Exfoliated Graphite Nanoplatelet-Carbon Nanotube Hybrid Composites for Compression Sensing.

In this study, we investigated the gauge factor and compressive modulus of hybrid nanocomposites of exfoliated graphite nanoplatelets (xGnP) and multiwalled carbon nanotubes (MWCNTs) in a polydimethylsiloxane matrix under compressive strain. Mechanical and electrical tests were conducted to investigate the effects of nanofiller wt %, the xGnP size, and xGnP:MWCNT ratio on the compressive modulus and sensitivity of the sensors. It was found that nanofiller wt %, the xGnP size, and xGnP:MWCNT ratio significantly affect the electromechanical properties of the sensor. The compressive modulus increased with an increase in the nanofiller wt % and a decrease in the xGnP size and xGnP:MWCNT ratio. However, the gauge factor decreases with a decrease in the nanofiller wt % and xGnP size and an increase in the xGnP:MWCNT ratio. Therefore, by investigating the piezoresistive effects of various factors for sensing performance, such as wt %, xGnP size, and xGnP:MWCNT ratio, the concept of one- and two-dimensional hybrid fillers provides an effective way to tune both mechanical properties and sensitivity of nanocomposites by tailoring the network structure of fillers.

Introduction

Carbon nanomaterials such as carbon nanotubes (CNTs), carbon black (CB), graphite, and graphene exhibit outstanding electrical properties and can easily form electrically conductive networks in various nonconductive polymer materials under the application of an external force.[1−18] These unique properties of carbon nanomaterials have been utilized for the development of sensitive strain sensors responding to external forces (compression, tension, etc.). For realizing the practical applications of carbon nanomaterials for the development of sensing devices, carbon nanocomposites consisting of a matrix and carbon nanofillers have been used. Especially, various efforts have also been made to functionalize carbon nanofillers for enhancing their electromechanical properties. The carbon fillers such as CNTs,[2,19,20] functionalized CNTs,[21,22] CB,[8,10,23,24] graphite,[11,25−27] and graphene[28−32] have been used to fabricate nanocomposites for sensing devices. The dispersion of nanofillers in the matrix and their properties and sensing capabilities can be optimized by controlling the processing conditions, raw material properties, nanofiller weight fraction (wt %), nanofiller conductivity, and barrier height of the polymer matrix.[33−37] The electrical properties of these carbon nanomaterials in nonconductive polymers have been investigated and verified by the tunneling effect to characterize the sensing capability of randomly distributed carbon nanoparticles in polymer matrices.[38−43] However, in all the previous studies carried out in this context, a single carbon nanomaterial was used in the polymer matrix. Recently, hybrid nanocomposites consisting of two types of carbon nanomaterials have been reported.[4,11,44,45] These hybrid nanocomposites showed better mechanical properties, optical energy densities, and piezoresistivities than nanocomposites consisting of a single carbon nanomaterial in the polymer. However, the material properties of these nanocomposites limit their practical applications due to their flexibility and sensitivity under compression. Sensing performance and mechanical properties of the hybrid nanocomposites consisting of carbon black and CNTs have been investigated, but there has been rather limited research on the electromechanical properties of flexible nanocomposites consisting of one-dimensional (1D) and two-dimensional (2D) fillers under compressive loading.[46−49] It would be of great research interest to investigate the physical interactions between the 1D and 2D nanofillers and how they affect the conductive path formation and piezoresistivity. It is also important to understand the correlation between the geometries, for example, length and lateral dimensions, of 1D and 2D conductive nanofillers and the electromechanical properties of hybrid nanocomposites. In this study, we fabricated a compressive strain sensor using a hybrid nanocomposite consisting of 1D multiwalled CNTs (MWCNTs) and 2D exfoliated graphite nanoplatelets (xGnPs) in a polydimethylsiloxane (PDMS) matrix. The MWCNTs and xGnPs interacted easily with each other under compression. We also investigated the piezoresistive effect of the composite sensor and the synergistic effect of the MWCNTs and xGnPs on the electrical and mechanical properties of the hybrid nanocomposites. We used xGnPs with different diameters (M5 or M15) and MWCNTs (CM250) at various concentrations (wt %) for the preparation of the hybrid nanocomposites. The relationship between the electrical properties and mechanical compressive strain of the sensors was also investigated. Therefore, this study will be helpful for developing hybrid nanocomposite compression sensors.

Results and Discussion

We fabricated flexible piezoresistive compression sensors with nanofiller wt %, two types of xGnPs with a difference in size, and various xGnP:MWCNT ratios (Table ). Fabricated samples are easily deformed in compression for their flexibility and show piezoresistive effect. The electrical properties of nanocomposites were optimized by controlling the dispersion and distribution of carbon nanofillers in the matrix (Scheme ). For dispersing carbon nanofillers in the polymer matrix, carbon nanofillers and PDMS are poured into a paste cup, and the resulting mixture was homogeneously dispersed using a paste mixer (500 revolutions and 400 rotations). After stirring it for 30 min, the mixture was further dispersed using the three-roll mill process with intense mechanical stirring. This mechanical stirring process was repeated more than 10 times. The well-dispersed mixture was then transferred to a mold. Mixtures consisting of carbon nanofillers in the polymer matrix were solidified at a high temperature (120 °C) and pressure (0.5 MPa) to finally obtain the hybrid nanocomposites with smooth surfaces and evenly dispersed carbon nanomaterials in the polymer matrix. The composites had a width, length, and height of 20, 20, and 1 mm, respectively. Finally, two Cu sheet electrodes were pasted at the top and bottom of the nanocomposites, and a mechanical force was applied to measure the z-axis resistance variations of the composites.

Table 1

Electromechanical Properties of the Hybrid Nanocomposites as a Function of the Size of xGnP, wt % of Fillers, and xGnP:MWCNT Ratio

filler type	total wt % of the filler	xGnP:MWCNT ratio	compressive modulus (MPa)	gauge factor
M5 and CM250	1	7:3	0.23	5.9
M5 and CM250	1	5:5	0.45	3.9
M5 and CM250	1	2:8	1.03	1.2
M5 and CM250	3	7:3	0.69	3.0
M5 and CM250	3	5:5	1.54	2.5
M5 and CM250	3	2:8	2.46	1.2
M15 and CM250	1	7:3	0.13	23.8
M15 and CM250	1	5:5	0.33	8.5
M15 and CM250	1	2:8	0.59	4.5
M15 and CM250	3	7:3	0.17	1.7
M15 and CM250	3	5:5	0.93	1.3
M15 and CM250	3	2:8	1.63	0.7
CM250	1	0:10	0.58	4.7
CM250	3	0:10	1.24	3.5

Scheme 1

Fabrication Process of the Hybrid Nanocomposite

To investigate the behaviors of the carbon nanomaterials (xGnPs and MWCNTs) in the polymer matrix under compression, we characterized the composite for investigating material interaction under compression by using Raman spectroscopy. Hybrid nanocomposites were placed between two cover glasses to measure the characteristic band changes of carbonaceous materials under compression as shown in Figure a. The sample surfaces were irradiated with a laser operating at 532 nm. To apply compressive stresses on the nanocomposites, two equal weights were placed on the cover glass symmetrically. The weights were increased in 200 g steps from 0 to 400 g, and Raman spectra were obtained in each step. Raman spectra exhibiting the G, D, and 2D peaks of pristine CNTs and xGnPs and the C–H peak of PDMS are shown in Figure b. The 2D peaks for the hybrid nanocomposites were recorded after each weight increasing step. As compressive loads are applied to a nanocomposite (xGnP:MWCNT ratio of 5:5 was chosen as an example), Raman spectra shows 2D peak shifts as evidenced in Figure c, which indicates load transfer from the polymer matrix to carbon nanomaterials.[50,51] As a reference, an identical measurement was performed on the MWCNT/PDMS composite film, and the respective Raman spectra are shown in Figure d, which shows a similar trend in 2D peak shifts as hybrid nanocomposites. The compressive strains induced altering the geometric configurations of CNTs and xGnPs due to the interfacial adhesion between the carbon nanomaterials and PDMS. As the compressive load increases, the 2D peak of nanocomposites shows a blue shift, that is, a decrease in wavelength. The results demonstrate the spatial rearrangements of the carbon nanomaterials under compressive loading, and the interfacial adhesion is the underlying mechanisms that contribute to the piezoresistive behavior, which enables the hybrid nanocomposites to serve as compression sensors.

Figure 1

Raman analysis of the hybrid composites. (a) The schematic of the Raman analysis setup. (b) Raman spectra of pristine MWCNT (CM250), xGnP, and PDMS. (c) Raman spectra of xGnP:MWCNT (5:5) nanocomposite as the applied load increases. (d) Raman spectra of MWCNT/PDMS composite as the applied load increases. Magnified images in panels (c) and (d) show the 2D peaks (near 2700 cm–1) in Raman spectra of nanocomposites showing the peak shift as the applied load increases.

To evaluate the electromechanical properties of the xGnP/MWCNT/PDMS nanocomposites, two copper electrodes were attached to the top and bottom surfaces of the samples using a conductive epoxy bond to obtain accurate electric signals during the compression test (Figure a). An acrylic plate was fixed to the aluminum compression zig to prevent the sensor output from electric signal interferences (Figure b).

Figure 2

Sample configuration setup and piezoresistivity setup. (a) Sample configuration showing the electric signal transfer method under compression. (b) Piezoresistivity measurement setup. The inset photo shows compression jig, sample, and multimeter (Photograph courtesy of “Changyoon Jeong”. Copyright 2019.).

To investigate the hybrid effect of the 1D MWCNTs and 2D xGnPs in hybrid nanocomposites, we used filler 1 and 3 wt %, various xGnP sizes, and xGnP:MWCNT ratios. Figure shows the conductivity of the hybrid nanocomposite consisting of the M15 xGnPs (large surface area (Figure a)) was higher than that of the composite consisting of the M5 xGnPs (small surface area (Figure b)) regardless of the xGnP:MWCNT ratio. The large surface area of M15 xGnP increases the number of contact points between MWCNTs and xGnPs, which facilitate the formation of conductive network under compression. In addition to the xGnP size effect, the conductivities of the xGnP M5 or M15/MWCNT/PDMS composites with various fillers ratios (7:3, 5:5, and 2:8) change with the composition of xGnP and MWCNT. An increase in the xGnP:MWCNT ratio resulted in a decrease in the conductivity of the hybrid composites, affecting compression sensing capabilities of the sensor due to a difference in initial resistance. In extremely low electrical conductivity, the electric responses of the sensor could hardly form conductive networks and were hampered by environmental (thermal or vibration) fluctuations. On the other hand, in very high electrical conductivity, the adverse effect of the contact resistance increases, and the sensitivity of the sensor decreases. Proper conductivity is very important in sensor design for sensing performance. The hybrid composite containing 1 wt % M15 xGnPs shows higher conductivity than that containing 1 wt % M5 xGnPs. This high conductivity facilitates the formation of new conductive networks under compression and influences sensitivity of the sensor. The 3 wt % hybrid nanocomposites, on the other hand, show too many conductive paths in the polymer matrix by lowering the sensitivity of the sensor. The generation of these conductive paths determines sensitivity of the sensor in response to compressive stress. This indicates that the proper conductivity of the composites affects the sensitivity of the sensor under compression. For the same reason, the effect of xGnP size is explained by the formation of conductive networks in the polymer matrix, affecting the sensitivity of the sensor. In addition to change of wt % and xGnP size, sensor sensitivity increases with an increase in their xGnP:MWCNT ratios. Nanocomposites with an xGnP:MWCNT ratio of 7:3 could readily form conductive networks under compression due to the strong interaction between the filler particles (Scheme ). This bridge effects induced by the combination of the 1D and 2D nanofillers are observed in all kinds of samples under compression regardless of the xGnP size and wt % depending on the xGnP:MWCNT ratio. In terms of flexibility, Figure shows that the rate of compressive stress increases with a decrease in the xGnP:MWCNT ratio regardless of the xGnP size and wt %. On the other hand, the sensitivity of the hybrid nanocomposites increases with an increase in the xGnP:MWCNT ratio (Figure ). Thus, changes of flexibility and conductivity depending on the xGnP:MWCNT ratio determine the sensor sensitivity under compression. The amount of filler (wt %) significantly affects the mechanical and electrical properties of the hybrid nanocomposites. With an increase in the filler weight fraction from 1 to 3 wt % (Figure a–d), compressive stress of the composites increased when samples were subjected to the same compressive strain because the compressive load could be easily transferred from the polymer matrix to the filler. The sensitivity of nanocomposites increased with a decrease in the filler wt % because saturated conductive networks disturb the formation of additional conductive networks under compression (Figure a–d).

Figure 3

Electrical conductivity of the hybrid nanocomposites as a function of the xGnP:MWCNT ratio and wt % of nanofiller.

Figure 4

Cross-sectional SEM images of the hybrid composites: (a) M15 xGnP/CM250 and (b) M5 xGnP/CM250. Lower images are low-magnification images.

Scheme 2

Compression-Sensing Mechanism of the Composites with Various xGnP:MWCNT Ratios (7:3, 5:5, and 2:8)

Figure 5

Compressive stress vs strain for (a) M15/CM250 7:3, 5:5, 2:8, and 0:10 (1 wt %), (b) M15/CM250 7:3, 5:5, 2:8, and 0:10 (3 wt %), (c) M5/CM250 7:3, 5:5, 2:8, and 0:10 (1 wt %), and (d) M5/CM250 7:3, 5:5, 2:8, and 0:10 (3 wt %).

Figure 6

Normalized change in the electrical resistance vs compressive strain for (a) M15/CM250 7:3, 5:5, 2:8, and 0:10 (1 wt %), (b) M15/CM250 7:3, 5:5, 2:8, and 0:10 (3 wt %), (c) M5/CM250 7:3, 5:5, 2:8, and 0:10 (1 wt %), and (d) M5/CM250 7:3, 5:5, 2:8, and 0:10 (3 wt %).

Figure shows five compressive loading–unloading cycles exerted on hybrid nanocomposites, strained up to ∼1%. The relative resistance decreased with increasing compressive strains and was fully recovered upon releasing unloading in both M15 (Figure a) and M5 (Figure b) xGnP hybrid nanocomposites.

Figure 7

Cycling compressive loading of hybrid nanocomposites strained up to 1%. (a) Change of electrical resistance vs time for M15/CM250 7:3 (1 wt %) and (b) M5/CM250 7:3 (1 wt %).

To better understand the effect of the amount of nanofiller on polymer, size of the nanofiller, and the ratio of the hybrid nanofiller, we compared compressive modulus depending on wt %, xGnP:MWCNT ratio, and xGnP size. Figure a shows that the compressive modulus increase is more obvious when wt % of filler in polymer increases. To illustrate the influence of xGnP size on mechanical properties, we used different sizes of xGnP, with 3 times the diameter difference, as the filler. Increasing compressive modulus can be observed with decreasing xGnP size in the nanocomposites containing hybrid fillers. The M5 xGnP could interact easily with the MWCNTs in the polymer matrix under compression as compared to the M15 xGnP in the same wt %. Regardless of xGnP size, we observed that, as the weight ratio of MWCNT in hybrid composites increases, the compressive modulus increases due to entanglement of MWCNT in fillers. Thus, the compressive modulus of the hybrid composites was determined by interaction between nanofillers.

Figure 8

Compressive modulus (a) and gauge factors (b) of the hybrid nanocomposites with different xGnP sizes and weight fractions (1 and 3 wt %), xGnP:MWCNT ratios (7:3, 5:5, 2:8, and 0:10).

For comparing sensitivity of the sensor under compression, the sensor sensitivity was quantified by the gauge factor defined aswhere ΔR is the relative change in electrical resistance, R0 is the initial electrical resistance, and ε is the compressive strain.

As shown in Figure b, wt %, xGnP:MWCNT ratio, and xGnP size influence the response to compressive strain. The gauge factor was mainly affected by the tunneling effect, which was strongly related to the distance between the nanofillers in the matrix under the compressive strain. The gauge factor of the sensor decreases with an increase in nanofiller weight fraction due to differences in flexibility. The flexibility of the nanocomposite depends on the CNT content in hybrid fillers, which also affects the sensitivity of the sensor. Increasing CNT content in the nanocomposites containing hybrid fillers weakens the sensitivity of the sensing in 1 and 3 wt % nanocomposites. Flexibility of the hybrid nanocomposite is also related to xGnP size. The hybrid composite with 1 wt % M15 xGnPs shows a higher gauge factor than the composite with 1 wt % M5 xGnPs under compression, showing the xGnP size effect. However, it can be seen that high conductivity due to high wt % affects the gauge factor. The results of changing CNTs and xGnP network structure, such as compressive modulus and gauge factor of nanocomposites, are summarized in Figure . As wt % of nanofillers increases, it reaches a point where additional conductive pathways do not significantly affect piezoresistivity under compression, while compressive modulus continues to increase. Figure also illustrates the hybrid effect that it enables the formation of a conductive pathway via bridging the gaps between neighboring CNTs, which exhibit high gauge factors in hybrid nanocomposites compared to that in CNT composites. Therefore, flexibility and conductivity, which are influenced by wt %, hybrid filler ratio, and filler size, affect the electromechanical properties of the hybrid nanocomposites.

Figure 9

Gauge factor vs compressive modulus curves for various nanocomposites showing the relation between flexibility and gauge factor.

Conclusions

The electrical and mechanical properties of hybrid nanocomposites consisting of xGnPs, MWCNTs, and PDMS polymer under compression were investigated. An increase in the weight fraction of the nanofillers increased the compressive modulus and decreased the gauge factor of the composites. This indicates that the filler weight fraction affected the mechanical properties of the composites, while their compression sensitivity was affected by the formation of additional conductive networks during compression. The xGnP size affects the electrical and mechanical properties of the hybrid nanocomposites. The nanocomposite with the M5 xGnPs shows better mechanical properties than that with the M15 xGnPs. However, the electrical properties of the composites depend on both the size and wt % of the nanofiller. In addition to these factors, the mechanical and electrical properties of hybrid nanocomposites were strongly related to the xGnP:MWCNT ratio regardless of the size and wt % of the nanofiller. The compressive modulus of the nanocomposites increases with a decrease in the xGnP:MWCNT ratio. This is because the dense 1D fillers strongly interlocked the 2D fillers and the polymer in low xGnP:MWCNT ratios. With an increase in the xGnP:MWCNT ratio, the gauge factor of nanocomposites increases because the “preplaced” 2D conductive fillers easily interacted with the 1D conductive fillers under compression. Hence, it can be stated that the xGnP size, xGnP:MWCNT ratio, and nanofiller wt % affect the compressive modulus and sensitivity of the nanocomposites under compression.

Experimental Section

Materials

The xGnPs (M5 and M15) used in this study were provided by XG Sciences. These xGnPs showed excellent electrical conductivity (parallel to surface: 107 S/m, perpendicular to surface: 102 S/m), a thickness of 6–8 nm, and average particle diameters of 5 (M5) and 15 μm (M15). The MWCNTs (CM250) used in this study were provided by Hanwha Chemical (Incheon, Korea). These MWCNTs had a length of 60–70 μm and a diameter of 3.5–4 nm. The MWCNTs showed a high aspect ratio (2 × 104), electrical conductivity (107 S/m), and dispersibility, which are important properties for sensor filler materials. The PDMS silicon-based organic polymer used as the matrix was purchased from SaeHwang Hi-tech. It showed a very low glass transition temperature (−125 °C), low shrinkage rate, ease of fabrication, and elastomer characteristics. The flexibility and sensitivity of the sensor fabricated in this study, which could be easily deformed by compressive strain, depended on the properties of the PDMS matrix.

Characterization

The morphologies of the hybrid nanofillers were characterized using field-emission scanning electron microscopy (FE-SEM; S-4800, Hitachi). Raman spectra were measured by scattered radiation of different wavelengths (confocal Raman; alpha300R, WlTec). For measurement of compression behavior of the sample, cover glass and weight enable us to measure Raman spectra at pressures down to several compression forces. The electrical conductivity of nanocomposites was measured by a four-point probe (CMT-SR1000N). The compression tests were carried out on a universal material testing system (Instron 5982) at a compressive strain speed of 1 mm/min at ambient temperature. The electrical signals were captured using a Keithley 2002 multimeter and a 7001-switching system, which were operated by programmed data acquisition software for compression.