*Article* **Replica of Bionic Nepenthes Peristome-like and Anti-Fouling Structures for Self-Driving Water and Raman-Enhancing Detection**

**Yen-Ting Lin <sup>1</sup> , Chun-Hao Wu <sup>1</sup> , Wei-Lin Syu <sup>1</sup> , Po-Cheng Ho <sup>1</sup> , Zi-Ling Tseng <sup>2</sup> , Ming-Chien Yang <sup>2</sup> , Chin-Ching Lin <sup>3</sup> , Cheng-Chen Chen 4,\* , Cheng-Cheung Chen 5,6,\* and Ting-Yu Liu 1,\***


**Abstract:** The flexible, anti-fouling, and bionic surface-enhanced Raman scattering (SERS) biochip, which has a Nepenthes peristome-like structure, was fabricated by photolithography, replicated technology, and thermal evaporation. The pattern of the bionic Nepenthes peristome-like structure was fabricated by two layers of photolithography with SU-8 photoresist. The bionic structure was then replicated by polydimethylsiloxane (PDMS) and grafting the zwitterion polymers (2-methacryloyloxyethyl phosphorylcholine, MPC) by atmospheric plasma polymerization (PDMS-PMPC). The phospholipid monomer of MPC immobilization plays an important role; it can not only improve hydrophilicity, anti-fouling and anti-bacterial properties, and biocompatibility, but it also allows for self-driving and unidirectional water delivery. Ag nanofilms (5 nm) were deposited on a PDMS (PDMS-Ag) substrate by thermal evaporation for SERS detection. Characterizations of the bionic SERS chips were measured by a scanning electron microscope (SEM), optical microscope (OM), X-ray photoelectron spectrometer (XPS), Fourier-transform infrared spectroscopy (FTIR), and contact angle (CA) testing. The results show that the superior anti-fouling capability of proteins and bacteria (*E. coli*) was found on the PDMS-PMPC substrate. Furthermore, the one-way liquid transfer capability of the bionic SERS chip was successfully demonstrated, which provides for the ability to separate samples during the flow channel, and which was detected by Raman spectroscopy. The SERS intensity (adenine, 10−<sup>4</sup> M) of PDMS-Ag with a bionic structure is ~4 times higher than PDMS-Ag without a bionic structure, due to the multi-reflection of the 3D bionic structure. The high-sensitivity bionic SERS substrate, with its self-driving water capability, has potential for biomolecule separation and detection.

**Keywords:** Nepenthes structure; bionic replica; zwitterion polymers; self-driving water; unidirectional water delivery; surface-enhanced Raman scattering (SERS) detection

#### **1. Introduction**

There are many micro- or nano-scale structures with different functions and special characteristics in nature, which have not yet been discovered because of the limitations of existing detection technologies. A benefit of the progress of manufacturing and the advancement in detection techniques in recent years has been that the micro-scale structure can now easily be observed, encouraging more researchers to invest in micro biomaterial structures. For example, lotus leaves possess a super-hydrophobic and self-cleaning ability

**Citation:** Lin, Y.-T.; Wu, C.-H.; Syu, W.-L.; Ho, P.-C.; Tseng, Z.-L.; Yang, M.-C.; Lin, C.-C.; Chen, C.-C.; Chen, C.-C.; Liu, T.-Y. Replica of Bionic Nepenthes Peristome-like and Anti-Fouling Structures for Self-Driving Water and Raman-Enhancing Detection. *Polymers* **2022**, *14*, 2465. https:// doi.org/10.3390/polym14122465

Academic Editor: Seong J. Cho

Received: 9 May 2022 Accepted: 15 June 2022 Published: 17 June 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

due to the unique nanostructure of their surface [1,2]. This nanostructure contains structural wax on the surface layer, which not only increases surface roughness, but also traps air between the solid and liquid interface [3]. Biomimetic technology [4–12] is a technology that imitates the 3.8 billion years of evolutionary experience of living things. Every lifestyle, growth process, and ecosystem is a source of inspiration for simulation. It has the two major contradictions of "nature but primitive" and "technology but pollution", which are currently significant [13]. Biomimetic technology has three advantages: low costs, high efficiency, and low pollution. With the proper utilization of this technology, we can achieve low resource and energy consumption while obtaining optimal production capacity and benefits. For example, aircraft coatings based on shark skin resistance [6,14–16], radar sonar based on bats [17], solar applications inspired by photosynthesis, bullet train heads based on kingfisher beaks [18,19], antifouling coatings based on frog skins [20], etc. Biomaterials have also been widely applied in SERS substrates due to their 3D periodic microstructures, such as butterfly wings, cicada wings, and rose petals [21–23]. The best case for the continuation, this study focuses on the surface structure of nepenthes.

Nepenthes lives in a barren environment. Uniquely, they can capture insects with their peristome to meet their fundamental nutrient needs [2]. Most species obtain their nutrients from trapped and food animals. The inner surface of the insect trap is covered with slippery and fragile wax [24], forming an effective trap. The peristome on the top aims to attract and capture prey by forming a smooth liquid film. Previous studies indicated that the peristome structure of Nepenthes consists of radially arranged ribs. Rain, dew, and honey form a layer of liquid film on the surface, transforming the peristome in a super-hydrophilic surface. The surface of the epidermal cells of Nepenthes peristome is smooth and wax-free. The addition of wax-free crystals and hygroscopic honey increases the capillary force and promotes the formation of the liquid film, further enhancing the liquid transport speed on the surface of the Nepenthes peristome [25], which even resists gravity. Nevertheless, the continuous radial arrangement of the peristome grooved structure exhibits liquid transmission characteristics without other external forces. Fabricating a surface with the same structural features as the replica molding method is ideal to achieve the one-way liquid transfer (self-driving water) capability, which has potential to apply in microfluidic devices without pumping. Directly using the natural peristome as a template, the peristome structure can be replicated by artificial poly (dimethylsiloxane) (PDMS); however, the surface is curved and difficult to reprocess. Therefore, we propose a high-precision photolithography process to imitate the Nepenthes peristome.

The plane structure is reproduced on PDMS through transfer-printing technology [13,26,27]. Most studies choose PDMS as the transfer material due to its excellent pattern reproduction accuracy, easy preparation, and ease of observation, etc. A microfluidic chip is a very promising analytical platform for the sample pre-treatment, separation, and detection [28]; for example, a microfluidic based surface-enhanced Raman scattering (SERS) chip was applied to detect creatinine in blood for 2 min [28,29]. The microfluidic SERS chip was proven to effectively detect bio-samples. However, the replica PDMS is an extremely hydrophobic material, whereas Nepenthes peristome has hydrophilic structure; to address this, we propose a hydrophilic modified material with 2-methacryloyloxyethyl phosphorylcholine (MPC), which contains a phospholipid structure that was polymerized and grafted on the PDMS surface by atmospheric plasma. MPC is a zwitterionic monomer with good antithrombotic and blood compatibility [30–32]. This polymer is highly hydrophilic due to the phospholipid polar tail in the MPC molecular structure. Moreover, the inhibitory effect on the adhesion of cells, platelets, enzymes, and proteins in the blood can also reduce the chances of the organism recognizing the material as a foreign object, and increases the biocompatibility of the material [33,34]. Therefore, the hydrophilic bionic structure was imitated to exhibit the same liquid transport capability as the Nepenthes peristome in this study.

peristome in this study.

**2. Materials and Methods**

#### **2. Materials and Methods** Figure 1. The silicon wafer substrate was consecutively cleaned with acetone, ethanol, and deionized water, and cleaned by vacuum plasma for 10 min after being dried by N<sup>2</sup> gas.

*Polymers* **2022**, *13*, x FOR PEER REVIEW 3 of 14

structure was imitated to exhibit the same liquid transport capability as the Nepenthes

The photolithography process of a Nepenthes peristome-like structure is shown in

#### *2.1. Photolithography*

The photolithography process of a Nepenthes peristome-like structure is shown in Figure 1. The silicon wafer substrate was consecutively cleaned with acetone, ethanol, and deionized water, and cleaned by vacuum plasma for 10 min after being dried by N<sup>2</sup> gas. After the plasma treatment, the surface achieved temporary hydrophilicity to increase the adhesion of the substrate and photoresist. SU-8 negative photoresist (KAYAKU SU8-2025) was spin-coated at a gradual acceleration speed of 500 rpm for 10 s, 1500 rpm for 10 s, and 3000 rpm for 30 s, which could make the homogenous coating for the SU-8 photoresist. Then, incubation occurred for 30 min before soft baking. Soft baking following gradual heating (65–90 ◦C) can avoid bubbles caused by an excessive heating rate. SU-8 negative photoresist was cross-linked with a mercury lamp (365 nm) for 25 s, with an exposure process of two layers of photomask patterns. After exposure, adhesion of the exposed pattern and the substrate were improved by post-exposure baking at 90 ◦C, and then immersed in the developer to remove the un-crosslinked SU-8 photoresist. Following developments, we rinsed with deionized water and hard baked for 3 min to remove the last remaining solvent. The SU-8 mold of the Nepenthes peristome-like structure was then used in the further replica of the PDMS membranes. The easy-cleaning coating solution, Naegix E720 (Senguan Tech Co., Ltd., Tainan City, Taiwan), was used as the control to evaluate the surface energy of a replica of a bionic Nepenthes peristome-like structure. After the plasma treatment, the surface achieved temporary hydrophilicity to increase the adhesion of the substrate and photoresist. SU-8 negative photoresist (KAYAKU SU8-2025) was spin-coated at a gradual acceleration speed of 500 rpm for 10 s, 1500 rpm for 10 s, and 3000 rpm for 30 s, which could make the homogenous coating for the SU-8 photoresist. Then, incubation occurred for 30 min before soft baking. Soft baking following gradual heating (65–90 °C) can avoid bubbles caused by an excessive heating rate. SU-8 negative photoresist was cross-linked with a mercury lamp (365 nm) for 25 s, with an exposure process of two layers of photomask patterns. After exposure, adhesion of the exposed pattern and the substrate were improved by post-exposure baking at 90 °C, and then immersed in the developer to remove the un-crosslinked SU-8 photoresist. Following developments, we rinsed with deionized water and hard baked for 3 min to remove the last remaining solvent. The SU-8 mold of the Nepenthes peristome-like structure was then used in the further replica of the PDMS membranes. The easy-cleaning coating solution, Naegix E720 (Senguan Tech Co., Ltd., Tainan City, Taiwan), was used as the control to evaluate the surface energy of a replica of a bionic Nepenthes peristome-like structure.

**Figure 1.** Diagram of the photolithography process of a Nepenthes peristome-like structure. **Figure 1.** Diagram of the photolithography process of a Nepenthes peristome-like structure.

#### *2.2. Replica of a Nepenthes Peristome-like Structure by PDMS*

*2.2. Replica of a Nepenthes Peristome-like Structure by PDMS* Polydimethylsiloxane (PDMS, Dow Corning® Sylgard 184, part A) and a crosslinking agent (Sylgard 184, part B) were mixed with a 10:1 PDMS base to a curing agent ratio. The stiffness of the PDMS can be manipulated by the addition of a curing agent. To avoid bubbles, the mixed PDMS must be placed in a vacuum desiccator until the bubbles disappear. The PDMS solution was then poured into a 35 mm dish and placed in the oven at 60 °C for 4 h until the structure stabilized. SU-8 based bionic (Nepenthes peristome-like) structure was cleaned with deionized water before plasma cleaning for 1 min. Moreover, Polydimethylsiloxane (PDMS, Dow Corning® Sylgard 184, part A) and a crosslinking agent (Sylgard 184, part B) were mixed with a 10:1 PDMS base to a curing agent ratio. The stiffness of the PDMS can be manipulated by the addition of a curing agent. To avoid bubbles, the mixed PDMS must be placed in a vacuum desiccator until the bubbles disappear. The PDMS solution was then poured into a 35 mm dish and placed in the oven at 60 ◦C for 4 h until the structure stabilized. SU-8 based bionic (Nepenthes peristome-like) structure was cleaned with deionized water before plasma cleaning for 1 min. Moreover, the platinum was coated on the surface of the SU-8 based bionic structure. After all these steps were completed, the mixed PDMS solution was poured into the SU-8 mold and

the platinum was coated on the surface of the SU-8 based bionic structure. After all these

vacuumed to remove extra bubbles. Then, it stood in the oven at 65 ◦C for 4 h. The bionic replica of the PDMS membrane was slit to generate the Nepenthes peristome-like structure (Figure 2). replica of the PDMS membrane was slit to generate the Nepenthes peristome-like structure (Figure 2). uumed to remove extra bubbles. Then, it stood in the oven at 65 °C for 4 h. The bionic replica of the PDMS membrane was slit to generate the Nepenthes peristome-like structure (Figure 2).

steps were completed, the mixed PDMS solution was poured into the SU-8 mold and vac-

steps were completed, the mixed PDMS solution was poured into the SU-8 mold and vacuumed to remove extra bubbles. Then, it stood in the oven at 65 °C for 4 h. The bionic

*Polymers* **2022**, *13*, x FOR PEER REVIEW 4 of 14

*Polymers* **2022**, *13*, x FOR PEER REVIEW 4 of 14

**Figure 2.** Diagram of the replicated Nepenthes peristome-like structure process by polydime-**Figure 2.** Diagram of the replicated Nepenthes peristome-like structure process by polydimethylsiloxane (PDMS). *2.3. MPC Immobilization by Atmospheric Plasma* The zwitterion polymer (2-methacryloyloxyethyl phosphorylcholine, MPC) modifi-

#### thylsiloxane (PDMS). *2.3. MPC Immobilization by Atmospheric Plasma* cation flowchart is shown in Figure 3. The bionic replica of the PDMS membrane was pre-

*2.3. MPC Immobilization by Atmospheric Plasma* The zwitterion polymer (2-methacryloyloxyethyl phosphorylcholine, MPC) modification flowchart is shown in Figure 3. The bionic replica of the PDMS membrane was pretreated by plasma cleaning for 1 min (Figure 3a) and then immersed in MPC solution (Figure 3b) [27,28]. In addition to activating the surface functional groups, the hydrophilic treatment can effectively improve the uniformity of the modification. Subsequently, the The zwitterion polymer (2-methacryloyloxyethyl phosphorylcholine, MPC) modification flowchart is shown in Figure 3. The bionic replica of the PDMS membrane was pre-treated by plasma cleaning for 1 min (Figure 3a) and then immersed in MPC solution (Figure 3b) [27,28]. In addition to activating the surface functional groups, the hydrophilic treatment can effectively improve the uniformity of the modification. Subsequently, the pre-treated PDMS was treated by oxygen atmospheric plasma (AP Plasma Jet, Feng Tien Electronic Co., Ltd., Taipei, Taiwan) (Figure 3c) to polymerize the MPC monomers to MPC polymer brushes (Figure 3d) at an operating power of 1.2 kW and oxygen flux rate of 10 slm (L/min) for 10 s, and a 1.2 cm distance was maintained from the plasma torch. treated by plasma cleaning for 1 min (Figure 3a) and then immersed in MPC solution (Figure 3b) [27,28]. In addition to activating the surface functional groups, the hydrophilic treatment can effectively improve the uniformity of the modification. Subsequently, the pre-treated PDMS was treated by oxygen atmospheric plasma (AP Plasma Jet, Feng Tien Electronic Co., Ltd., Taipei, Taiwan) (Figure 3c) to polymerize the MPC monomers to MPC polymer brushes (Figure 3d) at an operating power of 1.2 kW and oxygen flux rate of 10 slm (L/min) for 10 s, and a 1.2 cm distance was maintained from the plasma torch.

**(a) (b) (c) (d) Figure 3.** Schematic diagram of the replica of the PDMS-based bionic Nepenthes peristome-like structure with 2-methacryloyloxyethl phosphorylcholine (MPC) immobilization by atmospheric plasma. (**a**) The replica PDMS membrane was pre-treated by plasma cleaning for 1 min; (**b**) PDMS immersed in MPC solution; (**c**) MPC monomers were polymerized by oxygen atmospheric plasma and formed (**d**) MPC polymer brushes. **Figure 3.** Schematic diagram of the replica of the PDMS-based bionic Nepenthes peristome-like structure with 2-methacryloyloxyethl phosphorylcholine (MPC) immobilization by atmospheric plasma. (**a**) The replica PDMS membrane was pre-treated by plasma cleaning for 1 min; (**b**) PDMS immersed in MPC solution; (**c**) MPC monomers were polymerized by oxygen atmospheric plasma and formed (**d**) MPC polymer brushes.

#### *2.4. Anti-Bacterial Adhesion Capability*

croscope.

croscope.

*2.4. Anti-Bacterial Adhesion Capability* The antibacterial effect of PDMS and PDMS-PMPC against bacteria (*E. coli*) was inspected with the bacterial adhesion method. The samples were shaken with the lysate of *E. coli* at 37 °C. After 24 h, the samples were stained with SYTO 9 (green fluorescence The antibacterial effect of PDMS and PDMS-PMPC against bacteria (*E. coli*) was inspected with the bacterial adhesion method. The samples were shaken with the lysate of *E. coli* at 37 ◦C. After 24 h, the samples were stained with SYTO 9 (green fluorescence nucleic acid stain) and stood for 5 min. The results were obtained with fluorescence microscope.

nucleic acid stain) and stood for 5 min. The results were obtained with fluorescence mi-

**Figure 3.** Schematic diagram of the replica of the PDMS-based bionic Nepenthes peristome-like

plasma. (**a**) The replica PDMS membrane was pre-treated by plasma cleaning for 1 min; (**b**) PDMS immersed in MPC solution; (**c**) MPC monomers were polymerized by oxygen atmospheric plasma

The antibacterial effect of PDMS and PDMS-PMPC against bacteria (*E. coli*) was in-

spected with the bacterial adhesion method. The samples were shaken with the lysate of *E. coli* at 37 °C. After 24 h, the samples were stained with SYTO 9 (green fluorescence nucleic acid stain) and stood for 5 min. The results were obtained with fluorescence mi-

and formed (**d**) MPC polymer brushes.

*2.4. Anti-Bacterial Adhesion Capability*

#### *2.5. Anti-Protein Adhesion Capability*

The PDMS and PDMS-PMPC were incubated in a 10 mL phosphate-buffered solution (PBS) solution of albumin from human serum (HSA) in the 24-well tissue culture plate at 37 ◦C for 1 h. Afterward, the samples were gently washed 3 times using PBS in the 24-well tissue culture plate. Then, the samples were incubated with 1 wt% aqueous solution of sodium dodecyl sulfate (SDS). The BCA kit was used to determine the concentration of the proteins in the SDS solution, and the concentration was detected by UV-vis spectroscopy.

#### *2.6. Cell Attachment Tests*

The cell attachment was detected with cells adhesion. The PDMS and PDMS-PMPC were placed in the 24-well plate, injected with NIH 3T3 mouse embryonic fibroblast cells (3T3 cells) to the substrate surface, and incubated at 37 ◦C for 24 h. The cells' attachment behavior would be evaluated by fluorescent staining (nucleus staining and cell membrane staining) to observe the amount of cell adhesion by fluorescent microscopy.

#### *2.7. Biocompatibility*

Biocompatibility is defined as the ability of a material to perform with an appropriate host response in a specific application [28]. The biocompatibility of pristine PDMS and PDMS-PMPC was evaluated with the proliferation of 3T3 fibroblast. The substrates (1 cm <sup>×</sup> 1 cm) and 1 mL solution of 3T3 cells (10<sup>5</sup> cells/mL) were placed in the 24-well plate under a 5% CO<sup>2</sup> atmosphere The cell viability was determined at 37 ◦C for 24, 48, and 72 h by thiazolyl blue tetrazolium bromide (MTT) assay and absorbance at 570 nm.

#### *2.8. One-Way Liquid Transfer Capability*

One-way liquid transfer capability of PDMS and PDMS-PMPC substrates was measured by the stained deionized water dropped on the PDMS and PDMS-PMPC substrates. The distance of stained deionized water flowing was recorded after 4 min. The flowing distance would be evaluated between pristine PDMS and PDMS-PMPC substrates. The longer flowing distance shows the greater one-way liquid transfer capability.

## *2.9. Characterizations*

The morphology of the bionic SERS substrate was observed by scanning electron microscopy (SEM) (JEOL JSM-6701F, Tokyo, Japan). FT-IR spectroscopy (FT-IR, Perkin-Elmer Spectrum-One, Shelton, CT, USA) was used to differentiate the composition and chemical structure of the bionic SERS substrate. The chemical-binding energy of the bionic SERS substrate was carried out by a K-Alpha X-ray photoelectron spectrometer (XPS) (Thermo Fisher Scientific, Waltham, MA, USA). The hydrophilicity of the bionic SERS substrates was recorded by a contact angle goniometer (DSA 100, Krüss GmbH, Hamburg, Germany). Raman spectroscopy (HORIBA, LabRAM HR Evolution) was used to evaluate the SERS spectra of the bionic SERS substrate, with a 632.8 nm He-Ne laser, operated under a 10<sup>×</sup> objective lens with a detection range of 400–2000 cm−<sup>1</sup> .

#### **3. Results and Discussion**

#### *3.1. Optical Microscope (OM) Analysis*

The photomask patterns of the Nepenthes peristome-like structure were shown in Figure 4a–c. The arrow arrays (Figure 4a) were fabricated in the first (bottom) layer. Then, the straight pattern arrays (Figure 4b) were covered as the second layer to develop the bionic Nepenthes peristome-like structure (Figure 4c). Figure 4d–f show the OM images of the first layer (Figure 4d), the second layer (Figure 4e), and the overlaid layers pattern (Figure 4f) of SU-8 photoresist during the photolithographic process. The resulting three patterns are similar to the three photomask patterns. The line point A to point B (Figures 4d and 5) passes through the plane and the pattern represents the segmented plane of the cross-section from the first layer of SU-8 photoresist. Although the arrow array (Figure 4d) becomes broader during the photolithographic process, compared to the

photoresist.

the pristine Nepenthes peristome.

pristine photomask patterns (Figure 4a), the gap channels between the arrow arrays do not change too much. Furthermore, the cross-section of the first layer replicated by PDMS was observed in Figure 5. The depth of the gap channel could be clearly observed in the cross-section image, showing the pattern integrity after the photolithographic process. much. Furthermore, the cross-section of the first layer replicated by PDMS was observed in Figure 5. The depth of the gap channel could be clearly observed in the cross-section image, showing the pattern integrity after the photolithographic process.

image, showing the pattern integrity after the photolithographic process.

*Polymers* **2022**, *13*, x FOR PEER REVIEW 6 of 14

4d and 5) passes through the plane and the pattern represents the segmented plane of the cross-section from the first layer of SU-8 photoresist. Although the arrow array (Figure 4d) becomes broader during the photolithographic process, compared to the pristine photomask patterns (Figure 4a), the gap channels between the arrow arrays do not change too much. Furthermore, the cross-section of the first layer replicated by PDMS was observed in Figure 5. The depth of the gap channel could be clearly observed in the cross-section

4d and 5) passes through the plane and the pattern represents the segmented plane of the

tomask patterns (Figure 4a), the gap channels between the arrow arrays do not change too

*Polymers* **2022**, *13*, x FOR PEER REVIEW 6 of 14

**Figure 4.** Diagram of (**a**) the first and (**b**) the second layers of photomask patterns; (**c**) overlaid photomask patterns. Optical microscope (OM) image of (**d**) first layer pattern, (**e**) second layer pattern, and (**f**) overlaid layers pattern of SU-8 photoresist. The line point A to point B passes through the plane and the pattern indicates the segmented plane of the cross-section from the first layer of SU-8 **Figure 4.** Diagram of (**a**) the first and (**b**) the second layers of photomask patterns; (**c**) overlaid photomask patterns. Optical microscope (OM) image of (**d**) first layer pattern, (**e**) second layer pattern, and (**f**) overlaid layers pattern of SU-8 photoresist. The line point A to point B passes through the plane and the pattern indicates the segmented plane of the cross-section from the first layer of SU-8 photoresist. tomask patterns. Optical microscope (OM) image of (**d**) first layer pattern, (**e**) second layer pattern, and (**f**) overlaid layers pattern of SU-8 photoresist. The line point A to point B passes through the plane and the pattern indicates the segmented plane of the cross-section from the first layer of SU-8 photoresist.

*3.2. Scanning Electron Microscope (SEM) Observation* **Figure 5.** The OM image of the cross-section of the first layer of SU-8 photoresist. **Figure 5.** The OM image of the cross-section of the first layer of SU-8 photoresist.

#### From Figure 6a–c, the continuous radially arranged furrows were observed with *3.2. Scanning Electron Microscope (SEM) Observation*

grooves at the real (pristine) Nepenthes peristome structure. The width was 30–40 µm and the length was 130–160 µm. In comparison, the replica of the PDMS-based bionic Nepenthes peristome-like structure is shown in Figure 6d–f. The double layer structure was suc-*3.2. Scanning Electron Microscope (SEM) Observation* From Figure 6a–c, the continuous radially arranged furrows were observed with grooves at the real (pristine) Nepenthes peristome structure. The width was 30–40 µm and From Figure 6a–c, the continuous radially arranged furrows were observed with grooves at the real (pristine) Nepenthes peristome structure. The width was 30–40 µm and the length was 130–160 µm. In comparison, the replica of the PDMS-based bionic Nepenthes peristome-like structure is shown in Figure 6d–f. The double layer structure

the length was 130–160 µm. In comparison, the replica of the PDMS-based bionic Nepen-

the pristine Nepenthes peristome.

cessfully fabricated by the replicated process, and the detail structure is very similar to

*Polymers* **2022**, *13*, x FOR PEER REVIEW 7 of 14

was successfully fabricated by the replicated process, and the detail structure is very similar to the pristine Nepenthes peristome. *Polymers* **2022**, *13*, x FOR PEER REVIEW 7 of 14

**Figure 6.** SEM images of (**a**–**c**) the real (pristine) Nepenthes peristome and (**d**–**f**) the replica of the Nepenthes peristome-like structure of PDMS. **Figure 6.** SEM images of (**a**–**c**) the real (pristine) Nepenthes peristome and (**d**–**f**) the replica of the Nepenthes peristome-like structure of PDMS. **Figure 6.** SEM images of (**a**–**c**) the real (pristine) Nepenthes peristome and (**d**–**f**) the replica of the Nepenthes peristome-like structure of PDMS.

#### *3.3. Hydrophilicity by Contact Angle Measurements 3.3. Hydrophilicity by Contact Angle Measurements 3.3. Hydrophilicity by Contact Angle Measurements*

The contact angle of the pristine Nepenthes peristome exhibiting a super-hydrophilic surface leads to an unapparent water contact angle (almost 0°), as shown in Figure 7a,b. However, the replica of the bionic PDMS surface (without modified MPC) shows the super-hydrophobicity (CA: ~142°) in Figure 7c,d, which is higher than the coating of commercial self-cleaning products (Naegix E720) (~90°). The structure of the replica of bionic PDMS is similar to the pristine Nepenthes peristome, but the surface energy is completely different. The reason for that is because the hydrophilic functional group was found on the pristine Nepenthes peristome, whereas the replica of the bionic PDMS surface is more hydrophobic. Therefore, MPC polymer brushes (PMPC) were grafted on the PDMS substrate by atmospheric plasma to improve hydrophilicity. The contact angle displayed was ~28° after PMPC was immobilized on the PDMS-replicated Nepenthes peristome, which The contact angle of the pristine Nepenthes peristome exhibiting a super-hydrophilic surface leads to an unapparent water contact angle (almost 0◦ ), as shown in Figure 7a,b. However, the replica of the bionic PDMS surface (without modified MPC) shows the super-hydrophobicity (CA: ~142◦ ) in Figure 7c,d, which is higher than the coating of commercial self-cleaning products (Naegix E720) (~90◦ ). The structure of the replica of bionic PDMS is similar to the pristine Nepenthes peristome, but the surface energy is completely different. The reason for that is because the hydrophilic functional group was found on the pristine Nepenthes peristome, whereas the replica of the bionic PDMS surface is more hydrophobic. Therefore, MPC polymer brushes (PMPC) were grafted on the PDMS substrate by atmospheric plasma to improve hydrophilicity. The contact angle displayed was ~28◦ after PMPC was immobilized on the PDMS-replicated Nepenthes peristome, which is closer to the pristine Nepenthes peristome (Figure 7e,f). The contact angle of the pristine Nepenthes peristome exhibiting a super-hydrophilic surface leads to an unapparent water contact angle (almost 0°), as shown in Figure 7a,b. However, the replica of the bionic PDMS surface (without modified MPC) shows the super-hydrophobicity (CA: ~142°) in Figure 7c,d, which is higher than the coating of commercial self-cleaning products (Naegix E720) (~90°). The structure of the replica of bionic PDMS is similar to the pristine Nepenthes peristome, but the surface energy is completely different. The reason for that is because the hydrophilic functional group was found on the pristine Nepenthes peristome, whereas the replica of the bionic PDMS surface is more hydrophobic. Therefore, MPC polymer brushes (PMPC) were grafted on the PDMS substrate by atmospheric plasma to improve hydrophilicity. The contact angle displayed was ~28° after PMPC was immobilized on the PDMS-replicated Nepenthes peristome, which is closer to the pristine Nepenthes peristome (Figure 7e,f).

**Figure 7.** The contact angle of (**a**,**b**) the pristine Nepenthes peristome (almost 0°), (**c**,**d**) Nepenthes **Figure 7.** The contact angle of (**a**,**b**) the pristine Nepenthes peristome (almost 0°), (**c**,**d**) Nepenthes peristome-like structure replicated by PDMS without MPC modification, and (**e**,**f**) the Nepenthes peristome-like structure replicated by PDMS with MPC modification. **Figure 7.** The contact angle of (**a**,**b**) the pristine Nepenthes peristome (almost 0◦ ), (**c**,**d**) Nepenthes peristome-like structure replicated by PDMS without MPC modification, and (**e**,**f**) the Nepenthes peristome-like structure replicated by PDMS with MPC modification.

peristome-like structure replicated by PDMS without MPC modification, and (**e**,**f**) the Nepenthes

peristome-like structure replicated by PDMS with MPC modification.

#### *3.4. FTIR Spectrum Analysis* FTIR spectrum of PDMS shows the characteristic peaks at a stretching frequency of 789

*3.4. FTIR Spectrum Analysis*

Figure 8 shows the FTIR spectra of pristine PDMS and PDMS coated with PMPC. The FTIR spectrum of PDMS shows the characteristic peaks at a stretching frequency of 789 cm−<sup>1</sup> bending vibration modes of Si-CH<sup>3</sup> and Si-(CH2)n-Si, the characteristic peaks at the stretching frequency of 1020 cm−<sup>1</sup> due to the bending vibration modes of the asymmetric Si–O–Si stretching vibrations, and the characteristic peaks at 1259 cm−<sup>1</sup> due to the bending vibration modes of Si-CH<sup>3</sup> symmetric bending [35,36]. PDMS coated with PMPC has the characteristic peaks of 971 cm−<sup>1</sup> , 1090 cm−<sup>1</sup> , and 1247 cm−<sup>1</sup> due to P-O, R-N<sup>+</sup> (CH3)3, and the P=O stretching vibrational band [37,38]. The results confirmed that PMPC polymer brushes were successfully immobilized on the PDMS substrate. cm*−*<sup>1</sup> bending vibration modes of Si-CH<sup>3</sup> and Si-(CH2)n-Si, the characteristic peaks at the stretching frequency of 1020 cm*−*<sup>1</sup> due to the bending vibration modes of the asymmetric Si–O–Si stretching vibrations, and the characteristic peaks at 1259 cm*−*<sup>1</sup> due to the bending vibration modes of Si-CH<sup>3</sup> symmetric bending [35,36]. PDMS coated with PMPC has the characteristic peaks of 971 cm*−*<sup>1</sup> , 1090 cm*−*<sup>1</sup> , and 1247 cm*−*<sup>1</sup> due to P-O, R-N<sup>+</sup> (CH3)3, and the P=O stretching vibrational band [37,38]. The results confirmed that PMPC polymer brushes were successfully immobilized on the PDMS substrate.

Figure 8 shows the FTIR spectra of pristine PDMS and PDMS coated with PMPC. The

*Polymers* **2022**, *13*, x FOR PEER REVIEW 8 of 14

**Figure 8.** FTIR spectrum of PDMS and MPC polymer brushes (PMPC) were grafted on the PDMS substrate and PDMA-PMPC substrate. **Figure 8.** FTIR spectrum of PDMS and MPC polymer brushes (PMPC) were grafted on the PDMS substrate and PDMA-PMPC substrate.

#### *3.5. XPS Spectrum Analysis*

N<sup>+</sup>

*3.5. XPS Spectrum Analysis* The binding energy change between the PDMS and PDMS-PMPC substrates was measured by an XPS analysis (Figure 9). The full spectra (Figure 9a) show that the surface primarily contained O, Si, C, P, and N. Compared with the PDMS surface, the characteristic peaks of N-1s and P-2s orbitals were observed at the PDMS-PMPC substrates, which contribute to the phosphorus and nitrogen bonds of MPC polymer brushes. The element ratios of N and P element contents of PDMS-PMPC were 4.43% and 4.66%, whereas the characteristic peaks of N-1s and P-2s were barely observed on the pristine PDMS. From the XPS spectra of C-1s (Figure 9b), new characteristic peaks were observed at C-O (286 eV) and C=O (288.5 eV) for the PDMS-PMPC substrates. In addition, an additional peak of N-1s spectra (Figure 9c) was found at 402.4 eV, which is the functional group of The binding energy change between the PDMS and PDMS-PMPC substrates was measured by an XPS analysis (Figure 9). The full spectra (Figure 9a) show that the surface primarily contained O, Si, C, P, and N. Compared with the PDMS surface, the characteristic peaks of N-1s and P-2s orbitals were observed at the PDMS-PMPC substrates, which contribute to the phosphorus and nitrogen bonds of MPC polymer brushes. The element ratios of N and P element contents of PDMS-PMPC were 4.43% and 4.66%, whereas the characteristic peaks of N-1s and P-2s were barely observed on the pristine PDMS. From the XPS spectra of C-1s (Figure 9b), new characteristic peaks were observed at C-O (286 eV) and C=O (288.5 eV) for the PDMS-PMPC substrates. In addition, an additional peak of N-1s spectra (Figure 9c) was found at 402.4 eV, which is the functional group of N<sup>+</sup> (CH3)<sup>3</sup> on the PDMS-PMPC substrate. These results are similar to the FTIR analysis (Figure 8), indicating that the PMPC polymer brushes can grow on the replica of the bionic PDMS surface after atmospheric plasma polymerization.

(CH3)<sup>3</sup> on the PDMS-PMPC substrate. These results are similar to the FTIR analysis (Figure 8), indicating that the PMPC polymer brushes can grow on the replica of the bionic

PDMS surface after atmospheric plasma polymerization.

**Figure 9.** (**a**) XPS full spectra, (**b**) C-1s spectra, and (**c**) N-1s spectra of the PDMS and PDMS-PMPC substrates. **Figure 9.** (**a**) XPS full spectra, (**b**) C-1s spectra, and (**c**) N-1s spectra of the PDMS and PDMS-PMPC substrates.

#### *3.6. Antibacterial Adhesion Test*

bacterial properties [28].

*3.6. Antibacterial Adhesion Test* The antibacterial effect of PDMS and PDMS-PMPC against bacteria (*E. coli*) was assessed using the bacterial adhesion method. The fluorescence images of bacterial mor-The antibacterial effect of PDMS and PDMS-PMPC against bacteria (*E. coli*) was assessed using the bacterial adhesion method. The fluorescence images of bacterial morphology (Figure 10) demonstrate that the bacteria are attached to the sample surface. We

phology (Figure 10) demonstrate that the bacteria are attached to the sample surface. We noticed a higher number of bacteria on the pristine PDMS substrate (Figure 10a) than on

the steric hindrance, and the steric hindrance enables the PDMS surface to display anti-

noticed a higher number of bacteria on the pristine PDMS substrate (Figure 10a) than on the PMPC-coated PDMS substrate (Figure 10b). This is because the PMPC structure forms the steric hindrance, and the steric hindrance enables the PDMS surface to display antibacterial properties [28]. *Polymers* **2022**, *13*, x FOR PEER REVIEW 10 of 14

*Polymers* **2022**, *13*, x FOR PEER REVIEW 10 of 14

**Figure 10.** The anti-fouling capability of bacteria (*E. coli*) on the (**a**) pristine PDMS and (**b**) PDMS-PMPC substrate. **Figure 10.** The anti-fouling capability of bacteria (*E. coli*) on the (**a**) pristine PDMS and (**b**) PDMS-PMPC substrate. *3.7. Anti-Protein Adhesion Test* The anti-protein adsorption performance of the PDMS and PDSA-PMPC surfaces

#### *3.7. Anti-Protein Adhesion Test 3.7. Anti-Protein Adhesion Test* were quantitatively evaluated by a BCA protein assay (Thermo Fisher, USA). Figure 11 shows the amounts of human serum albumin (HSA), which adhered to the pristine and

The anti-protein adsorption performance of the PDMS and PDSA-PMPC surfaces were quantitatively evaluated by a BCA protein assay (Thermo Fisher, USA). Figure 11 shows the amounts of human serum albumin (HSA), which adhered to the pristine and PMPC-coated PDMS substrates. The adsorbed amount of HSA on the pristine PDMS was 60% higher than the PDMS-PMPC substrate. The PMPC was very effective in preventing the protein from adsorption. This result corresponded with the contact angle. The smaller the water contact angle after PMPC grafting, the greater the anti-fouling capability of pro-The anti-protein adsorption performance of the PDMS and PDSA-PMPC surfaces were quantitatively evaluated by a BCA protein assay (Thermo Fisher, USA). Figure 11 shows the amounts of human serum albumin (HSA), which adhered to the pristine and PMPC-coated PDMS substrates. The adsorbed amount of HSA on the pristine PDMS was 60% higher than the PDMS-PMPC substrate. The PMPC was very effective in preventing the protein from adsorption. This result corresponded with the contact angle. The smaller the water contact angle after PMPC grafting, the greater the anti-fouling capability of proteins [34]. PMPC-coated PDMS substrates. The adsorbed amount of HSA on the pristine PDMS was 60% higher than the PDMS-PMPC substrate. The PMPC was very effective in preventing the protein from adsorption. This result corresponded with the contact angle. The smaller the water contact angle after PMPC grafting, the greater the anti-fouling capability of proteins [34].

**Figure 11.** The anti-fouling capability of proteins (human serum albumin, HSA) on the pristine PDMS and PDMS-PMPC substrates (*n* = 3). **Figure 11.** The anti-fouling capability of proteins (human serum albumin, HSA) on the pristine PDMS and PDMS-PMPC substrates (*n* = 3). **Figure 11.** The anti-fouling capability of proteins (human serum albumin, HSA) on the pristine PDMS and PDMS-PMPC substrates (*n* = 3).

#### *3.8. Cells Attachment Test 3.8. Cells Attachment Test*

The results of 3T3 cells attachment on PDMS and PDMS-PMPC substrate are shown in Figure 12. Although the PDMS surface was hydrophobic, the amount of surface cell adhesion was concentrated and dense (Figure 12a). The PDMS grafted on PMPC (Figure 12b) made it difficult for the cells to attach to the substrate due to the change in surface charge, reduced the number of cells attached, and kept the cell adhesion survival rate. *3.8. Cells Attachment Test* The results of 3T3 cells attachment on PDMS and PDMS-PMPC substrate are shown in Figure 12. Although the PDMS surface was hydrophobic, the amount of surface cell adhesion was concentrated and dense (Figure 12a). The PDMS grafted on PMPC (Figure The results of 3T3 cells attachment on PDMS and PDMS-PMPC substrate are shown in Figure 12. Although the PDMS surface was hydrophobic, the amount of surface cell adhesion was concentrated and dense (Figure 12a). The PDMS grafted on PMPC (Figure 12b) made it difficult for the cells to attach to the substrate due to the change in surface charge, reduced the number of cells attached, and kept the cell adhesion survival rate.

12b) made it difficult for the cells to attach to the substrate due to the change in surface charge, reduced the number of cells attached, and kept the cell adhesion survival rate.

*Polymers* **2022**, *13*, x FOR PEER REVIEW 11 of 14

**Figure 12.** 3T3 cells attachment test on the (**a**) pristine PDMS and (**b**) PDMS-PMPC substrates. **Figure 12.** 3T3 cells attachment test on the (**a**) pristine PDMS and (**b**) PDMS-PMPC substrates. *3.9. Biocompatibility Test*

#### *3.9. Biocompatibility Test* Biocompatibility is defined as the ability of a material to perform with an appropriate

*3.9. Biocompatibility Test* Biocompatibility is defined as the ability of a material to perform with an appropriate host response in a specific application [31]. That is, the material will respond appropriately when it comes into contact with the host. In this study, the biocompatibility of pristine PDMS and PDMS-PMPC was evaluated with the proliferation of NIH 3T3 cells, which are mouse embryonic fibroblasts and have been widely used in biocompatibility tests. Figure 13 shows that 3T3 cells proliferated on the control, PDMS, and PDMS-PMPC substrates at 1–3 days. The results show that cell viability (%) at day 1 was almost the same in the three samples, but pristine PDMS and PDMS-MPC dropped to about 70% at day 2. At day 3, the cell viability (%) of 3T3 cells on the PDMS continued to decline, whereas that on the PDMS-PMPC increased to 85%, similarly with the control group. This demonstrates that Biocompatibility is defined as the ability of a material to perform with an appropriate host response in a specific application [31]. That is, the material will respond appropriately when it comes into contact with the host. In this study, the biocompatibility of pristine PDMS and PDMS-PMPC was evaluated with the proliferation of NIH 3T3 cells, which are mouse embryonic fibroblasts and have been widely used in biocompatibility tests. Figure 13 shows that 3T3 cells proliferated on the control, PDMS, and PDMS-PMPC substrates at 1–3 days. The results show that cell viability (%) at day 1 was almost the same in the three samples, but pristine PDMS and PDMS-MPC dropped to about 70% at day 2. At day 3, the cell viability (%) of 3T3 cells on the PDMS continued to decline, whereas that on the PDMS-PMPC increased to 85%, similarly with the control group. This demonstrates that PMPC immobilization on the surface can effectively improve biocompatibility at day 3, compared to the pristine PDMS. host response in a specific application [31]. That is, the material will respond appropriately when it comes into contact with the host. In this study, the biocompatibility of pristine PDMS and PDMS-PMPC was evaluated with the proliferation of NIH 3T3 cells, which are mouse embryonic fibroblasts and have been widely used in biocompatibility tests. Figure 13 shows that 3T3 cells proliferated on the control, PDMS, and PDMS-PMPC substrates at 1–3 days. The results show that cell viability (%) at day 1 was almost the same in the three samples, but pristine PDMS and PDMS-MPC dropped to about 70% at day 2. At day 3, the cell viability (%) of 3T3 cells on the PDMS continued to decline, whereas that on the PDMS-PMPC increased to 85%, similarly with the control group. This demonstrates that PMPC immobilization on the surface can effectively improve biocompatibility at day 3, compared to the pristine PDMS.

**Figure 13.** Biocompatibility tests of control, pristine PDMS, and PDMS-PMPC substrates (*n* = 3). **Figure 13.** Biocompatibility tests of control, pristine PDMS, and PDMS-PMPC substrates (*n* = 3).

#### *3.10. One-Way Liquid Transfer Capability*

the SERS technique.

the SERS technique.

**Figure 13.** Biocompatibility tests of control, pristine PDMS, and PDMS-PMPC substrates (*n* = 3). *3.10. One-Way Liquid Transfer Capability* The liquid channels (20 mm) were fabricated by imitating channels of the real Nepenthes peristome, as shown in Figure 14a. The stained deionized water (10 μL) was utilized for its one-way liquid transfer capability. The results indicated that the ability of selfdriving and unidirectional water delivery was found in the PDMS-PMPC substrate (redstained water) after 4 min incubation (running distance: ~10 mm), compared with the *3.10. One-Way Liquid Transfer Capability* The liquid channels (20 mm) were fabricated by imitating channels of the real Nepenthes peristome, as shown in Figure 14a. The stained deionized water (10 μL) was utilized for its one-way liquid transfer capability. The results indicated that the ability of selfdriving and unidirectional water delivery was found in the PDMS-PMPC substrate (redstained water) after 4 min incubation (running distance: ~10 mm), compared with the sticky blue stained water in the PDMS substrate (Figure 14b). The delivery rate of PDMS-PMPC was about 2.5 mm/min, which provides the chromatographic capability of the complex samples. The complex samples could be separated in the chromatographic process The liquid channels (20 mm) were fabricated by imitating channels of the real Nepenthes peristome, as shown in Figure 14a. The stained deionized water (10 µL) was utilized for its one-way liquid transfer capability. The results indicated that the ability of self-driving and unidirectional water delivery was found in the PDMS-PMPC substrate (red-stained water) after 4 min incubation (running distance: ~10 mm), compared with the sticky blue stained water in the PDMS substrate (Figure 14b). The delivery rate of PDMS-PMPC was about 2.5 mm/min, which provides the chromatographic capability of the complex samples. The complex samples could be separated in the chromatographic process (depending on the polarity or molecular weight of the samples), and then detected using the SERS technique.

sticky blue stained water in the PDMS substrate (Figure 14b). The delivery rate of PDMS-PMPC was about 2.5 mm/min, which provides the chromatographic capability of the com-

(depending on the polarity or molecular weight of the samples), and then detected using

(depending on the polarity or molecular weight of the samples), and then detected using

*Polymers* **2022**, *13*, x FOR PEER REVIEW 12 of 14

**Figure 14.** (**a**) The liquid channels (20 mm) of the PDMS-based substrate were used for the channels of the real Nepenthes peristome; (**b**) the stained deionized water (10 μL) was used for the one-way liquid transfer capability (blue water for PDMS and red water for PDMS-PMPC). **Figure 14.** (**a**) The liquid channels (20 mm) of the PDMS-based substrate were used for the channels of the real Nepenthes peristome; (**b**) the stained deionized water (10 µL) was used for the one-way liquid transfer capability (blue water for PDMS and red water for PDMS-PMPC). *3.11. SERS Detection by Raman Spectroscopy* To enable the SERS detection (Figure 15), the replica PDMS-PMPC with a bionic

#### *3.11. SERS Detection by Raman Spectroscopy 3.11. SERS Detection by Raman Spectroscopy* structure was deposited on 5 nm of Ag nanofilm (PDMS-PMPC-Ag (W/)) by thermal evap-

To enable the SERS detection (Figure 15), the replica PDMS-PMPC with a bionic structure was deposited on 5 nm of Ag nanofilm (PDMS-PMPC-Ag (W/)) by thermal evaporation, and pristine PDMS-PMPC with Ag (PDMS-PMPC-Ag (W/O)) deposition was used as the control. The 633 nm laser was used for Raman spectroscopy, and adenine (10−<sup>4</sup> M) was used as the model biomolecules analytes. The results indicated a characteristic peak of adenine at 733 cm−<sup>1</sup> , and the SERS intensity of PDMS-PMPC-Ag (W/) was ~4 times stronger than that of PDMS-PMPC-Ag (W/O), which indicated that multi-reflection by the 3D bionic structure enhanced SERS intensity. The pristine PDMS without Ag deposition and bionic structure (PDMS (W/O)) is displayed as a black line in the background. To enable the SERS detection (Figure 15), the replica PDMS-PMPC with a bionic structure was deposited on 5 nm of Ag nanofilm (PDMS-PMPC-Ag (W/)) by thermal evaporation, and pristine PDMS-PMPC with Ag (PDMS-PMPC-Ag (W/O)) deposition was used as the control. The 633 nm laser was used for Raman spectroscopy, and adenine (10−<sup>4</sup> M) was used as the model biomolecules analytes. The results indicated a characteristic peak of adenine at 733 cm−<sup>1</sup> , and the SERS intensity of PDMS-PMPC-Ag (W/) was ~4 times stronger than that of PDMS-PMPC-Ag (W/O), which indicated that multi-reflection by the 3D bionic structure enhanced SERS intensity. The pristine PDMS without Ag deposition and bionic structure (PDMS (W/O)) is displayed as a black line in the background. oration, and pristine PDMS-PMPC with Ag (PDMS-PMPC-Ag (W/O)) deposition was used as the control. The 633 nm laser was used for Raman spectroscopy, and adenine (10−<sup>4</sup> M) was used as the model biomolecules analytes. The results indicated a characteristic peak of adenine at 733 cm−<sup>1</sup> , and the SERS intensity of PDMS-PMPC-Ag (W/) was ~4 times stronger than that of PDMS-PMPC-Ag (W/O), which indicated that multi-reflection by the 3D bionic structure enhanced SERS intensity. The pristine PDMS without Ag deposition and bionic structure (PDMS (W/O)) is displayed as a black line in the background.

(W/O) bionic structure, and PDMS-PMPC-Ag with (W/) bionic structure. **Figure 15.** Raman spectrum of PDMS without (W/O) bionic structure, PDMS-PMPC-Ag without (W/O) bionic structure, and PDMS-PMPC-Ag with (W/) bionic structure. **Figure 15.** Raman spectrum of PDMS without (W/O) bionic structure, PDMS-PMPC-Ag without (W/O) bionic structure, and PDMS-PMPC-Ag with (W/) bionic structure.

#### **4. Conclusions 4. Conclusions**

In this study, we successfully fabricated a bionic Nepenthes peristome-like structure by photolithography with SU-8 photoresist. Furthermore, the bionic structure was replicated by a flexible PDMS substrate and then immobilized with MPC polymer brushes by an atmospheric plasma treatment to improve its hydrophilicity, antibacterial attachment, **4. Conclusions** In this study, we successfully fabricated a bionic Nepenthes peristome-like structure by photolithography with SU-8 photoresist. Furthermore, the bionic structure was repli-In this study, we successfully fabricated a bionic Nepenthes peristome-like structure by photolithography with SU-8 photoresist. Furthermore, the bionic structure was replicated by a flexible PDMS substrate and then immobilized with MPC polymer brushes by an

cated by a flexible PDMS substrate and then immobilized with MPC polymer brushes by

atmospheric plasma treatment to improve its hydrophilicity, antibacterial attachment, antiprotein adsorption, and biocompatibility. The self-driving capability of SERS detection and the one-way liquid transfer were demonstrated on the replica of the bionic PDMS-PMPC substrate. A benefit of the Nepenthes peristome-like structure with PMPC modification was that the delivery rate was about 2.5 mm/min, unlike the sticky stained water on the unmodified PMPC surface. This can provide the chromatographic capability to separate complex samples, and further detect them by Raman spectroscopy. The flexible and multifunctional SERS chips could potentially be applied in a wearable device for biomedical and environmental detection.

**Author Contributions:** Conceptualization, M.-C.Y., C.-C.L. and C.-C.C. (Cheng-Chen Chen); methodology, Y.-T.L., C.-H.W., W.-L.S., P.-C.H. and Z.-L.T.; software, W.-L.S. and P.-C.H.; validation, C.-C.C. (Cheng-Cheung Chen), C.-C.L. and T.-Y.L.; formal analysis, W.-L.S., P.-C.H. and Z.-L.T.; investigation, Y.-T.L. and C.-H.W.; resources, M.-C.Y. and C.-C.L.; data curation, Y.-T.L., C.-H.W. and Z.-L.T.; writing—original draft preparation, Y.-T.L., C.-H.W., W.-L.S., P.-C.H., C.-C.C. (Cheng-Chen Chen), C.- C.C. (Cheng-Cheung Chen) and T.-Y.L.; writing—review and editing, M.-C.Y., C.-C.C. (Cheng-Chen Chen), C.-C.C. (Cheng-Cheung Chen) and T.-Y.L.; visualization, W.-L.S. and P.-C.H.; supervision, C.- C.C. (Cheng-Chen Chen), C.-C.C. (Cheng-Cheung Chen) and T.-Y.L.; project administration, M.-C.Y. and T.-Y.L.; funding acquisition, C.-C.L. and C.-C.C. (Cheng-Cheung Chen) All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was financially supported by the Ministry of Science and Technology in Taiwan (MOST 109-2622-E-131-009, MOST 110-2221-E-131-009, MOST 111-2622-E-131-003, MOST 108-2623-E-016-003-D).

**Informed Consent Statement:** Not applicable.

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

## **References**


**Ming-Ze Gao <sup>1</sup> , Zhong-Yuan Li 2,\* and Wei-Feng Sun <sup>3</sup>**

	- Singapore 639798, Singapore; weifeng.sun@ntu.edu.sg

**Abstract:** To achieve a preferable compatibility between liquid silicone rubber (LSR) and cable main insulation in a cable accessory, we developed SiC/LSR nanocomposites with a significantly higher conductivity nonlinearity than pure LSR, whilst representing a notable improvement in space charge characteristics. Space charge distributions in polarization/depolarization processes and surface potentials of SiC/LSR composites are analyzed to elucidate the percolation conductance and charge trapping mechanisms accounting for nonlinear conductivity and space charge suppression. It is verified that SiC/LSR composites with SiC content higher than 10 wt% represent an evident nonlinearity of electric conductivity as a function of the electric field strength. Space charge accumulations can be inhibited by filling SiC nanoparticles into LSR, as illustrated in both dielectric polarization and depolarization processes. Energy level and density of shallow traps increase significantly with SiC content, which accounts for expediting carrier hopping transport and surface charge decay. Finite-element multiphysics simulations demonstrate that nonlinear conductivity acquired by 20 wt% SiC/LSR nanocomposite could efficiently homogenize an electric field distributed in high-voltage direct current (HVDC) cable joints. Nonlinear conductivities and space charge characteristics of SiC/LSR composites discussed in this paper suggest a feasible modification strategy to improve insulation performances of direct current (DC) cable accessories.

**Keywords:** cable joint; liquid silicone rubber; nonlinear conductivity; space charge

## **1. Introduction**

Recent electrical energy supply raises the requirements for delivering electricity over long distances in which traditional high-voltage alternating current (AC) transmission systems are facing inevitable challenges. Compared with conventional AC transmission, the major advantages of high-voltage direct current (HVDC) transmissions, such as long transmission distances, low losses, and large transmission capacities, need to competently provide long-distance transmissions [1–4].

HVDC cable accessories are essential connecting devices in power transmission and transformation systems. However, the multi-layer structures of composite insulation in cable accessories lead to great discrepancies and incompatibilities in the electrical conductivities of internal insulation materials. In particular, the conductivity of accessory insulation is far lower than that of the main insulation cross-linked polyethylene (XLPE), where the electric field is distributed inversely proportional to the material conductivity [5,6]. As a result, the poor accessory insulation results in local electric field distortions and space charge accumulations under long-term polarization electric fields, which degrades the insulation and breaks down the resistance of cable accessories. Therefore, cable accessories are the weakest insulation points in direct current (DC) cable lines.

**Citation:** Gao, M.-Z.; Li, Z.-Y.; Sun, W.-F. Nonlinear Conductivity and Space Charge Characteristics of SiC/Silicone Rubber Nanocomposites. *Polymers* **2022**, *14*, 2726. https:// doi.org/10.3390/polym14132726

Academic Editors: Ting-Yu Liu and Yu-Wei Cheng

Received: 5 June 2022 Accepted: 29 June 2022 Published: 3 July 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Recently, homogenizing the electric field by optimizing the geometry of the insulation structure and using nonlinear-conductivity composite materials has been found capable of alleviating local electric field concentration, improving the insulation performances of cable accessories [7,8]. With the development of large-capacity and high-voltage power equipment, nonlinear materials have shown significant advantages for improving practical efficiency and ensuring electrical compatibility in insulation equipment. The SiC/silicone rubber (SiR) composites with SiC content greater than 20 vol% exhibit evident conductivity nonlinearity which is attributed to carrier hopping conductance introduced by SiC fillers [9]. The dielectric properties of ZnO/SIR composites show significant conductivity nonlinearity when the volume fraction of spherical ZnO fillers is higher than 30 vol%, with relative permittivity being increased by five times [10]. It has been reported that calcium copper titanate (CaCu3Ti4O1) nanofibers/liquid silicone rubber composites present notable nonlinear conductivity and an improvement of space charge characteristics due to the introduced charge traps by CaCu3Ti4O<sup>1</sup> nanofillers [11]. SiC nanofillers with nonlinear conductivity can effectively suppress interface charge accumulations in SiC/polymer nanocomposites by introducing shallower traps [12]. In addition, coatings prepared by filling inorganic nonlinear nanoparticles (e.g., SiC and ZnO) into polymers exhibit self-adaptive conductivity in response to the applied electric field. The coating of SiC/Epoxy resin on insulator surfaces could promote the dissipation of surface charge and effectively improve the flashover voltage [13].

The previous studies concerning electric field homogenization using nonlinear additive fillers mainly focused on the influence of the type and content of fillers on the nonlinear conductivity of internal insulation, without elucidating clear pictures of nonlinear conductance in combination with space charge characteristics. Therefore, in this study, SiC nanoparticles filled into liquid silicone rubber (LSR) were used to render nanocomposites with appreciable nonlinear conductivities. The abnormally varying carrier transport and space charge suppression are analyzed to reveal the underlying mechanism of the comprehensive improvements in both nonlinear conductance and insulation strength, which provides a basic reference for the engineering application of nonlinear composites.

#### **2. Materials and Methods**

#### *2.1. Material Preparation*

Two-component (A and B) liquid silicone rubber (LSR, Wacker, Germany) was selected as the raw material for the composite matrix. SiC material (Deke Daojin, Beijing, China) with a particle size of 40 nm was adopted as the filling additive. SiC/LSR nanocomposites with SiC contents of 0, 5.0, 10.0, and 20.0 wt% were prepared by the mechanical stirring method. The preparation process was as follows: (1) same masses of A and B, and appropriate mass of SiC nanoparticles were accurately weighed according to filling content to be poured into an beaker and stirred continuously at room temperature; (2) the SiC/LSR mixture was poured into a 10 mm × 10 mm × 0.2 mold in a vacuum environment at a constant room temperature for 2 h; and (3) the mold containing the SiC/LSR mixture was placed into a flat vulcanizing instrument and vulcanized for 15 min under 15 MPa at 393 K, and then cured at 473 K for 4 h to finally obtain the material samples.

#### *2.2. Microstructure Characterization*

The insulation properties of nanodielectrics are closely related to the dispersion of the doped nano-phases in a polymer matrix. Thus, the dispersion of the SiC nanofillers in an EPDM polymer matrix was characterized by a scanning electron microscope (SEM) (SU8020, Hitachi High-Technologies Corporation, Tokyo, Japan). The specimens were cold-brittle ruptured in a vessel containing liquid nitrogen to acquire the cross section to be observed by SEM after being sprayed with a gold film.

#### *2.3. Mechanical Tensile Test*

According to the standard of ISO 37:2005, the stress–strain characteristics were measured with an elongation speed of 5 mm/min. The tested sample was made into dumbbell type sample 1 mm thick, with a total length of 75 mm and an effective test width of 4 ± 0.2 mm. The surface was smooth. without visible defects.

#### *2.4. Electric Conductivity*

Electric conductivities of the SiC/LSR nanocomposites in samples of 50 mm diameter and 300 µm thickness were measured with a three-electrode system consisting of a highvoltage DC power supply (HB-Z103-2AC, Tianjin Hengbo High Voltage Power Supply Co., Ltd., Tianjin, China), a picoammeter (Est122), three electrodes, and an oven at temperatures from 30 to 70 ◦C. The current values were predicted approach a quasi-steady sate after applying voltage for 30 min. Three aluminum electrodes were evaporated in a vacuum in which the measuring disc-shaped electrode with a 50 mm diameter was encircled by a protective annular-shaped electrode with a 75 mm diameter on one side of the film samples; a high-voltage circular electrode with a diameter of 78 mm was placed on the other side. Each group of samples was measured several times, and the average value was taken to ensure the accuracy and reliability of the tested results.

#### *2.5. Space Charge Distribution*

The pulsed electro-acoustic (PEA) method was used to measure the space charge characteristics of the SiC/LSR composites in samples with a dimension of 50 mm × 50 mm × 0.3 mm at room temperature in which the polarization electric field, the pulse voltage and width, and the input impedance were specified as 30 kV/mm, 400 V and 8 ns, and 1.368 Ω, respectively. After voltage was applied for 30 min to test the space charge accumulation in the polarization process, the tested sample was short-circuited to obtain depolarization decaying of the space charge distribution. The total space charge *Q*(*t*) was calculated as follows [14]:

$$Q(t) = \int\_{0}^{h} |\rho(\mathbf{x}, t)| \mathbf{S} \mathbf{d}x \tag{1}$$

where *Q*(*t*) denotes the total charge quantity internal material at time *t*, *h* signifies the sample thickness, *ρ*(*x*,*t*) symbolizes the space charge density at the position of *x* in the sample at *t*, and *S* is the electrode area.

Considering the small space charge injection under a low electric field, the waveform at a low electric field intensity of 3 kV/mm was adopted as the waveform reference for the space charge measurement system to restore the measured signal of a space charge distribution under a high electric field.

#### *2.6. Isothermal Surface Potential Decay*

The isothermal surface potential decay (ISPD) method can characterize the energy level distribution of charge traps and the charge transport in a detrapping process in which the corona-discharge method is used to charge the film sample, as shown by the schematic measurement system in Figure 1.

The distance between the needle tip and the grid was 5 mm, and the distance between the grid and the sample surface was 5 mm. The needle and grid electrodes were connected to a high-voltage source with an aluminum back electrode being grounded during measurement. The electrostatic probe was fixed on the sliding guide bracket about 3 mm from the sample surface. Under negative corona charging, the needle and grid electrodes were charged by −8 kV and −5 kV voltage, respectively, for 5 min. Then, the sample was transferred below an electrostatic probe to be measured for 25 min.

Double exponential function was used to numerically fit the decaying surface potential over time very well, as expressed by:

$$\mathbf{U} = \mathbf{A}e^{\frac{-t}{\tau\_1}} + \mathbf{B}e^{\frac{-t}{\tau\_2}} \tag{2}$$

where A, *τ*1, B, and *τ*<sup>2</sup> are fitting parameters by which the trap energy level distribution can be calculated by the surface potential decaying curve [14,15] which could be considered as the volume trap characteristics. The trap energy level and trap density are calculated by the following formulas:

$$E\_t = k\_\mathrm{b} T \ln(\nu t), \ N(E) = \frac{4\varepsilon\_0 \varepsilon\_1}{q k\_\mathrm{b} T d^2} \left| t \frac{\mathrm{d} \mathrm{d} \mathrm{I}}{\mathrm{d} t} \right| \tag{3}$$

where *E<sup>t</sup>* denotes trap energy level, *N*(*E*) indicate trap density, *d* represents sample thickness, *ε*<sup>0</sup> and *ε*<sup>1</sup> are vacuum permittivity and relative permittivity, respectively, *v* symbolizes the electron escape frequency, and *q* and *k*<sup>b</sup> signify the electron charge and Boltzmann constant, respectively.

**Figure 1.** Schematic test system of isothermal surface potential decay.

#### *2.7. Finite-Element Electric Field Simulation*

We investigates the electric field distributions in the 200 kV DC cable joint with the SiC/LSR composites used as reinforced insulation, as schematically shown by the geometrically modeled structure in Figure 2. The diameter of the cable core was 30 mm, and the thicknesses of the main insulation (XLPE material), inner shield, outer shield, and reinforced insulation were specified as 16, 2, 1, and 68 mm, respectively. The length of the simulation model was 2600 mm. To make the simulation more identical to the actual situation, an air environment domain was added to the simulation model. The convective heat transfer parameter at the interface of the cable joint and the air was set to 10 W/(m<sup>2</sup> ·K). DC high voltage was loaded to the edge of the cable core. The inner edge of the outer semi-conductive shield layer of insulation and the boundary of the stress–cone were set as the electric ground boundary. For thermal simulations, which were coupled to the electrical field based on the electrical properties of each conduction or insulation component, the core and ambient (outside outer protective layer) temperatures were set as 70 ◦C and 30 ◦C, respectively. The thermal–electrical coupling simulations were implemented by COMSOL Multiphysics according to reference [16], employing material parameters of electrical and thermal parameters for each constituent as listed in Table 1.

Numerical simulations with the finite-element method were performed to calculate the electric field distributed in the cable accessory using the modified electrical properties of reinforced insulation materials characterized by the prepared SiC/LSR nanocomposites, such as the nonlinear conductivity and the higher relative permittivity than cable main insulation materials as listed in Table 1. A free triangular cell was utilized for finite-element

meshing by the Delaunay triangulation algorithm, which was refined locally at the positions where the electric field strength varies significantly in the cable terminal. The maximum and minimum numbers of elements were adjusted until no obtuse angles in the triangulation meshing process appeared. The element growth rate was specified as 1.5, which means that the element size increased by about 50% from one element to another. The slack in narrow regions was set as 1.0 to prevent the triangular meshing from generating different sizes of elements.

**Figure 2.** Geometry model of cable joint: 1, inner shield; 2, outer shield; 3, XLPE; 4, connection fittings; 5-high-voltage shielding tube; 6, reinforced insulation; 7, outer protective layer; 8, stress cone; 9, conductor core.

**Table 1.** Electrical and thermal parameters of materials specified in the electric field simulations.


#### **3. Results and Discussion**

#### *3.1. Microstructure and Mechanical Property*

The microscopic morphologies of the SiC/LSR nanocomposites observed by SEM are shown in Figure 3 in which the highlighted areas indicate SiC nanofillers. For 5.0 wt% filling content, SiC nanofillers are evenly dispersed without any obvious agglomeration. With the increase of SiC content up to 20.0 wt% content, the highlighted parts inside the nanocomposites increase slightly in size, while retaining a high dispersivity of homogeneous spatial distributions, as shown in Figure 3. The SEM images verify that the SiC/LSR nanocomposites were successfully prepared as expected to maintain a high dispersion of SiC nanofillers even for the filling content approaching 20.0 wt%.

The tensile modulus and broken-elongation of the SiC/LSR nanocomposites, as illustrated by the stress–strain characteristics in Figure 4, increase and decrease, respectively, with the increase of SiC content. When SiC content approached 20.0 wt%, the tensile strength and broken-elongation of SiC/LSR nanocomposites were maintained at 6 MPa and 350%, respectively, which meet the mechanical requirement for cable accessories. However, due to the fact that the mechanical performances of cable accessories are greatly influenced by operating conditions and cable manufacturing technology, the applications of the SiC/LSR nanocomposites need to be further investigated for improving mechanical properties.

#### *3.2. Electrical Conductance*

Electrical conductivity was normalized as the dependence of the current density on the electric field (*J*-*E* variation curves) to investigate the charge transport characteristics of the SiC/LSR nanocomposites, as shown by the *J*-*E* curves of the SiC/LSR composites in Figure 5. The critical points arise around electric field strength of 10<sup>7</sup> V/m, as a threshold *E*th at which point the charge transport mechanism changes from Ohmic conductance to space charge limited conductance (SCLC). Nonlinear coefficients *β*<sup>1</sup> and *β*<sup>2</sup> distinguished by *E*th are obtained by the linear fitting for the logarithm *J*-*E* curves, as listed in Table 2.

**Figure 4.** Stress–strain characteristics of SiC/LSR nanocomposites.

**Figure 5.** Electric conductance characteristics in the logarithm *J*-*E* curves of pure LSR and SiC/LSR nanocomposites at diverse temperatures of: (**a**) 30 ◦C; and (**b**) 70 ◦C; hopping conductance characteristics of fitting conductivities for the SiC/LSR nanocomposites at (**c**) 30 ◦C; and (**d**) 70 ◦C.


**Table 2.** Nonlinear coefficients of the SiC/LSR nanocomposites.

When *E* < *E*th, the nonlinear coefficient *β*<sup>1</sup> < 2 complies with the Ohmic conductance mechanism that the charge transport is produced by impurity ionization in the SiC/LSR composites. In contrast, when *E* > *E*th, the nonlinear coefficient *β*<sup>2</sup> > 2 which means the carrier concentration and carrier mobility increase drastically. The process of carrier increment in polymers is a thermal excitation process. According to hopping conductance theory, most of the charges involved in conducting current are in local states, and the charge transfer between localized states is the main process. The process of charge passing from one localized state to another is graphically described as hopping conductance, the current density formula of the jump conductance model can be described as follows:

$$J\_n = 2ndve \exp\left(-\frac{\chi}{k\_b T}\right) \sinh\left(\frac{eIE}{2k\_b T}\right) \tag{4}$$

where *n* represents carrier concentration, *χ* denotes activation energy, *l* symbolizes hopping distance of the carrier, and *T* is thermodynamic temperature. To judge whether the conduction mechanism under a high electric field is dominated by hopping transport, the *J*-*E* curves of Figure 5a,b are fitted by Equation (4) according to the hopping conductance model, as shown in Figure 5c,d. The hopping distance of the SiC/LSR composites can be seen in Table 3. The fitting curves are highly consistent with the tested conductivities of the SiC/LSR nanocomposites. Based on the fitting curves in the hopping conductance model, the hopping distance of 5.0 wt%, 10.0 wt%, and 20.0 wt% for the SiC/LSR nanocomposites are calculated by Equation (4), as shown in Table 3. The hopping distance becomes evidently larger with the increase in SiC content, implying that SiC nanofillers facilitate the charge detrapping and thus expedite hopping transports to acquire or increase conductivity nonlinearity. Therefore, the higher the temperature, the higher the density of trap that can be disentangled, and the shorter the hopping distance. Moreover, according to the classical percolation theory [17], when the content of SiC nanoparticles in the LSR matrix exceeds the percolation threshold, the SiC nanofillers are close enough to form a random conductive network, resulting in the macroscopic performance of nonlinear conductivity of the SiC/LSR nanocomposites.

**Table 3.** Hopping distance of percolation conductance in the SiC/LSR composites.


#### *3.3. Space Charge Characteristics*

In high-voltage DC cable systems, the XLPE cable and its cable accessories bear a high electric field for long periods in which the insulation layer suffers space charge accumulation under electric field distortion at the interface of the multilayered composite insulation structure, accelerating dielectric aging and finally causing insulation failures. For cable accessories, the maximum electric field intensity resides at the stress-cone root of reinforced insulation. This is the part that requires high resistance to space charge accumulations. Space charge characteristics of pure LSR and SiC/LSR nanocomposites in polarization and depolarization processes are shown in Figure 6. Space charge accumulation appears inside the pure LSR sample in the polarization process, which can be significantly inhibited by filling SiC nanoparticles and becomes more obvious for higher SiC content, as comparatively illustrated by the left panels of Figure 6a. Due to the lower mass of negative carriers compared to positive carriers, it is more likely that the negative charge can approach the anode without neutralization in the electrical migration process, accounting for the evident heterocharge accumulations arising near the anode which can be inhibited by filling SiC particles and alleviated with the increase of SiC content.

Homocharge and heterocharge as two classes of space charges are accumulated by trapping the electrode-injected carriers and the impurity ionized carriers that form the applied electrode and the dielectric material interior, respectively [18–20]. Therefore, in the depolarization process, heterocharges decay more quickly than homocharges under a short-circuit as shown in the right panels of Figure 6. There, the space charge decaying near the anode is rather faster than that of the homocharges near the cathode. Meanwhile, the charge density of pure LSR at the initial stage of short-circuit depolarization is much higher than that of the SiC/LSR nanocomposites in which charge density has been further decreased by the increasing SiC content.

**Figure 6.** Space charge distribution in: (**a**) pure LSR; (**b**) 5 wt%; (**c**) 10 wt%; and (**d**) 20 wt% SiC contents in the polarization process of applying voltage (left panels) and in the depolarization process under a short-circuit (right panels).

Mean space charge densities in pure LSR and SiC/LSR nanocomposites during a depolarization process are calculated by Equation (1), with the results being shown in Figure 7. The space charges decaying with an almost constant rate in the initial depolarization stage of 0~300 s derives mainly from the detrapping of the charges captured in shallow traps, while the significantly lower decaying rate of the space charge density after depolarization for 300 s implies that the space charge decaying is derived from detrapping the charges captured in deeper traps. In particular, it is evident that the space charges density of the SiC/LSR nanocomposites decreases with the increase in SiC content. In comparison with pure LSR, SiC/LSR nanocomposites acquire a higher resistance to space charge accumulation, while persisting at a higher depolarization rate due to the higher electrical conductance derived from carrier hopping transport through the conductive channels formed between SiC nanofillers.

**Figure 7.** Space charge density of pure LSR and SiC/LSR nanocomposites during depolarization.

#### *3.4. Charge Trap Characteristics*

Surface charge dissipation is generally fulfilled by transporting along the material surface or from interior material toward electrode neutralization with charged particles in gas material. As indicated by Figure 8a, the SiC/LSR nanocomposites have achieved a notably higher rate of surface charge dissipation than pure LSR. For the electrode structure and the sample size used for our study, the strength of the normal electric field between the sample surface and the ground electrode is much higher than that of tangential electric fields along the sample and electrode surfaces in the process of surface charge dissipation. This process is dominated by the surface charge transports through the trapping and detrapping processes in the sample interior (volume conduction) to the ground electrode. The potential decay rate of the material surface increases significantly with the increase in SiC content due to the hopping transport of the charges percolated from the traps introduced by the SiC nanofillers. This is manifested by the higher threshold electric field of the SiC/LSR nanocomposites compared with pure LSR at which the electric conduction alters into the highly nonlinear region. Space charge characteristics or surface charge dissipation of the SiC/LSR nanocomposites is dominated by carrier trapping and detrapping processes. Meanwhile, the energy level and spatial density of trap distributions determine the activation energy (carrier concentration) and hopping distance (carrier mobility) of percolation conductance in the nonlinear region. According to ISPD curves, we calculate trap level distributions of pure LSR and the SiC/LSR nanocomposites by Equations (2) and (3), as shown by the results in Figure 8b.

**Figure 8.** (**a**) Isothermal surface potential decay (ISPD) curves; and (**b**) trap level distributions of SiC/LSR nanocomposites.

At about 0.86 and 0.93 eV of trapping depth, two peaks of trap density appear in the energetic spectra for the SiC/LSR nanocomposites, which could be referenced as the shallow and deep trap levels, respectively. It is noted that the SiC/LSR nanocomposites present distinctly lower and considerably higher densities of deep and shallow traps, respectively, compared with pure LSR, which is more obvious for higher SiC content. The multi-core model [21,22], which is competent to characterize the interface region between the nanofillers and the polymer matrix in nanodielectrics, is composed of a bonding layer, a binding layer, and a loose layer. In the bonding and binding layers, deep traps are derived from the imperfect chemical bonding in the molecular structure, while the amorphous structure in the loose layer mainly contributes to shallow traps. When the filling content is raised to a sufficient level, generally higher than 5.0 wt%, the interface areas as described by the loose layer between the SiC nanofillers overlap to form conductive channels for hopping transports of the charge carriers detrapped from shallow traps, as described by the percolation effect. The significant increase of shallow trap density by filling SiC nanoparticles results in the substantial percolation conductance at room temperature, which will be aggravated by increasing temperature or SiC content, accounting for the extraordinary conductivity nonlinearity given by the SiC/LSR nanocomposites.

#### *3.5. Electric Field Simulation*

For DC cable in operation, the temperature in the metallic cable core would be raised by Joule heating, leading to a specific temperature gradient from the cable core to the external insulation. The conductivity of polymer insulation materials, which determines the electric field distribution of the DC cable joint, depends greatly on operation temperature and heat production. For pure LSR or SiC/LSR composites as the reinforced insulation material, the electric field distributions in a 200 kV DC cable joint are simulated under a thermal–electrical fully coupling condition, as shown in Figure 9.

Electric field distribution is inversely proportional to conductivity under DC voltage, and the conductivity of pure LSR is much lower than that of XLPE. Thus, high electric field is distributed in LSR reinforced insulation, as shown in Figure 9a. Maximum electric field strength appearing at the stress-cone root approach 43.40 kV/mm when pure LSR is used for reinforced insulation, which will be effectively reduced by using SiC/LSR nanocomposites as reinforced insulation material, as shown in Figure 9b,c. The position of maximum electric field strength can be ameliorated by locating it in the main XLPE insulation when using the SiC/LSR nanocomposite with a high SiC content approaching 20 wt%, as shown in Figure 9d.

**Figure 9.** Steady-state electric field distributions in a cable joint with reinforce insulation of: (**a**) pure LSR and SiC/LSR nanocomposites of (**b**) 5 wt%; (**c**) 10 wt%; and (**d**) 20 wt% SiC contents.

#### **4. Conclusions**

The effect of SiC nano fillers on the conductivity and the charge trap characteristics of LSR composites was investigated in this study. Conductivity nonlinearity and threshold electric field strength of SiC/LSR nanocomposites increases and decreases, respectively, with the increase of SiC content, which will be intensified at a higher temperature. The highest density peak of trap level distribution shifts from deeper traps to shallow traps by introducing SiC nanofillers into the LSR matrix, which accounts for hopping conductance and impeding space charge injection. Based on the established cable joint model of finiteelement multiphysics simulations, it is concluded that the highest electric field strength at the stress-cone root can be effectively reduced by using SiC/LSR nanocomposites as reinforced insulation, compared with using pure LSR material, which can even be intensified by raising the SiC content. As a consequence, the approximate SiC particles doped LSR can suppress the accumulation of space charge and improve conductivity nonlinearity, which provides a possible method for the improvement of cable accessory performance.

**Author Contributions:** Conceptualization, W.-F.S.; data curation and formal analysis, M.-Z.G.; writing—original draft preparation, M.-Z.G.; writing—review and editing, Z.-Y.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by Research Start-up Fund of Mudanjiang Medical University (Grant No.: 2021-MYBSKY-057).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Theoretical methods and results are available from the authors.

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

## **References**

