Sustainable electromagnetic interference shielding materials from cellulose-grafted n-type polymers

Abstract

Electronic waste (e-waste) from ever-shorter device lifecycles is fueling a search for sustainable alternatives to conventional electromagnetic interference (EMI) shielding materials, which are typically non-degradable and hard to recycle. Addressing this challenge, here we present a durable, high-performance EMI shielding film that is both recyclable and biodegradable. Thanks to a simple fabrication process, our introduced composite blends renewable cellulose nanofibers, a conductive n-type polymer, and a small ionic liquid additive into a robust film. This all-organic film achieves tunable shielding effectiveness between 29.77 to 83.77 dB, comparable to traditional metal or carbon-based shields. The film demonstrates good recyclability, retaining 97.72% of its initial EMI shielding performance after ten reprocessing cycles. Furthermore, it undergoes complete biodegradation in soil at the end of its lifecycle, leaving no persistent waste, thereby offering a sustainable solution for EMI shielding applications. Combining strong EMI shielding with end-of-life degradability and reusability, this work offers a sustainable pathway to electronics that reduce e-waste and promote a circular economy.

Introduction

The rapid advancement of 5G/6G communication infrastructure, Internet of Things applications, and cutting-edge precision technologies underscores the pressing need for highly efficient electromagnetic interference (EMI) shielding films. Such materials are essential for preventing device malfunctions and signal disruption, as well as for mitigating potential health risks associated with electromagnetic (EM) pollution^1,2,3. Driven by growing demand across the consumer electronics, communications, and defense sectors, the global market for EMI shielding materials is projected to reach a value of USD 267 million by 2033, with a compound annual growth rate of 4.6% anticipated between 2025 and 2033⁴. Current reports on EMI shielding materials predominantly focus on metal-based composites, carbon-based materials, metal oxides, polymer-based systems, two-dimensional (2D) materials, and their hybrids^5,6. Although these materials exhibit fantastic electrical conductivity and shielding performance, they are typically non-biodegradable and difficult to regenerate at the end of their life cycle (Supplementary Table 1). This lack of degradability and the challenges associated with repair contribute to the accumulation of electronic waste. Globally, it’s estimated that around 56 million tonnes of electronic waste are generated each year. This staggering volume not only poses a serious threat to the environment but also raises significant concerns for human health and the economy. In the case of EMI shielding materials, which are an important component of electronic equipment, their mass production and eventual disposal have significantly contributed to the growing e-waste crisis. As demand for such materials continues to rise alongside the increasing use of electronics, the environmental impact becomes even more critical. Tackling these challenges is essential to protect the environment, public health, and economic resilience^7,8. The United Nations Sustainable Development Goals (SDGs) include a policy framework on responsible consumption and production, aiming to promote sustainable patterns of both⁹. Consequently, in light of the continuous accumulation of electronic waste and increasingly stringent environmental regulations, there is an urgent need to develop high-performance EMI shielding films that combine eco-friendliness, recyclability, repairability, a balanced trade-off between sustainability and performance, and scalable production¹⁰.

All-organic conductive polymers (CPs) are regarded as high-conductivity fillers with significant potential for applications in EMI shielding¹¹. However, most CPs reported to date for EMI shielding applications (e.g., polyaniline, polythiophene, and polypyrrole) are inherently p-type. When used as additives in EMI shielding materials, they face limitations arising from low hole mobility and modest intrinsic conductivity, often necessitating the incorporation of metallic fillers to boost electrical performance. This not only increases production costs but also complicates recycling efforts^12,13,14. n-type CPs are anticipated to exhibit high electron mobility, making them promising candidates for the development of cost-effective and sustainable EMI shielding materials. However, their application in this area remains scarcely explored, largely due to the limited availability of n-type CPs with sufficiently high electrical conductivity. Poly(benzodifurandione) (ⁿPBFDO), a recently reported n-type CP, enables efficient electron injection via in-situ reduction doping (~0.9 charges per repeat unit). Its strong electron-withdrawing groups stabilize negative charges, lowering the electron injection barrier and improving ambient stability. The rigid conjugated backbone facilitates electron transport, achieving conductivity above 2000 S/cm. Notably, ⁿPBFDO is also highly soluble in the green solvent dimethyl sulfoxide (DMSO)¹⁵. These attributes point towards a cost-effective and eco-friendly solution-processing route for fabricating ⁿPBFDO-based EMI shielding films¹⁵. In addition, to promote resource efficiency and environmental sustainability, there is a growing need for EMI shielding films that can be easily repaired and reused when damaged or defective. Integrating ⁿPBFDO with biodegradable and reusable substrates to produce repairable, recyclable films is therefore both timely and highly significant.

Biodegradable polymers are generally divided into two main types: natural polymers—such as polylactic acid, chitosan, gelatin, and cellulose—and synthetic ones, including polyvinyl alcohol, polyesters, and polyamides^16,17. Nevertheless, polymers sourced from microbes or animals tend to be expensive to produce. On the other hand, synthetic polymers often face challenges with recyclability due to strong molecular interactions and limited solubility. This has led to growing interest in biomass-based cellulose materials as a promising alternative. However, cellulose films typically suffer from poor mechanical properties, making it necessary to apply modification strategies to improve their strength and durability^18,19. Ionic liquid (IL) incorporating phenylboronic acid has been reported to markedly enhance the mechanical properties of silk fibroin through hydrogen-bonding interactions²⁰. As a proof of concept, IL presents strong potential as toughening agents in the fabrication of ⁿPBFDO/cellulose films.

Here, we present a simple solution-based vacuum filtration method for grafting ⁿPBFDO onto carboxylated cellulose nanofibres (^cCNFs), producing uniform ⁿPBFDO/^cCNFs composite films reinforced with trace amounts of IL. These (ⁿPBFDO/^cCNFs)^IL films offer efficient, tunable EMI shielding while enabling sustainable recycling, scalable production, and a balance between environmental responsibility and performance. In essence, the ⁿPBFDO is grafted onto wood pulp-derived ^cCNFs, with small amounts of [VPBAMIm]Br IL intercalated between the ^cCNFs and ⁿPBFDO, enhancing the (ⁿPBFDO/^cCNFs)^IL film’s mechanical strength. By adjusting the ⁿPBFDO content from 4.8 wt.% to 27.3 wt.%, the EMI shielding effectiveness can be tuned, reaching up to 80 dB. Thanks to the solubility of ⁿPBFDO in DMSO, damaged or used films can be easily reprocessed via a straightforward DMSO/H₂O redispersion followed by vacuum filtration, supporting a sustainable closed-loop approach. Even after ten reuse cycles, the films retain over 98% of their electrical conductivity and maintain shielding effectiveness above 97.72%. Crucially, when buried in soil, the (ⁿPBFDO/^cCNFs)^IL films biodegrade completely within 100 days, posing no risk to the environment. Overall, this work demonstrates a practical and scalable route to high-performance, recyclable, and biodegradable EMI shielding materials based on cellulose and n-type CPs.

Results

(ⁿPBFDO/^cCNFs)^IL films fabrication

Carboxylated cellulose fibers (^cCNFs), boron-containing IL, and n-type CP ⁿPBFDO were synthesized according to the literature (Supplementary Figs. 1–4)^15,21,22. The phenylboronic acid groups present in the IL undergo an addition reaction with the hydroxyl groups (–OH) of ^cCNFs, resulting in the formation of dynamic boronate ester bonds in ^cCNFs^IL22. Simultaneously, these groups formed hydrogen bonds with the carboxyl groups (–COOH) of ^cCNFs, which further enhanced the intermolecular interactions. Meanwhile, the imidazolium cations within the IL promoted the formation of hydrogen bonds with both the hydroxyl and carboxyl groups of ^cCNFs. Additionally, the double bonds in the IL gave rise to Van der Waals interactions with the carboxyl groups. It is worth noting that the [VPBAMIm]Br IL exhibits the best bendability compared to other typical ILs, such as [VEIm]Br, [C₁₂MIm]Br, and [APMIm]Br (Supplementary Fig. 5). Consequently, [VPBAMIm]Br IL was selected as a toughening agent for this study. These synergistic interactions led to a significant increase in the cross-linking density of the ^cCNFs network, consequently enhancing its mechanical strength. Based on this strong matrix, we dispersed ⁿPBFDO, which was dissolved in DMSO, into the aqueous ^cCNFs^IL suspension. Through solvent displacement, ⁿPBFDO self-assembled on the surface of the ^cCNFs^IL, resulting in the formation of a stable and homogeneous conductive network ((ⁿPBFDO/^cCNFs)^IL) (Fig. 1). Unlike traditional methods that employ conductive powder fillers, in this approach, a conductive solution is combined with vacuum filtration to fabricate uniform and freestanding all-organic conductive films, effectively addressing the problem of inhomogeneity caused by gravitational sedimentation during film formation²³.

Fig. 1: Preparation of (ⁿPBFDO/^cCNFs)^IL films.

Schematic illustrating the preparation of (ⁿPBFDO/^cCNFs)ᴵᴸ films, highlighting their good recyclability and biodegradability. Key interactions among ^cCNFs, ⁿPBFDO, and the ionic liquid underpin the effectiveness of the self-assembly approach.

(ⁿPBFDO/^cCNFs)^IL films characterization

Figure 2a schematically illustrates ^cCNFs toughened with [VPBAMIm]Br IL and further grafting ⁿPBFDO to form (ⁿPBFDO/^cCNFs)^IL. The boronic acid groups in [VPBAMIm]Br can form dynamic boronic acid bonds with carboxyl groups. Meanwhile, the cationic imidazole groups can form strong hydrogen bonds with hydroxyl and carboxyl groups, and the double bonds can generate Van der Waals forces with carboxyl groups. These interactions strengthen the intermolecular forces, causing the incorporation of [VPBAMIm]Br IL to trigger the aggregation of ^cCNFs. To this end, [VPBAMIm]Br is expected to significantly enhance the intermolecular interactions between ^cCNFs. Subsequently, the addition of ⁿPBFDO leads to the formation of interconnected conductive nanofiber networks, which show promise for large-scale production. The structural characterization of ^cCNFs, ^cCNFs^IL, ⁿPBFDO, ⁿPBFDO/^cCNFs, and (ⁿPBFDO/^cCNFs)^IL was carried out via transmission electron microscopy (TEM) and scanning transmission electron microscopy (STEM). As shown in Fig. 2b, c and Supplementary Fig. 6, ^cCNFs exhibit a uniform distribution of nanofibers, with diameters ranging from 5 to 40 nm. Upon the incorporation of IL, the ^cCNFs^IL displays significant aggregation. This phenomenon is mainly ascribed to the formation of dynamic phenylboronic acid bonds and hydrogen bonds between the IL and ^cCNFs. While ⁿPBFDO exhibits a large-scale network-like morphology (Supplementary Fig. 7). Upon dispersing ⁿPBFDO into the ^cCNFs^IL suspension, it encapsulates the nanofibers via solvent displacement, resulting in the formation of a dense and uniform conductive nanofiber network (Fig. 2d). STEM imaging unveils the conductive network structure of the (ⁿPBFDO/^cCNFs)^IL composite. Meanwhile, the secondary electron images (SEI) (Fig. 2e1, e2) confirm the uniform encapsulation of ^cCNFs^IL by the ⁿPBFDO films (Supplementary Fig. 8). The presence of sulfur (S) elements can be attributed to residual DMSO. DMSO remains on the surface of PBFDO because of the strong interactions between them. In contrast, the STEM images of ⁿPBFDO/^cCNFs indicate that PBFDO is unable to form a uniform coating on ^cCNFs. This is attributed to the lack of an interconnected conductive network (Supplementary Fig. 9). These observations underscore the crucial significance of IL in promoting the establishment of a homogeneous conductive nanofiber network.

Fig. 2: Structure and morphology of (ⁿPBFDO/^cCNFs)^IL.

a Schematic illustration of (ⁿPBFDO/^cCNFs)^IL fabrication and a photograph of the resulting suspension. SEM micrographs of ^cCNFs (b), ^cCNFs^IL (c), and (ⁿPBFDO/^cCNFs)^IL (d). HADDF-STEM image (e1) and SEI image (e2) of ^cCNFs^IL, respectively. f 2D-SAXS pattern of (ⁿPBFDO/^cCNFs)^IL. g Zeta potential values for ^cCNFs, ^cCNFs^IL, (ⁿPBFDO/^cCNFs)^IL, and ⁿPBFDO, respectively. h Particle-size distributions for ^cCNFs, ^cCNFs^IL, and (ⁿPBFDO/^cCNFs)^IL, respectively. i Diagram illustrating interactions between ^cCNFs, IL, and ⁿPBFDO obtained through RGD simulations.

Figure 2f depicts the 2D small-angle X-ray scattering (2D-SAXS) pattern of (ⁿPBFDO/^cCNFs)^IL. The pattern exhibits a circular morphology, which suggests the isotropic structure of the conductive network²⁴. The conductive network has no impact on the anisotropy of ^cCNFs^IL (Supplementary Fig. 10). Further analyses of particle characteristics and dispersion behavior are carried out via Zeta potential measurement and dynamic light scattering. As shown in Fig. 2g, the Zeta potential values of ^cCNFs and ^cCNFs^IL are −18.6 mV and −19.6 mV, respectively. Upon the incorporation of ⁿPBFDO, the Zeta potential of (ⁿPBFDO/^cCNFs)^IL decreases to −39.7 mV. This alteration stems from the n-type semiconducting characteristic of PBFDO. PBFDO, as an n-type semiconductor, can supply a large number of electrons, leading to the negatively-charged surface of the (ⁿPBFDO/^cCNFs)^IL network²⁵. The strengthened electrostatic repulsion effectively inhibits fiber aggregation, resulting in greater suspension stability. The incorporation of IL into ^cCNFs results in denser fiber packing because of the hydrogen-bonding interactions between IL and ^cCNFs. As a consequence, the average particle size of ^cCNFs^IL is smaller than that of ^cCNFs. Additionally, ^cCNFs^IL shows a narrower particle-size distribution compared to ^cCNFs. The encapsulation of ⁿPBFDO onto ^cCNFs^IL fibers leads to an increase in the overall particle size of the (ⁿPBFDO/^cCNFs)^IL fibers. All the samples display a narrow particle-size distribution, which indicates the uniform dispersion of ^cCNFs, ^cCNFs^IL, and (ⁿPBFDO/^cCNFs)^IL nanofibers in the solution.

To clarify the interactions between ^cCNFs, IL, and ⁿPBFDO, the reduced density gradient (RDG) approach is employed²⁶. RDG analysis, which is founded on pre-molecular approximations of electron density, unveils the nature of intermolecular forces. As shown in Fig. 2i, the green iso-surfaces signify hydrogen bonding and electrostatic interactions, whereas the blue iso-surfaces are associated with dispersion forces. The RDG analysis validates that the assembly of ^cCNFs, IL, and ⁿPBFDO is predominantly regulated by hydrogen bonding and electrostatic interactions. Moreover, ^cCNFs^IL involves dynamic phenylboronic acid bonds. In addition, short-range attractive forces originating from instantaneous and induced dipoles, namely dispersion forces, also play a role in enhancing the solution’s stability. These synergistic interactions serve as the foundation for the formation of a stable and uniform conductive nanofiber network. Geometry frequency noncovalent force field simulations and energy decomposition analysis reveal strong electrostatic interactions and notable induction effects among ^cCNF, IL, and ⁿPBFDO, indicating robust intermolecular forces (Supplementary Fig. 11; Supplementary Movie 1).

The ^cCNFs, ^cCNFs^IL, and (ⁿPBFDO/^cCNFs)^IL suspensions are subjected to vacuum filtration to form films. The incorporation of IL endows the resulting films with enhanced foldability (Fig. 3a, b and Supplementary Fig. 12a). 2D-SAXS analysis indicates that all three films display uniform isotropic structures (Fig. 3c1, c2 and Supplementary Fig. 13). Simulation of the radius of gyration (R_g) show that the R_g value of (ⁿPBFDO/^cCNFs)^IL is significantly higher than those ^cCNFs and ^cCNFs^IL. This implies that the assembly of ^cCNFs with ⁿPBFDO results in the formation of larger-sized conductive network units (Fig. 3d and Supplementary Table 2)²⁷. Further analysis via scanning electron microscopy (SEM) demonstrates that the incorporation of IL leads to denser packing of the ^cCNFs. When the ⁿPBFDO is incorporated, the ^cCNFs get encapsulated, giving rise to the formation of an aggregate network. Moreover, the films obtained through vacuum filtration display a layered stacking structure (Fig. 3e–j and Supplementary Fig. 12). Additionally, upon the introduction of IL, hydrogen bonding facilitates the further aggregation of the cellulose nanofibers (Supplementary Fig. 14). As PBFDO is incorporated, the original morphology of the ^cCNFs gradually vanishes, ultimately forming a continuous conductive network (Supplementary Fig. 15). Energy dispersive spectroscopy (EDS) analysis of the (ⁿPBFDO/^cCNFs)^IL film reveals a uniform distribution of elements across the film surface, indicating that this system effectively circumvents the inhomogeneities commonly associated with the addition of nanomaterials (Supplementary Fig. 16). Atomic force microscopy showed R_a values of 14.4 nm for ^cCNFs, 11.3 nm for ^cCNFs^IL, and 20.9 nm for (ⁿPBFDO/^cCNFs)^IL films (Supplementary Fig. 17). These results suggest that IL has little effect on CNF surface morphology, while ⁿPBFDO significantly increases roughness, likely due to its coating of the CNFs.

Fig. 3: Structure and morphology of (ⁿPBFDO/^cCNFs)^IL films.

Photographs of ^cCNFs^IL film (a) and an (ⁿPBFDO/^cCNFs)^IL film (b), respectively. The 2D-SAXS patterns of a ^cCNFs^IL film (c1) and (ⁿPBFDO/^cCNFs)^IL film (c2). d The 1D-SAXS intensity profiles with fitted curves for ^cCNFs film, ^cCNFs^IL film, and (ⁿPBFDO/^cCNFs)^IL film, respectively. e SEM image of the surface of a ^cCNFs^IL film. f, g Cross-sectional SEM images of the surface of a ^cCNFs^IL film. h SEM image and i, j Cross-sectional SEM images of the surface of an (ⁿPBFDO/^cCNFs)^IL film. k XRD patterns of ^cCNFs film, ^cCNFs^IL film, (ⁿPBFDO/^cCNFs)^IL film, and neat ⁿPBFDO polymer, respectively. l, m Tensile stress-strain curves and corresponding tensile moduli for each of the foregoing samples, respectively. Error bars show (mean values ±) s.d.; n = 3 repeats.

Figure 3k depicts the X-ray diffraction (XRD) patterns of the samples. The peaks at 16.1° and 22.8° are associated with the (101) and (002) planes of ^cCNFs, ^cCNFs^IL, (ⁿPBFDO/^cCNFs)^IL, and ⁿPBFDO films, respectively. Notably, the (002) plane of the ^cCNFs^IL film shifts to 22.9°. Based on the Bragg diffraction formula, this shift implies that the incorporation of IL decreases the spacing between the ^cCNFs, resulting in a more compact arrangement. This finding is in line with the strengthened hydrogen bonding interactions promoted by IL. These interactions reduce the inter-fiber distance and lead to a more densely packed and ordered structure. In the case of the (ⁿPBFDO/^cCNFs)^IL film, the (101) and (002) planes shift to 16.7° and 23.1°, respectively. The ester groups of ⁿPBFDO further reinforce the hydrogen-bonding interactions among ^cCNFs, facilitating the formation of a more tightly packed (ⁿPBFDO/^cCNFs)^IL film. The peak at 8.0° is associated with the (100) plane of ⁿPBFDO. Nevertheless, no distinct peaks of ⁿPBFDO are detected in the (ⁿPBFDO/^cCNFs)^IL film, indicating that ⁿPBFDO is arranged in a disordered fashion on the surface of the ^cCNFs^IL.

The Fourier-transform infrared (FTIR) spectra indicate that the peak at 1743 cm⁻¹ in the ^cCNFs film corresponds to the stretching vibration of the carboxyl group. When a small quantity of ILs is added, a new absorption peak emerges at 802 cm⁻¹. This peak can be ascribed to the out-of-plane bending vibration of the C-H bond in the benzene ring. Further encapsulation of ⁿPBFDO onto the ^cCNFs^IL film leads to the appearance of a new peak at 1780 cm⁻¹. This peak corresponds to the C=O stretching vibration of the lactone moiety in ⁿPBFDO, which verifies the successful synthesis of the (ⁿPBFDO/^cCNFs)^IL film (Supplementary Fig. 18). X-ray photoelectron spectroscopy (XPS) analysis reveals that all samples mainly comprise carbon (C) and oxygen (O) elements. Remarkably, when compared with that of ^cCNFs, the C1s peaks of ^cCNFs^IL show shifts in the binding energies of C–O, O–C–O, and O–C=O from 285.6 eV, 288.1 eV, and 289.1 eV to 286.5 eV, 288.9 eV, and 289.3 eV, respectively²¹. Moreover, the binding energy of O1s undergoes a shift, which implies that the IL ions interact with ^cCNFs, resulting in alterations in the chemical states of carbon and oxygen atoms. The C1s and O1s spectra of (ⁿPBFDO/^cCNFs)^IL predominantly display peaks characteristic of ⁿPBFDO, further validating its successful incorporation into the ^cCNFs^IL surface (Supplementary Fig. 19)¹⁵.

Physical properties of (ⁿPBFDO/^cCNFs)^IL films

Thermogravimetric analysis (TGA) and derivative thermogravimetric analysis curves show that the ^cCNFs and ^cCNFs^IL films exhibit similar thermal degradation behavior, with rapid pyrolysis occurring between 275 °C and 355 °C. The decomposition temperature of the (ⁿPBFDO/^cCNFs)^IL film is slightly higher, ranging from 303 °C to 404 °C. Eventually, all samples experience complete pyrolysis at approximately 800 °C, which validates the all-organic composition of the material system (Supplementary Fig. 20).

The mechanical properties of (ⁿPBFDO/^cCNFs)^IL film are of great significance for its practical application. The exploration of the films’ mechanical properties shows that the incorporation of a small amount of ILs (0.025 mL, 10 mg/mL) remarkably improves the tensile stress and tensile strain of the ^cCNFs film (Fig. 3l). However, an excessive amount of ILs results in overly strong intermolecular interaction, giving rise to a film that is excessively rigid. This, in turn, has a negative effect on its mechanical performance (Supplementary Fig. 21a). In contrast, the tensile strain of the ⁿPBFDO/^cCNFs film is merely one-third that of the ^cCNFs film. By incorporating IL simultaneously, both the tensile strain and tensile stress of the (ⁿPBFDO/^cCNFs)^IL film are significantly improved. Moreover, the mechanical properties of the (ⁿPBFDO/^cCNFs)^IL film can be precisely adjusted by regulating the ⁿPBFDO content (Supplementary Fig. 21b). The tensile modulus of ⁿPBFDO/^cCNFs is considerably higher than that of ^cCNFs, ^cCNFs^IL, and (ⁿPBFDO/^cCNFs)^IL. However, its tensile stress and strain are relatively lower, which implies that ⁿPBFDO/^cCNFs is a brittle material that undergoes limited deformation under high stress (Fig. 3m). The incorporation of IL notably enhances the film’s deformability.

EMI shielding performance of (ⁿPBFDO/^cCNFs)^IL films

After understanding the structure, morphology, and intrinsic physical properties of the prepared (ⁿPBFDO/^cCNFs)^IL films, the EMI shielding performance of (ⁿPBFDO/^cCNFs)^IL films was further studied. Conductivity is a critical factor in EMI shielding materials. Therefore, the conductivity of the prepared films was measured and evaluated using different methods. Figure 4a demonstrates that the (ⁿPBFDO/^cCNFs)^IL films are capable of lighting an LED bulb within an assembled circuit, highlighting their fantastic electrical conductivity. By tuning the mass ratio of ⁿPBFDO in the films (0.0%, 4.8%, 11.1%, 20%, 27.3%), a significant enhancement in the conductivity of the (ⁿPBFDO/^cCNFs)^IL films is achieved, with electrical conductivities of 1.47 × 10⁻⁴, 1.74 × 10², 2.39 × 10³, 5.50 × 10³, and 1.08 × 10⁴ S/m, respectively (Fig. 4b). The thickness of all film samples is approximately 0.17–0.21 mm. As shown in Fig. 4c, the EMI shielding performance of the samples in the 8.2–12.4 GHz (X band) range reveals a clear increase in the total shielding efficiency (SE_T) with higher ⁿPBFDO content. The ^cCNFs film almost has no EMI performance. For the sample with an ⁿPBFDO content of 4.8 wt.%, the EMI SE_T of the (ⁿPBFDO/^cCNFs)^IL-1 film reaches 29.77 dB, meeting the commercial EMI shielding requirement of 20 dB. At 11.1% ⁿPBFDO content, the EMI SE_T of the (ⁿPBFDO/^cCNFs)^IL-2 film is approximately 40 dB, which reduces the electromagnetic field intensity to 0.01% of its original value. When the ⁿPBFDO content is increased to 20% and 27.3%, the EMI SE_T of the (ⁿPBFDO/^cCNFs)^IL-3 and (ⁿPBFDO/^cCNFs)^IL-4 films reach 74.63 dB and 83.77 dB, respectively, indicating exceptional EMI shielding capability, with the ability to reduce electromagnetic field intensity by approximately 99.99999%. These remarkable properties position these films as ideal candidates for applications in high-precision laboratory environments, military communications, and aerospace devices²⁸.

Fig. 4: EMI shielding performance of (ⁿPBFDO/^cCNFs)^IL films.

a Photographs showing a non-conductive ^cCNFs^IL film and a conductive (ⁿPBFDO/^cCNFs)^IL film completing an LED circuit, respectively; schematic diagram of the setup used to assess film conductivity. b Electric conductivity of films with varying ⁿPBFDO content. Error bars show (mean values ±) s.d.; n = 5 repeats. c Frequency-dependent total EMI shielding effectiveness SE_T of (ⁿPBFDO/^cCNFs)^IL films with different ⁿPBFDO loadings and ^cCNFs film. d Shielding effectiveness contributions: The SE_T, SE_A, and SE_R values are plotted along with different ⁿPBFDO content. e Photographs of Tesla coil tests demonstrating the EMI shielding performance. f Proportion of SE_A to SE_T. g Reflection (R), transmission (T), and absorption (A) coefficients for films with increasing ⁿPBFDO content. h A comparison between the EMI shielding performances registered by the (ⁿPBFDO/^cCNFs)^IL films and other representative EMI shielding materials. i Schematic illustration of EMI attenuation mechanisms via a 3D conductive network and multi-reflection.

The reflection efficiency (SE_R) and absorption efficiency (SE_A), calculated based on the S-parameters, are shown in Fig. 4d. With an increasing content of ⁿPBFDO, both SE_R and SE_A exhibit notable improvements, with SE_A showing a more significant increase. The SE_A/SE_T ratios for the ⁿPBFDO at different mass ratios (4.8%, 11.1%, 20%, 27.3%) are 67.99%, 72.67%, 80.60%, and 82.05%, respectively (Fig. 4f). Further, the average scattering parameters of the reflection coefficient (R), absorption coefficient (A), and transmission coefficient (T) are presented in Fig. 4g. For all (ⁿPBFDO/^cCNFs)^IL films with ⁿPBFDO contents ranging from 4.8% to 27.3%, the T values approach zero, indicating effective shielding of the incident electromagnetic waves. Moreover, the R values of these films are close to 0.8, suggesting that electromagnetic waves are primarily attenuated by reflection upon contact with the (ⁿPBFDO/^cCNFs)^IL films, reducing the likelihood of wave penetration through the material. Meanwhile, the dielectric parameters of the film were measured through the waveguide method. It can be observed that as the content of ⁿPBFDO increases, the real part (ε‘) of the film’s dielectric constant becomes increasingly negative, indicating that electromagnetic waves cannot pass through. Simultaneously, the imaginary part (ε“) of the dielectric constant increases, indicating enhanced energy dissipation within the material (Supplementary Fig. 22).

The corresponding shielding mechanism of the (ⁿPBFDO/^cCNFs)^IL films is illustrated in Fig. 4i. Upon encountering the (ⁿPBFDO/^cCNFs)^IL films’ surface, the impedance mismatch between the air and the highly conductive film results in the majority of electromagnetic waves being directly reflected, while a small fraction penetrates into the film. On one hand, the conductive network formed by ⁿPBFDO provides abundant charge carriers, which interact with the electromagnetic waves through resistive losses, converting the waves into heat. On the other hand, the layered structure of the (ⁿPBFDO/^cCNFs)^IL films, with its numerous interfaces, promotes multiple reflections and scattering of the electromagnetic waves, further enhancing the absorption and attenuation of the waves.

The electromagnetic wave shielding performance of ^cCNFs^IL, (ⁿPBFDO/^cCNFs)^IL-1, (ⁿPBFDO/^cCNFs)^IL-2, (ⁿPBFDO/^cCNFs)^IL-3, and (ⁿPBFDO/^cCNFs)^IL-4 films in the frequency range of 8.2 GHz to 12.4 GHz was simulated using the High-Frequency Structure Simulator (HFSS) (Supplementary Figs. 23–27). The results indicate that the ^cCNFs^IL film offers minimal shielding, allowing most of the electromagnetic waves to penetrate. In contrast, the (ⁿPBFDO/^cCNFs)^IL-1 film effectively blocks most of the electromagnetic wave penetration, with only a small fraction passing through. For (ⁿPBFDO/^cCNFs)^IL-2, (ⁿPBFDO/^cCNFs)^IL-3, and (ⁿPBFDO/^cCNFs)^IL-4 films, nearly all electromagnetic waves are blocked, demonstrating significantly enhanced shielding performance. The EMI shielding performance of (ⁿPBFDO/^cCNFs)^IL films was further assessed using a Tesla coil circuit (Fig. 4e). When the ^cCNFs^IL film was placed between the coil and the lightbulb, the lightbulb remained illuminated. However, when the ^cCNFs^IL film was replaced with the (ⁿPBFDO/^cCNFs)^IL film, the lightbulb immediately turned off, indicating that the (ⁿPBFDO/^cCNFs)^IL film effectively shields the electromagnetic field within the coil. Additionally, the shielding efficiency-to-thickness ratio (SE_T/t) of the (ⁿPBFDO/^cCNFs)^IL films was compared to that of previously reported metal-based films, carbon-based films, and CP-based films (Fig. 4h and Supplementary Table 3). In comparison to previous studies, the (ⁿPBFDO/^cCNFs)^IL films presented in this work exhibited superior shielding performance even at ultrathin thicknesses. Specifically, when the ⁿPBFDO concentrations are 20% and 27.3%, the SE_T/t values of the (ⁿPBFDO/^cCNFs)^IL-3 and (ⁿPBFDO/^cCNFs)^IL-4 films are 392 and 399 dB/mm, respectively. These values are significantly higher than the SE_T/t values of 90.0 dB/mm for a 0.28 mm thick PANI/CNF film and the SE_T/t values of 215.0 dB/mm for a 0.35 mm thick Cellulose/MXene film. This clearly indicates that the EMI shielding performance of the (ⁿPBFDO/^cCNFs)^IL films is comparable to that of commonly used metal-based, carbon-based, and CP-based additives. The (ⁿPBFDO/^cCNFs)^IL films offer improved mechanical properties and lower cost than comparable samples, highlighting their strong potential for practical use (Supplementary Tables 4 and 5)^{29,30,31,32,33,34,35,36,37,38,39,40,41}.

Chemical resistance and sustainable recyclability of (ⁿPBFDO/^cCNFs)^IL films

(ⁿPBFDO/^cCNFs)^IL films against water and organic solvents were systematically studied. Initially, water contact angle measurements indicate that the incorporation of IL and ⁿPBFDO into the ^cCNFs matrix significantly enhances the hydrophobicity of the films (Supplementary Fig. 28). This suggests the potential of the (ⁿPBFDO/^cCNFs)^IL films to resist water. To assess their stability, the (ⁿPBFDO/^cCNFs)^IL films retained 94.5% EMI performance after 500 days under ambient laboratory conditions (Supplementary Fig. 29). Also, they were immersed in different solvents (e.g., tetrahydrofuran (THF), chloroform (CF), carbon tetrachloride (CB), acetonitrile (MeCN), and 2 M sulfuric acid (H₂SO₄)) for 10 min. Then, the films were rinsed with water and then dried. Interestingly, the conductivity of the films persisted at above 92% of its initial value (Supplementary Fig. 30). Furthermore, salt wet resistance tests were carried out to mimic the possible application scenarios of the films in marine environments. Impressively, the conductivity of the films was maintained at over 60% of its original value (Supplementary Fig. 31). Since ⁿPBFDO can be re-dissolved in DMSO, when the (ⁿPBFDO/^cCNFs)^IL films are damaged, they can be shredded and re-dispersed in a DMSO/H₂O solution via ultrasonication, yielding a (ⁿPBFDO/^cCNFs)^IL solution. Next, the film can be restored through filtration (Fig. 5a). Ten dispersion-filtration cycles were carried out on the (ⁿPBFDO/^cCNFs)^IL-2 film. Remarkably, its conductivity remained nearly constant, retaining 97% of its initial value (Fig. 5b). The EMI SE_T of the film stayed around 40 dB. After the 10th cycle, the SE_T10/SE_T0 ratio was 97.72% (Fig. 5c and Supplementary Fig. 32). Also, the tensile stress and strain remain at 93.30% and 93.46%, respectively, after 10 cycles (Supplementary Fig. 33). These findings indicate that the films sustain their electromagnetic wave interference shielding performance even after numerous cycles, underscoring the economic reusability of (ⁿPBFDO/^cCNFs)^IL films.

Fig. 5: Investigation of the reusability and biodegradability of (ⁿPBFDO/^cCNFs)^IL films.

a Photographs demonstrating the film’s reusability after ten dispersion-filtration cycles. b Electric conductivity of a (ⁿPBFDO/^cCNFs)^IL film monitored over successive dispersion-filtration cycles. c Total EMI shielding effectiveness (SE) of a (ⁿPBFDO/^cCNFs)^IL-2 film measured after each of the shown reprocessing cycles. d Biodegradation of a (ⁿPBFDO/^cCNFs)^IL film in moist soil, compared with polyvinyl chloride (PVC), highlighting the film’s superior biodegradability. e Radar chart illustrating key attributes of the (ⁿPBFDO/^cCNFs)^IL film.

When the (ⁿPBFDO/^cCNFs)^IL film reaches the end of its life cycle, it has the potential to be biodegraded by being buried, highlighting its environmental friendliness (Fig. 5d). To facilitate comparison, both (ⁿPBFDO/^cCNFs)^IL films and polyvinyl chloride (PVC) sheets were buried 5 cm deep in soil. Their morphological changes were monitored at regular intervals to evaluate the degradation process⁴². After 2 weeks, the surface of the (ⁿPBFDO/^cCNFs)^IL film started to turn rough, and microorganisms had formed a white biofilm on the material’s surface. One and a half months later, the film underwent fracture, presumably because microorganisms secreted hydrolytic enzymes. These enzymes interacted with the material surface, cleaving polymer chains through hydrolysis and thus leading to the formation of low-molecular-weight products. By the end of 3 months, the (ⁿPBFDO/^cCNFs)^IL film had been reduced to fine particulate residues, mainly as a result of the metabolic processes of microorganisms. Eventually, most of the material fragments were converted into water and CO₂. These findings suggest that (ⁿPBFDO/^cCNFs)^IL film is a fully biodegradable organic material. It not only demonstrates mechanical strength and remarkable electromagnetic shielding properties but also possesses biodegradability, recyclability, and environmental sustainability. This dual-benefit characteristic makes it both economically and ecologically advantages (Fig. 5e and Supplementary Table 6).

Discussion

In this work, we present a sustainable and scalable strategy for fabricating high-performance EMI shielding films using a prototypical n-type CP (ⁿPBFDO) grafted onto carboxylated cellulose nanofibres (^cCNFs), reinforced by a minimal amount of ionic liquid. The resulting (ⁿPBFDO/^cCNFs)^IL films demonstrate exceptional shielding effectiveness—reaching up to 83.77 dB—while offering key sustainability features including closed-loop recyclability, mechanical robustness, and full biodegradability within 100 days in soil. These films not only rival traditional carbon- and metal-based shielding materials in performance but also overcome their limitations in end-of-life recovery and environmental persistence.

Beyond their technical merit, this platform aligns with global efforts to reduce electronic waste, promote circular material use, and transition toward greener electronics manufacturing. By combining functional performance with eco-design principles, our approach offers a tangible step forward in developing next-generation electronic materials that support sustainable infrastructure, responsible consumption, and reduced environmental impact—key priorities under the SDG 12 (Responsible Consumption and Production) through providing a platform for reducing e-waste in electronics manufacturing.

Methods

Chemicals

Bleached kraft pulp was selected as a starting material. Benzo[1,2-b:4,5-b’]difuran-2,6(3H,7H)-dione (H-BFDO, >98%), [VEIm]Br (>99%), [C₁₂MIm]Br (>98%), and [APMIm]Br (>98%) were purchased from Adamas. Tetramethylquinone (TMQ, >98%), Citric acid (CA, >99%), Choline chloride (>98%), 4-(Bromomethyl)phenylboronic acid (BPA, >98%) and 1-Vinylimidazole (>98%) were purchased from Bidepharm Co., Ltd. DMSO, Dichloromethane (CH₂Cl₂, >99.5%), Ethyl acetate (EA, >99.5%) and Triethylamine (TEA, >99%) were purchased from Nanjing Chemical Reagent Co., Ltd. All chemical reagents were analytically pure without further purification.

Preparation of carboxylated cellulose nanofibers (^cCNFs) solution

Citric acid (CA, 90 g), Choline chloride (30 g), and water (30 g) were added to a 250 mL round-bottom flask, which was equipped with a magnetic agitation and condensation reflux. The flask was then mixed at 80 °C for 15 min to obtain a homogeneous H-DES solution. Next, a quantity of 4.5 g of bleached kraft pulp was added to the above flask and reacted at 130 °C for 1 h. The reaction pulp was then isolated and washed with deionized water; the precipitate was collected and dried at 60 °C for 24 h to obtain ^cCNFs. Following the pretreatment, the ^cCNFs (2 g) are mixed with water (100 mL) and subjected to an ultrasonic generator with a power of 800 W for 30 min, a uniform suspension of ^cCNFs (20 mg/mL) solution is obtained.

Preparation of n-type polymer ⁿPBFDO solution

3,7-dihydrobenzo[1,2-b:4,5-b’]dioxin-2,6-dione (H-BFDO, 3.20 g), tetramethylquinone (TMQ, 2.07 g), and 100 mL DMSO were added to a 250 mL round-bottom flask, which was equipped with a nitrogen inlet, magnetic agitation, and condensation reflux. The mixture was magnetically stirred at 100 °C for 6 h and cooled to room temperature. Next, 5 mL TEA and 50 mL CH₂Cl₂ were added to the above solution, accompanied by the precipitation of black solids. The resulting precipitate was washed three times with CH₂Cl₂ until the filtrate was colorless and dried at 60 °C for 24 h. Finally, ⁿPBFDO obtained was 3.02 g (with a yield of 94.4%). Following the pretreatment, the ⁿPBFDO (10 mg) is mixed with DMSO (1 mL) at 80 °C to obtain the ⁿPBFDO solution (10 mg/mL).

Preparation of [VPBAMIm]Br IL

4-(Bromomethyl)phenylboronic acid (BPA, 4.30 g), 1-Vinylimidazole (1.90 g), and 100 mL EA were added to a 250 mL round-bottom flask, which was equipped with a magnetic agitation and condensation reflux. The mixture was reacted at 40 °C for 24 h. The product was washed with 50 mL EA three times and dried at 60 °C for 1 h to obtain IL. Subsequently, the IL (100 mg) is mixed with 1 mL DMSO to obtain a 100 mg/mL IL.

Preparation of ^cCNFs^IL films

Take the preparation process of the ^cCNFs^IL-2 film as an example. 10 mL ^cCNFs (20 mg/mL) solution and 0.025 mL IL (100 mg/mL) were mixed through ultrasonic. The mixture is then subjected to vacuum filtration⁴³, oven drying, and hot pressing to obtain ^cCNFs^IL-2 film. ^cCNFs^IL-1 and ^cCNFs^IL-3 films were obtained by substituting 0.025 mL IL (100 mg/mL) for 0.010 mL and 0.050 mL, respectively. (Note: The ^cCNFs^IL sample mentioned in the article corresponds to the ^cCNFs^IL-2 sample.)

Preparation of (ⁿPBFDO/^cCNFs)^IL films

Take the preparation process of the (ⁿPBFDO/^cCNFs)^IL-1 film as an example. 10 mL ^cCNFs (20 mg/mL) solution and 0.025 mL IL (100 mg/mL) were mixed through ultrasonic. 1 mL ⁿPBFDO solution (10 mg/mL) was added to the above solution and mixed through ultrasonics. The mixture is then subjected to vacuum filtration⁴³, oven drying, and hot pressing to obtain (ⁿPBFDO/^cCNFs)^IL-1 film. (ⁿPBFDO/^cCNFs)^IL-2, (ⁿPBFDO/^cCNFs)^IL-3, and (ⁿPBFDO/^cCNFs)^IL-4 films were obtained by substituting 1 mL ⁿPBFDO solution (10 mg/mL) for 2.5 mL, 5 mL, and 7.5 mL, respectively. Reaction conditions for the synthesis of cellulose graft n-type polymers are labeled in Supplementary Table 7. (Note: The (ⁿPBFDO/^cCNFs)^IL sample mentioned in the article corresponds to the (ⁿPBFDO/^cCNFs)^IL-1 sample.)

Characterization

The XRD spectra were recorded on a Bruker D8 diffractometer using Cu Kα radiation. FTIR spectra were taken on a Thermo NICOLETIS 20 FTIR spectrometer. H¹ NMR spectra were taken on a Bruker AVANCE III 500 MHz spectrometer. Small-angle X-ray scattering (SAXS) patterns were obtained by Xeuss3.0 automatic small-angle X-ray scatterer produced by Xenocs. TGA measurements were performed on a TA Instruments Q50 thermogravimetric analyzer in a N₂ atmosphere at 25 °C to 1000 °C at a 5 °C min⁻¹ heating rate. The mechanical properties of the samples were tested via an Instron 5943 mechanical tester. The elemental composition and chemical valence states were analyzed by XPS with an Al Kα X-ray source at 150 W (Thermo Scientific K-Alpha). The morphology of the samples was observed with a Hitachi Regulus-8100 field-emission gun scanning electron microscope (SEM) (2–5 kV). TEM images were collected using a field-emission gun Hitachi JEM-1400plus microscope. EDS mapping was observed with a field-emission gun JEOL JEM-F200 microscope. Zeta curves were measured by a Malvern ZS90 nanoparticle size and zeta potential analyzer. Nanoparticle size curves were measured by a Malvern Mastersize 3000 laser particle size analyzer. The water contact angle of the samples was measured by a contact angle tester (JC2000D7M). The electrical conductivity was measured by the four-point probe meter (RTS-8). The surface roughness was tested by an Atomic Force Microscope (Bruker Dimension ICON).

Electromagnetic shielding measurements

The EMI shielding properties in the X band (8.2–12.4 GHz) were tested by a vector network analyzer (Keysight N5222B, USA) through a waveguide method. The test samples are carefully cut into 22.86 × 10.16 mm strips to fit the specific waveguide sample holders (8.2–12.4 GHz). EM waves can interact with EMI shielding materials via reflection, absorption, and transmission. According to the law of conservation of energy, the reflection coefficient (R), absorption coefficient (A) and transmission coefficient (T) can be expressed as²:

$$R+A+T=1\,$$

(1)

$$R={|{S}_{11}|}^{2}={|{S}_{22}|}^{2}$$

(2)

$$T={|{{{\rm{S}}}}_{12}|}^{2}={|{{{\rm{S}}}}_{21}|}^{2}$$

(3)

where S₁₁ and S₂₁ represent the reflection coefficient and the transmission coefficient, respectively. The total EMI shielding effectiveness (SE_T) can thus be divided into three aspects: the reflection effectiveness (SE_R) from the material surface, the absorption effectiveness (SE_A) originating from the magnetic and conduction losses, and the multiple reflection effectiveness (SE_MR) inside the materials. If SE_T exceeds 15 dB, SE_MR can be generally ignored. The shielding effectiveness of reflection and absorption can be rewritten using the following formulas:

$${{\mathrm{SE}}}_{R}=10\log \left(\frac{1}{1-R}\right)=10\log \left(\frac{1}{1-{\left|{S}_{11}\right|}^{2}}\right)$$

(4)

$${{\mathrm{SE}}}_{A}=10\log \left(\frac{1-T}{T}\right)=10\log \left(\frac{1-{\left|{S}_{11}\right|}^{2}}{{\left|{S}_{21}\right|}^{2}}\right)$$

(5)

$${{\mathrm{SE}}}_{T}={{\mathrm{SE}}}_{R}+{{\mathrm{SE}}}_{A}$$

(6)

Reduced density gradient (RDG) method simulation

The RGD simulations were conducted using the GROMACS software. To describe the (ⁿPBFDO/^cCNFs)^IL system, the Amber ff 99 force field was employed. The Avogadro software (Version 1.20. Avogadro: an open-source molecular builder and visualization tool. http://avogadro.cc/) was used to construct the model of the single chain ^cCNFs, ⁿPBFDO and IL. The three dimensions were subjected to periodic boundary conditions (PBC), and the particle mesh Ewald algorithm was used to calculate electrostatic interactions. To constrain the bond that involved hydrogen, the SHAKE algorithm was used. A cut-off value of 10 Å was used to treat noncovalent interactions. All the molecular dynamics simulations were carried out under the NPT ensemble, and the assembly process of ^cCNFs, ⁿPBFDO and IL was investigated for 20 ns. All systems achieved equilibrium after 10 ns.

HFSS simulation

In this work, all electromagnetic field simulations were performed by ANSYS HFSS 2023 R1 in the frequency domain⁴⁴. The size of the rectangular waveguide clamp is defined as a 22.86 × 10.16 × 29.12 mm (x × y × z). The model spaces are assigned the Floquet excitations and the master and slave boundaries. In brief, S1 and S2, corresponding to ports 1 and 2 for the vector network analyzer, are set up to generate electric field excitation and receive EM signals, respectively. The transmission mode of EM waves in the waveguide fixture is defined as the typical transverse electric wave, where the propagation directions of the electric field and magnetic field are perpendicular to each other. The XY plane of the cuboid is selected as the excitation face of the wave port and electromagnetic wave propagation along the Z direction. The rectangular waveguide boundary is set as a perfectly electric conductor (Perfect E). The mesh quality of the ^cCNFs^IL, (ⁿPBFDO/^cCNFs)^IL-1, (ⁿPBFDO/^cCNFs)^IL-2, (ⁿPBFDO/^cCNFs)^IL-3, and (ⁿPBFDO/^cCNFs)^IL-4 is the default value. The solution step is set to frequency sweep, where the frequency type is the linear step from 8.2 GHz to 12.4 GHz with an interval of 0.1 GHz. The detailed parameters of the HFSS EMI simulation, such as electrical conductivity, thickness, permeability, excitation type, boundary conditions, etc. (Supplementary Table 2), are then put into the created model. The simulation results are obtained after executing the calculations.

Salt wet resistance test

In order to verify the environmental suitability of the coating for combustible cartridges exposed to seawater, the specific operation of salt wet resistance test is as follows: 5 g NaCl was added into 150 mL distilled water to dissolve, then further distilled water was added to give a total mass of 200 g, reaching the theoretical salinity of 3.5% of seawater (according to standard seawater concentration). The temperature was controlled at 26 °C for 3 days. Afterwards, the samples were dried in the air.