1. Introduction

With the significant developments in electronic technology, including 5G technology, artificial intelligence (AI), soft robots, and wearable electronics, there is a growing demand for high-performance multi-functional materials for electromagnetic protection, thermal management, and intelligent sensing^[1-5]. These application scenarios increasingly require materials that can simultaneously offer mechanical flexibility, high signal sensitivity, thermal conductivity, and electromagnetic shielding efficiency^[6,7]. Over the last decades, many types of multi-functional materials have been proposed by developing new materials, hybrids and composites of multiple materials, and structural engineering of materials^[5,8-13]. Among them, the structural engineering strategy of functional components can efficiently achieve the property modulation and customization of functional materials^[14]. Oriented networks, due to their unique structural characteristics, have been widely applied in the preparation of high-performance multifunctional materials^[4,15,16].

Attributed to their unique anisotropic structural characteristics, materials with an oriented conductive network have distinctive advantages in the applications of electromagnetic interference (EMI) shielding, thermal management, and intelligent sensing^[4,15,17]. For example, in the field of EMI shielding that aims at diminishing or eliminating the damage of electromagnetic wave pollution to human health and device operation, materials with an oriented conductive network emerge as good candidates for the new-generation shielding materials with superior EMI shielding capability^[18]. This is because these oriented networks with an orderly arranged pore structure and anisotropic morphology can form more interfaces, resulting in more multiple reflection losses^[19,20]. Meanwhile, the composites with oriented filler conducting networks can efficiently utilize the filler/polymer interfaces to enhance the multiple reflections and attenuation of the incident waves^[18,21,22]. In addition, the oriented structure is beneficial for the formation of thermal and electrical conducting paths, which will lead to a significant improvement in electrical conductivity and thermal conductivity, thereby enabling applications in thermal management and smart sensors^[4,15]. Significantly, the unique oriented structure can also achieve special application scenarios. For example, in terms of thermal management, oriented conductive networks can endow materials with directional heat transfer capabilities, achieving efficient in-plane or out-of-plane thermal management^[15]. In the field of intelligent sensing, oriented conductive networks can greatly enhance sensitivity and can be used to fabricate multi-dimensional sensors^[4,23].

Due to their critical role in regulating structure–property relationships in functional materials, significant efforts have been devoted to developing the fabrication strategies for constructing oriented architectures over the last decade. Among them, directional freeze-casting has emerged as a feasible and versatile technique, where the oriented growth of ice crystals is exploited to guide the alignment of nanoscale building blocks into highly ordered networks^[24]. In addition, external-field-assisted methods, including electric-field, magnetic-field, and mechanical-force-induced alignment, have also been extensively reported as effective routes to generate oriented networks with tunable anisotropic structures^[4,25]. In recent years, inspired by the intrinsic structural features of precursor materials, several unconventional yet efficient strategies have also been proposed to construct oriented networks^[18,26]. For example, we have developed a rapid and scalable approach using commercially available graphite paper as the starting material, in which liquid-phase expansion combined with thermal expansion enables the formation of large-area, highly oriented graphite networks without complex processing steps^[26]. Such oriented architectures exhibit outstanding performance in directional thermal management, EMI shielding, intelligent sensing, and electrical switching^[18,26-28]. Collectively, these advances demonstrate that the construction of oriented conductive networks has become an effective and general strategy for enhancing the multifunctional performance of materials, providing a robust route for next-generation intelligent and energy-related applications.

In the last decade, intense research on multifunctional materials ranging from synthetic methods and performance enhancement to promising applications has been witnessed in a large number of publications. Up to now, there are many reviews on functional materials involving special materials, applications, and functions^[4,29]. Nevertheless, most of the current reviews focus on optimizing a single property or are based on a single material, while the topic of oriented structures for multi-functional applications is still omitted. As the inevitable trend of future functional materials, it is necessary to review and summarize the structural characteristics, structure–property relationships, and application prospects of multi-functional materials based on oriented conductive networks.

In this review, we briefly present the recent research progress in multi-functional materials based on the oriented conductive networks, including fabrication techniques, structural characteristics, and multi-functional applications. The techniques for fabricating structurally oriented networks are summarized. The design and working mechanisms of various oriented networks and their application prospects in EMI shielding, thermal management, and smart sensing are presented. Some of the exciting oriented structures and performance characteristics are highlighted. Finally, the challenges faced by multi-functional materials with oriented networks and the potential opportunities in future frontier technology for soft electronic devices are offered. It is believed that multi-functional materials based on the oriented conductive networks will have enormous potential for complex systems, including smart robotics, electronic skin, AI, and complex motion detection. Figure 1 summarizes the fabrication technologies of the oriented conductive networks and their applications.

Display Full Size

Download

Figure 1. Summary of fabrication methods and applications of oriented networks reported in recent years. Reproduced with permission from reference^[18]. Copyright © 2022 Royal Society of Chemistry; reference^[25]. Copyright © 2018 Elsevier; reference^[30]. Copyright © 2021 American Chemical Society; reference^[31]. Copyright © 2013 Royal Society of Chemistry; reference^[32]. Copyright © 2019 John Wiley & Sons; reference^[33]. Copyright © 2020 Elsevier. PECVD: plasma-enhanced chemical vapor deposition.

2. Fabrication Technologies of Oriented Networks

To design and construct oriented networks, researchers have developed many techniques. For the one-dimensional (tubes, nanofibers) and two-dimensional (2D) (sheets, layers) nanostructures, their special structural characteristics allow them to be easily assembled into the macro-oriented networks through simple and mature technologies^[34,35]. For example, boron nitride (BN), graphene, MXenes, and graphite, which possess an inherent layered structure, can be constructed into oriented networks via self-assembly, freeze-drying, or electric field/magnetic field/force induced orientation methods^[36,37]. Based on different mechanisms, the fabrication techniques can be divided into the following categories: template-assembly method, in-situ growth method, electric/magnetic field-assisted assembly method, force-induced assembly methods, and other methods.

2.1 Template-assembly method

The template-induced self-assembly method is a commonly used technology to fabricate oriented networks. The most commonly and widely used templates include ice templates and salt templates^[4,15]. As the sacrificial templates, ice and salt can be completely removed through subsequent processes.

The ice template method, also known as the freeze casting method, utilizes ice crystals as the structural template to construct the oriented filler networks^[38-40]. The ice template method is based on the directional growth process of ice crystals to repel fill particles and align them in the direction of ice crystal growth^[39]. Specifically, during the freezing process, ice crystals grow and arrange themselves along the temperature gradient direction, thereby compressing and repelling the nanoscale particles to align along the direction of ice crystal growth^[38]. The ice crystals can be completely removed during the subsequent freeze-drying process, resulting in a pure oriented filler network. The structural features of the oriented network prepared by the ice-freezing method are usually influenced by the process factors like freezing temperature, freezing direction, and freezing rate^[39]. For example, if the freezing process occurs uniformly around the mold and proceeds from outside to inside (isotropic freezing, Figure 2a), which means there is no preferential growth direction, the fabricated network will present an isotropic structural characteristic (Figure 2d)^[41]. Therefore, in order to fabricate an oriented network, it is necessary to make the ice crystals grow along a specific direction. Unidirectional freezing and bidirectional freezing technologies have been proposed to construct the oriented networks^[42]. As shown in Figure 2b, the unidirectional freezing method is conducted by setting up a unidirectional temperature gradient to make the ice crystals grow along the direction of the temperature gradient, which leads to the formation of an ordered arrangement network (Figure 2e)^[41,43]. For the bidirectional freezing method (Figure 2c), an extra insulated plate is usually used to restrict the growth of ice crystals along the direction parallel to the temperature gradient, leading to the directional extrusion of filler particles along the temperature gradient direction (Figure 2f)^[32,44]. The ice template method is suitable for large-scale preparation of 3D oriented networks. Since the ice template can be easily removed by freeze-drying without structural collapse, it is extraordinarily popular in the preparation of functional porous composites.

Display Full Size

Download

Figure 2. (a-c) Schematic diagram of (a) isotropic freezing mechanism, (b) unidirectional freezing mechanism, and (c) bidirectional freezing mechanism. Reproduced with permission from reference^[32]. Copyright © 2019 John Wiley & Sons; reference^[41]. Copyright © 2018, Elsevier; (d-f) Structural characteristics (SEM image) of three-dimensional networks prepared by different freezing methods: (d) isotropic freezing method, (e) unidirectional freezing method, and (f) bidirectional freezing method. Reproduced with permission from reference^[32]. Copyright © 2019 John Wiley & Sons. SEM: scanning electron microscopy; PDMS: polydimethylsiloxane.

Beyond ice templating, a variety of alternative sacrificial templates have been proposed for constructing directional and oriented filler networks^[45]. These include biomass-based templates such as cotton fibers, cotton-derived porous carbon fibers, and wood, as well as inorganic templates like sodium chloride (NaCl) and liquid film-assisted gas templates^[15,46-48]. Despite the diversity of template materials, the fabrication strategy is generally similar: the sacrificial templates with directional structures are first fabricated, followed by infiltration of pre-synthesized fillers, and finally the template is removed. This process can also yield a 3D interconnected framework with a pronounced anisotropic structure. For example, Tian and co-workers fabricated a hollow BN network with an anisotropic structure by using NaCl as the sacrificial template^[49]. After being composited with epoxy resin, the resultant BN/epoxy composites exhibit anisotropic thermal conductivity and much higher thermal conductivity than the BN/epoxy composites with randomly dispersed BN.

An alternative and straightforward strategy for constructing oriented networks is to use pre-existing anisotropic materials as structural support templates, including fabrics and papers^[3,4]. For instance, Kim et al. employed hierarchical textile fabrics as templates to fabricate stacked carbon nanotube (CNT) architectures through a dip-coating process^[50]. Similarly, Cheng et al. prepared graphene-based fibers by immersing double-wrapped yarns into a graphene oxide (GO) suspension^[51]. Despite the simplicity of this method and its suitability for large-scale fabrication of architectures with tailored anisotropic morphologies, the resulting networks are largely dependent on the anisotropic scaffold for particle support, and these templates usually are preserved in the composites and not removed, which severely limits their broader applicability.

2.2 In-situ growth

In contrast to approaches that rely on the induction of template structures, the in-situ growth strategy enables the fabrication of oriented three-dimensional (3D) and self-supporting architectures through catalytic reactions conducted at elevated temperatures^[52]. A representative application of this method is the preparation of oriented carbon-based networks, including CNTs, graphene, and graphite^[53]. Typically, chemical vapor deposition (CVD) techniques, particularly plasma-enhanced chemical vapor deposition (PECVD), are employed to directly grow highly ordered architectures, such as vertically aligned CNT arrays and vertical graphene frameworks^[31,54-57]. As a typical bottom-up strategy, CVD methods enable the assembly of carbon atoms into large-scale architectures, yielding graphene, CNTs, diamond, and other highly crystalline carbon materials^[58-60]. The growth behavior of the oriented networks is governed by a variety of parameters, such as the nature and dimensions of the catalysts, template geometry, carbon precursors, atmosphere conditions, and processing temperature^[3,61]. By rationally adjusting these templates and growth conditions, a wide range of anisotropic carbon architectures, including CNT arrays, CNT yarns and films, vertically aligned graphene, layered graphene films, and other directionally organized carbon structures, can be selectively produced^[62].

For the fabrication of macroscale CNT architectures with either in-plane or out-of-plane anisotropy, the precise tuning of CVD growth conditions is necessary. In-plane aligned CNT structures typically require post-growth assembly or spinning processes^[63]. For example, Feng et al. fabricated high-quality double-walled carbon nanotube (DWCNT) films using a horizontal CVD configuration, in which acetone and ferrocene were employed as the carbon precursor and catalyst, respectively^[64]. The continuously generated DWCNT aerogel was collected by winding onto a rotating spindle oriented perpendicular to the gas flow at the reactor outlet, resulting in horizontally aligned CNT films. By contrast, vertically aligned CNT arrays with pronounced out-of-plane anisotropy can be produced through floating-catalyst or fixed-catalyst CVD methods. (Figure 3a,b) As shown in Figure 3c, a “crowding effect” induced by van der Waals interactions and high-density catalyst particles restricts the available growth space, thereby promoting vertical alignment of CNTs^[61]. Notably, such vertically aligned CNT arrays can be further converted into in-plane aligned CNT fibers through mechanical pulling or dry-spinning processes, offering an effective route to tailor CNT orientation across different length scales^[66,67].

Display Full Size

Download

Figure 3. (a) Schematic illustrations of the experimental setup; (b) The process for the growth of CNT arrays by in-situ growth via CND. Reproduced from reference^[62]. CC BY 4.0; (c) Schematic of the preparation of CNT arrays by the “crowding effect”, and relationship between CNT alignment and catalyst density (SEM images). Reproduced with permission from reference^[61]. Copyright © 2012 American Chemical Society; (d) The structural characteristics (surface and cross-section SEM images) of vertical graphene networks prepared by inductively coupled PECVD. Reproduced with permission from reference^[31]. Copyright © 2013 Royal Society of Chemistry; (e) Schematic diagrams of the growth mechanism of the vertical graphene structures via PECVD. Reproduced with permission from reference^[65]. Copyright © 2016 Elsevier. CNT: carbon nanotube; CND: carbon nanodots; SEM: scanning electron microscopy; PECVD: plasma-enhanced chemical vapor deposition; SS: stainless steel; VACNTs: vertically aligned carbon nanotubes; VGNs: vertical graphene nanosheets.

CVD based methods also enable the fabrication of oriented graphene-based architectures. High-quality graphene can be grown with the assistance of metal catalysts, and anisotropic graphene structures can be directly obtained by the rational design of catalyst templates with directional morphologies^[68-70]. Furthermore, vertical graphene nanosheets (VGNs), featuring perpendicular orientation and open cellular frameworks, can be synthesized via PECVD^[71,72]. As illustrated in Figure 3d, Yang et al. fabricated an oriented network of vertically erected graphene walls by direct growth using the plasma-enhanced CVD at 900 °C^[31]. This fabrication mechanism of VGNs via the PECVD is schematically illustrated in Figure 3e. In the early stage of deposition, the thin carbon buffer layers are initially formed on the substrate surface, typically growing parallel to the substrate plane. Owing to the continuous bombardment of Ar ions during the plasma-assisted deposition process, structural defects are inevitably introduced into the buffer layers. These defect sites provide energetically favorable locations for the subsequent nucleation of vertical graphene sheets. As the growth continues, graphene nuclei preferentially emerge from these defective regions and gradually develop into vertically aligned nanosheets. Meanwhile, additional graphene layers tend to grow along the existing vertical sheets, resulting in a gradual thickening of the graphene structures. By carefully regulating the growth parameters, the number of graphene layers can typically be confined to fewer than ten, enabling the VGNs to retain distinctive physical and chemical properties comparable to those of few-layer graphene^[65]. The structural features of VGNs can also be controlled by applying an external electric field during the growth process. For example, Huang et al. prepared the parallelly aligned VGNs by using inductively coupled PECVD in the presence of metal inducers, where the introduced electric field plays a critical role in directing graphene growth^[73]. Through such approaches, vertically and horizontally aligned graphene architectures can be selectively constructed, enabling anisotropic properties along multiple axes.

Overall, a wide variety of anisotropic carbon architectures, including CNT arrays, CNT films and fibers, graphene sheets, VGNs, as well as more complex structures such as graphite networks and carbon nanofibers, can be realized through CVD-based strategies by tailoring reactor configurations and growth conditions. Nevertheless, despite the capability of producing highly crystalline carbon nanomaterials with minimal structural defects, the high processing cost and limited scalability of CVD techniques remain critical barriers for large-scale industrial implementation.

2.3 Electric and magnetic field-assisted assembly

Electric or magnetic field induced orientation is another important technological pathway for constructing an oriented network, which is suitable for the particles possessing intrinsic or tailored field responsiveness^[74-76]. Under the influence of an external field, the nanoparticles preferentially orient along a designated direction, thereby constructing oriented networks. Conductive nanostructures readily respond to external electric or magnetic fields, which exert directional forces that drive their alignment along the field direction^[77]. On this basis, well-organized conductive networks can be constructed through field-assisted assembly using either electric or magnetic fields. The degree of orientation can be effectively regulated by adjusting the field strength, with stronger fields generally leading to higher alignment levels^[78].

Representative studies have demonstrated the versatility of this strategy. For instance, Feng’s group fabricated a 3D hierarchical carbon framework by assembling carbon fibers and vertically aligned carbon nanotubes via an electrostatic flocking process (Figure 4a)^[25]. In another approach, Fischer et al. employed strong magnetic fields to induce the alignment of single-walled carbon nanotubes (SWCNTs), producing highly oriented SWCNT films through filtration or deposition from a suspension^[80]. Similarly, Tumpane et al. reported the alignment of SWCNTs under extremely weak magnetic fields, highlighting that a notable advance in simple electromagnetic setups can enable effective physical manipulation and tuning of the intrinsic properties of nanotubes (Figure 4b)^[74]. Moreover, by employing modified nanotubes with enhanced solubility, the system better preserves their inherent high aspect ratio, suppresses aggregation into bundles, and instead promotes the formation of more organized rope-like structures. In another example, Liu et al. realized patterned liquid-metal conductive networks by dispersing magnetic particles and applying a magnetic field to drive the controlled motion of liquid metal droplets^[81].

Display Full Size

Download

Figure 4. (a) Schematic illustrations of the preparation of oriented networks by an electric field-assisted method. Reproduced with permission from reference^[25]. Copyright © 2018 Elsevier; (b) Orientation of SWCNTs under magnetic fields. Reproduced from reference^[74]. CC BY 4.0; (c) Schematics of magnetic field induced graphene flake orientations. Reproduced with permission from reference^[79]. Copyright © 2018 John Wiley & Sons. SWCNTs: single-walled carbon nanotubes; CF-VACNTs: carbon fibers and vertically aligned carbon nanotubes.

Compared to one-dimensional materials (CNT, carbon fiber, etc.), the alignment of 2D structured particles by an electrical or magnetic field is more challenging because of their additional degrees of orientational freedom compared to those of rod-like nanostructures^[79,82]. To realize the planar alignment of graphene sheets, Lin et al. reported a magnetic field induced method by using a rotating magnetic field produced by a pair of small magnets based on neodymium, iron, and boron (NdFeB)^[79]. As illustrated in Figure 4c, the application of a rotating magnetic field can induce nearly ideal planar alignment in graphene suspensions. In comparison with the partially aligned sheets obtained under a static magnetic field, this dynamic alignment results in a substantially higher order parameter, as well as markedly enhanced birefringence and anisotropic absorption/transmission properties.

Notably, the alignment efficiency of nanostructures under magnetic fields can be significantly enhanced by incorporating magnetic components. For example, CNTs coated with magnetic nanoparticles can achieve pronounced orientation even under relatively low magnetic field strengths^[83]. Moreover, taking advantage of the electric/magnetic field-induced ordering of liquid crystal molecules, carbon nanomaterials dispersed in liquid crystal matrices can also be aligned under electric or magnetic fields^[84]. Importantly, as field-induced alignment does not involve chemical modification, this approach allows conductive nanomaterials to preserve their intrinsic structural integrity and functional properties.

2.4 Force-induced assembly

Force-induced alignment represents another widely adopted strategy for constructing oriented conductive networks (Figure 5). Among these approaches, pre-stretching followed by release has been extensively employed to generate anisotropic conducting architectures^[33,88]. As illustrated in Figure 5a, CNT networks preferentially orient along the applied stretching direction^[33]. During the stretching process, CNT films gradually transform from an initially random arrangement into a more compact and highly aligned microstructure along the tensile axis. This enhanced orientation is primarily associated with the intrinsic one-dimensional geometry of nanotubes. Consequently, pre-stretching techniques have been broadly extended to fabricate fiber-based oriented networks composed of CNTs, carbon fibers, and metallic nanowires.

Display Full Size

Download

Figure 5. (a) Schematics of the fabrication procedure of an oriented CNT film via the pre-stretching process. Reproduced with permission from reference^[33]. Copyright © 2020 Elsevier; (b) Shear force induced assembly of aligned CNT films by the sliding coating method. Reproduced with permission from reference^[85]. Copyright © 2018 Royal Society of Chemistry; (c) Preparation of an aligned AgNW network by the induction of centrifugal inertia force. Reproduced from reference^[86]. CC BY 4.0; (d) Schematics of the fabrication of layered structures by shear-flow-induced alignment of nanosheets. Reproduced with permission from reference^[87]. Copyright © 2020 Springer Nature; (e) Schematics of the fabrication of composites with an oriented graphene network via the 3D-printing method. Reproduced with permission from reference^[30]. Copyright © 2021 American Chemical Society. CNT: carbon nanotube; AgNW: silver nanowire; CSA: chlorosulfonic acid; PTFE: polytetrafluoroethylene.

In a similar manner, shrinkage-assisted strategies can convert initially isotropic conductive networks into anisotropic structures^[89,90]. For example, Ham et al. produced corrugated silver nanowire networks by pre-stretching a polydimethylsiloxane (PDMS) substrate followed by gradual relaxation to its original, unstressed state^[91]. Likewise, Khine et al. fabricated wrinkled CNT films using shape-memory polymers as shrinkage substrates^[92]. Thermally responsive shrink films, which undergo uniform contraction upon heating, have also been widely utilized to create distinctive wrinkled oriented architectures^[93]. Furthermore, uniaxial anisotropic shrinkage can be realized by fixing one side of the shrinking film, enabling the formation of graphene films with uniaxially crumpled morphologies^[94].

Shear force and centrifugal force induced assembly is also a commonly used method for constructing oriented networks. For example, Jiang’s group utilized the shear force provided by the liquid flow to fabricate GO based nanocomposite films with highly ordered layered structures^[87]. As shown in Figure 5d, shear-flow induces the oriented alignment of 2D nanosheets at an immiscible hydrogel/oil interface. With appropriate choices of flow rate, the distance between adjacent syringes, and the moving speed of the hydrogel, the multiple sources of solution from arrayed syringes rapidly spread and fused to form a uniform superspreading solution layer at the oil/hydrogel interface. This superspreading process provides a strong shear flow force, which is highly desirable for the in-plane stacking of 2D nanosheets. In another report, a scalable shear-induced alignment method was introduced for producing highly oriented CNT films and densely packed CNT fibers (Figure 5b)^[85]. In this process, shear forces generated between two glass microscope slides were employed to align CNTs along a single direction. Following immersion of the glass substrates into a coagulation bath to remove residual acid solvent, CNT films with uniform uniaxial alignment along the slide length were obtained. Importantly, the film thickness could be readily adjusted by modulating the CNT concentration in the precursor solution as well as the normal force applied between the glass slides. In another study, Zhao et al. reported a centrifugal inertia force induced method to construct the directionally aligned silver nanowire network^[86]. As shown in Figure 5c, the centrifugal inertial force generated from a high-speed rotating roller induces the alignment of silver nanowires (AgNWs) and results in a transparent and conductive film. This strategy effectively mitigates the high contact resistance commonly observed in randomly distributed networks, enabling high electrical conductivity even at relatively low AgNW network densities.

Beyond stretching, shrinking, shearing, and centrifugating, extrusion-based techniques such as spinning and 3D printing provide an additional route to induce orientation^[95,96]. In these processes, the shear and extensional forces generated at the nozzle promote the alignment of anisotropic conductive nanoparticles along the extrusion or fiber direction, offering a scalable pathway for constructing directionally ordered conductive networks. For example, Guo and co-workers fabricated an asymmetrically aligned structure of graphene filled thermoplastic polyurethane composites by the 3D printing technology (Figure 5e)^[30]. The as-printed composite with a vertically aligned structure demonstrates a high through-plane thermal conductivity (over 12 W·m^-1·K^-1 at 45 wt % graphene content), much higher than that of the horizontally printed composites and surpassing most traditional particle reinforced polymer composites. Another force-inducing process is melting spinning of composites. In this process, fillers with an aspect ratio significantly greater than 1 (such as CNTs) are oriented in the direction of the draw. This is reflected in the fact that electrical conductivity decreases significantly as the take-off velocity increases and is much higher than the electrical percolation threshold in the composite^[97-100]. Even during the compressing molding process of CNT/polymer composites, the filler becomes oriented in the in-plane direction, which is reflected in higher electrical conductivity in the in-plane direction compared to the through-plane direction^[101]. The effect is much more pronounced in melt film extrusion. Here, differences can even be observed in all three spatial directions. The electrical conductivity is highest in the direction of extrusion, followed by across the direction of extrusion and through the film, whereby the take-off velocity plays an important role in the filler orientation^[102,103].

2.5 Other methods

Besides the strategies discussed above, several additional methodologies have also been reported for organizing filler nanoparticles into directional architectures with diverse geometries. As an illustrative example, a simple yet scalable expansion-based approach has been proposed for constructing highly oriented graphite frameworks (HOGFs)^[26]. Drawing inspiration from conventional preparation routes for expanded graphite and GO, highly directional graphite networks were fabricated via a one-step acid expansion process conducted at room temperature, without requiring any chemical functionalization or auxiliary treatments (Figure 6a). Structural characterization using X-ray diffraction revealed pronounced anisotropy in different crystallographic planes, confirming the highly oriented nature of the resulting graphite networks. Building upon this work, the fabrication process was further refined into a two-step route (Figure 6b), integrating chemical intercalation through acid expansion with subsequent thermal expansion driven by high-temperature gasification^[18,27]. The thermal expansion process can increase the porosity of graphite networks while preserving their structural characteristics of orientation. Specifically, commercially available graphite paper, a low-cost graphite-based product characterized by a 2D stacked configuration, was employed as the precursor for fabricating these directional graphite frameworks. This work represents the first demonstration of large-scale fabrication of aligned graphite networks from commercial graphite paper using a facile expansion strategy. The method offers multiple advantages, including high expansion efficiency, low material cost, preservation of crystalline integrity, and excellent scalability. Overall, this straightforward and economical approach provides a promising pathway for constructing oriented filler networks.

Display Full Size

Download

Figure 6. (a) Schematic diagram of the process for preparing oriented graphite networks by the one-step expansion method. Reproduced with permission from reference^[26]. Copyright © 2022 Elsevier; (b) Mechanism and expansion result diagrams of the two-step expansion method for preparing an oriented graphite network. Reproduced with permission from reference^[18]. Copyright © 2022 Royal Society of Chemistry; (c) Schematic process of the HPF method to fabricate oriented graphene networks. Reproduced from reference^[104]. CC BY 4.0; Reproduced with permission from reference^[105]. Copyright © 2023 Springer Nature. HPF: hydroplastic foaming; HOG: highly oriented graphite; GO: graphene oxide; GAs: graphene aerogels.

In another report, Gao’s group developed a hydroplastic foaming (HPF) method to fabricate graphene aerogels with an oriented structure and 3D porous curvatures (Figure 6c)^[104,105]. Similar to the expansion method, tightly stacked and highly oriented GO films were used as the raw material. Firstly, the GO films were put into a hydrazine (30 wt%) solution to prepare GO aerogels by HPF. Then, the obtained GO aerogels were washed with ethanol to remove residual water and hydrazine. In the last step, the cleaned GO aerogels were chemically reduced at 90 °C for 12 h and annealed at 1,600 °C for 1 h under an argon atmosphere. Finally, a graphene network was obtained. In the fabrication process, the gas foaming mainly occurs between the graphene sheets without disrupting the oriented structure of intrinsically anisotropic GO films. The surface morphologies of the obtained graphene network exhibited dense structures with less vertical pores. Additionally, the highly ordered and oriented pore structure of the graphene aerogels can be controlled through the foaming time; thus, the pore size can be tailored from several micrometers to hundreds of micrometers. Attributed to the highly ordered and oriented porous structure, this graphene aerogel exhibited excellent electromagnetic shielding and sensing performance.

In addition to expansion-based strategies, anisotropic carbon conductive networks can also be generated on pre-existing carbon monoliths through laser engraving or oxygen plasma etching. For instance, Ziaie et al. fabricated directionally aligned carbon structures via laser-induced carbonization of polymer films, in which carbon flakes were preferentially oriented along the laser scanning direction^[106]. In another approach, Lu and co-workers deposited CNTs onto a self-adhesive elastomer substrate by spray coating, followed by rolling the film into a compact configuration to produce a spirally layered architecture within the composite^[107]. The resulting fibrous structure exhibited outstanding electrical conductivity along the axial direction. Furthermore, graphene sheets can self-assemble into films with directional alignment by exploiting their intrinsically anisotropic 2D geometry^[108,109]. Collectively, these techniques expand the toolbox for designing conductive networks with tailored anisotropic architectures, and continued advances in manufacturing technologies are expected to yield even more efficient routes for constructing directionally ordered structures.

2.6 Comparison of fabrication methods

Different fabrication technologies exhibit distinct advantages and limitations in terms of scalability, cost, structural controllability, and applicability. Template-assisted methods, particularly ice-templating, are among the most widely adopted approaches due to their relatively low cost, simple processing, and suitability for large-scale fabrication of 3D oriented architectures. However, the structural precision is largely governed by freezing conditions and template design. In contrast, in-situ growth methods such as CVD and PECVD offer excellent structural controllability and enable the fabrication of highly ordered architectures with superior crystallinity and minimal defects. Nevertheless, the high equipment requirements, elevated processing temperatures, and relatively low throughput significantly increase production costs and limit industrial scalability. Electric and magnetic field-assisted assembly techniques provide precise control over filler orientation and preserve the intrinsic properties of nanomaterials, but they are generally restricted by the requirement for field-responsive fillers and the difficulty of scaling up large-volume production. Force-induced assembly methods, including stretching, shearing, spinning, and 3D printing, are attractive owing to their compatibility with existing manufacturing processes and good scalability. However, achieving highly uniform orientation over large dimensions remains challenging. Other emerging approaches, such as expansion-based and HPF methods, offer a promising combination of low cost, structural anisotropy, and scalability, although their applicability is currently limited to specific material systems. Overall, the selection of fabrication methods should be determined by balancing structural controllability, production cost, and scalability according to the requirements of the targeted applications.

3. Applications of Oriented Networks in Intelligent Sensing

The rapid development of wearable electronics, soft robotics, and human–machine interfaces has stimulated increasing demand for flexible and multifunctional sensors^[110]. In this context, advances in materials science have enabled the design of diverse sensing structures tailored for different application scenarios^[111-113]. Among them, piezoresistive sensors have attracted extensive attention owing to their skin-like mechanical compliance, simple device configuration, low cost, and excellent compatibility with flexible substrates^[114-117]. These features make them promising candidates for applications such as health monitoring, motion tracking, human–computer interaction, and flexible electronics.

Piezoresistive sensing relies on monitoring resistance variations induced by mechanical deformation^[118]. Compared with conventional isotropic conductive networks, introducing oriented or anisotropic microstructures as the sensing layer has been demonstrated to be an effective strategy to enhance sensing performance^[27,119,120]. Such oriented architectures not only amplify resistance changes under external stimuli, leading to higher sensitivity, but also endow sensors with directional selectivity, enabling discrimination of deformation direction and magnitude^[4,43,121]. Therefore, oriented conductive networks represent a powerful design paradigm for next-generation intelligent sensors.

3.1 Improved sensitivity

Over the past decades, piezoresistive sensors based on electronically conductive fillers have achieved remarkable progress and widespread application. The sensitivity of these sensors is commonly quantified by the dimensionless gauge factor (GF), defined as^[4,122]:

(1)$GF=\left(\Delta R/R_0\right)/\epsilon$

where ΔR is the resistance change, R₀ is the initial resistance, and ε is the applied strain. Improving GF has long been a central goal in the development of high-performance strain sensors.

For electronically conductive networks, high sensitivity is generally achieved by tailoring the microstructure of conductive fillers^[118]. The underlying mechanisms mainly involve electron tunneling effects and the formation or evolution of microcracks during deformation, which induce pronounced resistance variations even under small strains. Sensors with ultra-high GF can even function as flexible electromechanical switches, enabling reversible ON/OFF control of electrical circuits through mechanical deformation^[27,28,123]. Compared with conventional mechanical switches that rely on physical contacts or gates, such flexible switches contain no moving parts and can be seamlessly integrated into soft robotic systems and wearable intelligent control platforms^[123].

Recent studies have further demonstrated that orientation engineering of conductive networks provides an additional and effective degree of freedom for regulating both sensitivity and sensing range^[124]. In our recent work, the strain-sensing behavior was systematically tuned by controlling the alignment of graphite layers within the conductive network as well as the macroscopic dimensions of the sensing element^[27]. As depicted in Figure 7a, when graphite sheets are preferentially oriented parallel to the stretching direction, neighboring layers overlap within the plane, forming continuous conductive networks with enlarged contact interfaces. This configuration effectively lowers junction resistance and enables gradual modulation of interlayer contacts under applied strain, resulting in an extended sensing range. In contrast, when the graphite layers are aligned perpendicular to the stretching direction, electrical conduction depends on contacts between vertically arranged, wrinkled sheets. Under relatively low strain, adjacent layers can readily separate, causing a sharp rise in resistance and yielding an exceptionally high ON/OFF ratio. This switching behavior arises from a “contact–disconnection” mechanism: upon stretching, the increasing interlayer distance progressively disrupts conductive pathways, leading to a transition from the “ON” to the “OFF” state. Upon release, the elastic PDMS matrix restores structural integrity and re-establishes electrical connections, enabling recovery to the conductive state. Moreover, the macroscopic width of the conducting strips plays a critical role in determining the performance of sensors and switches. These results demonstrate that sensing performance can be precisely tailored through the synergistic design of oriented microstructures and macroscopic geometry, offering a scalable and versatile strategy for performance optimization in printed or spun sensing architectures.

Display Full Size

Download

Figure 7. (a) Schematic diagram of the process for preparing highly sensitive sensors and switches based on highly oriented graphite networks; (b) Application of the obtained sensors in motion monitoring; (c) Application of the obtained switches in back posture monitoring. Reproduced with permission from reference^[27]. Copyright © 2023 Elsevier; (d) Application of the switch to soft reconfigurable logic gates to control an led and heart-shaped DEAs; (e) Mechanism of the soft oscillator, and the equivalent electrical circuit of the soft oscillator at a controlled input voltage. Reproduced from reference^[28]. CC BY 4.0. DEAs: dielectric elastomer actuators; HAGF: highly-anisotropic graphite framework; PDMS: polydimethylsiloxane; LED: light-emitting diode; MOSFET: metal-oxide-semiconductor field-effect transistor.

Building on these characteristics, the flexible and highly sensitive sensors and switches enable a wide range of applications in wearable electronics, including motion tracking, human-machine interaction, and intelligent control systems. As shown in Figure 7b,c, the sensor based on an oriented graphite network can be integrated with a smartphone and allows real-time monitoring of human activities such as resting, walking, and running. The developed switch can be employed for posture correction by providing timely feedback on improper sitting behaviors. In another demonstration (Figure 7d,e), such flexible switches were further utilized to construct soft, reconfigurable logic gates, which can effectively control light-emitting diodes (LEDs) and drive heart-shaped dielectric elastomer actuators, highlighting their potential in soft robotics and adaptive electronic systems.

Despite these advantages, sensors relying on electronically conductive networks often suffer from pronounced nonlinearity. The rapid initiation and propagation of cracks, together with strong tunneling-dominated conduction, frequently result in nonlinear output–strain relationships, especially at large deformations. Such nonlinearity necessitates complex signal processing for accurate strain quantification and restricts practical applicability. In addition, many high-GF sensors exhibit a limited working strain range, further constraining their use in scenarios involving large or repeated deformation.

3.2 Directional sensing ability

Most conventional flexible sensors are designed with isotropic conductive networks, resulting in strongly coupled responses to external stimuli^[4,122,125]. As a consequence, resistance changes induced by deformation in different directions are often indistinguishable, limiting the ability to resolve multidirectional strain. This shortcoming significantly restricts their application in advanced motion detection, soft robotics, and intelligent human–machine interfaces. In contrast, anisotropic conductive networks with controlled orientation and morphology offer an effective solution for directional sensing^[33,121]. Directional selectivity is commonly evaluated by comparing GF values measured along different directions. Table 1 summarizes some representative multi-directional sensors based on oriented conductive networks. A larger discrepancy in GF indicates stronger anisotropy and superior capability to discriminate the direction of applied mechanical stimuli. Oriented networks possess unique characteristics for preparing directional sensors due to their special decoupling structure, which can avoid the interference among different deformations for realizing multiaxial sensing^[4].

A wide range of oriented conductive structures, including aligned AgNWs, oriented CNT films, and graphene-based cellular frameworks, have been employed to fabricate in-plane strain sensors with direction-dependent responses (Table 1)^[4,33]. In such systems, deformation parallel and perpendicular to the alignment direction induces fundamentally different microstructural evolution, leading to distinct resistance changes. Generally, a higher degree of network orientation corresponds to greater directional selectivity. For example, highly aligned conductive arrays often exhibit pronounced sensitivity to strain applied parallel to the alignment direction, while showing negligible response to transverse deformation^[23]. Lee et al. fabricated flexible strain sensors based on overlapped aligned CNT arrays that displayed stable and repeatable resistance variations under parallel stretching over thousands of cycles^[137]. Anisotropic honeycomb-like architectures reinforced with soft matrices also have been demonstrated to enable directional discrimination in bending and stretching modes^[138]. In another representative report, aligned fibrous conductive networks exhibit sharply contrasting responses under perpendicular and parallel bending^[127]. In this sensor, perpendicular deformation disrupts the continuity of conductive pathways, resulting in a dramatic resistance increase, whereas parallel bending preserves network connectivity due to the corrugated morphology along the fiber direction. The realization of sensing multidirectional strains is essential to understanding the nature of complex motions. As shown in Figure 8a,b,c,d, Lee et. al fabricated a multidirectional strain sensor by constructing highly aligned, anisotropic carbon nanofiber films^[23]. Based on these highly oriented conducting networks, stretchable sensors with in-plane anisotropic sensing capability can be realized. As illustrated in Figure 8b, applying tensile strain along the fiber orientation causes progressive fracture of the carbon fibers, disrupting the conductive network; upon release, the fractured interfaces reconnect and restore conductivity. In contrast, deformation perpendicular to the fiber alignment exerts a much weaker influence on the effective conductive pathways. Such anisotropic deformation mechanisms provide a clear physical basis for directional sensing. Moreover, by integrating this anisotropic sensor in a cross-ply configuration, simultaneous brightness modulation of two LEDs (Figure 8c) and discrimination of complex, multi-degree-of-freedom human joint motions (Figure 8d) can be realized. Highly oriented networks have also been employed to fabricate uniaxial strain sensors that are ultrasensitive along one direction while remaining nearly insensitive along the orthogonal direction (Figure 8e)^[130].

Display Full Size

Download

Figure 8. (a) Schematic diagram of the process for the fabrication of multidirectional strain sensors based on anisotropic carbon nanofiber films; (b) Schematic illustration of the sensing mechanism of the sensor with the fiber parallel and perpendicular to the loading direction; (c) Utilization of the multidirectional sensor to control the brightness of two LEDs simultaneously; (d) Detection of multi-degrees-of-freedom neck motions using the multidirectional sensor. Reproduced with permission from reference^[23]. Copyright © 2019 John Wiley & Sons; (e) Schematic diagram of the morphology evolution of the strain sensor based on aligned CNTs networks. Reproduced with permission from reference^[129]. Copyright © 2018 Elsevier. LEDs: light-emitting diodes; CNTs: carbon nanotubes; ACNF: anisotropic carbon nanofiber; PDMS: polydimethylsiloxane.

Beyond uniaxial sensing, in-plane multiaxial strain sensors can be realized by stacking or cross-aligning multiple oriented layers, forming sandwich-like architectures capable of independently resolving strain components along different axes^[66,139]. More sophisticated designs further enable detection of arbitrary in-plane strain directions by integrating multiple anisotropic sensing elements with different orientations^[66]. Additionally, pre-stretching strategies can introduce controlled microcracks within oriented networks, allowing sensors to detect both tensile and compressive strains with enhanced sensitivity^[140]. Through rational design of oriented conductive architectures, strain sensors can thus achieve accurate sensing of both the magnitude and direction of complex deformation vectors.

Oriented and anisotropic conductive networks provide a powerful framework for advancing intelligent sensing technologies. By simultaneously enhancing sensitivity and enabling directional discrimination, these architectures address key limitations of conventional isotropic sensors. Future efforts focused on integrating diverse oriented microstructures, improving linearity and signal decoupling, and expanding multifunctionality will further broaden the application scope of flexible sensors in complex, real-world environments.

Overall, the sensing performance of oriented conductive networks is strongly dependent on their structural organization. Compared with randomly distributed conductive fillers, aligned conductive pathways provide more efficient and predictable electron transport channels, resulting in improved signal stability and sensitivity. More importantly, the anisotropic architecture can amplify resistance variations induced by external stimuli, such as strain, pressure, bending, or deformation. During mechanical loading, the directional conductive network undergoes more pronounced changes in interparticle contact, tunneling distance, and conductive pathway continuity, leading to enhanced signal responsiveness. Furthermore, the orientation degree and network density directly influence the balance between sensitivity and sensing range. Therefore, the sensing behavior of oriented conductive networks is not solely determined by material composition but is strongly governed by the structural arrangement of conductive fillers and the evolution of conductive pathways under external stimuli.

4. Application of Oriented Conductive Networks in Electromagnetic Shielding

With the rapid advancement of electronic technologies such as the Internet of Things, 5G communication, and AI, EMI and electromagnetic pollution have become increasingly severe issues over the past decades^[141-144]. The proliferation of compact, high-frequency, and densely integrated electronic devices places stringent demands on electromagnetic shielding materials. In particular, lightweight, flexible, and low-cost materials with high EMI shielding effectiveness (SE) are highly desired for next-generation portable, wearable, and flexible electronic systems^[142].

To mitigate undesirable electromagnetic radiation, a wide variety of EMI shielding materials have been explored^{[18,20,145-147]}. Compared with conventional metallic shields, polymer-based conductive composites have emerged as attractive alternatives due to their low density, mechanical flexibility, corrosion resistance, scalable fabrication, and cost-effectiveness^[148-150]. These advantages make them well-suited for the miniaturization and structural integration required in modern electronic and communication equipment. A diverse range of conductive fillers, including carbonaceous materials (e.g., graphene, CNTs, graphite, carbon fibers, and carbon black)^{[18,20,151-153]}, metallic conductors (e.g., AgNWs, metal nanoparticles, metal frameworks, and liquid metals)^[154-156], and 2D MXenes (e.g., Ti₃C₂T_x, Mo₂TiC₂T_x, and Mo₂Ti₂C₃T_x)^[157,158], have been incorporated into polymer matrices to construct EMI shielding composites. These fillers can be assembled into various morphologies, such as films, foams, aerogels, arrays, fibers, textiles, and braided architectures^[95,159]. In conventional designs, emphasis is often placed on achieving uniform dispersion and forming percolated conductive networks, which are essential for enabling effective EMI shielding.

Beyond filler type and loading, it has been well established that electrical conductivity, microstructural organization, and morphological architecture play decisive roles in determining EMI shielding performance^[18,160,161]. In general, the total EMI SE arises from the combined contributions of reflection, absorption, and multiple reflections within the material^[20]. The presence of abundant internal interfaces in porous conductive networks significantly increases the path length of incident electromagnetic waves, leading to enhanced attenuation. In the last decades, anisotropic and oriented conductive networks have attracted growing interest as a powerful strategy to further improve EMI shielding performance^[162-164]. Compared with isotropic networks formed by randomly dispersed fillers, oriented structures with orderly arranged conductive pathways and pores provide a higher density of internal interfaces and reflective planes^[18]. These anisotropic features promote multiple scattering and repeated reflections of electromagnetic waves, thereby enhancing energy dissipation through ohmic loss and interfacial polarization. Moreover, oriented conductive networks facilitate more efficient utilization of filler–polymer interfaces, which further strengthens electromagnetic wave attenuation^[165].

Representative examples have demonstrated the advantages of oriented architectures in EMI shielding. For instance, Wei et al. fabricated polyimide composite films containing highly aligned graphene networks, which exhibited an EMI SE of up to 63 dB at a thickness of only 60 μm, highlighting the efficiency of orientation-induced conductive pathways (Figure 9a,b,c,d)^[166]. In these systems, highly oriented cell-wall-like architectures simultaneously enhanced electrical conductivity and interfacial polarization, enabling effective regulation of EMI shielding behavior. The benefits of orientation are closely related to the formation of continuous and preferentially aligned conductive pathways. Oriented networks reduce charge transport barriers and improve electrical conductivity along specific directions, which directly enhances reflection and absorption losses. At the same time, the anisotropic arrangement of conductive layers or pores increases the probability of multiple internal reflections, leading to exponential decay of electromagnetic wave intensity within the material^[166]. In another study, Nyström’s group utilized 1D cellulose nanofibrils to assist in the fabrication of microhoneycomb-like MXenes aerogels with oriented biomimetic cell walls. This hybrid aerogel showcased an EMI SE of up to 74.6 dB, specific shielding effectiveness (SSE) as high as 30,660 dB cm³ g^-1, and SSE/t achieving 189,400 dB cm² g^-1, exceeding that of other MXene-based or other shielding architectures. In particular, they found that the presence of highly oriented biomimetic cell walls effectively governs the EMI shielding performance through the adjustment of their orientation angle relative to the electric field direction of the incident EM waves (Figure 8e,f,g,h)^[167].

Display Full Size

Download

Figure 9. (a) Schematic diagram of the process for the fabrication of graphene films based on the SCC method; (b) Prepared freestanding graphene film; (c) Cross-sectional SEM image of graphene films; (d) Schematic of the EMI shielding mechanism based on the aligned graphene film and the EMI shielding performance of PG films at different thicknesses. Reproduced with permission from reference^[166]. Copyright © 2020 John Wiley & Sons; (e) EMI shielding mechanism of the MXene/CNF hybrid aerogels with oriented cell walls; (f) Longitudinal plane structure of the MXene/CNF hybrid aerogels; (g) Transverse (SE_T, SE_A, and SE_R) and longitudinal (L-SE_T, L-SE_A, and L-SE_R) EMI shielding performance changes at a 10 GHz frequency; (h) EMI shielding performance at 10 GHz for the honeycomb-like porous architectures at various angles between the oriented cell direction and the electric field direction of incident EM waves. Reproduced from reference^[167]. CC BY 4.0. SCC: scanning centrifugal casting; SEM: scanning electron microscopy; EMI: electromagnetic interference; PG: pristine graphene; CNF: carbon nanofiber; SE: shielding effectiveness.

The superior EMI shielding performance of oriented conductive networks originates from their ability to regulate electromagnetic wave propagation through structural engineering. Unlike isotropic conductive networks, aligned architectures introduce anisotropic electrical conductivity and dielectric characteristics, which significantly affect reflection, absorption, and multiple-reflection processes. The orientation direction relative to the incident electromagnetic wave can alter the interaction between electromagnetic fields and conductive pathways, thereby influencing SE. In addition, highly ordered conductive networks facilitate the formation of continuous electron transport channels, enhancing conduction loss and dielectric polarization. Hierarchical pore structures and aligned interfaces can further increase electromagnetic wave scattering and prolong propagation pathways within the material, resulting in greater energy dissipation. Consequently, the enhanced EMI shielding performance arises from the synergistic effects of electrical conductivity optimization, anisotropic architecture, and multiscale structural regulation.

In our recent work, the influence of network orientation on EMI shielding was systematically investigated using graphite/polymer composites with HOGF^[18]. By cutting the HOGF/PDMS composites along different directions, samples with vertically and horizontally oriented graphite layers were obtained. Due to the intrinsic anisotropy of the conductive framework, the EMI shielding performance strongly depended on the incident direction of electromagnetic waves. When electromagnetic radiation impinged on composites with vertically stacked graphite layers, the surface impedance mismatch with air was reduced owing to the lower surface conductivity, resulting in moderate reflection. Once entering the material, electromagnetic waves were repeatedly reflected between parallel graphite layers, generating multiple coherent reflections that dramatically attenuated wave intensity. Although vertically oriented layers allowed a small fraction of electromagnetic waves to transmit through conductive channels, the overall SE reached as high as 96 dB, significantly outperforming the counterpart with horizontally oriented graphite layers.

The EMI shielding performance of materials is closely related to their electrical conductivity, structural characteristics, and geometric parameters, such as density, thickness, and porosity. Effective EMI shielding is achieved through the combined contributions of reflection, absorption, and multiple internal reflections of incident electromagnetic waves. Therefore, tailoring the internal architecture of materials provides an effective strategy for regulating electromagnetic wave propagation and enhancing shielding performance. The introduction of oriented conductive architectures can significantly improve the utilization efficiency of conductive fillers by promoting directional charge transport, increasing electromagnetic wave attenuation pathways, and optimizing impedance matching. When designing oriented conductive architectures, multiple factors, including the intrinsic characteristics of the fillers, such as their dimensionality (e.g., one-dimensional or 2D structures), size, aspect ratio, morphology, and electrical conductivity, as well as the arrangement of the filler particles within the aligned framework, play a great role in the EMI shielding performance of materials. Furthermore, the structural features of the resulting oriented networks, including the orientation direction relative to the incident electromagnetic wave, orientation degree, network density, pore structure, and hierarchical microstructural characteristics, can substantially influence the reflection, absorption, and multiple-reflection processes of electromagnetic waves. Through the rational selection of conductive fillers and precise engineering of oriented architectures, it is possible to maximize EMI SE while minimizing filler loading and material thickness.

These results clearly demonstrate that oriented conductive networks can fundamentally alter the balance between reflection, absorption, and multiple reflections, enabling superior EMI shielding efficiency compared with isotropic counterparts. By rationally designing oriented architectures based on filler characteristics and structural parameters, it is therefore possible to achieve high-performance EMI shielding with reduced filler loading and minimal material thickness.

5. Application of Oriented Network Structures in Thermal Management

With the continuous miniaturization, high integration, and increasing power density of modern electronic devices, efficient thermal management has become a critical factor governing device reliability, performance stability, and service lifetime^[2,168,169]. In nano- and micro-electronic systems, localized heat accumulation can easily lead to thermal runaway, accelerated material degradation, and premature failure if the generated heat is not dissipated effectively.

In most electrically insulating materials, especially polymers, heat transport is dominated by phonons, which are highly sensitive to lattice disorder and interfacial scattering^[3,15]. Although polymers possess attractive characteristics such as electrical insulation, low density, mechanical flexibility, ease of processing, and low cost, their intrinsically amorphous molecular structures induce severe phonon scattering, resulting in very low thermal conductivities (typically 0.1-0.5 W·m^-1·K^-1)^[170]. Consequently, neat polymers behave as thermal insulators and are inadequate for advanced thermal management applications. Constructing oriented thermally conductive networks within polymer matrices has emerged as an effective strategy to overcome this limitation^[15,171]. By aligning high-thermal-conductivity fillers along preferred directions, phonon transport pathways with reduced interfacial resistance can be established, thereby maximizing heat conduction along the alignment direction. In practical thermal management systems, such anisotropic heat transport can be exploited in two major application scenarios:

(1) in-plane heat spreading, where rapid lateral heat dissipation is required (reduces the temperature of the hot spot and quickly equalizes the heat), and,

(2) through-plane heat conduction, which is essential for efficient heat transfer across interfaces (between heat-generating devices and heat sinks).

Accordingly, oriented networks play a pivotal role in both heat spreaders (in-plane high thermal conductivity) and thermal interface materials (TIMs) (through-plane high thermal conductivity).

5.1 In-plane high thermal conductivity: Heat spreaders

Heat spreaders are designed to redistribute localized heat from hot spots over a larger area, thereby reducing peak operating temperatures^[172-174]. In integrated circuit packaging, conventional heat spreaders typically rely on metals such as copper or artificial graphite foils. However, the increasing demand for lightweight, flexible, and high-performance thermal materials has driven extensive interest in new materials with ultrahigh in-plane thermal conductivity^[173]. Flexible oriented films with high in-plane thermal conductivity are particularly attractive for advanced electronics, where lateral heat spreading is required without introducing excessive heat transfer to adjacent components^[175]. Such materials can efficiently dissipate heat along the plane of the device while maintaining thermal isolation in the thickness direction, thereby mitigating thermal cross-talk between neighboring hot spots.

2D materials, including graphene, hexagonal boron nitride, and MXenes, exhibit intrinsically anisotropic thermal transport owing to their strong in-plane covalent bonding and weak interlayer interactions^[42,176,177]. When these nanosheets are assembled into well-aligned layered architectures, phonon transport along the in-plane direction can be significantly enhanced, while cross-plane heat transfer remains relatively suppressed^[175]. The 2D layer structure is easily assembled into horizontally aligned films by filtration, hot pressing, casting, and superspreading, etc. Graphite papers and graphene films are representative examples of this class of heat spreaders^[173,178].

Among them, graphene stands out due to its exceptionally high intrinsic in-plane thermal conductivity, reported to range from approximately 2,000 to over 5,000 W·m^-1·K^-1 near room temperature^[179]. By assembling GO sheets into highly oriented films followed by reduction and graphitization, freestanding graphene or reduced graphene oxide papers with ultrahigh in-plane thermal conductivity can be obtained^[172,178]. Typical fabrication strategies include vacuum filtration, evaporation-induced self-assembly, electrospray deposition, and mechanical pressing, all of which promote the parallel stacking of graphene sheets. For example, Zhang et al. fabricated ultrathick graphene films (thickness up to 200 μm) with high thermal conductivity (1,224 ± 110 W m^-1 K^-1) based on the self-fusion character between GO sheets (Figure 10a,b,c)^[180]. This ultrathick graphene film has been proven to render high heat flux during the heat transfer process, much better than that of the metal Al and Cu films (Figure 10d,e). In another work, Xin et al. developed large-area, freestanding graphene papers through direct electro-spray deposition, combining a continuous roll-to-roll process with a simple water-assisted exfoliation from highly hydrophilic aluminum substrates (Figure 10f,g). After mechanical densification and thermal annealing, the resulting films exhibit remarkably high thermal conductivities in the range of 1,238-1,434 W m^-1 K^-1[181]. Owing to their ultralight nature, these graphene papers show superior heat dissipation capability compared with conventional high-thermal-conductivity Cu or Al foils, effectively mitigating localized hot spots (Figure 10h).

Display Full Size

Download

Figure 10. (a) Schematic diagram of the fusion between the two nano-thickness films; (b) Cross section of the pasted graphene films with 6 layers; (c) Thermal conductivity values of the three kinds of films with various thickness; (d) Infrared thermal image of graphene film and metallic foils; (e) Temperature profile of the PGF, Cu foil, Al foil, and single thin GF, according to the infrared thermal image. Reproduced with permission from reference^[180]. Copyright © 2020 Elsevier; (f) Cross-sectional SEM image of the pristine graphene paper and the highly ordered and crystalline graphene paper by thermal annealing at 2,200 °C for 30 mins followed by mechanical pressing; (g) Thermal properties of different graphene papers; (h) Images recorded of the steady state temperature distributions of Al, Cu, and annealed graphene paper. Reproduced with permission from reference^[181]. Copyright © 2014 John Wiley & Sons. PGF: pyrolytic graphite film; GF: gauge factor; SEM: scanning electron microscopy; GP: graphene paper.

The in-plane thermal conductivity of such graphene-based films is strongly dependent on both the lateral size of the nanosheets and the degree of structural perfection achieved during reduction or high-temperature annealing^[172]. Large-area graphene sheets reduce inter-sheet junction density, while high-temperature treatment suppresses defect- and boundary-induced phonon scattering. As a result, in-plane thermal conductivities exceeding 1,000 W·m^-1·K^-1 have been widely reported, surpassing those of conventional copper and aluminum foils and even rivaling commercial graphitized polyimide films^[182-184]. For example, Liu et al. prepared graphene film with excellent thermal and mechanical properties by engineering the film structures in terms of grain size, film alignment, and thickness, and interlayer binding energy^[185]. The resulting smooth and firm (≈2.1 g cm^-3), thin (0.8 µm), large grain (13.3 µm), and partly turbostratic-stacking (37%) graphene film is greatly beneficial for the transfer of both high-frequency diffusive phonons and low-frequency ballistic phonons and demonstrates a superior thermal conductivity value of 3,200 W·m^-1·K^-1.

The highly anisotropic nature of layered 2D materials often results in limited cross-plane thermal conductivity. To address this issue, 3D interconnected architectures have been explored, in which vertically bridging structures are introduced between adjacent graphene layers to partially enhance cross-plane heat transport while retaining excellent in-plane performance. These hybrid designs offer a promising pathway toward balanced thermal transport for next-generation heat spreading applications.

5.2 Through-plane high thermal conductivity: TIMs

TIMs are essential components that fill the microscopic gaps between heat-generating devices and heat sinks, eliminating thermally resistive air pockets and enabling efficient heat transfer across interfaces (Figure 11a)^{[1,186,187,189-192]}. Polymers such as epoxy resins and silicone gels are widely used as TIM matrices because of their mechanical compliance, low cost, and good processability^[193,194]. However, their inherently low thermal conductivity severely restricts heat dissipation efficiency, particularly in high-power-density electronic systems.

Display Full Size

Download

Figure 11. (a) Schematic configuration of the TIMs in microelectronics thermal management, and the principle and mechanism of TIMs. Reproduced with permission from reference^[186]. Copyright © 2024 John Wiley & Sons; reference^[187]. Copyright © 2019 American Chemical Society; (b) The through-plane thermal conductivity of the HOG/PDMS composites with different HOG loadings; (c) Comparison of the application of HOG/PDMS and commercial thermal conductive silicone gasket as the TIMs in the thermal management of LED lamps. Reproduced with permission from reference^[26]. Copyright © 2022 Elsevier; (d) The thermal immune mechanism of the prepared PCCs in battery thermal management; (e) The digital image of the cell pack wrapped with PCCs and finite element analysis of heat transfer in a naked cell and a PCC wrapped cell. Reproduced with permission from reference^[188]. Copyright © 2023 Elsevier. TIMs: thermal interface materials; HOG: highly oriented graphite; PDMS: polydimethylsiloxane; LED: light-emitting diode; PCCs: phase-change composites; TCA: thermal conductive adhesive; DBC: direct bonded copper; AMB: active metal brazed.

Introducing thermally conductive fillers into a polymer matrix is the most widely adopted and scalable strategy to enhance through-plane thermal conductivity^[195]. When fillers are randomly dispersed, however, the resulting heat conduction pathways are often discontinuous, and phonon transport is hindered by numerous filler–filler and filler–matrix interfaces^[1,196,197]. In contrast, constructing vertically oriented and interconnected filler networks can dramatically reduce thermal resistance by forming continuous heat conduction channels across the thickness direction^[15]. One-dimensional and 2D fillers, such as CNTs, metal nanowires, graphene, BN, MXenes, carbon fibers, and silicon carbide frameworks, are particularly suitable for building anisotropic through-plane networks^[1,4,195]. By directional stacking, template-assisted growth, or freeze-casting techniques, these fillers can be assembled into vertically aligned architectures that facilitate efficient phonon transport from the heat source to the heat sink.

Recent studies have demonstrated that composites incorporating vertically aligned graphene foams, lamellar graphene aerogels, or ceramic scaffolds can achieve through-plane thermal conductivities exceeding 10 W·m^-1·K^-1, far outperforming conventional polymer-based TIMs^[198-201]. Such performance enhancements are attributed to the synergistic effects of reduced interfacial thermal resistance, continuous filler networks, and minimized phonon scattering along the alignment direction. Notably, vertically oriented graphite or graphene networks can achieve exceptionally high through-plane thermal conductivity even at relatively low filler loadings, highlighting the effectiveness of directional network design compared with random dispersion^[26,202]. As shown in Figure 11b, we developed soft TIMs by constructing a highly oriented graphite network. The resultant composites presented a remarkably high through-plane thermal conductivity of 35.4 W·m^-1·K^-1 at a filler content of 17.6 vol.%. When this material was also used as a TIM for the heat dissipation system of an LED lamp, it exhibited a much better heat dissipation effect than the commercial thermal conductive silica gasket (Figure 11c). In another work, Zhou et al. prepared multifunctional phase-change composites by immersing polyethylene glycol (PEG) and PU into the highly oriented graphite network. The thermal conductivity of PEG@PU/graphite composite reached 36.52 C at a graphite loading of 40 wt.%. These obtained composites can be utilized as an active preheating or passive cooling system in a battery thermal management system for all-climate demands (Figure 11d,e). These results underscore the critical role of filler orientation and network topology in determining the overall thermal performance of TIM composites.

The thermal management performance of oriented conductive networks is closely associated with the directional transport of phonons and heat carriers. In conventional composites with randomly dispersed fillers, heat transfer is often hindered by discontinuous pathways and high interfacial thermal resistance. In contrast, oriented conductive networks create continuous and highly interconnected thermal transport channels, significantly reducing phonon scattering and thermal transport resistance along the alignment direction. The orientation degree, filler aspect ratio, network connectivity, and interfacial interactions collectively determine the efficiency of heat conduction. Furthermore, anisotropic thermal transport enables preferential heat dissipation in targeted directions, which is particularly advantageous for thermal management applications in electronic devices. Therefore, the remarkable enhancement in thermal conductivity is primarily attributed to the formation of continuous directional heat-transfer pathways and the effective reduction of thermal transport barriers through structural alignment.

Overall, oriented network structures provide a powerful design paradigm for thermal management materials. By rationally tailoring the alignment direction and connectivity of thermally conductive fillers, heat transport can be selectively enhanced either in-plane for heat spreading or through-plane for interfacial heat transfer. Such anisotropic thermal management strategies offer significant advantages over isotropic systems and are expected to play a key role in addressing the thermal challenges of next-generation electronic and optoelectronic devices.

6. Conclusions

This review systematically summarizes the recent progress in multifunctional materials based on oriented conductive networks and highlights the critical role of structural engineering in regulating material performance. Specifically, this work emphasizes oriented conductive networks as a universal design strategy capable of simultaneously optimizing electrical transport, thermal transport, and electromagnetic wave attenuation through controlled structural alignment. Compared with randomly dispersed or isotropic networks, such anisotropy of the oriented conductive networks provides unique performance advantages and unlocks application scenarios that are difficult to achieve with conventional random networks.

From a fabrication perspective, a variety of strategies have been developed to construct oriented networks. Representative approaches include template-assisted methods, such as directional freeze-casting, which allow precise control over macroscopic alignment and hierarchical porosity; field- or force-induced assembly, including electric-field, magnetic-field, shear, and stretching-induced alignment, which offer tunable orientation under relatively mild conditions; and in situ growth techniques, exemplified by CVD and PECVD, which can generate highly ordered and interconnected networks with low interfacial resistance. Beyond these established routes, a growing number of emerging methods, such as hybrid multi-field coupling, additive manufacturing–assisted alignment, and bio-inspired assembly, are rapidly expanding the design space of anisotropic networks, providing greater flexibility in structure–property tailoring.

Oriented networks offer an efficient strategy to maximize functional performance at reduced filler loadings, which is extremely important from a material design perspective, especially for functional composite. By concentrating conductive pathways and interfaces along preferred orientations, high thermal conductivity and electrical conductivity can be achieved without relying on excessive conductive filler content, which is beneficial for maintaining mechanical flexibility, lightweight characteristics, and processability. Through the analysis of intelligent sensing, EMI shielding, and thermal management applications, it is demonstrated that the performance enhancement of oriented conductive networks originates from the synergistic interactions between filler characteristics, network architecture, and anisotropic transport mechanisms. Key structural parameters, including filler orientation, orientation degree, network density, pore structure, and hierarchical organization, play decisive roles in determining functional properties. Establishing these structure–property relationships provides important guidance for the rational design of next-generation multifunctional materials.

7. Perspectives and Outlooks

Looking forward, oriented conductive networks are rapidly emerging as an important research direction in advanced composites due to their unique capability to optimize multifunctional properties through structural engineering. Compared with conventional composites with randomly distributed filler networks, oriented architectures enable more efficient transport pathways and anisotropic functional responses, thereby significantly enhancing EMI shielding, thermal conductivity, electrical conductivity, and sensing performance while maintaining low filler loading. The ability to precisely regulate the propagation of electrons, phonons, and electromagnetic waves through directional architectures provides unprecedented opportunities for achieving high-performance multifunctional materials.

Recent years have witnessed the rapid development of various advanced fabrication technologies for constructing oriented conductive networks. In addition to conventional methods such as directional freezing, mechanical stretching, and shear-induced alignment, emerging techniques including magnetic/electric-field-assisted assembly, template-assisted assembly, additive manufacturing (3D/4D printing), and self-assembly strategies have enabled the fabrication of increasingly complex and highly ordered architectures. These approaches offer new possibilities for tailoring filler orientation, network density, hierarchical pore structures, and anisotropic transport pathways, thereby facilitating precise control over material performance. As fabrication technologies continue to evolve, increasingly sophisticated oriented architectures are expected to be developed for next-generation multifunctional composites.

Despite these promising advances, the industrial implementation of oriented conductive networks remains at an early stage. Compared with conventional composite manufacturing processes, the fabrication of highly ordered architectures often requires additional processing steps, specialized equipment, and strict control of processing parameters, which inevitably increase manufacturing cost and process complexity. Furthermore, many current fabrication methods are limited to laboratory-scale production and face challenges in achieving large-area, high-throughput, and cost-effective manufacturing. Therefore, future research should focus on developing simplified, scalable, and economically viable fabrication technologies that can bridge the gap between laboratory demonstrations and industrial production.

From the perspective of future materials innovation, several important directions deserve particular attention. First, material design will continue to play a central role in the development of high-performance oriented conductive networks. Beyond conventional carbon-based fillers, emerging conductive nanomaterials, including MXenes, transition-metal dichalcogenides, liquid metals, conductive polymers, and hybrid nanostructures, provide new opportunities for constructing multifunctional conductive pathways. Second, structural design is expected to become increasingly important for optimizing material performance. AI and data-driven methodologies are expected to significantly accelerate the development of oriented conductive networks. The performance of these materials is governed by a large number of interconnected variables, including filler type, filler dimensions, orientation parameters, processing conditions, and microstructural characteristics. Traditional trial-and-error approaches are often time-consuming and inefficient for exploring such high-dimensional design spaces. Machine learning, high-throughput simulations, and AI-assisted optimization can help establish quantitative structure–property relationships, predict optimal architectures, and guide the rational design of materials with targeted multifunctional performance.

Overall, oriented conductive networks offer a powerful platform for simultaneously optimizing multiple functional properties through structural engineering. Continued progress in fabrication technologies, structural design strategies, and fundamental understanding of anisotropic transport mechanisms will further accelerate their development. As challenges related to scalability, cost, and structural precision are gradually overcome, oriented conductive networks are expected to play an increasingly important role in next-generation EMI shielding, thermal management, sensing, and intelligent electronic systems.

Authors contribution

Zhang F: Conceptualization, investigation, visualization, writing-original draft, writing-review & editing.

Staudinger U, Gültner M, Krause B, Voit B: Writing-review & editing.

Conflicts of interest

Fei Zhang serves as a Youth Editorial Board Member of Smart Materials and Devices. The other authors declare no conflicts of interest.

Ethical approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

Not applicable.

Funding

This work was financially supported by the Alexander von Humboldt Foundation (This program does not have a grant number).

Copyright

© The Author(s) 2026.