Green preparation of carbon fiber/liquid silicone rubber composites for flexible electrode

Stretchable flexible conductive polymer composites (flexible electrodes) had become a research hot spot. In this paper, two-component room-temperature vulcanized liquid silicone rubber (LRTV) and short carbon fibers (CFs) were mixed by mechanical blending without solvent to prepare a tensile self-reply composites with high conductivity. The relationships between the average length, length distribution and content of CFs and the performance of CFs/LRTV composites were investigated. When the CFs length was 100 μm, the composites achieved a high conductivity. The composites conductivity threshold was reached when the CFs content was 3 wt%. In addition, the composites could be used as a conductor to light the bulb when the CFs content reached 8 wt%. The conductivity remained stable during cyclic stretching with a strain of 8%. The breaking and reconstruction of the internal 3D conductive network in the composites during the stretching process were discovered. The obtained results revealed that CFs/LRTV composites can be used as highly effective, flexible, stretchable electrode materials for stretchable displays, electronic skin, personalized healthcare.


Introduction
Flexible conductive materials had attracted a lot of attention, which could make up for the defects of rigid conductors of application scenarios such as stretching and bending, and thus had a wide range in application prospects such as stretchable displays [1][2][3][4], electronic skin [5], personalized healthcare [6][7][8][9].
Conductive composites made of polymer elastomers and conductive fillers were an important research direction. Hydrogels and rubber were the most commonly used flexible materials. However, hydrogels would lose strength and conductivity due to dehydration during use [10,11]. Rubber could be divided into carbonbased and silicon-based according to the main chain structure. Vulcanizing agents, vulcanization accelerators and antioxidants were added to the carbon-based rubber elastomer during the vulcanization process, and those additions would cause certain harm to the human body and the environment. Furthermore, carbon-based rubber was incompatible with the human body and was not accepted by the concept of green development [12][13][14][15]. Therefore, the application of carbon-based rubber was limited.
In recent years, many studies had been conducted on the preparation of conductive materials for silicone rubber [14][15][16][17][18][19][20][21]. Shao et al [16] prepared highly conductive silicone rubbers by forming a hyperbranched structure of MWCNTs with long-chain silicone rubber and co-crosslinking the hyperbranched structure with short-chain silicone rubber. Zhang et al [19] used PDA to treat silicone rubber surface, Ag was deposited on the silicone rubber surface due to the adsorption and reduction of Ag + by PDA, the silicone rubber achieved high conductivity. Yeo et al [21] and Jeong et al [20] investigated the application of silicone rubber in electronic skin. Huang et al [22] added CFs to methyl vinyl silicone rubber and used castor oil as a thixotropic agent, and the prepared flexible conductive material showed good conductivity (σ = 0.13 S cm −1 ) after the addition of 10 wt% CFs. However, the high viscosity of traditional silicone rubbers made fillers difficult to mix and would cause CFs to break during mixing. Two-component liquid room temperature vulcanized silicone rubber (LTRV) was a Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. new type of low-viscosity, self-vulcanising liquid silicone rubber. The low viscosity of LRTV allows the filler to be easily mixed into the matrix and reduces the possibility of breakage during mixing of fillers with longer lengths.
Conductive fillers were another important component of flexible conductive materials. They could be divided into metals [23,24] and carbon-based [25][26][27] (carbon fibers (CFs), carbon black (CB), graphene (G), carbon nanotubes (CNT)). Metal fillers were easily oxidized during use, resulting in a decrease in conductivity. Carbon based fillers had stable chemical properties, which could be more suitable for use as conductive fillers compared to metals. The size of CB, G, CNT was too small, and a large amount of filler needed to be added to catch a better conductivity. The study [17] proved that when the CB content was 23 wt%, the conductivity reached 10 −2 S m −1 . Among the carbon fillers, CFs has a large aspect ratio. A small amount of addition can achieve good conductivity effect. The small amount of filler could better maintain the material flexibility of materials. Huang et al [22] prepared composites with a conductivity of 0.13 S cm −1 by adding 10 wt% CFs to silicone rubber.
In this paper, new flexible conductive composites with stable conductivity during repeated deformation were prepared by direct mechanical blending of LRTV and CFs. The effects of CFs length and its distribution on the conductivity and mechanical properties were studied. The mechanism of CFs conductive bridge disconnection and reconstruction in the microstructure of composites was studied.

Materials
Two-component liquid room temperature vulcanized silicone rubber (LRTV) (A part (vinyl capped dimethyl silicone oil, alkyl platinum complex)/B part (methyl hydrogen containing silicone oil , vinyl capped dimethyl silicone oil) (viscosity 1.9 P.s) was purchased from Shenzhen Red Leaf Jie Technology Co. The bundled short-cut carbon fibers (5 mm×7 μm) were purchased from Guangwei Composites Co.

Preparation of CFs/SR composites
The preparation of composites was carried out according to the following steps: (1) The bundled CFs were broken up with a high-speed stirrer (32000 rpm). (2) Broken CFs were added to the mortar. (3) Component A of the silicone rubber was added to the mortar several times, keeping it stirred at all times during the addition process. (30 min) (4) Material B was added to the mortar. (5) The mixed silicone rubber was placed in a lowtemperature vacuum drying oven (10 min) to remove air bubbles in the silicone rubber. (6) The bubbled silicone rubber was added to the cylinder, and samples of different shapes were prepared by an injection molding machine. (injection pressure 150 bar, injection time 3 s, holding pressure 120 s, mold temperature 85°C, barrel temperature 35°C).

Characterization
The microstructure of the composites was recorded using a microscope (Leica). To obtain clear images of the microstructure, the composites were prepared into thin sheets and then placed under the microscope for observation. The microstructural changes during the stretching process were recorded using a computer.
Scanning electron microscopy (SEM, JSM-6390LV) had been used to obtain the composites formation. The sample was cryo-fractured in liquid nitrogen. In order to obtain clear and stable images, the sample surfaces needed to be sprayed with gold before being watched by SEM.
To test conductive of composites, electrochemical workstations were used to measure the current, the bulk conductivity of the material is calculated according to the equation (1) [17,28] ( ) Where L was the thickness (m) of the sample, S was the contact area of the sample with the electrode sheet, I was the current (A) through the material, U was the applied voltage (V). To achieve better contact between the material and the electrode sheet, the sample surface was coated with conductive adhesive and then placed between the two electrode sheets. To ensure the reliability of the experimental results, each sample was tested five times, and the average conductivity was taken as the conductivity of that sample. The mechanical properties of composites were tested by the A1-7000 tensile testing machine (Taiwan, China, range 0-30000 N), according to the national standard (GB/T 33429-2016). The samples were made into standard tensile specimens of 25´4´2 mm dumbbell type. The tensile rate was 500 mm min −1 . Each experiment was repeated five times at ambient temperature to ensure repeatability of the test results.
The Shao A hardness of composites were tested according to the Chinese National Standards JJG 304-2003. Five samples were used for testing.
Glass transition temperature (T g ) of composites was tested by differential scanning calorimetry (2°C min −1 , N 2 ).
The swelling resistance of the composites was characterized by the quality of the solvent absorbed by the material. According to S = (m 2 -m 1 )/m 1 × 100% alculate the mass change rate (S) of the sample. Where m 1 was the quality of clean samples, m 2 was quality of the sample after soaking in toluene solution for 72 h. Five samples were selected for each test, and the final results were averaged.

The length statistics of CFs in the composites
The CFs length and length distributions in the samples prepared by coating the mixed composites on slides and coverslips were observed by microscopy. To prevent statistical errors due to uneven distribution of fibers on the slide, each sampling was performed within a set 9 regions ( figure 2(a)). The number of statistics for each sampling was controlled between 500-600. The average length of CFs in the composites was 100 μm. And the length distribution of CFs was concentrated between 75 ∼250 μm. The longest CFs (550 μm) in the composites was 533 μm longer than the shortest CFs (17 μm). And the longest CFs was 1/10 of the length of commercially available chopped CFs, as shown in figure 2(b). When the puffing time was 300 s ( figure 3(c)), the statistical average length was 40 μm and the distribution was relatively centralized. The average length was reduced by half when the puffing time was 30 s compared with the puffing time of 300 s. And the longest CFs length at 300 s just reached the average CFs length at 30 s. The length reduction could reduce the material conductivity.    Compared with the bridge point between CFs ( figure 3(d)), electrons were more easily transferred through the graphitized CFs, so CFs had better conductivity. The electrical resistance inside the CFs was lower than the resistance of the bridge connection point formed between the CFs, making it easier for electron transfer.

Different content of CFs
The influence of CFs content on the composites conductivity was shown in figure 3(b). With the increase content of CFs, the conductivity of the composites increased rapidly. When the CFs content reached 3 wt%, the conductivity improvement decreased significantly. The content of 3 wt% was regarded as the composite conductivity threshold in this study. In our study only 3 wt% CFs was added, the conductivity threshold could be reached. Compared with the work (figure 3(c)), our study reduced the use of CFs. The two accepted theories of conductivity were as follows:(1) Gurland [33] suggested that electron transfer between the fillers in contact is the key reason for the conductivity of the composites, but the theoretical model is only suitable at high filler content.
(2) Voet [34] pointed out that electrons could be transferred across the insulating layer between fillers with the presence of an external electric field, and this theoretical model was applicable at both higher and lower contents. The Voet model had higher energy barriers across the insulating layer for electron transfer than the Gurland model. When the content of CFs was higher than 1 wt%, the bridging points increased rapidly, but the conductive network were not completely formed ( figure 1). The Voet model was dominated and higher energy was required for electron transfer. Until when the content of CFs reached 3 wt%, complete conductive network formed. Both models existing at the same time, less energy required compared to low filler content. After 3 wt% CFs content, the fully conductive network was constructed and the conductivity was limited by the CFs. The conductivity of the composites no longer increased rapidly with increasing content. The conductivity threshold could be reached by adding a small amount of CFs in two main ways. On the one hand, heavy proportion of long CFs in the filler could provide better electron channels. On the other hand, well-dispersed CFs allowed for easier formation of conductive networks. In this study, the bulb was lighted when the content of CFs in the composites was 8 wt% (figure 3(b)).

Mechanical properties of composites
The mechanical properties of the composites were shown in figure 4. The surfaces of the untreated CFs and LRTV could not form good interfacial wettability, and stress concentrations were easily formed at the interface of CFs and LRTV during the tensile process. The stress concentration led to the degradation of mechanical properties, as shown by the decrease of tensile strength and elongation at break (figures 4(a), (b)). CFs formed rigid 3D networks inside the composites, and these 3D networks led to an increase in the composites hardness. The tensile modulus of the composites was proportional to the CFs content when the tensile deformation was below 50%. However, the modulus was nearly constant. when the deformation exceeded 50% as shown in figures 4(a), (c). At tensile deformation below 50%, the three-dimensional framework of CFs inside the composite was not destroyed, so the composite modulus increased with the increasing CFs content. However, when the tensile deformation exceeded 50%, the CFs framework inside the composite was broken and the external forces were mainly carried by the LRTV, so the modulus almost remained unchanged with the increase of CFs content. The tensile fracture permanent deformation of the composite increased with increasing CFs content, as shown in figure 4(d). The tensile elongation at break of the composites ranged from 150% to 350%. When the composites were subjected to tensile fracture, the framework network formed by the CFs inside LRTV was broken, the rigid CFs was oriented in the tensile direction and the elastic recovery of the LRTV molecular chain was limited. Therefore, the permanent deformation of the composites increased.
The solvent resistance of the composites was characterized by studying the swelling of the composites by toluene as shown in figure 4(e). The swelling force of toluene on LRTV is fixed, and the 3D network formed by CFs inside LRTV could effectively improve the elastic shrinkage of the composites. The 3D network prevented further infiltration of toluene, which lead to the resistance of the composites to organic solvents as the CFs content increased. DSC curve of composites was shown in figure 4(f). The glass transition temperature (T g ) of composites increased with the increase of content of CFs, but the change was not significant. Because the addition of CFs would hinder the sliding of LRTV molecules and block their continuous movement.

The composites conductivity under cyclic tensile deformation
The change in resistance of the composite in the cyclic stretching condition was shown in figure 5. Composites conductivity reduced with increase deformation. Composites conductivity reduced for tensile deformation 8% ( figure 5(a)) was less than that for tensile deformation 20% ( figure 5(b)). Considering that the elastic return of the LRTV molecular chain was limited by rigid CFs inside the LRTV, the few oriented CFs could not return to their original positions leading to the reduction of the bridge linkage, and therefore the channels that could transmit electrons were reduced, which eventually lead to the increase of the resistance. The disconnected bridging point increased with tensile deformation, so large tensile deformations led to a greater increase in resistance of the composites.
3.6. The failure and reconstruction mechanism of CFs bridging Microscope images of the composites at different stages of stretching were shown in figure 6. The images of composites structure before and after stretching were shown in figures 6(a)-(f), respectively. The CFs originally lapped together were dislocated during the tensile stress process due to tensile deformation, as shown in figures 6(a), (d) and figures 6(b), (e). With the bridging points were broken, new bridging points would formed as shown in figures 6(c), (f). Fewer new bridging points formed than broken, which led to a reduction in the conductivity of the composites during tensile. The interruption and reconnection of CFs bridge points in the composites was better demonstrated in the video (see video).

Conclusion
This work provided a green preparation scheme of flexible electrode materials by mechanical mixing without solvents, LRTV was used as a flexible elastomer and CFs as a conductive filler that were harmless to humans and had high conductivity. The composites conductivity was related to the length and length distribution of CFs. The composites with longer CFs would have a better conductivity. The filler was easy to disperse in the low viscosity LRTV, which could reduce the mixing time, thus maintaining a better CFs length. The composites reached conductive threshold by adding a small amount of filler (3 wt%), and used as conductor to light bulb when the content of CFs was 8 wt%. ΔR/R 0 remained stable during cyclic stretching with 8% strain.

Data availability statement
No new data were created or analysed in this study.