Carbon fiber doped thermosetting elastomer for flexible sensors: physical properties and microfabrication

We have developed conductive microstructures using micropatternable and conductive hybrid nanocomposite polymer. In this method carbon fibers (CFs) were blended into polydimethylsiloxane (PDMS). Electrical conductivities of different compositions were investigated with various fiber lengths (50–250 μm), and weight percentages (wt%) (10–60 wt%). Sample composites of 2 cm × 1 cm × 500 μm were fabricated for 4-point probe conductivity measurements. The measured percolation thresholds varied with length of the fibers: 50 wt% (307.7 S/m) for 50 µm, 40 wt% (851.1 S/m) for 150 µm, and 30 wt% (769.23 S/m) for 250 μm fibers. The conductive composites showed higher elastic modulus when compared to that of PDMS.

More recently, conductive polymeric composites have been extensively employed in the fabrication of energy storage devices, batteries, 3D printing materials, electromagnetic shielding, flexible displays and smart sensors [30][31][32][33][34] . Compact network of conductive fillers in a polymeric matrix is essential for the development of the composites with high conductivity [35][36][37] . The characteristics of carbon fibers, such as, high aspect ratio, high conductivity and non-spherical shape are advantageous to achieve percolation path at lower concentration. Most widely used carbon-based polymer composites include polyamides, polymethylmethacrylate (PMMA), acrylonitrile butadiene styrene (ABS), and polyvinylalchol (PVOH). However, many of these are either not flexible and/or difficult to fabricate in micro-scaled features. Therefore, exploration of a flexible materials with better mechanical strength are of great interest.
PDMS is a silicone-based elastomer, which has been widely used as the elastomeric substrate for stretchable electronics. It has been easily and extensively employed for micromolding and fast prototyping of microdevices, such as in the fabrication of microfluidic chip and the stamping for pattern transfer via soft lithography [38][39][40] . Owing to its diverse properties such as transparency, biocompatibility and flexibility, it has been widely used for fabricating micromixers, microchannels, pumps etc. However due to its non-conductive behavior, the incorporation of electrical functionality has lagged. PDMS is very difficult to pattern conductive pattern lines on it, due to its poor adhesion with metal and conductive polymers 41,42 . Therefore, the integration of the conducting patterns on PDMS is a critical issue, for the fabrication of microsensors, actuators and micropumps, which need electrodes for signal detection. The incorporation of conductive lines is extremely important for signal routing, signal processing electronics and in some cases to supply power. Therefore, in this paper we present a flexible, conductive PDMS composites by incorporating micrometer size pitch-based milled CF. Moreover, we could construct 3D conducting microstructures with excellent mechanical strength incorporated into the PDMS. Fabrication. Conductive hybrid polymer was prepared using PDMS from elastomer kit by mixing base elastomer and curing agent in ratio of 10:1 weight ratio and degassed to get rid of the bubbles formed during the mixing using a vacuum desiccator for 30 min at room temperature ( Fig. 1). Carbon fibers with different wt% (10,15,25,30,40,50, and 60) were added and stirred manually in the PDMS matrix. Further, sample were made in the rectangular shape of 2 cm × 1 cm size, and heat pressed at 120 °C at 10 Pa for 3 min to achieve a final thickness of 500 μm.

Microfabrication of Carbon Fiber PDMS Nano composites. PDMS-based electrically conductive
composite were micropatterened using hybrid fabrication process shown in Fig. 2. A 3D-printed master with negative tone features, was fabricated using an Objet30 Pro 3D printer.

Results and Discussion
CF with different wt% (10, 15, 40 and 50) were added and stirred manually in the PDMS matrix. Samples were made in the rectangular shape of 2 cm × 1 cm size, and heat pressed at 120 °C at 10 Pa for 3 min. Figure 3 shows scanning electron microscopic (SEM) images of the carbon fibers in the PDMS matrix with different wt% of carbon fiber (average length of 50-250 μm).
Hybrid fabrication process involves pouring electrically conductive nanocomposites on the 3D printed mold and using spin coater at 1000 rpm for 4 min settling the composites on the mold evenly. The excess amount of the composite was wiped off the master using IPA wipes. Undoped PDMS was poured on the master to transfer the conductive features. Hybrid composite was cured at 110 °C under heat plate for 30 min and peel off from the master. This gives the electrically conductive features on the nonconductive PDMS as shown in the Fig. 4.

Conductivity Measurements.
Sample size of 500 µm thickness was fabricated and tested with the 4-point terminal sensing to eliminate contact resistance. 3D printed jig was used to place the sample to measure resistance. Known current was passed through the sample using the input leads, and sense leads measured the voltage through the sample 43 . Resistance between the input and sense leads was calculated using the Ohm's law. Since no current is withdrawn from the sense leads, contact resistance could be eliminated thus giving bulk resistance. The distance between the leads and sample geometry was known resistivity of the sample and was calculated using Resistivity (ρ) was calculated by bulk resistance multiplies thickness (t) for variant weight percentage of filler. Samples with different length of the fibers were fabricated and resistance was test was performed as shown in the Fig. 5. From Fig. 5 it can be seen that the resistance values decrease with the increase in the wt% of carbon fibers. Further, Table S1 shows the different values of resistivity as a function of the length of the fibers with the wt% remaining constant.
The electrical conductivity of CFs in PDMS matrix as a function of wt% was measured using the setup shown in Fig. 6. From Fig. 7, it is observed that the conductivity increases slowly up to 30 wt% for 250 μm length fiber, 40 wt% for 150 μm and 50 wt% for 50 μm. However, conductivity increases swiftly after these points. This variation can be explained in terms of percolation theory. The percolation path is not set up at lower concentration of carbon fibers, as the concentration of fibers increases conducting percolation paths are setup 44,45 . Insulator to conductor transition point on fibers weight percentage plot is referred as percolation threshold. The percolation limit for carbon fiber-PDMS composites for 250 μm, 150 μm and 50 μm length fibers were at 30 wt%, 40 wt% and 50 wt%.     results of PDMS and carbon fibers shows that due to the incorporation of fibers in the PDMS matrix enhance the thermal stability of composite. However, thermal degradation of PDMS and PDMS-Carbon fibers 15 wt% were left with around 25% of the total weight. Whereas 77-80% weight was remaining for PDMS-Carbon fibers 30 wt% and 50 wt% at around 780 °C as shown in Fig. S1. Furthermore, in PDMS-Carbon fibers composites starts degrading at <170 °C and weight loss of 5-7% occurred at 250 °C. The first derivative of thermo-gravimetric (DTG) shows the small shoulder on the TGA. DTG plot confirms the thermal degradation at 480 °C and 780 °C for the composites 46 .
Differential Thermal Analysis (DTA). Further, (DTA-Heat flow) results shows the PDMS and PDMS-carbon fiber samples exhibits small endothermic peak at 180 °C showing loss of adsorbed water as shown in Fig. S2. After a gradual weight loss, a sharp weight loss was observed accompanied by the large broad exothermic peak shows in the DTA curve suggesting the decomposition of some organic compounds. After ∼620 °C the weight loss remained approximately half of the initial weight and the DTA analysis proved endothermic reaction. This could be ascribed to the PDMS weight rapidly decreased after 400 °C which accompanied by large shrinkage and even cracks in the composite (Fig. 8). The overall weight loss is higher in the samples with lower CFs content.
Dynamic Mechanical Analysis (DMA). To investigate the mechanical properties of the hybrid polymer at varying temperature conditions, DMA was carried out from room temperature to 150 °C. DMA results shown in Fig. 9 for pure PDMS and PDMS-carbon fibers composites demonstrates storage modulus (E′) and loss modulus (E′′) against temperature at frequency of 10 Hz. The storage modulus of pure PDMS has been found to be about 45 times lower than that of PDMS 50 wt% CFs composites. Storage modulus for both cases has been found to remain almost constant throughout the whole temperature region suggesting good mechanical performance at elevated  temperature on the other hand loss moduli for both cases showed monotonous decrease with increase in temperature signifying decreased elasticity at elevated temperature. For both samples loss moduli are lower than the storage moduli where loss modulus of PDMS 50 wt% CF has been found higher than the pure PDMS. So, it can be said that doping CFs in PDMS matrix the mechanical properties can be improved to a large extent.

Conclusion
In conclusion, we demonstrated the microfabrication of electrically conductive polymer. Further, we investigated the electrical and mechanical properties as well as the carbon fibers dispersion of carbon fibers/PDMS composites for varying carbon fiber weight ratios. SEM analysis shows that the carbon fibers were dispersed with little agglomeration. It was observed that the percolation threshold for carbon fibers with length 250 μm, 150 μm and 50 μm were observed at 30, 40 and 50 wt% respectively. We compared the thermal stability of PDMS and PDMS-carbon fibers with various wt%. TGA results showed the thermal stability improved with the increase in wt% of carbon fibers and DTA analysis showed the better thermal degradation properties of PDMS-carbon fibers. Hybrid PDMS nanocomposites showed enhancement in elastic modulus than that of pure PDMS. Thus, the developed micropatternable composites can be used to develop flexible and stretchable micro-sensors (chemical and physical), conductive lines or traces, elastomeric printed circuits and flexible sensors for Internet of Things (IoT) and for lab-on-chip devices.