*2.3. Measurements*

#### 2.3.1. Thermal Characterization

The experiments on the melting point of LMs were used a DCS 214 (NETZSCH, Selb, German) with high purity alumina ceramic crucible that can withstand 100 ◦C and bear the corrosion of Ga. The measurement temperature range was from −40 ◦C to 40 ◦C, and the heating rate was 10 K/min.

#### 2.3.2. Rheological Characterization

The experiments on rheology were conducted by Discovery HR-20 (TA, New Castle, DE, USA), equipped with a 20 mm parallel plate geometry. To minimize the effect by measuring, all samples were pre-sheared and tested for three times. Additionally, the specific steps have been discussed in further detail in the ESI.

#### 2.3.3. Mechanical Characterization

Mechanical experiments were conducted using Dynamic Mechanical Analyzer DMA Q800 (TA, New Castle, DE, USA) and universal testing machine UTM Roell Z030 (Zwick, Ulm, German). Both instruments have been equipped with the heating function, and the detail settings have been shown in the ESI.

#### 2.3.4. Electrical Characterization

A high-precision LCR digital bridge TH2827c (Tonghui, Changzhou, China) was used to confirm whether the elastomer is conductive. The change of resistance during the stretch progress was monitored, and related electrical data has been synchronized directly to the computer.

#### **3. Results and Discussions**

#### *3.1. Rheological Properties of Matrix*

In the progress of DIW, the slurries pass through the narrow constriction of a needle which generates high shear force; once extruded, such shear force disappears instantly. Thus, it is required that the printed slurries can be smoothly extruded from the nozzle and perform well self-supporting after extrusion. In other words, the proportioning material should meet the characterization in both shear thinning [36] that viscosity decreases with shear strain, and viscoelastic inversion [37] that the changes of storage modulus and viscoelastic modulus show an intersection with the increase in shear strain. Inspired by Sangchul et al. [38], the above characterization can be obtained by adding other polymers. Here, SE 1700 was mixed with Ecoflex 00-30 and Sylgard 184, respectively, at a ratio of 1:1 and 1:1.5 in weight. Prior to combining, rheological tests have been carried out on these three silicones, all of which occurred shear thinning, but merely SE 1700 has the

characteristic of viscoelastic inversion (Figure 2a–c). The plateau value of storage modulus of SE 1700 is two orders of magnitude larger than that of the other two silicone rubbers, while its loss modulus is larger as well (Figure 2d,e). Furthermore, two types of mixed silica gels were also conducted systematic rheological investigations that reflect direct write printability (Figure 2f). The combination of SE 1700 and Sylgard 184 at a weight ratio of 1:1.5 (as shown in Figure S2) has both the properties of shear thinning and viscoelastic inversion. Additionally, the above materials have been mixed in a 1:1 ratio (consistent with Ecoflex 0030), and G' and G" have no intersecting trajectories (Figure S3), that is, they cannot be applied to DIW printing. 1700 is two orders of magnitude larger than that of the other two silicone rubbers, while its loss modulus is larger as well (Figure 2d,e). Furthermore, two types of mixed silica gels were also conducted systematic rheological investigations that reflect direct write printability (Figure 2f). The combination of SE 1700 and Sylgard 184 at a weight ratio of 1:1.5 (as shown in Figure S2) has both the properties of shear thinning and viscoelastic inversion. Additionally, the above materials have been mixed in a 1:1 ratio (consistent with Ecoflex 0030), and G' and G'' have no intersecting trajectories (Figure S3), that is, they cannot be applied to DIW printing.

A high-precision LCR digital bridge TH2827c (Tonghui, Changzhou, China) was used to confirm whether the elastomer is conductive. The change of resistance during the stretch progress was monitored, and related electrical data has been synchronized directly

In the progress of DIW, the slurries pass through the narrow constriction of a needle which generates high shear force; once extruded, such shear force disappears instantly. Thus, it is required that the printed slurries can be smoothly extruded from the nozzle and perform well self-supporting after extrusion. In other words, the proportioning material should meet the characterization in both shear thinning [36] that viscosity decreases with shear strain, and viscoelastic inversion [37] that the changes of storage modulus and viscoelastic modulus show an intersection with the increase in shear strain. Inspired by Sangchul et al. [38], the above characterization can be obtained by adding other polymers. Here, SE 1700 was mixed with Ecoflex 00-30 and Sylgard 184, respectively, at a ratio of 1:1 and 1:1.5 in weight. Prior to combining, rheological tests have been carried out on these three silicones, all of which occurred shear thinning , but merely SE 1700 has the characteristic of viscoelastic inversion (Figure 2a–c). The plateau value of storage modulus of SE

*Micromachines* **2022**, *13*, x FOR PEER REVIEW 5 of 12

2.3.4. Electrical Characterization

**3. Results and Discussions**

*3.1. Rheological Properties of Matrix*

to the computer.

**Figure 2.** Viscoelastic inversion measurement of (**a**) PDMS Sylgard 184, (**b**) Ecoflex 0030 and (**c**) PDMS SE 1700. (**d**) The storage modulus and (**e**) loss modulus versus strain for silicone elastomer. (**f**) The viscoelastic inversion characteristics of the combination of Ecoflex 0030 and PDMS SE 1700. **Figure 2.** Viscoelastic inversion measurement of (**a**) PDMS Sylgard 184, (**b**) Ecoflex 0030 and (**c**) PDMS SE 1700. (**d**) The storage modulus and (**e**) loss modulus versus strain for silicone elastomer. (**f**) The viscoelastic inversion characteristics of the combination of Ecoflex 0030 and PDMS SE 1700.

The shear thinning behavior can be derived from a power−law variant of Herschel and Bulkley model [39], in detail that if the value of *n* is between 0 and 1 in the fitted linear function relationship based on the Equation (1), the material possesses the characteristics of shear thinning.

$$
\pi = \pi\_0 + k\dot{\gamma}^n \tag{1}
$$

where τ is shear stress, τ<sup>0</sup> > 0 is the yield stress, *k* > 0 is the consistency parameter, and *n* > 0 is the power index. Figure 3b can be obtained by fitting a linear function by taking a logarithmic relationship to Equation (1), and all value of *n* below 1. As the increasing of *n*, the phenomenon of shear thinning becomes more obvious. So, the joint of a certain amount of SE 1700 can effectively make the substrate printable.

amount of SE 1700 can effectively make the substrate printable.

**Figure 3.** (**a**) Shear thinning of the selected silicone slurries (**b**) The fitting linear function of selected silicone slurries by Herschel and Bulkley model. **Figure 3.** (**a**) Shear thinning of the selected silicone slurries (**b**) The fitting linear function of selected silicone slurries by Herschel and Bulkley model.

The shear thinning behavior can be derived from a power−law variant of Herschel and Bulkley model [39], in detail that if the value of *n* is between 0 and 1 in the fitted linear function relationship based on the Equation (1), the material possesses the characteristics

τ = τ<sup>0</sup> + ̇

where τ is shear stress, τ<sup>0</sup> >0 is the yield stress, *k* > 0 is the consistency parameter, and *n* > 0 is the power index. Figure 3b can be obtained by fitting a linear function by taking a logarithmic relationship to Equation (1), and all value of *n* below 1. As the increasing of *n*, the phenomenon of shear thinning becomes more obvious. So, the joint of a certain

(1)

#### *3.2. Mechanical Properties of Matrix*

of shear thinning.

*3.2. Mechanical Properties of Matrix* The stiffness is a vital metric for defining the forces changing that the composites can support. After the printed splines are cured (Figure 4a, and the detail dimensions has been labeled in the engineering drawing), a series of DMA measurements have been established that all the tested LM composites are elastomeric in nature. As shown in Figure S4a, the storage modulus of silicone elastomer mixed with Ecoflex 0030 and SE 1700 is smaller than that of Sylgard 184 but greater than that of SE 1700, which is more conductive to highlight the effect of LM additive. So, the former mixed matrix has been focused on, while the latter analysis can be referred to in the Supporting Information section. Despite that LMs possess fluidity in the liquid state at room temperature, the storage modulus of LM composites increases with LM volume fraction (Figure 5a), compared to the unfilled LM composites with 0.429 ± 0.01 MPa, storage modulus at 60 vol% experienced increasing by a factor of around 11 to 4.71 ± 0.1 MPa. The incorporation of liquid inclusions enables to improve the stiffness of polymer composites to a certain extent, which has been demonstrated [40,41]. In addition, the stiffness is also influenced by the interfacial tension between LM and silicone matrix [42]. Taking example of the composites with 60 vol%, its storage modulus decreases gradually as temperature rises, but a plunge has been occurred when the temperature reaches the melting point T<sup>m</sup> of LM. Such a dip becomes more pronounced as the LM volume content increases, and the change of storage modulus can be over 200% during the transformation between the rigid and soft state. Moreover, the relationship between deformation strain and load bearing of LM composites has been explored by UTM, as shown in Figure 5b. As a result of the solid-liquid transformation tak-The stiffness is a vital metric for defining the forces changing that the composites can support. After the printed splines are cured (Figure 4a, and the detail dimensions has been labeled in the engineering drawing), a series of DMA measurements have been established that all the tested LM composites are elastomeric in nature. As shown in Figure S4a, the storage modulus of silicone elastomer mixed with Ecoflex 0030 and SE 1700 is smaller than that of Sylgard 184 but greater than that of SE 1700, which is more conductive to highlight the effect of LM additive. So, the former mixed matrix has been focused on, while the latter analysis can be referred to in the Supporting Information section. Despite that LMs possess fluidity in the liquid state at room temperature, the storage modulus of LM composites increases with LM volume fraction (Figure 5a), compared to the unfilled LM composites with 0.429 ± 0.01 MPa, storage modulus at 60 vol% experienced increasing by a factor of around 11 to 4.71 ± 0.1 MPa. The incorporation of liquid inclusions enables to improve the stiffness of polymer composites to a certain extent, which has been demonstrated [40,41]. In addition, the stiffness is also influenced by the interfacial tension between LM and silicone matrix [42]. Taking example of the composites with 60 vol%, its storage modulus decreases gradually as temperature rises, but a plunge has been occurred when the temperature reaches the melting point T<sup>m</sup> of LM. Such a dip becomes more pronounced as the LM volume content increases, and the change of storage modulus can be over 200% during the transformation between the rigid and soft state. Moreover, the relationship between deformation strain and load bearing of LM composites has been explored by UTM, as shown in Figure 5b. As a result of the solid-liquid transformation taking place below 60 ◦C, it can be stretched much more than that keeping at 0 ◦C, particularly, up to 3.5 times for LM (60 vol%) composites mixed with Ecoflex 0030 and SE 1700. Meanwhile, the corresponding carrying load has been weakened due to the soft state of LM. Similar changes occurred in the LM composites with Sylgard 184 and SE 1700 as the carrier (Figure S5).

Apart from the simple splines, the flower-shaped and Poisson structure have been designed (shown in Figure 4b–e). The modeling and fabrication of relatively complex structures in a short period of time further confirm the high efficiency of additive manufacturing. As the temperature rises, the changes on tensile strength of the above designed structures have been recorded in Movies S1 and S2. In terms of Poisson structures, the connection between bearing capacity and deformation degree at high temperature has been displayed in Figure S6.

carrier (Figure S5).

**Figure 4.** (**a**) The detail engineering diagram of sample splines, (**b**) the design structure, (**c**) the printed real part, (**d**) hollow flower structure and (**e**) Poisson structure. **Figure 4.** (**a**) The detail engineering diagram of sample splines, (**b**) the design structure, (**c**) the printed real part, (**d**) hollow flower structure and (**e**) Poisson structure. *Micromachines* **2022**, *13*, x FOR PEER REVIEW 8 of 12

ing place below 60 °C, it can be stretched much more than that keeping at 0 °C, particularly, up to 3.5 times for LM (60 vol%) composites mixed with Ecoflex 0030 and SE 1700. Meanwhile, the corresponding carrying load has been weakened due to the soft state of LM. Similar changes occurred in the LM composites with Sylgard 184 and SE 1700 as the

As the typical types of PCMs, LMs possess good electrical conductivity. The state of

 

(2)

(3)

(4)

of composites. In theoretical, the resistance of LM composites can be referred as the stand-

=

in which is the electrical resistivity, *L* is the length and *A* is the cross-sectional area of conductor. It can be seen that the resistance varies with the geometry for a certain material. During the stretching process at high temperature, the samples will not only undergo thermal expansion but also length elongation, and the length increases faster than the

To characterize the change in resistance, the samples have been clamped on UTM, connecting with LCR digital bridge simultaneously. The relative resistance change has

*R* and *R*<sup>0</sup> are corresponding to the resistance values with and without deformation, respectively. The Δ*R*/*R*<sup>0</sup> increases with stretching deformation, which proves that the external strain has a certain influence on the relative resistance change. To investigate the sensing performance in terms of tensile strain, its sensitivity can be defined by gauge fac-

 − <sup>0</sup> 0

∆/<sup>0</sup> ∆/<sup>0</sup>

∆ 0 =

=

where, *L*<sup>0</sup> represents the initial size of the splines, and *ΔL* indicates the size change. As shown in Figure 6a, with the good corresponding consistency, the highest strain can reach 9400% approximately, and *GF* value is around 60 at this point. Moreover, the relationship between resistance changes and deformation of LM composites with Sylgard 184 and SE 1700 has been explored (Figure S8a), which can be up to around 120% with the GF value

According to the maximum strain obtained from the measurement, the sample of LM composites mixed with Ecoflex 0030 and SE 1700 have been applied to the strain from 0%

*3.3. Resistance Changes in the Process of Stretching*

ard equation for the wire:

cross-sectional area.

been introduced:

tor (GF):

of 1.28 approximately.

#### *3.3. Resistance Changes in the Process of Stretching*

As the typical types of PCMs, LMs possess good electrical conductivity. The state of LM can be judged by monitoring changes in resistance, thereby determining the softness of composites. In theoretical, the resistance of LM composites can be referred as the standard equation for the wire:

$$R = \frac{\rho L}{A} \tag{2}$$

in which *ρ* is the electrical resistivity, *L* is the length and *A* is the cross-sectional area of conductor. It can be seen that the resistance varies with the geometry for a certain material. During the stretching process at high temperature, the samples will not only undergo thermal expansion but also length elongation, and the length increases faster than the cross-sectional area.

To characterize the change in resistance, the samples have been clamped on UTM, connecting with LCR digital bridge simultaneously. The relative resistance change has been introduced:

$$\frac{\Delta R}{R\_0} = \frac{R - R\_0}{R\_0} \tag{3}$$

*R* and *R*<sup>0</sup> are corresponding to the resistance values with and without deformation, respectively. The ∆*R*/*R*<sup>0</sup> increases with stretching deformation, which proves that the external strain has a certain influence on the relative resistance change. To investigate the sensing performance in terms of tensile strain, its sensitivity can be defined by gauge factor (GF): *Micromachines* **2022**, *13*, x FOR PEER REVIEW 9 of 12 to 120% at a constant rate of 10 mm/min, and then released until they return to the initial state. The relative resistance change versus time has been achieved in Figure 6b by applying and releasing pressure several times repeatedly, and the spline takes place fracture

$$GF = \frac{\Delta R / R\_0}{\Delta L / L\_0} \tag{4}$$

where, *L*<sup>0</sup> represents the initial size of the splines, and ∆*L* indicates the size change. As shown in Figure 6a, with the good corresponding consistency, the highest strain can reach 9400% approximately, and *GF* value is around 60 at this point. Moreover, the relationship between resistance changes and deformation of LM composites with Sylgard 184 and SE 1700 has been explored (Figure S8a), which can be up to around 120% with the GF value of 1.28 approximately. force is released. The sample can basically return to its original shape when the external force disappears. However, the stretching process will cause a certain degree of permanent loss for no contract of internal LMs fracture due to elastomer stretching, which will make the value of resistance become larger when returning to the origin point. At the tenth stretch, the sample has broken. Moreover, the reproducibility test graph for LM composites with Sylgard 184 and SE 1700 have been demonstrated in the Supporting Information (Figure S8b).

**Figure 6.** (**a**) At high temperature (60 °C), the relative resistance changes with the stretchable strain, and (**b**) the stretch repeatability over time. **Figure 6.** (**a**) At high temperature (60 ◦C), the relative resistance changes with the stretchable strain, and (**b**) the stretch repeatability over time.

**4. Conclusions** We have developed a variable stiffness composite that consists of LM and silica gel with different mixing ratios, which can change properties in response to the thermal stimuli. In DIW printing, the process of slurry extrusion can be regarded as material prepared by a microfluidic channel, so the related rheological properties are necessary for the combined slurries. With a certain printability, dual material printing alternately has been applied to fabricate LM composites in one step for the relative complex structures. The samples presented here illustrate the stiffness change of greater than 1900% from a stiff to soft state, while the storage modulus decrease from 4.75 MPa and 0.2 MPa after heating up. According to the maximum strain obtained from the measurement, the sample of LM composites mixed with Ecoflex 0030 and SE 1700 have been applied to the strain from 0% to 120% at a constant rate of 10 mm/min, and then released until they return to the initial state. The relative resistance change versus time has been achieved in Figure 6b by applying and releasing pressure several times repeatedly, and the spline takes place fracture when the number of cycles is about 10 times. With the increase in strain, ∆*R*/*R*<sup>0</sup> improves gradually; and after reaching the peak value, ∆*R*/*R*<sup>0</sup> deceases as a result that

Furthermore, by changing the inner structure design or volume fraction between LMs and silicon elastomer, different stiffness values for these two steady states can be achieved.

the elastomer to a certain extent; the spline fracture is generated after about ten repetitions of the tensile test. Overall, this work has demonstrated the LM composites undergo the changes in mechanical and electrical properties under temperature stimuli. With the tuning capability, LM composites are expected to be used in the field of soft sensing actuators,

**Supplementary Materials:** The following supporting information can be downloaded at: www.mdpi.com/xxx/s1; Figure S1: The thermal properties of LM by DSC measurements; Figure S2: a. Shear thinning and b. Viscoelastic inversion measurement of the combination of PDMS Sylgard 184 and PDMS SE1700 with 1:1.5 in weight ratio; Figure S3: a. Shear thinning and b. Viscoelastic

even towards artificial muscle applications after enhancing adhesion.

the external force is released. The sample can basically return to its original shape when the external force disappears. However, the stretching process will cause a certain degree of permanent loss for no contract of internal LMs fracture due to elastomer stretching, which will make the value of resistance become larger when returning to the origin point. At the tenth stretch, the sample has broken. Moreover, the reproducibility test graph for LM composites with Sylgard 184 and SE 1700 have been demonstrated in the Supporting Information (Figure S8b).

#### **4. Conclusions**

We have developed a variable stiffness composite that consists of LM and silica gel with different mixing ratios, which can change properties in response to the thermal stimuli. In DIW printing, the process of slurry extrusion can be regarded as material prepared by a microfluidic channel, so the related rheological properties are necessary for the combined slurries. With a certain printability, dual material printing alternately has been applied to fabricate LM composites in one step for the relative complex structures. The samples presented here illustrate the stiffness change of greater than 1900% from a stiff to soft state, while the storage modulus decrease from 4.75 MPa and 0.2 MPa after heating up. Furthermore, by changing the inner structure design or volume fraction between LMs and silicon elastomer, different stiffness values for these two steady states can be achieved. Owing to the electrical conductivity of LMs, the composite exhibits electrical resistance that changes with stretching. However, each stretch will lead to irreversible damage in the elastomer to a certain extent; the spline fracture is generated after about ten repetitions of the tensile test. Overall, this work has demonstrated the LM composites undergo the changes in mechanical and electrical properties under temperature stimuli. With the tuning capability, LM composites are expected to be used in the field of soft sensing actuators, even towards artificial muscle applications after enhancing adhesion.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/mi13081343/s1; Figure S1: The thermal properties of LM by DSC measurements; Figure S2: a. Shear thinning and b. Viscoelastic inversion measurement of the combination of PDMS Sylgard 184 and PDMS SE1700 with 1:1.5 in weight ratio; Figure S3: a. Shear thinning and b. Viscoelastic inversion measurement of the combination of PDMS Sylgard 184 and PDMS SE1700 with 1:1 in weight ratio; Figure S4: The storage modulus of splines composed by different silicone elastomer a. Comparison with Sylgard 184, SE 1700 and their mixture based on 1:2 in volume fraction. b. Comparison with Ecoflex 0030, SE 1700 and their mixture on the basis of volume fraction with 1:1; Figure S5: Th storage modulus of LM composites that subtracted by Sylgard184 and SE1700 with the LM increasing volume fraction; Figure S6: The load capacity versus the stretchable strain at low and high temperature for the Poisson structure with 60 vol% in LM; Figure S7: The Schematic diagram of real-time monitoring resistance measurements; Figure S8: At high temperature (60 ◦C), the relative resistance of LM composites (that based on the mixture of Sylgard184 and SE 1700) changes with the stretchable strain, and b. the stretch repeatability over time; Movie S1: Flower; Movie S2: Poisson Structure.

**Author Contributions:** Conceptualization, F.L., Y.R. and Y.C.; Data curation, F.L. and Y.C.; Formal analysis, F.L. and Y.C.; Funding acquisition, G.X., J.W. and Y.R.; Investigation, F.L. and Y.R.; Methodology, F.L., Y.R. and Y.C.; Project administration, G.X. and Y.C.; Resources, G.X., J.W., Y.R. and Y.C.; Supervision, Y.R. and Y.C.; Validation, F.L.; Writing—original draft, F.L.; Writing—review & editing, G.X., J.W., Y.R. and Y.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by Zhejiang Provincial Natural Science Foundation of China under grant No. LZ22E030003, LY19E060001 and LQ19F050003, Ningbo Science and Technology Bureau under Service Industry Science & Technology Programme with project code 2019F1030, and Zhejiang Provincial Department of Science and Technology under its Provincial Key Laboratory Programme (2020E10018). F.L. acknowledges the Ph.D. scholarship of Doctor Training Program between University of Nottingham Ningbo China and Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences.

**Data Availability Statement:** The data that support the findings of this study are available from the corresponding author upon reasonable request.

**Acknowledgments:** This work was financially supported by Zhejiang Provincial Natural Science Foundation of China under grant No. LZ22E030003, LY19E060001 and LQ19F050003, Ningbo Science and Technology Bureau under Service Industry Science & Technology Programme with project code 2019F1030. The Zhejiang Provincial Department of Science and Technology is also acknowledged for this research under its Provincial Key Laboratory Programme (2020E10018).

**Conflicts of Interest:** The authors declare no conflict of interest.

## **References**

