Design and characterization of electro-active soft actuator based on phase transition of microencapsulated silicon-ethanol composite

Materials

In this study, a two-component room temperature vulcanized (RTV) silicone elastomer was used as the matrix, which is environmentally friendly, safe, and has excellent processability (Haotian, China). Ethanol (>99.5%) was used as PCM, with physical, chemical, and economic requirements, namely suitable phase change temperature, non-toxicity, and low price. Ethanol was purchased from Merck (Germany). A 0.25mm-diameter nickel-chromium resistance wire was used for electrically-driven actuation of the composite by Joule heating.

Preparation of liquid-silicone elastomer composite

The functional composite was synthesized by mechanical mixing of 5, 10, and 15 wt% of ethanol with the silicone elastomer. The findings indicated that the preparation of the composite with 20 wt% of ethanol and more, in addition to the difficulty in the mixing process, led to phase separation in the chamber. After mixing the silicone with ethanol for 5 min, a hardener with a weight ratio of 1 to 50 was added to the mixture. Note that the media were weighted regularly during the process to retrieve ethanol volatilized. In this research, the molding process was performed in a polymethyl methacrylate (PMMA) mold (15 × 15 × 90 mm dimensions) with a pre-installed thermistor and Ni-Cr wire in a spiral shape inside it. Ni-Cr resistance wire with a length of 75 cm was wound into a spiral by twisting around a shaft. After molding, the actuator was cured for 4 h. Figure 1 shows a schematic of the PCM-elastomer composite fabrications method, the shape of the mold, and the as-cured sample.OPEN IN VIEWER

Characterization

The nanocomposite microstructure was evaluated by field emission scanning electron microscopy (FESEM, MIRA3 TESCAN, Czech Republic) at an accelerating voltage of 10 kV. Samples were cryogenically fractured in liquid nitrogen and coated with a thin layer of platinum before the observation. Olympus BX41M-LED (Japan) optical microscope was used for optical characterization of the actuators. Tensile tests were carried out at room temperature (20°C) using a universal testing machine (UTM, Zwick 250, Germany) at 500 mm/min crosshead speed. The specimens were prepared according to ASTM D-412, type A. The force generated by the actuator was measured using a 500N load sensor. The cylindrical actuator samples (50 mm long, 2.5 mm diameter) were prepared for the blocked force experiment. First, a suitable casing was designed to constrain the motion of the actuator in all directions except at the actuator tip. Another flat rectangular surface was prepared and attached to the load sensor. Once the actuator was electrically actuated, the tip was allowed to apply force on the sensor through the flat surface; the force generated was then monitored using a computer. Thermogravimetric analysis (TGA) was applied at 25°C–300°C temperature range and a heating rate of 10°C/min. Sample decomposition was studied in the air environment, and measurements were performed using the TGA instrument (Shimadzu, Japan). The Joule heating generated by resistive wire was measured using pt1000 and lm35 sensor (Japan), capable of measuring up to 300°C. The NEC Avio Infrared compact thermal camera (Japan) measured the thermal response of actuator and heater design. The actuation performance was evaluated for different input energies (10, 15, 20, and 25 W) by capturing the deformed shape using a video camera (Canon, Japan). The image processing was conducted by MATLAB software, similar to Zamyad et al. ⁴⁶ Composite performance was evaluated under four different powers (10, 15, 20, 25 W) in the first, second, and third periods corresponding to days 1, 3, and 6. Samples are named SEx which means silicone-ethanol composite and x is a number showing the actuation power in watts.

Selection of the optimal value of ethanol

As shown in Figure 2(a), the internal pressure inside the soft body actuator is almost zero before electrical actuation. Composite performance becomes faster as heating continues beyond the boiling point of ethanol (T= 78.32°C). The electrical actuation of the actuators by Joule heating provides a vast volume expansion due to the vapor pressure induced by the liquid-vapor transition in the microcapsules. The micropores distributed throughout the polymer matrix are visible in the optical image of Figure 2(a).OPEN IN VIEWER

Thermal stability is a key factor in the thermo-active cyclic performance of PCM, which determines the availability and working durability of phase change materials.^47–49 For this reason, the thermal stabilities of the silicone-ethanol composites in different weight ratios (0, 5, 10, 15 wt%) were detected by TGA (Figure 2(b)). For SE15, a drastic weight loss takes place between 45 and 100°C because of the evaporation of ethanol out of the microcapsules. At around 120°C, ethanol completely evaporated from the microcapsules (overall, about a 15% decrease in weight was noted). The weight loss value corresponding to 45°C for SE15 is about 4.8%, while 1.8% and 1.2% for SE10 and SE5, respectively. The latter may be attributed to the smaller weight ratio of ethanol/silicone, which encapsulated less ethanol by a greater amount of silicone. The TGA spectrum pertaining to the 10 wt%-ethanol composite showed similar behavior where the total weight loss was measured at about 10%. The TGA curve of 5 wt%-ethanol composite exhibited a minor change around 45°C. This may be attributed to the relatively low initial ethanol amount. No significant weight loss occurred after 125°C in all composite samples. In contrast, the TGA curve of the pure silicone elastomer had no changes in the 50°C–100°C temperature range, and the total weight loss was measured for this specimen via TGA analysis of about 1.98%. As shown in Figure 2(b), it is notable that the residual ethanol in silicone microcapsules is significantly affected by the weight ratio of ethanol/silicone. When the samples were synthesized with the higher weight ratio of ethanol to silicone, the microcapsules obtained had lower residual ethanol than others.

In most animals, muscle contraction is restricted by the joints, which causes less than 20% fiber tension. ⁵⁰ Therefore, achieving an expansion strain value equivalent to at least 20% of the initial length was considered to select the optimal value of transformable liquid to meet the purpose of the study. Accordingly, three composite samples comprised of 5, 10, and 15 wt% ethanol were prepared by embedding 75cm of nickel-chromium wire. All samples were triggered under similar electrical actuation conditions (0.5A×20V=10 W). Figure 2(c) shows the heat-triggered elongation as a function of ethanol content in composites. As can be seen, there is ascending proportion of axial expansion by increasing ethanol content, while the heat-triggered elongation of 15 wt% ethanol sample achieved the least 20% of the initial length. Therefore, 15 wt% was selected as the optimal measured value of phase change fluid in the composite.

Thermal properties assessment

Figure 3 shows the temperature response of composite samples in three actuation periods as a function of driving power. As shown in Figure 3(a), there is a noticeable decrease in total working time by increasing applied power from 10 to 25 W. While it takes about 9300 s to complete 10 working cycles for the SE10, the latter decreases to 3600, 2600, and 2100 s for the SE15, SE20, and SE25, respectively. This can be attributed to the value of input heat flux, which plays a critical role in the thermo-responsive performance of soft actuator and affects the liquid-vapor phase transition rate.OPEN IN VIEWER

As seen in Figure 3(a), there is a descending trend in the frequency of average maximum value of surface temperature in the first actuation period in which by increasing of applied power from 10 to 25 W, the latter decreased from 45.6 to 34°C. The slow rate of input heat flux at low powers (10 and 15 W) provides more time to transfer thermal energy from internal layers of the actuator to the surface. The latter may result in homogeneous heat distribution throughout the sample. The findings reveal that the actuators show a stable temperature response in which the highest recorded internal temperature was ⁓73°C after applying 10 working cycles.

The samples were tested on the third day (second actuation period) under electrical actuation conditions similar to the first day. As can be seen in Figure 3(b), the total working time of all samples increased on the third day, which can be attributed to the depletion of ethanol during the first prolonged actuation and formation of passive-elastic outer layers around them. As the capsuled liquid in each lateral micron-scale pocket is evaporated, it leaves an ethanol-free residual layer that acts as one- a barrier to reach the heat-triggered length, which extends the total working time, and two- a protective layer against more depletion of ethanol located in inner microcapsules. As shown in Figure 3(b), longer actuation time caused receiving more input heat flux by samples, which increased the internal temperature of all samples compared to the first day. The average maximum value of internal temperature for all samples increased in the second actuation period, and the latter is the most for SE25, which changed from 68.8 to 86°C (up to 20% growth).

The ethanol content is the lowest in the third actuation period (sixth day). The escape of ethanol in the first and second actuation periods provides extreme conditions, resulting in a significant increase in total working time and internal temperature of composite actuators (Figure 3(c)).

Chemical composition analysis

The effect of multi-cyclic electrical actuation on the chemical composition of composite samples was investigated by weighing the samples before and after each working period. Figure 4(a) shows the residual ethanol in composite samples at the end of the first and second electrical actuation periods. As shown in Figure 4(a), more than 20 wt% of the active phase was released from the composites at the end of the first period. Compared to other samples, the residual ethanol is lowest in SE10 due to the slow input heat flux rate and longer total working time. The same level of ethanol evaporation in SE15 and SE20 may be explained by receiving equal thermal energy (actuation time * power ⁓ 12.8Kj). As can be seen in Figure 4(a), there is a marked decrease (over 30%) in residual ethanol in the second actuation period, which can be attributed to extended total working. As can be seen, the proportion of ethanol escape is similar to the first actuation period, while SE25 possesses the lowest ethanol index. This may be explained by the higher internal temperature, around 10°C more than other samples.OPEN IN VIEWER

Findings reveal that the actuator’s residual ethanol is significantly affected by the value of driving power. The microcapsules actuated at power 10 and 25 had lower residual ethanol than the others, which can be attributed to the slow rate of input heat flux and thermal shock at low and high power, respectively.

Mechanical properties

Figure 5(a) shows the typical stress-strain curves of pure silicone and silicone-ethanol composite at room temperature. The samples exhibited rubbery behaviors without yield, necking, or strain hardening. It may be seen that the optimal measured value of ethanol (15 wt%) affects the mechanical properties of the composite. Findings reveal a significant decrease in elastic modulus (1.11 to 0.2 MPa, Figure 5(b)) and tensile strength (3.72 to 1.05, over 70% reduction), while the elongation at break increased up to 200%. The incorporation of ethanol reduces the entanglement of the silicone chains. It provides more free volume, which causes the polymer chains to stretch easily at a specific level of tension. Also, the weaker mechanical properties of the silicone-ethanol composite can be attributed to a decrease in the mass of the solid phase (silicone) replaced by ethanol with no reinforcing effect.OPEN IN VIEWER

Figure 5(c) illustrates information regarding the blocked-force characteristics of soft actuators prepared in a cylindrical shape (Figure 1(b)). In each cycle, the composite actuator was triggered via Joule heating to exert a temperature of 78°C (boiling point of ethanol) and then de-actuated by ambient cooling until the exerted temperature decreased to 45°C. A further decrease in room temperature was avoided, which is significantly time-consuming. As can be seen in Figure 5(c), in the first cycle, the achievable force corresponding to T=78°C reached a peak of ⁓80N, then faced a gradual decrease for the subsequent nine cycles, and finally dropped to ⁓30 N in the tenth cycle. This may be attributed to ethanol's loss, which resulted in weaker actuator functionality. It is notable that the actuator succeeded to produce heat-triggered force 100 times greater than its own (25 gr) in the last cycle.

Figure 5(d) shows the force generated by the soft actuator equal to 1000 times of its own weight (⁓25 gr) to characterize force-temperature correlation. Under 15 W, the pristine actuator could generate a ⁓250 N force in 763 s, but with the preheated system, it took 611 s to generate the same force. In the first cycle, the internal temperature stood at the lowest point (115°C), which was 5°C less than the third cycle. The required time for achieving the maximum force value increased from 611 to 638s in the third cycle, which may be explained by the evaporation of ethanol from capsule structure of actuator.

Strain characteristic

The thermo-responsive behavior of the soft actuator is shown in Figure 6-a-b. We used a thermal camera to characterize the ability of the Ni-Cr wire to heat the surrounding actuator composite and heat distribution throughout the sample. As shown in thermal and digital images of Figure 6(a), the current is forced to flow directly from one end to the other through the spiral path by implementing the Ni-Cr wire heater in the composite, which brings heat through the majority of the composite, allowing for even actuation. The temperatures in the tips of the actuator (⁓55°C) are less than the middle (⁓80°C), due to the more sidewalls in touch with environment and connecting to clamp.OPEN IN VIEWER

As shown, the performance of actuator is slow at the beginning of the actuation (45°C, 150 s), but as the actuation continues and the internal temperature rises from 45 to 75°C, the vapor pressure inside the microcapsules increases. This is followed by a better electrically actuation response and rapid movement. The actuator shape recovery process occurs dramatically within the initial seconds of relaxation time due to the fast strain energy dissipation of the specimen. This can be seen in relaxation images of Figure 6(a) when the shape recovery becomes approximately complete over a 150 s period (between 300 and 450 s).

For further evaluation, we measured the velocity of the tip of the actuator at similar locomotion power (10 W) in two different states: one- pristine actuator (first cycle, T_initial=20°C) and two- preheated actuator (second cycle, T_initial=45°C). Figure 6(a) presents information regarding to velocity of the actuator and the effect of preheating on this value. As shown, in the first 150 s, the pristine actuator shows negligible forward motion, and the velocity is stabilized at 0.018 mm/s due to the gradual boiling of ethanol. Then, followed by a steady temperature increase to 78.3°C, the soft actuator reaches the maximum heat-triggered elongation (25% of initial length) when it moves at = 0.066 mm/s. In the reverse direction (relaxation time), the velocity of the actuator dramatically changed from 0 to 0.3 mm/s during the first 15 s, which can be attributed to the fast elastic energy dissipation. A similar trend is observed over time when the variation of the shape recovery becomes inconspicuous. The further decline in velocity from 0.3 to less than 0.05 mm/s is due to the termination of strain energy. At the end of the relaxation time, the internal temperature of the sample is approximately 45°C (Figure 6(b)). In the second actuation cycle, the pre-heated system needs less actuation time (215 s, Figure 6(b)), and the slope of velocity graph increases sharply in the forward direction due to the part of the supplied energy was stored in the system as internal energy (Figure 6(c)). After this, the value of the maximum velocity of the actuator in forward direction increased 40%. The slight decrease in maximum value of velocity in reverse direction (from 0.3 to 0.28 mm/s, 6%) can be attributed to the more induced vapor pressure in the preheated system that acts as a barrier to reaching the initial length.

Figure 6(d) shows strain characteristics of soft actuator (15 wt% ethanol) at cyclic actuation (locomotion powered at 10 W). A visual comparison between strain-time curvatures reveals the limitation of repeatable expansion in different actuation periods. The multiple working cycles increase the actuation durability by two key factors, 1-ethanol escape, which resulted in a reduction of excitability of the actuator and limitation of extending the life cycles, 2-formation of passive outer layers acts as a barrier against soft actuator heat-triggered elongation.

Material microstructure

To investigate the effect of long-term electrical actuation and applied power value on silicone-ethanol composite microstructure, SE10 and SE25 samples were cryogenically fractured after 10 working cycles in the third period. The fracture surface was observed by field emission electron microscopy (Figure 7). It is obvious that hand-mixing of 15 wt% ethanol with silicone elastomer matrix has resulted in a homogeneous distribution of active phase across the entire composite volume. The separation of micron-scale pockets allows the distribution of encapsulated liquid in small quantities across the entire matrix volume. As shown in Figure 7(a-I-II), the number of ethanol encased in the micro-pockets near the outer layers of SE10 is more than SE25, which can be attributed to the higher surface temperature in SE10 (⁓60°C) and heat concentration in the core of SE25.OPEN IN VIEWER

As seen in Figure 7(b-c), the microcapsules located in the core of SE10 are smaller and distributed more uniformly than SE25. To verify these observations, the diameter-size distribution of microcapsules located in the inner layers of composite samples was evaluated, and the results are presented in Figure 8 and summarized in Table 1. According to Figure 8, the number of microcapsules in an equal scanned surface area in SE10 is more than SE25. Also, the main diameter size distribution of microcapsules varies in the range of 2< D (µm)< 8 and 4< D (µm)< 10 for SE10 and SE25, respectively. The data presented in Table 1 reveals a much more significant surface area (more than twice) and higher average diameter (more than 25%) in SE25 than in SE10. High ethanol depletion and thermal shock in SE25 limited its actuation functionality; therefore, higher vapor pressure was needed to achieve determined elongation, which causes a large plastic deformation within the matrix and merging the small microcapsules. The optical and FE-SEM images of Figure 7(d-e) shows large interconnected cavities (60μm) next to small capsules (3μm) in SE25.OPEN IN VIEWER

OPEN IN VIEWER

Table 1.

Microcapsules characteristics.

Samples	Average value of diameter (µm)	Average value of surface area (µm)
SE10	4.6	16.1
SE25	6.3	32.6