Multilayer Three‐Dimensional Woven Silver Nanowire Networks for Absorption‐Dominated Electromagnetic Interference Shielding

Secondary electromagnetic reflection is a great challenge in the field of high conductivity electromagnetic interference (EMI) material due to the poor electromagnetic absorption ability. In this work, a multilayer stack strategy compositing of the absorption‐dominated thin films and the reflection‐dominated thick film is proposed to realize both high shielding effectiveness (SE) and low reflection. The film comprises 3D woven silver nanowire (AgNWs) networks embedded in the polydimethylsiloxane (PDMS) substrate, and the shielding film thickness can be adjusted via controlling coating cycles. It can be found the three‐ply stack of two thin films and a thick film, (B:F)10(B:F)10(B:F)40, (B:F)5(B:F)10(B:F)40, and (B:F)5(B:F)5(B:F)40, can achieve EMI SEs up to 40.8, 40.7, and 36.3 dB, and the absorptions up to 0.59, 0.65, and 0.69, respectively. Results from verification investigation and theoretical simulation indicate that the multilayer can regulate the shielding mechanism of AgNWs‐based films from reflection‐dominated to absorption‐dominated, with no apparent deterioration of EMI SE. The synergistic effect of a multilayer stack is regarded to originate from a special “absorb‐absorb‐reflect‐absorb‐absorb” course. This finding suggests a promising approach to acquiring both high‐efficient and absorption‐dominated EMI shielding material.


Introduction
With the rapid development of wireless electronic devices, electromagnetic interference (EMI) pollution has become a www.advmatinterfaces.de a great challenge to develop shielding materials with both superior EMI SE and absorption. During past decades, more and more researchers have attempted to resolve the contradiction through optimized structural designs. [21][22][23] For instance, Qi et al. prepared a PVDF/GNP-Ni-CNT composite with a thickness of 2.4 mm which shows an EMI SE of 46 dB and a low absorption of 0.15. [24] Sun et al. combined a conductive silver layer with TPU/CNT composite to create an asymmetric layered structure with a thickness of 3 mm showing an EMI SE of 79.4 dB and an absorption coefficient of 0.46. [25] Using asymmetric shielding construction, EM is absorbed by the non-metallic layer and reflected by the metallic layer successively, which forms the "absorb-reflect-reabsorb" EMI attenuation mechanism. However, the absorption coefficient is still lower than 0.5 and the contribution of reflection exceeds 50%, indicating that reflection is still the dominant shielding mechanism.
Except for the structural design, thickness is considered to be another main factor affecting EMI shielding performance. [26][27][28][29][30][31] Gao et al. prepared a 3D thermal annealed graphene aerogel/AgNWs conductive networks by vacuumassisted injection. It is found that EMI SE increases from 14 to 70 dB, while the absorption coefficient decreases from 0.6 to 0.15 when the density of AgNWs increases from 22.5 to 75 mg mL −1 . [32] Yang et al. prepared UV-curable shielding layer overcoated AgNWs. The EMI SE increases from 6.5 to 30 dB and the absorption coefficient decreases from 0.75 to 0.3 when the density of AgNWs increases from 0.6 to 2.4 mg mL −1 . [33] However, these results show that reflection is still the dominant shielding mechanism in the above samples. It is worth noting that the EMI shielding type develops from absorption type to reflection type with increasing thickness.
In this work, we propose the multilayer strategy and an "absorbabsorb-reflect-absorb-absorb" EM attenuation mechanism.
A key note is that only an effective shielding material of AgNWs is utilized in our design. The multilayer consisted of the absorption-dominated thin shielding films and the reflectiondominated thick films. The two-ply stack composited of a thin film and a thick film reaches an EMI SE of more than 35 dB and an absorption coefficient exceeding 0.4, which is much higher than that of the reference thick film. Furthermore, the three-ply stack of two thin films and a thick film achieves an EMI SE of 40 dB and an absorption coefficient of 0.6, suggesting an absorption-dominated shielding mechanism. In addition, a comprehensive study of the attenuation mechanism, the influence of film thickness, and the ply laminated sequence have been performed. The multilayer strategy has shown great potential for absorption-dominated EMI shielding applications.

Result and Discussion
As well known, secondary electromagnetic reflection is a challenge in reflection-type EMI materials. Here, we report a 3D woven metal nanowire network for EMI film. The detailed structure of the film sample is depicted in Figure 1a. The film is composed of two parts, a PDMS film with two different surface structures including the plate side (back side) and the square pyramid array (face side), and a 3D woven metal nanowire network. The networks are buried into the square pyramid side or groove among the square pyramids (as shown in insert i of Figure 1a). The 3D woven metal nanowires network is prepared by using a solution-assistant self-assembly method, and the details can be found in the Experimental section. The thickness of the network can be adjusted by controlling the coating cycle of AgNWs ink. The employment of PDMS is aimed at enhancing its flexibility. The dimension of the testing sample is 23 (length) × 23 (width) × 0.5 (thickness) mm 3 unless mentioned otherwise. In addition, the film in insert i of Figure 1a can be referred to as (B:F) x or (F:B) x , where B represents the back side of the film, F represents the face side, and subscript x represents the coating cycles of the conductive network. Besides, the first letter of B or F implies the EM wave incidence direction as well. To find a strategy for enhancing the absorption of reflective type EMI film, the EMI shielding and absorption performances of three kinds of groups (a thin film (low coating cycles sample), two thin films stack, and two thin films adding a thick film as shown in Figure 1b) have been investigated by using simulating method, the as-obtained results are shown in Figure 1c. The simulation model and boundary conditions have been summarized in the supplementary material. It is found that the thin film shows relatively poor EMI performance and high absorption, while the thick film shows higher EMI performance but poor absorption. A two-ply stack with two thin films shows enhanced EMI SE without significantly deteriorating its absorption. Meanwhile, a three-ply stack composited of two thin films and a thick film further improves the overall EMI SE and the absorption is similar to that of a two-ply stack. Based on the above results, here we propose a multilayer strategy for enhancing the EMI and absorption ability in metallic shielding film, simultaneously.
Generally, increasing the conductivity of the network can enhance the EMI SE and weaken absorption. The nanowire density increases with the increase of the coating cycle, and the conductivity depends on the nanowire density. It can be seen from Figure 2a that the resistance decreases sharply from 96 to 3.6 Ω □ −1 with increasing coating cycles of AgNWs from 10 to 50, indicating that the resistance decreases with the increase of the coating cycle. Here, the relationship between the coating cycle and EMI has been investigated. Figure 2b shows the www.advmatinterfaces.de internal structure of the networks with coating 50 cycles on the silicon mold. Most of the AgNWs are oriented along the edge of the silicon mold and physically interconnected with each other, which leads to the network being structurally and electrically robust. Due to the asymmetric feature of the 3D network, the EMI performance of the two propagation paths is probably different from each other. The different propagation paths have been summarized in Table S2, Supporting Information. In order to investigate the EMI performance of the two propagation paths, EMI SE and absorption are measured and the obtained results are shown in Figure S1, Supporting Information. The vector network analyzer (E5071C, Keysight Technology) is used to measure the S 11 and S 21 which can be used to calculate the EMI SE of the sample by the following equations: [35] where R, T, and A are the reflection, transmission, and absorption coefficients, respectively. SE Total represents the total EMI SE. It can be seen from Figure S1a, Supporting Information that the EMI SE of the face side to the back side (i.e., F-B) is the same as that of the back side to the face side (i.e., B-F). It is worth noting that the absorption of the B-F is better than that of the F-B, which suggests the asymmetric feature of the absorption performance. Based on the above results, the EMI SE and absorption of the samples with different coating cycles (B-F) have been investigated systematically, from the aspects of the experiment and simulation. Figure 2c shows the simulation EMI SE of the samples with different coating cycles, which indicates that the EMI SE is closely related to the network thickness. The increased EMI SE is mainly attributed to the increased conductivity of the sample at higher AgNWs cycles. The thickness and continuity of the conductive network influence the EMI SE. To reveal the EMI performance, the data of the absorption, reflection, and EMI SE performance at 10 GHz have been summarized in Figure 2d. The red, blue, and orange curves represent the absorption, reflection, and EMI SE, respectively, which indicates that the absorption decreases slightly, and the reflection and EMI SE increase with increasing coating cycle. Moreover, the absorption coefficient increases to 0.35 when coating cycles are lower than 10. Such phenomenon indicates that the EMI performance of the sample gradually develops from absorption to reflection type with increasing coating cycle.
To further elucidate the thickness-absorption relationship, the samples with different coating cycles have been prepared and tested, as shown in Figure 2e. A significant increase of EMI SE can be found from 17 to 37 dB at 10 GHz when coating cycles increase from 10 to 50. A detailed analysis of the contribution of absorption and reflection to the EMI SE at 10 GHz as a function of network thickness in Figure 2f indicates that the reflection coefficient decreases from 0.7 to 0.55 and the absorption coefficient increases from 0.3 to 0.45 when coating cycles decrease from 50 to 10. It is clear that absorption increases with thickness decrease while reflection decreases. The total EMI SEs for different samples suggest that the reflection might only happen at the incident surface of the AgNWs network and dominates in compact AgNWs network samples, while the absorption prefers to happen in loose AgNWs network samples. However, the sample with the loose network (coating 10 cycles) exhibits poor total EMI SE. To fulfill the demand for practical application, the multilayer strategy is proposed to enhance the EMI shielding property and absorption simultaneously. Considering the asymmetrical feature of this film, the EMI performance and absorption of the multilayer probably depend on the propagation path. Obviously, the propagation path depends on the arrangement sequence. There are three arrangement sequence probabilities in the two-ply stack, that is,  Table S2.
The section of the two-ply stack involves two aspects, the effect of the arrangement sequences (two single-layers with the same coating cycle, and different coating cycles), and the coating cycle on shielding properties. The influence of the arrangement sequence of the stack composed of two singlelayers with coating 10 cycles is shown in Figure 3. It can be seen from Figure 3a that the simulation results of the EMI SE performance decrease in order from (F:B) 10 (B:F) 10 , (B:F) 10 (B:F) 10 , and (B:F) 10 (F:B) 10 to the single-layer with equivalent coating 20 cycles. The experiment results also show a similar tendency to the simulation results, except the EMI SE performance of the single-layer with equivalent coating 20 cycles is higher than that of all two-ply stacks, seeing Figure 3b,c. Generally, with an increasing coating cycle, the bonding force between the AgNWs network and silicon mold gradually drops, which can prevent the conductive network from being destroyed during transferring to PDMS. [35] In other words, the damage degree of the film with large coating cycles is lower than that of the film with low coating cycles. Therefore, the EMI SE performance of the single-layer with equivalent coating 20 cycles is higher than that of all two-ply stacks, which can be attributed to the damage degree of the conductive network. In order to classify the universality of the EMI SE performance tendency with the arrangement sequence, the EMI SE performance of the two-ply stack composed of two single-layers with different coating cycles is summarized in Figure S2, Supporting Information, exhibiting a similar tendency. When the coating cycle is more than 10 in single-layer, the EMI SE performance of the arrangement sequence of the (B:F) x (B:F) x and (F:B) x (B:F) x are better than that of the (B:F) x (F:B) x , and the single-layer with equivalent coating 2x cycle (i.e., (B:F) 2x and (F:B) 2x ). Moreover, it should be noted that the total EMI SE of the two-ply stack (single-layer with coating 10 cycles) is higher than 20 dB, which exceeds the applicable standard of commercial EMI materials. Figure 3d shows the simulation results of the absorption of the arrangement sequence of the (B:F) 10  . As a consequence, the two-ply stack of (B:F) 10 (B:F) 10 shows a comprehensive performance with sufficient EMI SE and absorption compared with other samples, which indicates that the arrangement order of (B:F) x (B:F) x maybe suit for preparing high absorption and EMI SE performance film.
Based on the results of the arrangement sequence, here the effect of the coating cycles on the EMI SE performance and absorption of the two-ply stacks is investigated from both theory and experiment. Figure 4a shows the simulation results of the two-ply stacks with different single-layer coating cycles, suggesting that the EMI SE increases with increasing the coating cycle. The absorption of two-ply stacks is summarized in Figure 4b, which indicates that the absorption decreases slightly with the increasing coating cycle. The absorption coefficient is close to 0.3 when the coating cycle is lower than 10 for each layer. As shown in Figure 4c, the dominant shielding mechanism of the two-ply stacks tends to develop from absorption into reflection with increasing the single-layer coating cycle. The shielding performances of two-ply stacks with different coating cycles have also been experimentally investigated, as shown in Figure 4d,e. It is noted that the EMI SE of the experimental samples is lower than those of the simulation results. Oppositely, the absorptions are higher than those of simulation results. It can be seen from Figure 4f that the absorption of two-ply stacks decreases from 0.59 to 0.3, while its EMI SE increases from 20 to 47 dB when the single-layer coating cycles increase from 10 to 50. In addition, Figure 3c,f illustrate that the EMI SE and absorption of the single-layer with equivalent coating 20 cycles ((B:F) 20 ) are 26 dB and 0.3, respectively. These results indicate that the two-ply stack can enhance the absorption and maintain the EMI SE performance at a level of application standard, leading to a high ability of secondary electromagnetic pollution abatement.
The effects of the two-ply stack (B:F) x1 (B:F) x2 on the EMI SE performance and absorption are investigated (x1≠x2 represents the two single-layers in the stack are different coating cycles). Figure 5a shows the EMI SEs of (B:F) 10 (B:F) 10 , (B:F) 10 (B:F) 40 , and (B:F) 10 . The blue and orange curves represent the EMI SEs of (B:F) 10 and (B:F) 10 (B:F) 10 , respectively, which indicates that the stack of two thin single layers is helpful to enhance the EMI SE exceeding 20 dB. Moreover, a stack with a thin layer and a thick layer can obtain a higher EMI SE than that of the thick single layer, that is, (B:F) 40 . Figure 5b shows the absorptions of (B:F) 10 (B:F) 10 , (B:F) 10 (B:F) 40 , and (B:F) 10 . An interesting phenomenon should be noted the absorption of the stack with a thick layer and a thin layer has not decreased to a level lower than that of a thick layer. Therefore, the above results indicate that the EMI SE performance of the thick layer and the absorption of the thin layer will be complemental in superiority for both of them.
To further illustrate the synergistic effect of asymmetry stack architecture, the effect of the three-ply stack on shielding properties is investigated. Figure 6a shows the simulation results of the EMI SEs and the absorption of the three-ply stacks, (B:F) 10 40 , which shows that the EMI SE increases and the absorption decrease with the increasing of the equivalent coating cycles of the three-ply stack. Figure 6b shows the corresponding experiment results, which show a similar tendency to simulation results. In addition, the experiment results show that the threeply stacks can further enhance absorption. It can be seen that almost unchanged from 40.7 to 40.8 dB, respectively. It indicates that the stack of two thin layers and a thick layer can acquire both high absorption and high EMI SE, and absorption turns out to be the dominant shielding mechanism. Compared with traditional iron or carbon-based absorption materials, the multilayer's advantages include the number of shielding materials, thickness, and absorption coefficient. The corresponding results are shown in Table S3, Supporting Information. To confirm the effects existing in other frequency ranges, here we show the experiment results of the EMI SE and the absorption coefficient in the frequency range of 12-18 GHz, which shows

www.advmatinterfaces.de
a similar tendency to that in the frequency range of 8-12 GHz. In addition, the EMI SEs are higher than those in the frequency range of 8-12 GHz.
The excellent absorption of the as-prepared sample is attributed to two aspects, first is the contribution of the electronic transport and dipole polarization in the single conductive layer. Second is the stack's synergistic effects, composed of two thin films and a thick film. As the above depiction, the conductive network's EMI performance has transferred from absorption to reflection with the coating cycle increasing, which can be attributed to the enhancement of the conductivity and the electronic transport ability of the network. [36] The enhanced conductivity and electronic transport ability result in serious impedance mismatching in the interface between the network and free space. On the other hand, there are many wire-to-wire contact points in the network. Some electrons tend to accumulate in the area near the contact points when EM waves pass through the network, which leads to an integral dielectric polarization

www.advmatinterfaces.de
and generates polarization loss. [13,37] Therefore, the absorption mechanism of the network is involved in the electric loss and dielectric polarization loss. Figure 6e depicts the mechanism of the excellent absorption of the stack composed of two thin films and a thick film. When EM waves incident from air to the thin film I, a small amount of EM waves are reflected, whereas most of them enter into the thin film I due to a good impedance match between the thin film and free space. The entered EM waves are divided into two parts, absorption, and transmission. The part of absorption is attributed to the electrical loss and dielectric polarization loss of the thin film I. When the part of the transmission moves to thin film II, it will encounter the same attenuation process as that of thin film I. It is worth noting that the reflection part will experience round-trip reflections between thin films I and II, which results in multiple absorptions on the two thin films. Obviously, the EM wave would be significantly attenuated as EM waves pass through two thin films. When the part of the transmission from thin film II move to thick film III, EM waves would be reflected by the thick film III due to serious impedance mismatch, and only less than 0.05% of EM waves could pass through thick film III. The most reflected EM waves are absorbed by the former two thin films again. Finally, only less than 34% of EM waves are reflected to the original free space. The multilayer design of the two high-absorption thin films and a high-reflection thick stack realizes a special "absorb-absorb-reflect-absorb-absorb" process. As a result, 66% of the EM wave can be absorbed, demonstrating an auspicious absorption-dominated EMI shielding performance with high EMI SE.

Conclusion
In summary, the multilayer strategy is proposed to realize both high EMI SE and high absorption. Its EMI SE performance and absorption have been predicted by using a simulation method. The simulation results demonstrate that high EMI SE performance and excellent absorption ability can be achieved simultaneously by employing the multilayer, which is composed of two thin films and a thick film. Based on the above results, a series of samples have been prepared by using a solution-assistant self-assembly method, and their EMI SEs and absorptions have been measured by VNA. The results show that the EMI performance of the single layer can be adjusted from absorption type to reflection type by controlling the coating cycle. The two-ply stack of a thin layer and a thick layer referred to as (B:F) 10 (B:F) 40 , can obtain an EMI SE of more than 35 dB and an absorption coefficient of more than 0.4 in superiority for that of the single-layer of (B:F) 40 . Furthermore, the three-ply stack of two thin films and a thick film, referred as (B:F) 10  In addition, the EMI SE and absorption of the multilayer in the frequency range of 12-18 GHz are better than those in the frequency range of 8-12 GHz. In addition, the synergistic effect of two thin films and a thick film stack results in a special "absorb-absorb-reflectabsorb-absorb" course. Therefore, the multilayer strategy has great potential for reducing secondary pollution.

Experimental Section
Fabrication of AgNWs 3D Networks: A double oxidized (001) silicon wafer with 1 µm oxide layer was washed using the ammonia and hydrogen solution and then was washed using DI water to remove impurities. Secondly, the positive photoresist (RZJ-304, Suzhou, China) with a thickness of 1.5 µm was patterned on the top side of the silicon wafer. After tricky baking, the naked SiO 2 on the substrate was etched in an etching solution (BOE ∼ 6:1). Then, the photoresist mask was washed using the SYS9070 solution purchased from Shanghai Xinyang. Afterward, a hot potassium hydroxide solution (concentration ≈10 wt%, temperature ≈70 °C) was used to etch the as-etched silicon wafer, and the inverted square pyramid cavity arrays are formed on the silicon mold.
Then, an aluminum film with a thickness of ≈30 nm was deposited on the silicon mold as a sacrificial layer to decrease the surface bonding force between PDMS and the silicon mold. A commercial AgNWs ink (nanowire length 50-150 µm, diameter 70-105 nm, concentration ≈1 mg/mL −1 , alcohol as dispersant, New Material Technology Company, Zhejiang, China) was dripped on the silicon mold forming a uniform ink film. After the solvent drying, repeat the coating process to obtain the sample with different thicknesses. To transfer networks from the silicon template to the PDMS substrate, PDMS solution (model: SYLGARD 184, Dow Corning Company, Midland, USA., the weight ratio between PDMS prepolymer and curing agent ≈10:1) was spin-coated on the template and heated at the temperature of 105 °C for 21 min. The as-prepared sample was peeled off from the silicon template.
In this work, AgNWs are dispersed uniformly in the alcohol, and their concentration is 1 mg mL −1 . The volume of the AgNWs ink per coating one cycle on the silicon mold is about 0.1 mL. Therefore, the dosage of AgNWs is ≈0.1 mg per coating one cycle. The maximum dosage of AgNWs in the multilayer is about 6 mg ((B:F) 10 (B:F) 10 (B:F) 40 ).
Characterization: The morphologies of the AgNWs 3D networks were observed by using a field emission scanning electron microscope (SEM; S4700, Hitachi). The square resistance was measured by using a four-point probe method (FT-340, RICO Company, Ningbo, China). The square resistance of each film should be measured at 10 different locations, thus at least 10 values of the square resistance sample were calculated.
Simulation of the EMI SE effect: Frequency domain simulation was employed with Computer Simulation Technology Microwave Studio software (CST Studio Suite) to calculate S 11 and S 21 . The EMI SE and absorption can be calculated by Equations (5) and (6)

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.