[14–16%(chloropropyl)methylsiloxane]-dimethylsiloxane, LMS-152 (M_n = 8,750 g·mol⁻¹, ∼14 chloropropyl groups per polymer) was obtained from Gelest Inc. Silanol-terminated polydimethylsiloxane, C2T (M_n = 22,000 g·mol⁻¹) was purchased from Wacker Chemie AG. Toluene, chloroform, ethyl acetate, and 1-ethylimidazole were acquired from Sigma-Aldrich. Trimethoxylsiloxane (crosslinker) and dibutyltin diacetate (Sn catalyst) were obtained from Sika Technology AG. Silicon dioxide amorphous hexamethyldisilazane-treated particles (SIS6962.0) were acquired from Fluorochem. All products were used as received.

LMS-EIL was prepared by nucleophilic substitution of chloropropyl groups with 1-ethylimidazole, then purified by extraction. For synthesis, 1-ethylimidazole (1.50 g, 15.60 mmol), LMS-152 (10.00 g, 1.14 mmol) with 15.60 mmol chloropropyl groups, and toluene (10 ml) were mixed together in a single-neck round-bottom flask. The molar ratio of 1-ethylimidazole-to-chloropropyl groups on LMS-152 was 1, as in previous work on IL synthesis (Jiang et al., 2008; Zhang et al., 2017). The reaction mixture was deoxygenated by bubbling nitrogen through it for 10 min before sealing with a rubber stopper. The reaction was then carried out at 85°C for 24 h, after which it was stopped by cooling the flask to room temperature. Purification was carried out as follows: the initial product was obtained by removing toluene in a rotary evaporator; the product was then washed three times with ethyl acetate to extract the unreacted 1-ethylimidazole; the final product was then dried in a vacuum oven for 24 h.

We reported the optimal formulation for preparing condensation-cured elastomers in a previous study on the influence of stoichiometric imbalance (Jurásková et al., 2020). C2T (5.00 g, 0.22 mmol), cross-linker (0.10 g, 0.76 mmol), and treated silica particles (10 phr) were mixed together using a dual asymmetric centrifuge at 3,500 rpm for 3 min (FlackTek Inc. DAC 150.1 FVZ-K SpeedMixer). The catalyst dibutyltin diacetate (0.5 wt%) and LMS-EIL were then added, and the mixture was speed mixed again. The uniform mixture was cast with a coating gap of 300 μm onto a polyethylene terephthalate (PET) substrate using a film applicator (3,540 bird, Elcometer, Germany). The coated samples were subsequently cured for 48 h in a humidity oven at 25°C and 80% relative humidity.

The sensing layer for the pressure sensor was comprised of pure PDMS elastomer or a PDMS composite with either LMS-152 or LMS-EIL added (dimensions: 10 mm length × 5 mm width × 1 mm thickness). Silver electrodes were then sputtered onto both top and bottom layers at a thickness of 30 nm to form a sandwiched plate capacitor using a sputter coater (Cressington, model 208HR) under vacuum conditions. Copper wires were used to connect the impedance analyzer to the specimens.

The dispersion of LMS-EIL additive within the elastomer matrix was investigated using FE-SME and FE-SEM coupled with EDS. Figure 1 shows the FE-SEM surface images of pure elastomer, elastomer with 10 phr LMS-152, and elastomers modified with different amounts of LMS-EIL. Magnified FE-SEM and optical cross sectional images of these elastomers are shown in Supplementary Figures S2, S3. The elastomers with 5 and 10 phr LMS-EIL clearly display uniform morphologies comparable to that of both the pure elastomer and the elastomer with LMS-152. Furthermore, the EDS images (Supplementary Figure S4) of the elastomer with 10 phr LMS-EIL quite clearly demonstrate the presence of N and Cl uniformly dispersed throughout the sample. However, LMS-EIL droplets are clearly visible in the elastomers with 15 and 20 phr LMS-EIL—indicating that aggregation has taken place; the size of these droplets increased with increasing LMS-EIL concentration, indicating more serious aggregation in the resultant elastomer.

The effect of LMS-EIL concentration on the dielectric properties of composite elastomers was investigated using dielectric spectroscopy. As the comparison with the pure elastomer in Figure 3A clearly illustrates, the elastomer modified with LMS-EIL displays a higher ε_r across the whole test frequency range. The ε_r at 10⁻¹ Hz of the elastomers increased initially from 3.7 to ∼1,000 with increasing LMS-EIL content from 5 to 20 phr, then decreased significantly to 280 for the 25 phr sample, as detailed in Supplementary Table S2. The increasing relative dielectric permittivity of the LMS-EIL-modified elastomers is due to the increased concentration of high-permittivity IL, while the subsequent decrease results from the aggregation of LMS-EIL in the elastomer matrix, as shown in Figure 1. Moreover, the elastomers with 10 and 15 phr LMS-EIL both display noticeable ε_r plateaus at frequencies below 10² Hz and 10⁴ Hz, respectively—consistent with the dielectric behavior of the LMS-EIL additive shown in Figure 2D. As noted in Figure 3A and Supplementary Table S2, the ε_r of the elastomer with 10 phr of LMS-EIL increased 6.7-fold at 10⁻¹ Hz compared to that of the elastomer with LMS-152—indicating that, as expected, it is more efficient to modify elastomers with high-permittivity LMS-EIL than with the pristine chloropropyl silicone oil LMS-152.

At 10⁻¹ Hz, the tanδ values of elastomers with 5 and 10 phr LMS-EIL were 0.07 and 0.06, respectively—lower than that of both the pure elastomer and the elastomer with LMS-152. At 10⁶ Hz, however, elastomers with LMS-EIL showed higher tanδ values compared to both the pure elastomer and the elastomer with LMS-152. This can be explained by the fact that, as seen in Figure 2E, LMS-EIL displays a low loss tangent at low frequency, thus decreasing the tanδ of the resulting elastomers. Nevertheless, when the LMS-EIL content was further increased to more than 10 phr, the tanδ of the resulting elastomers increased above that of both the pure elastomer and the elastomer with LMS-152 at 10⁻¹ Hz.

Figure 3C shows the complex conductivity of the studied elastomers. It is notable that σ^* of the elastomers modified with LMS-EIL increased with increasing LMS-EIL content for the range of 0–15 phr; above 15 phr, however, the elastomers showed inconsistent σ^* values due to the lack of homogeneity of the elastomers (Figure 1). Due to the higher σ^* of LMS-EIL compared to LMS-152, the σ^* of the elastomer with 10 phr LMS-EIL was 2.6 and 7.6 times higher than that of the corresponding elastomer with LMS-152 at 10⁻¹ and 10⁶ Hz, respectively, as shown in Figure 2F. Importantly, while the elastomer with 10 phr LMS-EIL had the highest σ^* at 10⁻¹ Hz, both the pure elastomer and the elastomer with LMS-152 displayed σ^* values on the same order of magnitude, indicating that the elastomer with 10 phr LMS-EIL possesses extremely low complex conductivity and thus significant potential as a dielectric material.

In flexible capacitive pressure sensors, the relative dielectric permittivity and softness of the dielectric layer both decisively influence capacitance and sensitivity. Tensile testing was therefore performed to evaluate the Young’ modulus as well as tensile strength and strain of PDMS elastomers with different LMS-EIL contents. As shown in Figure 4A and Supplementary Table S3, both Young’ modulus and tensile strength were highest in the pure elastomer, as expected, since both of these values tend to decrease with increasing LMS-EIL content due to softening effect of the additive on the elastomer/dilution of the network (Shivapooja et al., 2016); this effect is evident from the decrease in shear modulus with increasing LMS-EIL content shown in Supplementary Figure S5. A decrease in both Young’s and shear moduli—i.e., softening—makes it easier to alter the shape of dielectric layer in order to improve sensor sensitivity. The elastomer with 10 phr LMS-EIL possessed only a slightly higher Young’s modulus (0.78 MPa) and slightly lower tensile strength (0.54 MPa) compared to the elastomer with the same concentration of LMS-152 (0.76 and 0.56 MPa, respectively), indicating no significant difference in softening effect between the two additives with and without IL. Figure 4B shows the strain at break for all elastomers studied. The elastomer with 10 phr LMS-EIL exhibited a strain at break of ∼321%, higher than that of the pure elastomer as well as the elastomers with 5 or 15–25 phr LMS-EIL. These results indicate that gains in elastomer strain at break resulting from the addition of LMS-EIL are mitigated at concentrations above 10 phr due to the stress concentrations at aggregation areas in the elastomer matrix. Compared to the other composites with additives (Table 1), the excellent combination of dielectric and mechanical properties achieved by the elastomer with 10 phr LMS-EIL can be attributed to the high permittivity, softness and high compatibility of the IL-grafted functional oil.

As shown in Figure 4C, the cyclic loading-unloading curves indicate that the elastomer with 10 phr LMS-EIL has a low mechanical hysteresis; further, the fact that the mechanical hysteresis loop becomes more stable after the first cycle suggests excellent long-term mechanical stability. Indeed, the elastomer with 10 phr LMS-EIL maintained 94% elastic energy after 100 loading-unloading cycles, as shown in Supplementary Figure S6 and Supplementary Table S4. Overall, this elastomer displays excellent mechanical properties due to a combination of inherent softness, sufficiently high strain at break and elasticity, and long-term stability.

Gel fractions of LMS-EIL-modified elastomers were investigated via swelling experiments in order to determine the amount of bonded (gel fraction) and non-bonded species in the elastomer networks. Figure 4D and Supplementary Table S5 list the gel fractions of the elastomers studied: gel fraction decreased with increasing LMS-EIL content, indicating that a significant amount of unreacted PDMS chain remained in the PDMS matrix in elastomers with high concentrations of LMS-EIL. This may be due to a slight inhibitory effect of IL on the Sn catalyst (Skov and Yu, 2018; Liu et al., 2019), attenuating its catalytic efficiency in the condensation curing reaction. The elastomer with 10 phr LMS-EIL had a gel fraction of 92%, within acceptable limits for silicone networks and comparable to our previous results for soft elastomers (Madsen et al., 2018; Liu et al., 2019). This elastomer also displayed a relative viscous loss of ∼2% at 10² Hz (Supplementary Figure S5), indicating that it maintained a high degree of crosslinking and mechanical integrity. Therefore, the silicone composite presented a little viscoelastic effect in cyclic loading and unloading (Figure 4C).

Pressure sensors were prepared using LMS-EIL-modified PDMS elastomers as dielectric layers and their sensing performance was evaluated. The procedure for preparation and characterization is shown in Supplementary Figure S8. The photographs for elastomer with 10 phr LMS-EIL and the obtained capacitor are shown in Supplementary Figure S9. Figure 5A shows the relative capacitive change (ΔC/C₀) as a function of applied pressure (P) for sensors utilizing pure elastomer as well as elastomers with various LMS-EIL and LMS-152 additions: the higher ΔC/C₀ observed for the sensor with 10 phr LMS-EIL compared to the sensor with LMS-152 can be attributed to the LMS-EIL elastomer’s higher relative dielectric permittivity (Figure 3A). The ΔC/C₀ of sensors utilizing LMS-EIL initially increases as a result of the increased ε_r and decreased Y of the elastomers before decreasing once a concentration threshold is reached due to LMS-EIL aggregation (Figures 1D,E).

The sensitivity of a capacitive pressure sensor (S) is defined by the equation: S = ((∆C/C₀))/P, where ΔC is the change of capacitance and C₀ is the capacitance without applied pressure. According to this equation, sensitivity increases with greater variation in the relative capacitive change per unit pressure. Figure 5A shows the sensors’ relative capacitive change at low and high pressure (Stage 1 and Stage 2). At Stage 1, ΔC/C₀ changes rapidly with increasing pressure in the range of 0–7.5 kPa (highlighted in purple). Pressure sensor sensitivities were calculated by fitting ΔC/C₀ linearly from 0 to 7.5 kPa, and the resulting sensitivities and R-squared values are summarized in Figure 5B and Supplementary Table S6. Sensors with LMS-EIL displayed increasing sensitivity before beginning to decrease at LMS-EIL concentrations as a result of the more inhomogeneous composites. The combination of low Y, high ɛ_r and morphology homogeneity in the elastomer with 10 phr LMS-EIL produced the sensor with the highest sensitivity of 0.51 kPa⁻¹, which is 100 and 51 times higher, respectively, compared to the sensors based on pure elastomer (0.005 kPa⁻¹) and a 10 phr LMS-152 addition (0.01 kPa⁻¹). The promising sensitivity and linear sensing pressure range of sensor with 10 phr LMS-EIL compare well against other recent reports of sensors with additives (Table 2). In addition, the sensor with 10 phr LMS-EIL shows 51-fold higher sensitivity as compared to that of the sensor with 10 phr LMS-152, while the relative permittivity just presents a 6.7 times increase (Figure 3A). The significant increase in pressure sensitivity may be due to the higher relative permittivity change of the potentially nanoporous structure with compression. The R-squared value from the linear regression for the sensor with 10 phr LMS-EIL is 0.99, indicating that the sensor displays almost perfectly linear response in the pressure range of 0–7.5 kPa due to the excellent dispersion of LMS-EIL additive in the dielectric layer (Figure 1C) and the robust mechanical properties (Figure 4C). The importance of dispersion quality and robust mechanical properties is further stressed by the sensor with 15 phr LMS-EIL which exhibits a really low R-squared value (R ² = 0.67) compared to the 10 phr sample. At Stage 2 (highlighted in olive green), sensors were subjected to higher pressure (7.5–17.5 kPa). The sensors with both LMS-EIL and LMS-152 exhibited much lower sensitivity under higher pressure—typical behavior for flexible capacitive sensors that experience an increase in the elastic resistance (Yin et al., 2018; Li et al., 2020).

The relative capacitance change of elastomers with LMS-EIL became saturated with increasing pressure beyond a certain pressure, such as 10 kPa for elastomer with 10 phr LMS-EIL. This may be due to the elastomer with LMS-EIL behave like a nano-porous foam that has the included oil being squeezed out above a certain pressure threshold. However, the SEM images in Figure 1 do not show any foam-like structure, most likely because the LMS-EIL and PDMS chains are fully miscible, and thus they cannot be observed on the micro-scale. In addition, the saturation pressure (10 kPa) of elastomer with 10 phr LMS-EIL is relatively small as compared to the Young’s modulus (0.78 MPa, Figure 4A) of the silicone elastomeric composite. The Young’s modulus was determined by tension testing, while the applied pressure when sensor testing is related to the compressive modulus.

The cyclic relative capacitive change for the pressure sensors with 5, 10 and 15 phr LMS-EIL at various pressures ranging from 0 to 18 kPa is shown in Figure 5C. The amplitude of this change is nearly unaltered after 100 loading-unloading cycles at two stages for the sensors with 5 and 10 phr LMS-EIL, while the sensor with 15 phr LMS-EIL now exhibits serious deviation. This non-coincident cyclic response behavior is the result of LMS-EIL aggregation in the elastomer (Figure 1D).

The response of the sensor constructed with 10 phr LMS-EIL under 1, 4, and 9 kPa is shown in Figure 5D. This sensor shows a stable capacitive response and excellent repeatability under nine loading-unloading cycles at each pressure. It also exhibits a rapid corresponding capacitive response to a step pressure, indicating outstanding pressure recognition performance.

The durability of this same sensor was evaluated via a cyclic test of 100 cycles at a pressure of 4 kPa. As shown in Figure 5E, due to the excellent elasticity of the PDMS elastomer (Figure 4C), no obvious fatigue or degradation occurred during the 100 loading-unloading cycles. This continuously stable and rapid response under cyclic testing indicates that the sensor with 10 phr LMS-EIL has sufficient durability for real-time applications. However, due to the limited testing instruments, we just provide general data to support the long-term stability of the sensor. Some tests, such as thermal effects including diffusion and radiation, are also important factors to the long-term stability of the sensor, and they deserve further study in the future.

Figure 5F shows the relative capacitive change curve of the sensor with 10 phr LMS-EIL at a pressure of 4 kPa in a single loading-unloading cycle. The ΔC/C₀ presents negligible fluctuation during both loading and unloading. Furthermore, ΔC/C₀ increases immediately from 0 to 2.0 when pressure is applied to the sensor. Conversely, ΔC/C₀ returns quickly to baseline upon removal of the applied pressure. As shown in the magnified response and relaxation curve in Figure 5F, both response and recovery times are around 60 ms—faster than those of sensors with complicated, prefabricated structure to increase their sensitivity (around 100 ms) (Ma et al., 2018). This combination of stability with fast response and recovery means that the 10 phr LMS-EIL sensor possesses high potential as healthcare and wearable applications.