. Carbon 2005; 43: 727–733.

51.

Zhang

Xie

Varadan

. Functionalization of carbon nanotubes by potassium permanganate assisted with phase transfer catalyst. Smart Mater Struct 2002; 11: 962.

52.

Kim

J-K

Tang

. Functionalization of carbon nanotubes using a silane coupling agent. Carbon 2006; 44: 3232–3238.

53.

Zschoerper

Katzenmaier

Vohrer

, et al. Analytical investigation of the composition of plasma-induced functional groups on carbon nanotube sheets. Carbon 2009; 47: 2174–2185.

54.

Lipińska

Rebelo

Pereira

MFR

, et al. New insights into the functionalization of multi-walled carbon nanotubes with aniline derivatives. Carbon 2012; 50: 3280–3294.

55.

Okpalugo

Papakonstantinou

Murphy

, et al. High resolution XPS characterization of chemical functionalised MWCNTs and SWCNTs. Carbon 2005; 43: 153–161.

56.

Sezer

Koç

. Oxidative acid treatment of carbon nanotubes. Surf Interfaces 2019; 14: 1–8.

57.

Zhang

Q-Q

, et al. One-step preparation of water-soluble single-walled carbon nanotubes. Appl Surf Sci 2009; 255: 7095–7099.

58.

Datsyuk

Kalyva

Papagelis

, et al. Chemical oxidation of multiwalled carbon nanotubes. Carbon 2008; 46: 833–840.

59.

Tarfaoui

El Moumen

Boehle

, et al. Self-heating and deicing epoxy/glass fiber based carbon nanotubes buckypaper composite. J Mater Sci 2019; 54: 1351–1362.