Recent advances in the thermal management performance of polymer-based composite materials

Abstract

Technological advancements have created an urgent demand for effective thermal regulation, including heat generation, transfer, storage and dissipation, in fields such as communication electronics, vehicles and electrochemical energy storage. This drives the development of high-performance thermal management materials. Compared to traditional materials like ceramics and metals, polymer-based materials are regarded as highly promising matrix materials due to their superior electrical insulation properties, flexibility and processability. This review summarizes the types and regulation methods of polymer-based thermal management materials. It focuses on recent research progress in efficient thermal insulation, rapid heat dissipation, radiative cooling and phase change energy storage materials. Furthermore, this review analyzes the role of research strategies such as bio-inspiration, rational engineering, simulation and machine learning in fostering innovation within this field. Finally, current challenges and potential future research directions for polymer-based thermal management materials are outlined.

---

Jia Li

Jia Li obtained her undergraduate degree from the Ocean University of China in 2024 and is currently a master's student at Tianjin University. Her main research direction is thermally conductive polymer composite materials.

Mengmeng Qin

Mengmeng Qin is a professor at Tianjin University and the Secretary General of the Thermal Conductive Materials Branch of the Chinese Society for Composite Materials. He obtained his doctoral degree from Tianjin University in 2017. His research focuses on thermally conductive polymer composite materials for device thermal management.

Wei Feng

Wei Feng is a professor at Tianjin University. He obtained his Ph.D. from Xi'an Jiaotong University in 2000 and worked at Osaka University and Tsinghua University as a JSPS fellow and a Postdoctoral Researcher. He received the National Science Fund for Distinguished Young Scholars and served as the Executive Director of the China Composites Society, the First Chairman of the thermally conductive composites committee, and the Vice Chairman of the smart composites committee of SAMPE China. His research focuses on functional organic carbon composites, high-thermal-conductivity composites, photothermal energy conversion and storage materials, fluorinated carbon materials, and smart responsive functional composites.

---

Wider impact

This review highlights key developments in polymer-based thermal management materials, focusing on breakthroughs in efficient thermal insulation, rapid heat dissipation, radiative cooling and phase change energy storage. We also analyze how bio-inspiration, rational engineering, simulation and machine learning accelerate material innovation. This field holds broad significance as efficient thermal regulation is critical for sustainable technological progress: it enhances the safety and longevity of electronics, boosts energy efficiency and reduces global cooling energy demands. Polymer-based solutions uniquely integrate scalability, lightweight design and electrical safety, making them vital for next-generation devices and carbon-neutral goals. Future developments will focus on intelligent materials with multifunctional, adaptive thermal responses driven by predictive design tools. This review provides a strategic roadmap for this evolution, demonstrating how interdisciplinary approaches can overcome current challenges in thermal conductivity, stability and integration. By consolidating cutting-edge advances and emerging methodologies, this work aims to catalyze cross-disciplinary collaborations, bridging the gap between materials science and advancing thermal management systems.

1. Introduction

Throughout the history of human civilization, the utilization of thermal energy has continuously evolved alongside technological advancements, progressing from rudimentary applications like cooking and heating in primitive societies to sophisticated multi-dimensional thermal management technologies addressing complex modern demands. Industrial smelting requires sustained high temperatures to melt metals, whereas semiconductor devices must maintain junction temperatures below 85 °C for safe operation.¹ In construction engineering, cold regions rely on thermal barrier construction using insulating materials with low thermal conductivity,² while tropical areas depend on radiative cooling coatings for passive cooling.³ For transportation scenarios lacking continuous energy supply, such as vaccine cold chain logistics, phase change materials (PCMs) enable passive isothermal control through latent heat storage.⁴

These fundamental differences reflect the diverse demands across domains for the precise control of heat acquisition, maintenance and dissipation, driving thermal management technology beyond simple heating towards sophisticated regulation. Within this context, polymer-based thermal management materials have emerged as a crucial pathway for overcoming thermal management bottlenecks. Leveraging advantages like lightweight nature, excellent processability, electrical insulation and cost-effectiveness,⁵ these materials demonstrate significant application value in emerging fields.

Polymer-based thermal management materials represent a class of composites where a polymer matrix is utilized and functional fillers or specific structural designs are incorporated to achieve on-demand heat regulation. Polymers inherently possess a low thermal conductivity, typically below 0.5 W m⁻¹ K⁻¹,⁶ making them naturally ideal candidates for thermal insulation substrates. Furthermore, porous structures fabricated via processes such as freeze-drying or supercritical foaming can further reduce the thermal conductivity to below 0.02 W m⁻¹ K⁻¹,² exhibiting unique application potential in building insulation and spacecraft thermal protection.⁷

For thermal conduction applications, the inherently low thermal conductivity of polymers can be overcome by incorporating highly conductive fillers.⁸ The inherent flexibility and low modulus of polymer materials enable them to conform tightly to interfacial micro-/nano-pores, thereby minimizing interfacial thermal resistance and meeting the heat dissipation requirements of electronic devices.⁹ These materials can also achieve optimized mid-infrared emissivity through the manipulation of molecular functional groups. When combined with nanostructured networks¹⁰ or periodically arranged micropillar arrays,¹¹ they can modulate the interaction between light and the structure to achieve efficient solar reflectance and high infrared emittance, thereby enhancing radiative cooling performance. Additionally, the excellent compatibility of polymers with organic PCMs makes them ideal encapsulation matrices.¹² Blending and cross-linking enable precise tuning of PCM phase transition temperature and enthalpy. Meanwhile, polymer melt fluidity and solution spinnability facilitate processing into diverse components for thermal energy storage, critically supporting thermal management in renewable energy systems. The integration of these strategies enables polymer-based composites to deliver multifunctional and high-performance thermal management, proving their indispensable role in meeting complex thermal demands.

Realizing the full potential of polymer-based thermal management materials across these diverse thermal management scenarios hinges critically on innovative material design and processing strategies. To overcome intrinsic limitations and achieve the precise, multifunctional and often contradictory thermal properties demanded by advanced applications, researchers are increasingly leveraging cutting-edge approaches.¹³ These include bioinspired design, rational engineering, computational modeling and simulation and machine learning. The effective integration of these strategies is pivotal for tailoring polymer-based thermal management materials with unprecedented performance.

This article systematically reviews recent advances in polymer-based thermal management materials, focusing on four key performance aspects: high-efficiency thermal insulation, thermal conductivity enhancement, radiative cooling and phase-change energy storage. We further analyze how research strategies, including bioinspired design, rational engineering, computational modeling and machine learning, drive innovation in these materials. Through multiscale structure–property relationship analysis, this work aims to establish a theoretical framework and technological pathway for developing next-generation intelligent thermal management materials.

2. Research strategies for advanced polymer-based thermal management materials

Facing increasingly complex thermal management demands across heterogeneous and extreme environments, conventional material development approaches are constrained by their inability to achieve precise structure–property customization. Research and development in polymer-based thermal management materials demand a paradigm shift beyond traditional empirical trial-and-error approaches.¹⁴ Achieving precise control from structural characterization to functional customization necessitates interdisciplinary strategies targeting core requirements of thermal transport efficiency, energy storage density and environmental adaptability. Current research has established four interconnected methodological frameworks: bioinspired structural design through mimicking efficient biological heat-transfer mechanisms, rational engineering approaches incorporating molecular design and directional filler control, multiscale simulation spanning molecular dynamics to device-level thermal analysis and data-driven machine learning for accelerating composition–process–property correlation modeling.

These methodologies transcend independent design dimensions to form a synergistic innovation paradigm.¹⁵ Their cross-integration provides systematic solutions to critical challenges¹⁶ including minimizing polymer–filler interfacial thermal resistance, optimizing thermal conduction networks and balancing multifield-coupled performance. This section subsequently examines how these strategies advance material design theories and fabrication technologies, detailing their foundational principles, technical implementation pathways and recent research progress.

2.1. Bioinspired design

The bionic approach is highly innovative and inspirational in the research of polymer-based thermal management materials. Through long-term evolution, numerous biological structures in nature exhibit excellent thermal management properties, providing a rich source of inspiration for material design.¹⁵

The unique hollow structure of polar bear fur effectively traps air to form an insulating layer that reduces heat loss while simultaneously absorbing and converting infrared radiation from sunlight for efficient thermal utilization (Fig. 1a).¹⁷ Inspired by this structure, Cui et al. employed the freeze-spinning technique to achieve continuous and large-scale fabrication of fibers mimicking polar bear hair. Precise control of solution concentration/viscosity, extrusion speed and freezing temperature enabled accurate tuning of the fibers' hollow architecture. Textiles woven from these biomimetic fibers exhibit excellent thermal insulation and mechanical properties, demonstrating significant potential for applications in personal thermal management and building insulation.¹⁸

Fig. 1 (a) SEM images of the hollow core and aligned shell of a polar bear hair. Reproduced with permission.¹⁸ Copyright 2018, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim. (b) Directional water transport in natural wood; schematic illustration of the fabrication process and the heat-transfer mechanism in CFS/EB composites. Reproduced with permission.²⁰ Copyright 2024, Elsevier. (c) Cross-sectional view of the hairs of Saharan silver ants milled with FIB and the interaction between visible and NIR light and a hair at small (I), intermediate (II), and large (III) incidence angles. Reproduced with permission.²¹ Copyright 2015, American Association for the Advancement of Science.

Vertically aligned sieve tube structures in tree trunks form efficient and continuous long-range pathways that prioritize axial transport of water and nutrients.¹⁹ Inspired by this wood microstructure, Luo et al.²⁰ developed an epoxy-based electronic packaging material (CFS/EB) with biomimetic vertically aligned thermal channels (Fig. 1b). Their methodology employed directional freezing to construct an aerogel skeleton, which was subsequently infiltrated with the boron nitride/epoxy (BN/EP) hybrid material to replicate vertically aligned bioinspired conduits for axial-selective thermal transport. This bioinspired design achieves a through-plane thermal conductivity of 1.51 W m⁻¹ K⁻¹ while maintaining electromagnetic wave absorption capability and effectively directing waste heat from electronic components to external heat sinks.

In the bionic prototype of the Saharan silver ant, radiative cooling is achieved through triangular hairs. These hairs enhance reflection of visible and near-infrared solar radiation via Mie scattering and total internal reflection, minimizing heat gain. Simultaneously, in the mid-infrared range (>2.5 μm) where solar radiation is negligible, the hairs act as an anti-reflective layer, boosting emissivity to maximize heat dissipation through blackbody radiation (Fig. 1c).^21,22 Based on this mechanism, Zhang et al. engineered biomimetic short silk fiber (SSF) arrays with triangular cross-sections. When coated onto polydimethylsiloxane (PDMS) films, these bioinspired polymer composites reduced the substrate temperature by 7.6 °C under solar exposure, demonstrating efficient radiative cooling for advanced thermal management materials.²³

2.2. Rational engineering

Rational engineering entails precise design and regulation across molecular-to-macroscopic scales, tailored to specific thermal management requirements for optimized material performance.

Molecular design employs chemical techniques like copolymerization,^24,25 grafting²⁶ and cross-linking⁷ to modify polymer chain structures, imparting targeted material properties. For instance, thermal interface materials simultaneously require high thermal conductivity and flexibility. As shown in Fig. 2a, Zhang et al.²⁷ grafted liquid-crystalline mesogenic groups onto a PDMS backbone via ring-opening copolymerization. The high grafting density facilitated ordered mesogen alignment, reducing phonon scattering induced by lattice defects, while π–π stacking between mesogens enhanced phonon coupling. Through molecular grafting alone, the liquid crystal PDMS achieved a thermal conductivity of 0.56 W m⁻¹ K⁻¹ while retaining the flexibility and low density of pristine PDMS, satisfying comprehensive requirements for advanced thermal interface materials.

Fig. 2 (a) Schematic diagram of the synthesis route for LC-PDMS. Reproduced with permission.²⁷ Copyright 2025, Wiley-VCH GmbH. (b) Schematic representation of core-sheath 3D printing of the bean-pod-structured OD/GO phase change microlattice. Reproduced with permission.³² Copyright 2021, Wiley-VCH GmbH. (c) Liquid nitrogen-assisted ice templating enables the directional alignment of fillers in aerogels. Reproduced with permission.³³ Copyright 2024, Wiley-VCH GmbH.

At the macrostructural scale, technologies such as 3D printing^28,29 and template methods³⁰ enable fabrication of polymer-based composites with precisely controlled architectures. 3D printing facilitates the construction of complex geometries with tailored internal structures.³¹ Yang et al.³² fabricated octadecane/graphene phase-change composites with microlattice architectures via extrusion-based core–sheath 3D printing (Fig. 2b). Precise control of printing parameters enabled tailored pore distribution and interfacial structures, achieving synergistic enhancement of thermal conduction efficiency (the transversal thermal conductivity is 1.67 W m⁻¹ K⁻¹) and energy storage stability. Alternatively, template methods construct ordered thermal pathways or hierarchical pore structures, forming continuous conduction channels that reduce interfacial thermal resistance and promote directional heat transfer (Fig. 2c).^33,34

2.3. Simulation and modeling

Simulation and modeling, leveraging computer technology and numerical calculation methods, serves as a “bridge” connecting the microscopic structure of materials with their macroscopic thermal properties. It overcomes the limitations of experimental trial-and-error approaches, which are often time-consuming and costly. The mechanisms of thermal management in polymer-based materials vary significantly across different scales. A key challenge in current simulation-based design is achieving effective integration between microscopic structures and macroscopic properties while avoiding the limitations of single-scale modeling. Commonly used simulation methods include molecular dynamics (MD), finite element analysis (FEA) and the Boltzmann transport equation (BTE).³⁵

MD is a method based on Newtonian mechanics that simulates the physical trajectories and states of atoms and molecules under a molecular force field, thereby investigating the energy and properties of a system.³⁶ It primarily includes equilibrium molecular dynamics (EMD) and non-equilibrium molecular dynamics (NEMD) simulations. Compared to NEMD, EMD exhibits less dependence on the size of the simulated system.³⁷ The EMD method, rooted in the fluctuation–dissipation theorem, provides insights into the contributions of different types of interactions to thermal conductivity, though it requires more time for the system to reach equilibrium. In contrast, NEMD relies on establishing a constant temperature gradient and heat flux within the modeled system, applying Fourier's law of heat conduction when heat transfer is diffusive.³⁸ Through MD simulations, it is possible not only to control molecular structures and related parameters in practical studies but also to model the thermal behavior of complex materials,^39,40 demonstrating strong controllability, guidance and cost-effectiveness.

Zhang et al.⁴¹ proposed an interfacial welding strategy to construct graphite-structure welded carbon nanotube (GS-w-CNT) networks for enhancing thermal conductivity in composite materials. Building on experimental data, MD simulations investigated thermal transport in GS-w-CNT/PDMS composites (Fig. 3a). These simulations elucidated how the GS welding degree critically governs interfacial phonon scattering reduction and suppresses thermal resistance between carbon nanotubes. Experimental results indicate that GS welding transforms the discontinuous CNT network into a 3D interconnected GS-w-CNT framework, resulting in thermal conductivity that increased to 4.1 times that of the original CNT/PDMS composite.

Fig. 3 (a) Molecular models of GS-w-CNTs and GS-w-CNTs with increasing weight ratio of GS; GS-w-CNT/PDMS and GS-w-CNT model used in the RNEMD method. Reproduced with permission.⁴¹ Copyright 2023, Wiley-VCH GmbH. (b) FEA results of heat transfer for a polymer filled with different structures of GNPs and LM particles; FEA images of composites with filled and half-filled wavy surfaces at different pressures. Reproduced with permission.⁵⁰ Copyright 2024, the Royal Society of Chemistry.

Compared to EMD and NEMD, the BTE represents a more complex simulation approach that is particularly suitable for crystalline solids with weak anharmonicity. The BTE is a complex nonlinear integro-differential equation that describes the probability distribution of phonons in both position and momentum space.^42,43 By tracking changes in the phonon distribution function, it quantifies the contribution of various scattering mechanisms to thermal conductivity, thereby going beyond the diffusion assumption of Fourier's law.⁴⁴

FEA primarily simulates heat transfer processes in materials at macroscopic scales. It discretizes the material into finite elements and analyzes the local behavior of these elements to approximate solutions to partial differential equations, enabling the prediction of macroscopic thermal properties and the optimization of material structures.^45,46 FEA is well-suited for solving heat transfer problems in irregularly shaped domains or 3D scenarios.⁴⁷ For highly thermally conductive polymer-based composites, FEA can simulate the temperature distribution in practical applications, optimize material layout⁴⁸ and dimensions,⁴⁹ and prevent the formation of local hot spots. A comparison of the key characteristics of MD, FEA and the BTE is provided in Table 1.

Table 1 Comparison of the key characteristics of representative simulation methods

Method	Core principle	Applicable scenarios
MD	Newtonian mechanics describing atomic motion	Studying molecular chain dynamics and filler/matrix interfacial interactions
BTE	Describes the non-equilibrium transport processes of phonons	Predicting thermal transport characteristics at micro/nano scales
FEA	Discretization and solution of the heat conduction equation	Predicting the macroscopic thermal performance of components and optimizing structural design

---

He et al.⁵⁰ prepared a thermal interface material of liquid metal–graphene aerogel-brush-shaped polydimethylsiloxane (LM–VGA/BPDMS). COMSOL-based FEA revealed bicontinuous phonon pathways through vertically aligned graphene aerogels and patterned liquid metal channels, while simultaneously modeling stress distribution and thermal contact resistance under dynamic loading on wavy interfaces. Simulations confirmed that the material's low modulus enables complete interfacial conformity during compression, with compressive strain enhancing the liquid metal-VGA contact to reduce thermal resistance (Fig. 3b). Experimental results demonstrate that LM–VGA/BPDMS exhibits a high thermal conductivity (7.11 W m⁻¹ K⁻¹), an ultralow elastic modulus (10.13 kPa) and a low interfacial thermal resistance (14.1 K mm² W⁻¹).

Furthermore, it is essential to validate the reliability of simulation results. Experimental comparison serves as the most direct validation method. It is also feasible to use interfacial thermal resistance and filler thermal conductivity obtained from MD simulations as input parameters for FEA, thereby establishing a macroscopic model that accounts for filler distribution. This model can be used to calculate the overall thermal conductivity of the material for various filler contents and morphologies. The calculated values can then be compared with experimental data to determine whether the deviation falls within an acceptable range.

2.4. Machine learning

As a powerful data analysis tool, machine learning plays a crucial role in the research of polymer-based thermal management materials.⁵¹ It can mine potential laws from large amounts of experimental and simulation data and rapidly screen and optimize material formulations and preparation processes.¹⁴

Machine learning algorithms enable predictive modeling of structure–property relationships in materials science.⁵¹ Trained on performance datasets, including the thermal conductivity and coefficient of thermal expansion across diverse polymer matrices, filler types, filler concentrations and processing parameters, these algorithms establish robust performance predictors. Commonly employed approaches include neural networks,⁵² reinforcement learning,⁵³ transfer learning⁵⁴ and support vector machines. The trained models rapidly and accurately predict material properties from structural input parameters, significantly accelerating research and development cycles.

Huang et al.⁵⁵ developed a hybrid framework integrating automated physical feature engineering with symbolic regression to discover high-thermal-conductivity amorphous polymers. As shown in Fig. 4a, beginning with 325 initial physical descriptors, their methodology employed statistical filtering and random forest algorithms to optimize the feature set to 25 key descriptors. Subsequent multilayer perceptron modeling established high-accuracy structure–thermal conductivity relationships (R² = 0.83). Combining feature importance ranking with explicit analytical models fitted by symbolic regression reveals that conjugated molecular structures, chain rigidity and the highest relative atomic mass are key factors that significantly influence the thermal conductivity of polymers.

Fig. 4 (a) Polymer descriptor down-selection process; calculation and prediction of thermal conductivity by the MLP model; average SHAP importance for optimized descriptors (the blue and red bars indicate positive and negative importance); the relationships between the radius of gyration Rg and the thermal conductivity of 104 polymers. Reproduced with permission.⁵⁵ Copyright 2023, the Royal Society of Chemistry. (b) TRC concept and the QA-assisted active learning scheme; comparison of computational times of schemes leveraging QA and the exhaustive enumeration for different numbers of layers in the PML. Reproduced with permission.⁶¹ Copyright 2022, American Chemical Society.

Additionally, Yuan et al.⁵⁶ developed machine learning models to predict radiative cooling aerogel performance by integrating multidimensional parameters including the material composition, optical properties and environmental conditions. Performance comparisons identified the optimized XGBoost model as the most effective predictor. SHapley Additive exPlanations analysis quantified ZnO modifiers and environmental parameters (ambient temperature and solar irradiance) as dominant factors governing cooling performance while revealing nonlinear interactions between material constituents and external conditions.

In material optimization, machine learning enables inverse design by determining optimal material compositions and structures based on target properties. Coupled with advanced optimization algorithms, including genetic algorithms, particle swarm optimization⁵⁷ and Bayesian optimization,⁵⁸ this approach efficiently navigates high-dimensional parameter spaces to achieve intelligent material design.⁵⁹

Huang et al.⁶⁰ developed an AI-assisted framework for efficient inverse design of high-thermal-conductivity polymers, integrating deep neural networks with multi-objective optimization algorithms. This approach leverages data analysis capabilities of machine learning to overcome limitations of traditional trial-and-error methods. The framework unifies thermal conductivity prediction and synthesizability assessment within a single optimization workflow. Subsequently, MD simulations validated 50 designed polymers, revealing Rg–thermal conductivity correlation and rigid aromatic structures' key role. The designed pentablock polymer Pen_01 exhibits a thermal conductivity of 1.32 W m⁻¹ K⁻¹, while the triblock polymer Tri_01 reaches 1.01 W m⁻¹ K⁻¹.

Similarly, in transparent radiative cooler (TRC) design, quantum annealing (QA) and active machine learning enable an inverse design scheme that advances this technical paradigm (Fig. 4b). This approach integrates the global optimization capability of quantum computing with machine-learned surrogate models to precisely balance visible light transmittance and infrared radiative performance during photonic structure optimization. Experimental validation demonstrates that the optimized TRC, comprising a multilayered stack of four common dielectric materials, exhibits superior transparency and cooling performance compared to commercial glass. This design achieves an annual energy saving of 86.3 MJ m⁻² and its computational efficiency surpasses that of traditional exhaustive enumeration by several orders of magnitude, fully underscoring the transformative potential of this technology.⁶¹

Machine learning fundamentally advances the prediction and optimization of structure–property relationships in polymer-based thermal management materials by establishing high-dimensional mappings between molecular architectures and macroscopic performance metrics.⁶² Through surrogate modeling and multi-objective optimization frameworks, it navigates complex compositional spaces to simultaneously maximize target thermal properties, while satisfying manufacturability constraints. Active learning strategies accelerate the discovery of non-intuitive designs by prioritizing high-value computational or experimental evaluations.⁶² Crucially, interpretability tools decode hidden physical mechanisms governing thermal transport, resolving intrinsic trade-offs between conflicting material requirements. This paradigm transforms optimization from empirical iteration to physics-informed inverse design while accelerating development cycles through intelligent parameter space navigation.⁶³

3. Research progress on advanced polymer-based thermal management materials

3.1. Thermal insulation materials

In extreme high-temperature applications, thermal insulation materials mitigate energy loss and thermal damage risks by suppressing heat transfer pathways.^64,65 While traditional metal or ceramic insulators offer high-temperature resistance, their high density, low flexibility and complex processing present challenges for lightweight designs, complex geometries and large-scale applications. Polymer-based composites, due to their light weight, corrosion resistance and facile processability, hold significant strategic importance for advanced thermal insulation.⁶⁶ Within this domain, polymer matrices with a cellular structure and polymer composites with hollow-structured fillers represent two pivotal technological pathways for lightweight insulation. The former achieve ultralow thermal conductivity through nano/micro-scale pore engineering, while the latter enhance overall performance by leveraging the “thermal barrier” effect of hollow fillers coupled with interfacial optimization.

3.1.1. Polymeric cellular structures for thermal insulation. Aerogels constitute a class of porous, lightweight materials exhibiting ultra-low thermal conductivity.⁶⁷ While inorganic aerogels typically demonstrate high thermal stability and flame retardancy, their inherent brittleness critically compromises structural integrity under mechanical stress. In contrast, polymer-based aerogels not only inherit the intrinsically low thermal conductivity of polymeric materials, endowing them with exceptional insulation performance, but also maintain superior flexibility,⁶⁸ thereby substantially enhancing overall practicality.

The unique nano/micro-scale porous network of polymer-based aerogels constructs a 3D thermal resistance barrier. This pore structure synergistically suppresses heat transfer through a triple mechanism (Fig. 5a), encompassing thermal convection, thermal radiation and thermal conduction.⁶⁹ Convection involves the macroscopic movement of fluids, such as liquids and gases. Air convection at a pressure of 1 atm requires pore diameters larger than 10 mm.⁶⁸ In contrast, the typically nanoscale pore diameter of aerogel materials significantly attenuates conductive heat flow through air. Xue et al.⁷⁰ uniformly dispersed SiO₂ nanoparticles within a dual-crosslinked PI nanofibrous skeleton, creating internal pores approximately 4 nm in size. The resulting PI@SiO₂ aerogel achieved a minimum thermal conductivity of 0.0258 W m⁻¹ K⁻¹.

Fig. 5 (a) Schematic diagram of the thermal-insulating mechanism in aerogels and foams. Reproduced with permission.⁶⁹ Copyright 2024, Wiley-VCH GmbH. (b) Schematic illustration of the preparation of the all-natural wood-inspired aerogel; comparison of the thermal conductivity of the all-natural wood-inspired aerogel and other materials; SEM images of the natural wood of the longitudinal section and the cross-section; SEM images of the all-natural wood-inspired aerogel of the longitudinal section and the cross-section. Reproduced with permission.² Copyright 2022, Wiley-VCH GmbH. (c) The schematic diagram of the change in thermal conductivity in the porous aerogel with the density increases. Reproduced with permission.⁹⁰ Copyright 2019, Elsevier Inc. (d) WCAs and SEM images of the aerogel's surface. Reproduced with permission.⁸¹ Copyright 2024, the Author(s). (e) Schematic illustration of the proposed flame-retardancy mechanism of KGM/SA/PA. Reproduced with permission.⁸⁴ Copyright 2023, Elsevier B.V. (f) Schematic diagram showing the thermal transfer mechanism of hollow PI nanofibers. Reproduced with permission.⁸⁸ Copyright 2024, the Author(s).

Thermal radiation is the phenomenon where atoms within an object, undergoing complex and intense motion above 0 K, emit electromagnetic waves and thermal rays. This mode of heat transfer depends highly on the material's optical response to electromagnetic waves. In the infrared region at temperatures of several hundred Kelvin, the contribution of radiative heat transfer becomes significant. As thermal radiation passes through an aerogel, it may be absorbed, reflected or refracted at the pore-wall interfaces, effectively dissipating infrared radiation energy.^71,72 Shi et al.⁷³ developed entirely biomass-derived core–shell structured aerogel fibers with a cellulose acetate shell and a silk fibroin core. The irregular porous sheet-fiber hybrid architecture of the silk fibroin core enhanced infrared reflectivity to inhibit thermal radiation.

Heat transfer through thermal conduction in aerogels involves both gas conduction within the pores and solid conduction through the skeletal framework. The porous structure introduces a large volume of low-thermal-conductivity air, where heat is transferred in the gas phase through collisions between molecules. When the pore size of the aerogel is smaller than the mean free path of gas molecules (69 nm),⁶⁹ the Knudsen effect occurs, significantly reducing the thermal conductivity of air inside the pores.⁷⁴ Solid conduction, on the other hand, occurs through lattice vibrations (phonons) and free electron movement. In aerogels, phonon-mediated energy transfer is the dominant mechanism of heat conduction through the solid phase. The extremely high porosity and nanoscale solid skeleton of aerogels greatly reduce the effective solid heat transfer path and weaken the efficiency of phonon transport. Guo et al.⁷⁵ developed cation–π cross-linked PI aerogels, in which the cation–π interactions promote a structural transition from lamellar stacking to tubular pore arrangements. This results in reduced density and a smaller pore size, effectively suppressing heat conduction and leading to outstanding thermal insulation performance (0.0427 W m⁻¹ K⁻¹).

Liu et al.⁷ proposed a “double-phase-networking” strategy to fabricate a polyimide hybrid aerogel (PIA-S/Z series). The average pore diameter of PIA-Z/S_1.2 is as small as 16 nm, which is less than the mean free path of air molecules under standard conditions (≈70 nm). This significantly restricts the free movement of gas molecules. Concurrently, owing to the aerogel's 3D network structure, heat must traverse through numerous tortuous solid pathways, thereby extending the heat transfer path and increasing thermal resistance. This results in an ultralow thermal conductivity of 0.0213 ± 0.001 W m⁻¹ K⁻¹.

Wood's natural hierarchical architecture provides a biological template for thermal insulation development. By adopting a bottom-up strategy to remove lignin from wood, porous materials with a wood-like aligned structure can be obtained.⁷⁶ Han et al.² developed all-natural aerogels featuring wood-mimetic directional channels (Fig. 5b). These axially continuous tubular pores exhibit radial dimensions below the critical distance for natural convection, effectively suppressing convective heat transfer. Additionally, micro/nano-scale pores within channel walls, formed by cellulose nanofibers and wood tracheid structures, restrict solid-phase thermal conduction. Combined with nanoclay-derived ceramic barriers at high temperatures, this biomimetic architecture achieves ultralow radial thermal conductivity (0.0174 W m⁻¹ K⁻¹), surpassing most commercial insulators.

Aerogel density critically governs thermal insulation performance. With >99% porosity,⁶⁹ heat transfer occurs primarily via gas/solid conduction and radiation.⁷⁷ An initial density increase reduces the pore size, suppressing gas conduction and radiation while slightly increasing solid conduction, resulting in a net decrease in thermal conductivity. Beyond a critical density, however, significant skeleton densification dramatically intensifies solid conduction, dominating heat transfer and degrading insulation. Consequently, thermal conductivity forms a characteristic U-shaped curve versus density, reaching a minimum at an optimum point (Fig. 5c).⁹⁰ This enables precise insulation optimization through density control.

While high porosity endows aerogels with exceptional thermal insulation, it concurrently renders them vulnerable to moisture infiltration. Since water exhibits substantially higher thermal conductivity than air, absorbed moisture severely compromises the insulating efficacy of porous aerogels.⁷⁸ Implementing surface chemical modification,⁷⁹ low-surface-energy polymer compositing⁸⁰ or graded-pore structuring results in hydrophobic aerogel interfaces. This hydrophobic barrier prevents liquid intrusion and suppresses vapor condensation/adsorption within pores. By ensuring thermal stability under high humidity and freeze–thaw cycles, it enables reliable deployment in demanding applications like cold-chain logistics and marine engineering.

For instance, Sui et al.⁸¹ engineered a hydrophobic aerogel using methyl-functionalized Si–O–Si binders to reduce surface energy. Directional freeze-drying created hierarchical micro/nano-rough pores that trap air (Cassie–Baxter state), achieving water contact angles >134.5°. This effectively blocks conductive media intrusion, preserving thermal insulation performance (Fig. 5d).

The inherent flammability of organic polymers poses substantial fire hazards, making flame-retardant engineering essential for practical insulating aerogel applications.⁸² Incorporating phosphorus/nitrogen-based flame retardants or layered silicates into the aerogel's 3D network creates a synergistic system, enabling gas-phase inhibition, condensed-phase barrier formation and thermal conduction suppression. This structural modification preserves low-density characteristics while delaying heat transfer, suppressing flammable volatile release, eliminating melt-dripping behavior and achieving UL94 V-0 or superior ratings.⁸³

Phytic acid (PA), a natural biobased phosphorus compound with high P content (28 wt%), significantly enhances the flame retardancy of aerogels. As shown in Fig. 5e, PA decomposes into phosphoric acid and polyphosphoric acid during combustion, catalyzing protective char formation that blocks heat/oxygen transfer while scavenging ˙H/˙OH radicals to interrupt combustion chain reactions.⁸⁴ In a fully biobased flame-retardant aerogel, synergistic effects between PA and chitosan further enhance performance. Chitosan-derived inert gases (e.g., NH₃ and N₂) dilute oxygen and flammable vapor concentrations, suppressing gas-phase combustion. Crucially, despite a marginal increase in thermal conductivity to 0.038–0.042 W m⁻¹ K⁻¹, the incorporation of chitosan, PA and citric acid endowed the aerogel with significantly enhanced fire safety (with an LOI value of up to 31.2%) and notable antibacterial properties (exhibiting antibacterial activities of 81.98% and 67.43% against Staphylococcus aureus and Escherichia coli, respectively).⁸⁵

Conventional polymer aerogels exhibit limited thermal stability, failing to meet demanding thermal insulation requirements under extreme conditions.⁸⁶ In contrast, aerogel fibers leverage entangled architectures to form elastic skeletons, significantly enhancing mechanical robustness and pore stability. The tortuous phonon pathways created by disordered stacking of high-aspect-ratio fibers substantially suppress solid-phase conduction, collectively improving both thermal insulation and extreme-environment adaptability.⁸⁷

Hollow PI nanofibers fabricated via coaxial electrospinning achieve ultralow thermal conductivity (0.0206 W m⁻¹ K⁻¹) with exceptional temperature resilience across −196 to 350 °C. This performance originates from a high-porosity network where submillimeter inter-fiber gaps suppress convection, while internal cavities immobilize insulating air (Fig. 5f).⁸⁸ Carbon nanofiber aerogels (CNFAs) derived from aramid precursors extend operational limits to −196 to 1000 °C with constant ultralow conductivity (0.01993 W m⁻¹ K⁻¹), where hierarchical graphitized structures attenuate radiative heat transfer. Crucially, the inherent crystallinity of Kevlar precursors underpins simultaneous optimization of thermal insulation and mechanical durability under extreme thermomechanical stress.⁸⁹

3.1.2. Thermal insulating polymer composites with hollow-structured fillers. Hollow-structured fillers leverage their unique internal cavities to establish low-thermal-conductivity insulation systems by simultaneously suppressing conductive, convective and radiative heat transfer.⁶⁸ Their integration into polymer matrices enhances interfacial phonon scattering while preserving lightweight flexibility, exhibiting significant academic value and engineering potential for advanced thermal insulation applications.

Hollow glass microspheres (HGMs) leverage nanoscale cavities with dimensions approaching the mean free path of air molecules to form quasi-vacuum environments, suppressing gaseous conduction through gas rarefaction.⁹¹ Additionally, SiO₂ exhibits intrinsically low thermal conductivity.⁹² By leveraging polydopamine's adhesive properties, HGMs were embedded onto glass fibers. HGMs extend the heat transfer path, while the polymerization-induced dark coloration of polydopamine inhibits radiative heat transfer (Fig. 6a). Compared to the pure glass fiber reinforced polymer, the composite exhibited a 14% reduction in thermal conductivity, while maintaining excellent flame retardancy.⁹³

Fig. 6 (a) Thermal insulation mechanism of HGM-PDA-GFRP. Reproduced with permission.⁹³ Copyright 2024, Elsevier Ltd. (b) Schematic illustration of SOS (or SOS-MH) microspheres; SEM images of SOS-MH particles and the PAN/SOS-MH nanocomposite fibrous membrane. Reproduced with permission.⁹⁴ Copyright 2021, American Chemical Society (c) Thermal insulation mechanism of HPM@ATO/WPU coating. Reproduced with permission.⁹⁶ Copyright 2025, Elsevier B.V. (d) Schematic diagram of heat conduction when hollow microspheres are uniformly distributed and agglomerated. Reproduced with permission.⁹¹ Copyright 2024, Elsevier Ltd.

Researchers have further engineered sphere-on-sphere microspheres (SOS-MH) featuring a unique multiscale air-entrapping architecture (Fig. 6b). This design integrates a core cavity with protruding secondary hollow structures to create multiple air barriers that encapsulate static air, effectively suppressing heat conduction and convection. The multiscale hollow structures synergistically increase the gas fraction and prolong the heat conduction path. Concurrently, the capillary channels within the mesoporous outer shell layer further suppress solid-phase heat conduction through enhanced phonon scattering. Ultimately, this results in a thermal conductivity of 0.0307 W m⁻¹ K⁻¹ for the PAN/SOS-MH composite membrane, representing a reduction of approximately 40% compared to that of the pure PAN membrane.⁹⁴

Beyond suppressing thermal conduction, hollow microspheres significantly reduce radiant heat penetration and absorption by creating numerous highly reflective interfaces within the material that effectively reflect and scatter infrared radiation.⁹⁵ Yang et al.⁹⁶ synthesized dual-shell HPM@ATO composites via a novel coprecipitation approach, coating hollow polystyrene microspheres (HPMs) with antimony-doped tin oxide (ATO) nanoparticles. As shown in Fig. 6c, the HPM amplifies light scattering through multilevel refraction and reflection at air–solid interfaces. Simultaneously, the ATO shell scatters and absorbs solar radiation while promoting multiple internal reflections to reduce direct transmittance, with its high thermal conductivity enabling rapid dissipation of absorbed heat. These synergistic mechanisms integrate optical regulation with active/passive thermal management, endowing the composite with superior insulation performance exceeding that of single-component systems.

Crucially, the homogeneous dispersion of hollow particles within the polymer matrix is crucial for effectively reducing the composite's thermal conductivity, particularly at higher particle loadings.⁹¹ At elevated loadings, high concentrations of hollow particles tend to agglomerate and form continuous thermal conduction networks (Fig. 6d). The thermal conductivity of such networks substantially exceeds that of the polymer matrix itself, creating additional heat transfer pathways. This percolation effect can significantly diminish or even negate the low thermal conductivity benefit achieved by the high-density porosity provided by the hollow particles.⁹⁷

3.2. Thermally conductive materials

The rapid evolution of electronic components toward higher integration, increased frequency, and elevated power density has positioned thermal management as a critical bottleneck constraining device reliability, operational safety and service lifetime.^98,99 Localized overheating accelerates material aging and performance degradation, potentially triggering catastrophic electronic system failure. Empirical evidence indicates that for every 2 °C temperature rise in advanced precision electronics, safety stability decreases by 10%.¹⁰⁰ Conventional metal heat sinks face inherent limitations (excessive weight and rigidity), rendering them incompatible with modern lightweight and flexible device architectures. While conventional polymers (e.g., epoxy resins and silicone rubbers) offer advantages including low density, corrosion resistance and processability, their intrinsically low thermal conductivity proves to be inadequate for efficient heat dissipation.⁶

The emergence of polymer-based high-thermal-conductivity composites represents the pivotal pathway to resolving this materials paradox. The core design philosophy centers on implementing functionalized structural innovations that transcend conventional thermal transport limitations while preserving the intrinsic benefits of polymer matrices.

Heat transfer in solids primarily occurs through two carriers: phonons and electrons. In metallic materials, thermal conduction is dominated by free electrons, whereas in non-conductors such as polymers, heat is mainly carried by phonons via lattice vibrational waves.¹⁰¹ When heat is applied to a polymer surface, phonons in the superficial layer transmit energy to adjacent groups through various modes including stretching, rocking, and torsional vibrations. Furthermore, studies have shown that atoms in polymers undergo disordered vibrations or rotations around their equilibrium positions. Phonon transport involves extensive anharmonic coupling interactions and proceeds at relatively low velocities, which collectively contribute to the generally low thermal conductivity of polymeric materials.^102,103

3.2.1. Intrinsic thermal conductivity regulation. In polymeric materials, monomeric units are connected by covalent bonds to form long chains, through which phonons experience minimal scattering and can propagate with relatively high efficiency along the backbone. In contrast, intermolecular forces between polymer chains are weak, and the chains are often entangled, disordered, and separated by gaps. As a result, phonons undergo frequent scattering and even energy dissipation during interchain transport, leading to significantly lower thermal conductivity between chains than along them.¹⁰⁴ However, in polymers with well-ordered crystalline structures, phonon scattering is substantially reduced within the crystalline regions. Phonon transport pathways may even form along the crystal orientation, thereby resulting in notably higher thermal conductivity compared to conventional amorphous polymers.^105,106

Stretching and orientation processing of polymeric materials represent a critical strategy for modulating their aggregated state structure and enhancing thermal conductivity.¹⁰⁷ During tensile deformation, molecular chains align preferentially along the stretching direction, establishing ordered domains that significantly reduce phonon scattering and facilitate unimpeded phonon transmission pathways.¹⁰⁸

Rastogi et al.¹⁰⁹ achieved a metallic-like thermal conductivity of 18 W m⁻¹ K⁻¹ in low-entanglement UHMWPE through biaxial stretching, which produced a two-dimensionally oriented crystalline network. This process prevented the fibrillation typical of uniaxial stretching by enabling dense chain packing, thereby reducing interfacial thermal resistance. Meanwhile, uniform chain distribution within the XY plane and ordered alignment of (110) crystal planes minimized phonon scattering at grain boundaries. The resulting enhancement in lattice vibration coupling yielded a 60-fold thermal conductivity increase compared to the unstretched polymer.

The team further engineered an end-linked star-shaped thermoset (ELST) composed of tetra-armed polyethylene glycol, achieving reversible thermal conductivity tuning through elastic deformation cycles (Fig. 7a). When stretched above its melting temperature, the material develops highly oriented amorphous chains along the strain direction while simultaneously inducing aligned crystalline domains and enhancing overall crystallinity. Subsequent cooling under fixed strain preserves these configurations, optimizing phonon transport pathways by reducing phonon scattering and improving chain segment ordering through expanded intercrystalline spacing. This molecular-scale ordering increases the thermal conductivity by up to 11.5-fold.¹¹⁰ This work demonstrates rapid, reversible thermal conductivity cycling in macroscale polymers for the first time, establishing a new paradigm for intelligent thermal management materials.

Fig. 7 (a) Schematics of the thermal transport mechanism for enhanced thermal conductivity in the ELST by strain-induced crystallization and the strain effect on the ELST structure characterized via X-ray scattering. Reproduced with permission.¹¹⁰ Copyright 2024, the Author(s). (b) Synthesis of PSBNP-co-PTNI; schematic and TEM image of the ordered self-assembled structures of PSBNP-co-PTNI_0.02. Scale bar, 10 nm. Reproduced with permission.²⁴ Copyright 2023, the Author(s). (c) Proposed thermally conductive mechanism. Reproduced with permission.¹¹⁶ Copyright 2025, Science China Press.

In addition, chemical synthesis strategies, such as monomer design, copolymerization and functional group modification, can directly regulate the chemical composition, bonding patterns, topological architecture and side-chain structures of polymer chains. This enhances the intrinsic thermal conductivity of polymers by strengthening intermolecular supramolecular interactions and promoting the formation of crystalline or liquid crystalline structures.¹¹¹

Chen et al.²⁴ engineered a dual-chain ladderized alkane copolymer based on polynorbornene (Fig. 7b). This material self-assembles into highly ordered arrays via π–π stacking, reducing phonon scattering and achieving a thermal conductivity of 1.96 W m⁻¹ K⁻¹ perpendicular to the assembly plane.

Similarly, Yu et al.¹¹² incorporated poly-2-[[(butylamino)carbonyl]oxy]ethyl ester soft segments into PDMS. Dynamic short-chain hydrogen bonds create phonon-transmitting pathways through non-covalent interaction-coordinated vibrations, nearly doubling the copolymer's intrinsic thermal conductivity versus the baseline. This mechanism leverages the directionality and reversibility of hydrogen bonding to optimize energy transfer efficiency while preserving chain flexibility, making it particularly suitable for thermal management in electronic packaging applications.

The orientational order of liquid crystalline molecules provides long-range pathways for phonon transmission, where the synergistic interplay between their intrinsic ordered structure and polymer chains emerges as a pivotal mechanism for enhancing thermal conductivity, thereby conferring unique value in thermal management applications.^113,114

Qin et al.¹¹⁵ synthesized a novel vanillin-based liquid crystalline epoxy monomer (BEP) and three curing agents (ICA) containing conjugated aromatic imine structures. Compared with the traditional epoxy resin, the performance of the prepared liquid crystal epoxy resin has improved by 65%. Experimental and MD simulations revealed that thermal conductivity scales proportionally with the length of conjugated structures in ICAs. This phenomenon arises because imine bonds form π–π conjugated systems with aromatic groups, while the liquid crystalline ordering of BEP synergizes with the structured arrangement of ICAs to extend phonon mean free paths, thereby increasing thermal conductivity.

Zhang et al.¹¹⁶ developed biphenyl-based liquid crystalline epoxy resins exhibiting both high intrinsic thermal conductivity and flexibility. Key to this performance is the ordered microstructure formed by the liquid crystalline mesogenic units through π–π stacking and self-assembly within the crosslinked network (Fig. 7c). This molecular ordering significantly reduces phonon scattering and increases phonon mean free paths, leading to an intrinsic thermal conductivity of 0.40 W m⁻¹ K⁻¹. Critically, the stable orientation of the liquid crystalline phase remains compatible with the inherent flexibility of the polymer network. Such liquid crystalline epoxy systems offer vital solutions for thermal management in flexible electronics and wearable devices.

3.2.2. Filler-enhanced thermal conductivity. The intrinsically limited thermal conductivity of polymers necessitates strategic regulation of thermal enhancers as the pivotal approach for further advancing the thermal performance of polymer-based composites.¹¹⁷ In filled polymer composites, the fillers are isolated and uniformly dispersed within the polymer matrix. When heat is applied to the material surface, phonons propagate along a matrix–filler–matrix path.¹¹⁸ Due to the significant mismatch in the phonon vibrational spectra between the filler and the polymer matrix, considerable energy is consumed in coupling vibrations at the polymer/filler interface, leading to substantial phonon energy attenuation.¹⁰² A high filler loading can promote the formation of a continuous phonon transmission network through interconnecting and overlapping fillers, which is why highly thermally conductive composites often require a high filler content. However, introducing large amounts of fillers inevitably introduces numerous defects, resulting in enhanced phonon scattering and increased interfacial thermal resistance. Therefore, achieving high thermal conductivity at low filler loading remains a key focus of current research. This section focuses on three core strategies: preconstructed 3D networks, oriented thermal conduction architectures and synergistic filler systems. These strategies center on optimizing the spatial distribution, alignment orientation and interfacial interactions of fillers.
Three-dimensional networks. Template-assisted methods, 3D printing¹¹⁹ and self-assembly techniques¹²⁰ enable precise spatial arrangement of thermally conductive fillers within polymer matrices, forming continuous, interpenetrating 3D thermal pathways. This architecture effectively reduces interfacial thermal resistance and phonon scattering, substantially enhancing composite thermal conductivity.^121–124

For instance, Wu et al.¹²⁵ fabricated a polydopamine (PDA) and silver (Ag)-modified boron nitride (BN-PDA-Ag) porous network via a salt-templated method. EP was subsequently infiltrated into the 3D interconnected BN-PDA-Ag architecture, yielding a high-thermal-conductivity 3D BN-PDA-Ag/EP composite (Fig. 8a). The 3D skeletal structure provides continuous pathways for phonon transport, reducing phonon scattering. Moreover, Ag nanoparticles loaded on BN surfaces act as thermal bridges between adjacent BN platelets, significantly decreasing the thermal contact resistance at filler interfaces. This composite achieves an isotropic thermal conductivity of 1.37 W m⁻¹ K⁻¹, representing a 7.61-fold enhancement compared to pure EP.

Fig. 8 (a) Schematic illustration of preparation and the thermally conductive mechanisms for 3D BN-PDA-Ag/EP composites. Reproduced with permission.¹²⁵ Copyright 2023, Elsevier Ltd. (b) Schematic diagram of the structural design concept of the VSCG network. Reproduced with permission.¹²⁷ Copyright 2024, the Author(s). (c) Schematic of the manufacture procedure of BN/TPU films through the blade coating process and the lamination process. Reproduced with permission.¹³⁴ Copyright 2025, the Author(s). (d) Fabrication of the PBC composites with different compression ratios by using a bidirectional freezing strategy and the schematic diagram of the compression process. Reproduced with permission.¹³⁵ Copyright 2024, Elsevier Ltd.

Meanwhile, foams with continuous 3D networks and abundant interconnected pores serve as excellent templates. Yao et al.¹²⁶ created a 3D hybrid nanofiller (P-PANI@f-Cu) by depositing conductive nanofibers onto functionalized copper foam. In the resulting epoxy composite, the 3D copper network establishes primary thermal pathways, while the nanofibers enhance interfacial compatibility. This synergistic architecture minimizes phonon scattering and interfacial thermal resistance, resulting in a thermal conductivity of 5.467 W m⁻¹ K⁻¹.

As shown in Fig. 8b, Yu et al.¹²⁷ constructed an orthotropic 3D hybrid carbon network (VSCG) by integrating vertically aligned carbon nanotubes (VACNTs) with horizontally oriented graphene films (HOGF). This symmetrically designed framework with an orthogonal alignment maximizes the anisotropic thermal conductivity of graphene and CNTs, establishing macroscopically uniform phonon pathways.¹²⁸ The in-plane and out-of-plane thermal conductivities of VSCG/PDMS composites are 113.61 and 24.37 W m⁻¹ K⁻¹, respectively. Similarly, by welding vertically aligned carbon fiber (CF) arrays with self-assembled high-quality graphene networks, Lu et al.¹²⁰ fabricated a graphene/CF framework. Following polymer backfilling and curing, the resulting graphene/CF/epoxy (G/CF/EP) composite exhibited an excellent through-plane thermal conductivity of 262 W m⁻¹ K⁻¹.

Directional structure. During fabrication, techniques such as the mechanical machining process,¹²⁹ magnetic/electric field alignment¹³⁰ or template guidance induce ordered orientation of nanofibers or lamellar fillers along specific directions. This creates one-dimensional linear or two-dimensional layered thermal pathways, enabling composites to possess anisotropic thermal conductivity that meets directional heat transfer requirements across diverse applications.³⁰

For instance, Zhang et al.¹³¹ fabricated a vertically oriented boron nitride film (BNF) filled silicone rubber (SR) composite via the stacking-cutting method. At 37 vol% BN loading, this composite achieved a record-high through-plane thermal conductivity of 19.1 W m⁻¹ K⁻¹. Xu et al.¹³² fabricated vertically aligned carbon fibers (VCF) with >93% orientation via magnetic induction. The resulting VCF/PDMS composite achieved an ultrahigh in-plane thermal conductivity of 141.57 W m⁻¹ K⁻¹. Dai et al.¹³³ employed a mechanical alignment strategy to fabricate vertically aligned graphene monoliths (VAGM). Subsequent micro-scale gallium LM deposition on both surfaces formed a sandwich-structured thermal interface material (LM-VAGM) with interfacial buffer layers. This architecture exhibits an exceptional through-plane thermal conductivity of 176 W m⁻¹ K⁻¹.

Zhou et al.¹³⁴ fabricated aligned h-BN/thermoplastic polyurethane (TPU) composites using a two-step method involving blade coating and lamination (Fig. 8c). The aligned h-BN platelets form continuous thermal pathways, significantly reducing interfacial thermal resistance and phonon scattering. The combined effect of large h-BN platelets and the lamination process not only optimize filler packing density but also further reduce thermal resistance by minimizing grain boundary defects. The enhanced h-BN orientation yielded a high through-plane thermal conductivity of 43 W m⁻¹ K⁻¹ at an h-BN content of 67 vol%.

Utilizing a bidirectional freeze-casting strategy, Peng et al.¹³⁵ fabricated vertically aligned BNNS/CF skeletons (BCF). After compression to varying ratios and impregnation with polyethylene glycol (PEG), phase-change PBC composites were obtained. The vertically aligned BNNS and CF formed an ordered bridge-structured network (Fig. 8d). Compression reduced the interlayer spacing to 49 μm, constructing dense thermal pathways and minimizing phonon scattering. At a compression ratio of 0.4, the through-plane thermal conductivity of the composite material reached 4.7 W m⁻¹ K⁻¹. When applied for LED chip cooling, this material reduced the operating temperature by 36.3 °C.

Filler synergy. Individual fillers within polymer matrices tend to agglomerate due to high surface energy or polarity mismatch, forming island-like structures. This causes thermal path discontinuity and substantially increases interfacial thermal resistance, paradoxically reducing the efficiency of thermal conductivity enhancement.¹³⁶ Moreover, surpassing the thermal percolation threshold typically demands high-volume fractions of single-component fillers, which compromise matrix flexibility and exacerbate processing challenges, thereby limiting practical applicability.

Combining thermally conductive fillers of different dimensions (e.g., 0D nanoparticles, 1D nanowires, and 2D sheets),^137,138 sizes and various compositions (metallic, ceramic, and organic) leverages size complementarity, interfacial coupling and synergistic thermal conduction effects.³⁴ This strategy constructs multilevel thermal networks that fully utilize the advantages of diverse fillers while overcoming the limitations of single filler systems, thereby providing more flexible design principles for developing high-performance thermal composites for diverse applications.¹⁰¹

BNNSs are promising 2D materials with a theoretical thermal conductivity up to 1700–2000 W m⁻¹ K⁻¹¹³⁹ and are widely employed in high-thermal-conductivity composites.¹⁴⁰ However, their tendency toward planar orientation creates discontinuous out-of-plane thermal pathways and substantial interlamellar interfacial resistance, fundamentally limiting through-plane thermal conductivity enhancements even at high filler loadings.¹⁴¹ Hybridizing BNNSs with other high-aspect-ratio fillers has been widely recognized as an effective strategy to lower thermal percolation thresholds.¹⁴²

For example, the growth of CNTs within the aligned BNNS structure formed a continuous thermal conduction pathway. The resulting composite achieved a through-plane thermal conductivity of 4.83 W m⁻¹ K⁻¹.¹⁴³ Utilizing chemically bonded local networks constructed from 2D BN and 1D glass fibers effectively reduced phonon scattering. At 50 wt% filler loading, the composite exhibited a thermal conductivity of 2.121 W m⁻¹ K⁻¹, representing a 624% enhancement over the pure PA6 matrix.¹⁴⁴

Zhou et al.¹⁴⁵ constructed an aligned heterostructure comprising 2D functionalized BNNSs (FBN) and 1D polydopamine-modified BN nanotubes (FBT) within a polyamide-imide (PAI) composite film, achieving a remarkably high in-plane thermal conductivity of 71.1 W m⁻¹ K⁻¹. As shown in Fig. 9a, the FBT acts as a bridge connecting adjacent FBN sheets, establishing dual-path phonon transport channels characterized by a line-to-plane configuration. This design not only increases the thermal contact area but also reduces interfacial thermal resistance through the continuous nature of the 1D structures.

Fig. 9 (a) Schematic diagram of the heat transmission mechanism in the PBBP composite film. Reproduced with permission.¹⁴⁵ Copyright 2024, Elsevier B.V. (b) Schematic diagram of the interface interaction and internal heat transfer in LM@GN/ANF films. Reproduced with permission.¹⁴⁸ Copyright 2024, Wiley-VCH GmbH. (c) Schematic diagram of hydrogen-bond engineering and phonon transport under low and high interfacial HBD. Reproduced with permission.¹⁵⁹ Copyright 2025, the Royal Society of Chemistry. (d) Schematic diagram of the thermal conduction mechanism and interface covalent bonding and crosslinked network. Reproduced with permission.¹⁶⁴ Copyright 2025, Elsevier B.V.

Dimensional synergy among fillers enables the construction of continuous thermal pathways through geometric matching, simultaneously reducing interfacial thermal resistance.¹⁴⁶ Building upon this, the complementary thermal conduction mechanisms (electron conduction and phonon conduction), polarities and functional properties (electrical insulation, flexibility, and high-temperature resistance) of fillers made from different materials create a synergistic effect. This ultimately enhances the material's ability to meet the demands of complex thermal management scenarios. For instance, thermal interface materials typically require both high thermal conductivity and flexibility. LM, as an emerging class of thermal fillers, offers the advantageous combination of low Young's modulus while maintaining high thermal conductivity, but it is prone to leakage.¹⁴⁷

Luo et al.¹⁴⁸ employed ultrasonication-assisted surface coating to construct a core–shell structure (LM@GN) featuring graphene nanoplatelet (GN)-encapsulated LM nanodroplets, effectively suppressing LM agglomeration and leakage. The GN coating serves as a two-dimensional thermal skeleton, while the LM droplets fill the network voids, creating three-dimensional thermal bridges (Fig. 9b). An interfacial oxide layer between them reduces phonon scattering, yielding a composite film thermal conductivity of 5.67 W m⁻¹ K⁻¹. Concurrently, the LM nanodroplets can radially elongate under external force, enhancing energy dissipation and endowing the composite film with excellent flexibility. The synergistic effects among fillers not only resolve the LM leakage challenge but also achieve simultaneous enhancement of both thermal conductivity and flexibility, providing a new paradigm for flexible thermal management materials.¹⁴⁹

It should be noted that incorporating hybrid fillers into polymer-based thermal composites can reduce the total filler loading, suppress agglomeration, and establish synergistic thermal networks to enhance thermal conductivity.⁸ However, introducing multiple fillers also increases the number of interfaces. If interfacial interactions are weak, this can exacerbate phonon scattering and lead to increased thermal resistance.¹⁵⁰ Optimizing the interfacial contact through covalent bonding or alignment can mitigate thermal resistance issues. Simultaneously, it is crucial to balance filler ratios, interfacial compatibility, and structural orderliness to achieve concurrent improvements in both thermal performance and processability.

3.2.3. Interfacial modification. Thermal resistance at material interfaces arises from lattice vibration mismatches and contact gaps, impeding heat transfer across boundaries.³⁴ Constructing effective connections between fillers and between the filler and the matrix through covalent/non-covalent modification mitigates phonon spectral mismatch, thereby reducing interfacial thermal resistance and enhancing heat conduction efficiency.^101,151
Noncovalent modification. Interfacial modification of thermally conductive enhancers and polymer matrices utilizing non-covalent interactions, such as hydrogen bonding, electrostatic forces¹⁵² and π–π stacking,¹⁵³ is a mild yet effective strategy. These secondary bonding mechanisms strengthen interfacial adhesion between fillers and matrices while preserving material flexibility and processability.¹⁵⁴

Hydrogen bonding serves as a critical interfacial interaction mechanism, where directional bonding between polar groups (–OH and –NH₂) effectively reduces phonon scattering and void defects at interfaces, forming thermal bridges with quasi-covalent character that significantly accelerate heat transfer.^155,156 Reversible hydrogen-bonded crosslinks at the interface also facilitate the self-healing of damage in the composites.¹⁵⁷

Yao et al.¹⁵⁶ utilized hydroxy-rich cellulose to securely anchor BNNSs via a dense hydrogen bonding network. This strong interfacial bonding effectively reduced phonon scattering, resulting in an 8-fold increase in the thermal conductivity of the resulting composite compared to pure epoxy resin. Zhang et al.¹⁵⁸ employed graphene oxide (GO) as an inorganic binder, constructing a GNP-BN skeleton through hydrogen bonding and π–π interactions formed with both graphene nanoplatelets (GNPs) and BN. The final GNP-BN/PDMS composite exhibited a 104.7% enhancement in thermal conductivity over pure PDMS.

Lin et al.¹⁵⁹ regulated the interfacial interaction between BNNSs and the polymer matrix by tuning the hydrogen bond density (HBD). Designated PVA-DX with tunable HBD was prepared by grafting 3,4-dihydroxyphenylalanine (DOPA) onto poly(vinyl alcohol) (PVA). BNNS fillers were then incorporated into matrices containing groups with varying hydrogen bonding capabilities, enabling HBD control from 0.50 to 2.14 mmol cm⁻³. Both experiments and MD simulations revealed that the hydrogen bonding network enhances phonon coupling at the BNNS/matrix interface, reduces interfacial phonon scattering and establishes efficient thermal conduction pathways (Fig. 9c). At a BNNS loading of 70 vol%, the composite with high HBD achieved a thermal conductivity of 51.01 W m⁻¹ K⁻¹. This represents a 45% increase over the system with an HBD of 0.5 mmol cm⁻³ and corresponds to a reduced interfacial thermal resistance of 0.60 × 10⁻⁸ m² K W⁻¹.

Electrostatic interactions enhance thermal conductivity in composites by leveraging attractive forces between charged functional groups on fillers and ionic moieties in polymer matrices. This mechanism promotes uniform filler dispersion and increases the interfacial contact area.¹⁵⁵

Wei et al.¹⁶⁰ utilized the pH-dependent tunable surface charge of PDA to drive the electrostatic attraction-driven alignment of PDA-modified BNNSs (BNNS@PDA) within the interlayers of aramid sheets (ANFS), optimizing the filler topology. The electrostatic force-induced interfacial charge transfer enhanced hydrogen bonding and π–π conjugation between PDA and ANFS, reducing phonon scattering. The composite film exhibits an excellent thermal conductivity of 36.9 W m⁻¹ K⁻¹ and good flexibility (33.2 MJ m⁻³), demonstrating promising potential for application in flexible thermal management.

In addition, studies demonstrated that imidazolium-based ionic liquids with hexadecyl chains form cation–π and CH–π non-covalent bonds with BNNSs, effectively reducing interfacial thermal resistance while establishing continuous thermal pathways. This interfacial engineering enhanced the conductivity to 15.2 W m⁻¹ K⁻¹, a value substantially higher than that of unmodified systems.¹⁵²

Building on this foundation, Li et al.¹⁶¹ implemented sulfonated ionic liquid (s-IL) modification. The imidazolium rings in s-IL maintain cation–π interactions with BNNSs, while their hydroxyl groups simultaneously form OH⋯π bonds with BNNS π-electron systems and hydrogen bonds with ANF. This synergistic multi-interfacial bonding network minimized thermal resistance and established highly efficient continuous conduction pathways, achieving an in-plane conductivity of 23 W m⁻¹ K⁻¹ in s-IL@BNNS₄₀/ANF composites.

Covalent modification. Introducing reactive functional groups into thermally conductive enhancers via chemical reactions to establish covalent bonding with polymer matrices represents a robust interfacial modification strategy.¹⁶² This covalent bonding significantly strengthens the interfacial bonding between the filler and the matrix, effectively reducing interfacial thermal resistance and enhancing thermal conductivity efficiency.³⁴

Wan et al.¹⁶³ employed in situ polymerization to form amide covalent bonds between amine-modified BN (m-BN) and anhydride groups in PI, anchoring m-BN firmly onto PI molecular chains. Meanwhile, during the blade-coating process, m-BN aligned within the plane, forming a scale-like ordered structure. This directional arrangement increases the contact probability between m-BN sheets, which helps to form a complete heat conduction path within the composite material and promotes rapid heat transfer. The in-plane thermal conductivity of the m-BN/PI composite reached 4.43 W m⁻¹ K⁻¹, when the filler content is 30 wt%.

Furthermore, covalent bonding suppresses filler agglomeration, significantly reducing interfacial thermal resistance. The resulting uniform filler dispersion prevents the stress concentration caused by particle clustering, maintaining structural stability during thermal cycling. Liu et al.¹⁶⁴ covalently functionalized BNNSs with benzocyclobutene (BCB) and embedded them into an allyl-functionalized polyarylene ether nitrile (DPEN) matrix. Thermal crosslinking formed covalent bonds between BCB and allyl groups within the DPEN matrix, firmly anchoring the BNNS within the polymer network. This suppressed filler agglomeration and promoted uniform dispersion. The covalent network synergistically enhanced BN's inherent high thermal conductivity, optimizing phonon transfer pathways (Fig. 9d). Consequently, the composite achieved an in-plane thermal conductivity of 1.92 W m⁻¹ K⁻¹.

Covalent bonding at the interface not only significantly enhances filler dispersion and thermal conduction efficiency but also enables multifunctional synergy. Covalent bond-driven interface engineering enhances composite material properties by tailoring interfacial chemical bonding. This approach simultaneously optimizes mechanical performance, electrical insulation and flame retardancy, producing functional materials with high thermal conductivity, toughness, insulation and flame resistance.^165,166 These materials hold significant promise for integrated circuit thermal management.

For example, Zeng et al.¹⁶⁷ anchored fillers firmly within the polybutadiene matrix via borate ester covalent bonds. The interfacial chemical bonding induced the formation of an ordered filler network, establishing efficient thermal pathways. Furthermore, covalent connections between the filler and the matrix effectively suppressed crack propagation, achieving an adhesion fatigue threshold of 1377.23 J m⁻² and endowing the composite with remarkable fatigue crack resistance. Similarly, Liu et al.¹⁶⁸ modified the interface between soy protein isolate (SPI) and GNPs using dynamic borate ester covalent bonds. The ordered dispersion and physical barrier effects of the GNPs, along with the borate ester bond-induced char formation mechanism, yielded a multifunctional composite exhibiting high thermal conductivity, high strength and flame retardancy.

3.3. Radiative cooling materials

Radiative cooling materials represent a revolutionary advancement in thermal management. By leveraging passive radiation, they achieve efficient cooling, offering an innovative approach to address global energy crises and thermal pollution.¹⁶⁹ This technology provides cooling through two mechanisms. On one hand, it reflects the incoming solar radiation (0.3–2.5 μm) to reduce heat absorption. On the other hand, it emits thermal radiation within the 8–13 μm atmospheric transparency window (ATW), dissipating heat directly into the cold universe (∼3 K) (Fig. 10a).¹⁷⁰ This zero-energy, pollution-free cooling mechanism overcomes the energy dependence and carbon emission constraints inherent in traditional active cooling technologies (e.g., compression refrigeration). It possesses critical strategic value in building energy efficiency, electronics thermal management and personal thermal comfort. This section overviews three approaches to enhance the radiative cooling performance of polymer-based composites: tailored structural design, filler incorporation and synergistic cooling.

Fig. 10 (a) The spectra for normalized solar irradiation (AM 1.5G), blackbody thermal irradiation (300 K), and atmospheric transmittance. The ideal emittance (absorbance) spectra of coolers are shown as dashed lines. Reproduced with permission.¹⁷⁰ Copyright 2024, the Author(s). (b) Schematic diagram of PMF–TiO₂–EP; solar reflectance and emissivity of PMF–TiO₂–EP, TiO₂–EP, and EP. Reproduced with permission.¹⁷⁶ Copyright 2024, American Chemical Society. (c) Schematic diagram of the bioinspired concept from fluffs into hybrid films with photonic architectures; reflectivity (black line) in the normalized ASTM G173 global solar spectrum and absorptivity/emissivity in the TASW (red line). Reproduced with permission.¹⁷⁸ Copyright 2020, PNAS. (d) SEM images of metafoam and scattering efficiency of pores with a size of 0.5–9 μm over the solar spectrum. Reproduced with permission.¹⁸⁵ Copyright 2023, the Royal Society of Chemistry. (e) Design of full-scale structure cooling fabric; thermal emission spectrum of the PDMS film and FP-12-6; thermal radiation transmission spectrum of the PE film and FP-12-6. Reproduced with permission.¹⁸⁷ Copyright 2025, the Author(s).

3.3.1. Structural design. Specific C–H/O–H vibrational modes of polymer backbones absorb incident IR photons, inducing lattice vibrations that relax as mid-infrared radiation (predominantly within the 8–13 μm ATW) and thus facilitating passive radiative cooling.¹⁷¹ With inherent advantages including lightweight nature, flexibility, facile processability and robust chemical stability, polymer-based materials serve as ideal matrices for radiative cooling. However, natural polymers typically struggle to simultaneously fulfill the dual optical requirements of high solar reflectance and strong thermal infrared emissivity.¹⁷² To overcome this limitation, recent studies have utilized architected nanostructures, including multilayer films, periodically patterned surfaces and porous scattering networks, to synergistically modulate the spectral properties of polymer composites. This structural engineering enhances both broadband solar reflectance and mid-infrared thermal emissivity, thereby optimizing the radiative cooling performance of the materials.
Multilayer structures. Multilayer structures enhance radiative cooling performance by alternately stacking nanoscale to microscale dielectric layers (polymer layers) with high and low refractive indices, leveraging the principle of constructive interference.¹⁷³ Within the solar spectrum, each pair of high/low refractive index layers is designed with an optical thickness equal to one-quarter or odd multiples of the target center wavelength. This design ensures that light incident from the low-index to high-index medium undergoes a 180° phase shift upon reflection at the interface. Simultaneously, the optical path difference within the layers introduces an additional 180° phase difference, resulting in a total phase difference of 360°. Consequently, constructive interference occurs in the reflection direction, significantly boosting broadband solar reflectance. For the thermal infrared band, thermally infrared-transparent polymers are selected as spacer layers to avoid scattering or absorption losses that would diminish thermal radiation.^174,175

A composite material (PMF–TiO₂–EP) featuring this multilayer design was developed. Its upper section consists of a 128-layer polymer multilayer film (PMF) with alternating PC (high refractive index) and PMMA (low refractive index) layers, achieving an average reflectance of 92.3% in the ultraviolet band (300–400 nm) through optical interference effects. The lower section comprises EP doped with 400 nm TiO₂ particles (TiO₂–EP). As shown in Fig. 10b, the composite exhibits exceptionally high solar reflectance (93.52%) and mid-infrared emissivity (93.21%), demonstrating its potential for widespread applications such as building roofs and automotive transportation.¹⁷⁶

Patterned structures. Patterned micro–nano structures enhance solar band scattering efficiency and suppress radiative heat absorption through geometric design that controls light propagation paths.¹⁷⁷

Inspired by the triangular microtrichia of the longhorn beetle Neocerambyx gigas, Zhang et al.¹⁷⁸ developed a bio-inspired radiative cooling film (Bio-RC). Its core structure features a periodic array of PDMS pyramids with an approximate period of 8 μm, internally dispersed with randomly distributed 2 μm Al₂O₃ particles (Fig. 10c). The pyramid array utilizes total internal reflection to redirect incident light multiple times towards the atmosphere, thereby reducing solar radiation heat absorption. Concurrently, the triangular cross-sectional design overcomes the wavelength selectivity limitations inherent in periodic structures by promoting multiple light scattering within the cavities.^179,180 These synergistic effects enable the film to achieve 95% solar reflectance. Furthermore, the graded refractive index formed by the air–PDMS–Al₂O₃ interface effectively suppresses Fresnel reflection losses, ensuring efficient directional thermal radiation emission into outer space. Tests demonstrated a film cooling power of 90.8 W m⁻², significantly outperforming conventional white paints.

Additionally, the periodic PDMS pyramid array effectively reduces the droplet contact area through densely arranged microprotrusions. The incorporation of Al₂O₃ nanoparticles further results in a micro–nano dual-scale roughness, significantly increasing the gas–liquid interface ratio. As a result, the Bio-RC film achieves a high contact angle of approximately 138°, along with excellent anti-fouling and anti-corrosion properties. These features collectively enhance the material's adaptability in harsh environments, ensuring long-term surface cleanliness and stable radiative cooling performance.^11,181

Moreover, micro-/nano-fabrication of specific surface patterns enables efficient directional radiative cooling through geometric asymmetry.¹⁸² Kirchhoff's law dictates the equality of directional spectral absorptivity and emissivity. Accordingly, Zhou et al.¹⁸³ designed structures with angle-selective absorption/emissivity. Their design features PDMS wedge units coated with aluminum. PDMS exhibits high intrinsic emissivity (∼0.9) near 10 μm, while aluminum is highly reflective with low emissivity (<0.1) in the mid infrared band. Radiation incident toward the wedge opening enters the gap, inducing strong absorption via wedge-gap resonance and enhanced local electromagnetic fields, thereby facilitating thermal emission. Conversely, radiation incident on the aluminum-coated side is largely reflected, suppressing absorption. This creates significant broadband emissivity contrast (Δε ≈ 0.8) between opposing directions. When vertically deployed, the structure directionally emits thermal infrared radiation (8–13 μm) skyward while reflecting ground radiation. Furthermore, embedding iron microparticles enables dynamic adjustment of the wedge rotation angle, modulating the thermal emission angular range and imparting adaptability to environmental changes.

Porous structures. Polymer-based porous films containing abundant nano–micro hierarchical pores can synergistically enhance radiative cooling performance through multi-scale scattering.¹⁸⁴ Within these films, micrometer-scale pores (1–100 μm) serve as efficient scatterers. Their size, comparable to or larger than the wavelengths in the solar spectrum (0.4–1.8 μm), predominantly induces strong scattering of visible to near-infrared light via Mie scattering or geometric optical scattering, while nanopores (<100 nm) primarily enhance scattering of ultraviolet and visible light, particularly blue-violet light, through Rayleigh scattering.¹⁰ The coupling of these dual-scale micro-nano pores synergistically covers the broad solar spectrum from the ultraviolet to near-infrared region. This significantly enhances the solar reflectance of the polymer-based composite material within the solar spectral band. For instance, Fig. 10d demonstrates that the PC–PDMS blend foam featuring hierarchical pores spanning 0.5–9 μm exhibits high scattering efficiency in the solar spectral region, achieving a solar reflectance of up to 97%.¹⁸⁵

By designing pore size distributions concentrated within the solar spectrum (0.3–2.5 μm) while avoiding the mid-infrared band (2.5–25 μm), Mie scattering interference with thermal radiation emission can be effectively prevented. Li et al.¹⁸⁶ fabricated hierarchically porous cellulose acetate membranes (HAPM) with an 82.6% porosity via solvent-template-assisted evaporation-induced phase separation. The HAPM integrate micro-macropores (0.8–2.5 μm) and nano-micropores (200–800 nm). Chemical bond vibrations in cellulose acetate yield 64.1% emissivity within the 8–13 μm band. Simultaneously, the pore structure enables optical transparency, achieving 74.5% transmittance in non-atmospheric windows (4–8 μm and 13–25 μm). This spectral engineering achieves sub-ambient cooling of 14.0 °C under strong solar irradiance and 4.2 °C below ambient temperature under cloudy conditions.

Open-pore structures prevent light trapping within closed pores, enabling direct mid-infrared (2.5–25 μm) thermal radiation transmission/emission. This significantly enhances heat dissipation in cloudy or indoor environments. For instance, Yu et al.¹⁸⁷ fabricated a PVDF–PVP fabric with full-scale hierarchical porosity via electrospinning and ultrasonic etching. The structure integrates 20–170 nm semi-through nanopores, 0.6–1.5 μm microfibers and 1.5–2.5 μm inter-fiber voids (Fig. 10e). This hierarchical design establishes a refractive index gradient between the polymer matrix (PVDF, n = 1.42) and air (n = 1), thereby minimizing interfacial reflection losses. Additionally, C–F bond vibrations in PVDF enable high emissivity (ε ≈ 81%) within the 8–13 μm ATW while permitting ∼25% thermal transmittance. This dual functionality supports both indoor and outdoor radiative cooling. By synergizing scattering, emission and transmission mechanisms, the fabric achieves net cooling powers of 60 W m⁻² (outdoor) and 26 W m⁻² (indoor).

Additionally, porous structures further enhance cooling performance by combining ultralow thermal conductivity with radiative dissipation.¹⁸⁸ Air trapped within pores interacts with the solid matrix, forming a thermal resistance network that impedes environmental conductive heat ingress. This ultralow conductivity is critical for maintaining radiative cooling efficiency under high-temperature conditions by preventing conductive heat gain.¹⁸⁹

3.3.2. Filler introduction. Specific C–H/O–H vibrational modes of polymer backbones absorb incident IR photons, inducing lattice vibrations that relax as mid-infrared radiation (predominantly within the 8–13 μm ATW), thus facilitating passive radiative cooling.

Introducing functional fillers is an effective strategy to enhance the radiative cooling performance of polymer-based composites. Specifically, high-refractive-index inorganic fillers (e.g., TiO₂, Al₂O₃, and ZnO) create optical interfaces with lower-index polymer matrices. This leverages Mie scattering to broaden solar reflectance, minimizing solar heat absorption.¹⁹⁰ Meanwhile, synergistic interactions between fillers and the polymer matrix ensure high emissivity within the critical 8–13 μm ATW, efficiently dissipating heat as infrared radiation to cold outer space.

For instance, Park et al.¹⁹¹ developed Al₂O₃-assisted organic composite (AOC) cooling sheets featuring randomly dispersed Al₂O₃ particles within a polymeric microfiber network. The composite demonstrated exceptional optical performance, achieving 97.9% solar reflectance and 95.2% emissivity within the ATW, which delivered a cooling power of 120.1 W m⁻². During outdoor experimentation with an average solar flux of 612.6 W m⁻², the AOC sheet demonstrated a cooling performance of 4.9 °C below the ambient temperature.

SiO₂ fillers not only reduce solar heat absorption through optical scattering but also, as polar dielectric materials, exhibit strong Fröhlich resonance when their phonon vibration frequencies match infrared photon frequencies.¹⁹² This coherent phonon–electromagnetic coupling induces a peak in the dielectric function's imaginary component (ε′′). Emissivity significantly increases at this resonant peak due to its direct correlation with ε′′.¹⁹³

Zhang et al.¹⁹⁴ fabricated a PTFE/SiO₂@Cotton textile by in situ synthesizing SiO₂ nanoparticles on cotton followed by PTFE coating (Fig. 11a). Through synergistic phonon-enhanced Fröhlich resonance in SiO₂ and C–F bond vibrations in PTFE, this material exhibits high emissivity within the dual ATW (8–13 μm, 16–25 μm), achieving an average emissivity of 97.8% across the 4–25 μm range. Micro-/nano-rough interfaces between SiO₂ and PTFE further enhance infrared scattering. This dual-window emission enables effective heat dissipation in humid environments via the 16–25 μm window, broadening the application potential. Meanwhile, the in situ synthesized SiO₂ forms hydrogen bonds with the cotton fabric surface and the PTFE coating is solidified during the preparation process. As a result, the PTFE/SiO₂@cotton fabric demonstrates robust laundering durability and self-cleaning functionality.

Fig. 11 (a) Working mechanism and emissivity of PTFE/SiO₂@Cotton. Reproduced with permission.¹⁹⁴ Copyright 2025, Elsevier B.V. (b) Composition structure of P-ZIF and schematic diagram of the backscattering enhancement effect of a hollow sphere; comparison of scattering efficiency between the hollow sphere and the solid sphere; the solar reflectance of P-ZIF@PS and P-ZIF. Reproduced with permission.¹⁹⁷ Copyright 2025, Elsevier B.V. (c) Structure of HPCC. Scale bar, 5 μm; spectral reflectance and emittance of the HPCC presented against processed the AM1.5 solar spectrum and the atmospheric transmittance spectrum. Reproduced with permission.²⁰² Copyright 2024, Wiley-VCH GmbH. (d) Schematic illustration of the performance and reflectivity of the hierarchical metafabric; comparison of the temperature measured before and after sweat evaporation under convection. Reproduced with permission.²⁰⁵ Copyright 2022, American Chemical Society.

Beyond traditional inorganic fillers, metal–organic frameworks (MOFs) optimize radiative cooling composites through tunable pore structures, refractive indices and optical bandgaps. Wide-bandgap MOFs effectively reflect energy in the 0.3–2.5 μm range, minimizing heat absorption.¹⁹⁵ Additionally, functional groups and crystalline structures within MOFs enhance light scattering.¹⁹⁶ When dispersed in a polymer matrix, MOFs synergistically enhance solar reflectance via dense interfacial networks and complex microstructures.

Qi et al.¹⁹⁷ synthesized phosphated MOF hollow spheres (P-ZIF) via coordination bonding and sacrificial templating. Because the shell material ZIF has a higher refractive index than air, the hollow spheres exhibit enhanced backscattering, enabling P-ZIF to achieve an exceptionally high solar reflectance of 97.1% (Fig. 11b). The P-ZIF/PDMS composite coating delivered a cooling power of 63.3 W m⁻². Under an average irradiance of 550 W m⁻², the P-ZIF/PDMS composite coating demonstrated a cooling effect of 3.7–4.8 °C compared to commercial coatings across different seasons.

3.3.3. Synergetic cooling. In practical applications, the efficacy of standalone radiative cooling is constrained by complex ambient conditions.¹⁹⁸ Therefore, synergistic integration with other heat dissipation pathways is imperative to overcome the efficiency limitations in thermal management.

Hygroscopic–evaporative cooling removes heat via evaporation latent heat,¹⁹⁹ while radiative cooling dissipates energy through the ATW. Their combination enhances the cooling magnitude and rate, enabling all-weather operation. Moisture stored at night/high humidity enables daytime evaporation without external water,²⁰⁰ extending applicability to arid regions. This renewable-powered synergy provides efficient, sustainable thermal management through passive/low-energy operation, demonstrating transformative potential for universal cooling.²⁰¹

Liu et al.²⁰² fabricated a hybrid passive cooling composite (HPCC) with an MOF-based layered porous structure (Fig. 11c). The bottom adsorption layer uses MOF micro-/mesopores to retain water vapor, while LiCl enhances moisture capture under medium-low humidity. Inter-fiber meso-/macropores reduce vapor transport resistance. The HPCC adsorbs 2.7 kg m⁻² atmospheric moisture at night/high humidity. During daytime/thermal loads, water desorbs and evaporates for cooling. Simultaneously, the top layer provides radiative cooling with 98% solar reflectance and 0.96 mid-infrared emissivity. This synergy delivers an evaporative cooling power of 740 W m⁻², achieving 25.3 °C sub-ambient cooling for electrical equipment like transformers.

Particularly in the field of personal thermal management, sweat evaporation accounts for approximately 30% of total human heat transfer.²⁰³ Its synergistic effect with radiative cooling is crucial for enhancing comprehensive performance in complex environments.

Liu et al.²⁰⁴ developed a three-layer meta-fabric. A silicone elastomer is used as the skin-facing back layer to eliminate insulating air gaps and enhance heat conduction. The intermediate patterned bamboo fiber layer creates asymmetric hydrophilicity with second-level sweat transport. This design synergizes radiative cooling, sweat evaporation and heat conduction. The average temperature of metafabric-covered sweating skin measured only 27.7 °C, representing reductions of 26.2 °C versus bare skin and 10.9 °C relative to sweat-free skin.

Zhang et al.²⁰⁵ designed a bilayer fabric enabling rapid directional sweat transport (100 μL diffusion in <90 s) via the wettability gradient. The bottom layer exhibited a 35° contact angle, while the top layer achieved 0° superhydrophilicity. This configuration allowed the hydrophilic bottom layer to absorb sweat, while the top layer simultaneously provided high solar reflectance (99.16% in visible light and 88.60% in the near-infrared range) and high ATW emissivity. Experimental results demonstrated that the hierarchical metafabric reduced skin overheating by 16.6 °C compared to traditional textiles, with humidity management enhancing the cooling effect by 8.2 °C (Fig. 11d).

3.4. Phase change materials

Applications ranging from heat dissipation in high-power electronics and battery safety to renewable energy utilization all require precise heat flow control.²⁰⁶ However, inherent mismatches in the temporal, spatial and intensity dimensions between thermal energy supply and demand persist, challenging conventional thermal strategies.²⁰⁷ Integrating thermal energy storage, particularly phase change thermal energy storage, has thus emerged as a vital approach for enhancing thermal management efficiency and enabling intelligent regulation.

A key advantage of PCMs is their ability to leverage the substantial latent heat of solid–liquid or solid–solid phase transitions for high-density energy storage/release near constant temperature.²⁰⁸ This enables instantaneous absorption of excess heat to prevent overheating and controlled release on demand to maintain temperature or supplement heat sources, thereby dampening temperature fluctuations and extending device operating ranges.

Organic solid–liquid PCMs (e.g., PEG, paraffin wax (PW), and fatty acids (FA)) offer comprehensive advantages including high energy density and chemical stability, demonstrating promising potential for thermal management applications.²⁰⁹ However, their low intrinsic thermal conductivity and liquid-phase leakage during transition significantly hinder practical deployment.²¹⁰

3.4.1. Thermal conductivity enhancement. Organic phase change materials (OPCMs) are predominantly composed of long-chain or complex functional group compounds held together by weak intermolecular forces, such as van der Waals forces and hydrogen bonding. These materials exhibit inherently low thermal conductivity, primarily due to phonon-dominated heat transfer mechanisms.²¹¹ Even crystalline states display significant thermal anisotropy, with markedly reduced conductivity perpendicular to molecular chains. Crucially, phase transition to the liquid state increases molecular disorder and phonon scattering, further lowering conductivity below solid-phase levels. This low thermal conductivity critically limits OPCM heat charging/discharging rates, particularly in applications demanding a rapid thermal response or high heat flux handling.²¹² Similar to the approaches discussed in Section 3.2, strategies such as constructing thermal networks^213–216 and directional heat transfer pathways^217,218 can be employed to accelerate thermal energy storage/release.

Chen et al.²¹⁹ engineered a 3D porous Co/TLC framework by pyrolyzing the CoZn-ZIF. This process catalytically grew 1D CNTs bridging 2D carbon layers via Co nanoparticles, shortening thermal pathways and minimizing interfacial resistance through high graphitization. After PEG20000 infiltration, the resulting Co/TLC@PEG20000 composite achieved enhanced thermal conductivity (0.48 W m⁻¹ K⁻¹) and accelerated heat charging/discharging rates, yielding photothermal and electrothermal efficiencies of 93.01% and 81.41%, respectively.

Building upon directional architectures, Zhao et al.²²⁰ fabricated a radially aligned graphene aerogel (G-CGA) via bidirectional freeze-casting, inducing long-range graphene sheet alignment with uniform 100–200 μm interlayer spacings. Graphitization at 2800 °C significantly reduced defects (I_D/I_G = 0.12). The G-CGA/PEG composite achieved a thermal conductivity of 2.90 W m⁻¹ K⁻¹. These directional pathways accelerated thermal response, exhibiting excellent photothermal (93.01%) and electrothermal (81.41%) conversion and storage ability.

3.4.2. Leakage suppression. The leakage risk of OPCMs is intrinsically linked to their solid–liquid phase transition process. Exceeding the phase change temperature melts the material, where weak intermolecular forces fail to maintain structural integrity. Low melt viscosity in PCMs like PW and FA allows easy flow under gravity, capillary action or pressure. Furthermore, OPCMs with a low molecular weight or crystallinity show reduced solid-state binding energy and narrow transition ranges, leading to rapid softening/melting and shape instability. Repeated cycling further degrades the microstructure through molecular chain slippage, exacerbating leakage. Molten PCM leakage can contact and corrode electronics, causing short circuits,²⁶ while material loss reduces heat storage capacity and increases interfacial resistance, compromising thermal management and reliability. Consequently, leakage mitigation is critical for maintaining thermophysical functionality.²²¹
Porous medium confinement. 3D porous networks serve dual critical functions: providing high-speed phonon transmission pathways for thermal conduction while simultaneously acting as effective carriers for molten PCMs through robust capillary adsorption forces, thereby effectively suppressing leakage.²²²

Shi et al.²²³ engineered a PW-encapsulated system within a bidirectional carbonized polyimide/Kevlar/graphene oxide@ZIF-67. Its unique bidirectional lamellar structure synergized capillary forces and physical confinement to achieve 81.7% PW loading. The composite delivered high latent heat (>209 J g⁻¹) and retained >99% enthalpy after 200 thermal cycles, demonstrating exceptional reliability under harsh conditions.

Ding et al.²²⁴ impregnated PEG into an interconnected microporous single-walled carbon nanotube nonwoven fabric (SWNWF). As shown in Fig. 12a, the woven micron-scale fibers formed a 3D network, enabling efficient PEG uptake and robust leakage resistance. The SWNWF@PEG membrane achieved an 81.7% PEG loading, maintained structural integrity without leakage at 200 °C, and exhibited outstanding thermal shape stability.

Fig. 12 (a) SEM image of the SWNWF and the surface and the cross-section of SWNWF@PEG. Reproduced with permission.²²⁴ Copyright 2025, Elsevier Ltd. (b) Schematic diagrams of the MEPCM structure; four distribution configurations (S1–S4) and schematic thermal conduction paths of S1–S4. Reproduced with permission.²²⁷ Copyright 2025, Elsevier Ltd. (c) Optically controlled thermal energy storage and release cycle; schematic energy diagram of PCMs containing trans-isomers or cis-isomers, showing enthalpy and temperature changes along the (1)–(4) stepwise process for the thermal energy storage and release cycle. T₁ and T₂ are crystallization points of the composite system, respectively, and ΔT_c is the difference between T₁ and T₂. ΔH_total represents the total energy density of the composite system; chemical structures of PCMs and dopants. Reproduced with permission.²³¹ Copyright 2017, the Author(s). (d) Chemical structure of a-g-Azo PCMs and schematic of the structural transformation during the charging and discharging processes; a structural diagram of the novel three-branch distributed energy recycling LMCD and the schematic diagram of the cyclic transmission process. Mode I: closed loop flow to the low-temperature zone; Mode II: a closed-loop flow toward the room-temperature region; Mode III: a closed-loop flow toward the high-temperature region. Reproduced with permission.²³² Copyright 2024, the Authors.

Microencapsulation. Microencapsulation encapsulates PCMs within micrometer-scale shells via physical/chemical methods, forming a continuous dense barrier.²²⁵ This effectively suppresses liquid leakage during phase change and increases the specific surface area. Moreover, modifying shell materials can significantly enhance the thermal conductivity of microencapsulated PCMs (MEPCMs), accelerating heat storage/release rates and broadening the application scope of OPCMs in thermal management.²²⁶

Wang et al.²²⁷ synthesized MEPCMs with n-octadecane cores and melamine-formaldehyde shells via in situ polymerization. Microencapsulation inherently mitigates liquid PCM leakage. By strategically distributing GO and octadecylamine-grafted GO (GO–ODA) within the microcapsules (Fig. 12b), this approach simultaneously reduced shell defects and established continuous thermal conduction networks. At 0.2 wt% filler loading, the modified MEPCMs exhibited a 480% thermal conductivity increase and an 87.5% leakage reduction versus unmodified samples, achieving synergistic enhancement of leakage resistance and thermal conductivity.

Al-Shannaq et al.²²⁸ developed a photo-induced polymerization process for microencapsulating PCMs within polymer shells using various acrylate-based monomers in a novel thin film UV reactor at room temperature. This technique overcomes limitations of conventional methods requiring high temperatures and a long time. In particular, shells formed with MMA and EAA exhibited smooth surfaces and structural integrity, significantly reducing leakage risks from shell defects.

3.4.3. Photo-triggered phase transition regulation. Organic solid–liquid PCMs store latent heat via solid–liquid phase transitions but lack an effective energy barrier during thermal energy storage. This results in phase change processes governed solely by thermodynamics (occurring spontaneously when the ΔG < 0), precluding autonomous control over heat storage/release behavior.^229,230

Grossman et al.²³¹ pioneered photoresponsive PCMs by integrating azobenzene (Azo) derivatives into tridecanoic acid. As illustrated in Fig. 12c, ultraviolet light (365 nm) triggers trans-to-cis Azo isomerization. The cis isomer disrupts tridecanoic acid molecular packing through steric hindrance and dipole interactions, depressing its crystallization temperature from 38 °C to 28 °C to form a metastable liquid state. This state retains latent heat (∼200 J g⁻¹) for over 10 hours below the original crystallization point. Subsequent visible light (450 nm) reverses isomerization, restoring intermolecular forces to trigger crystallization and on-demand heat release. This photoregulation overcomes spontaneous crystallization limitations in conventional PCMs, enabling molecular-level control of thermal storage for portable thermal management.

Jiang et al.¹⁰⁹ developed a photo-regulated phase change composite (PCC) by covalently integrating long-chain Azo into tetradecanol (TA). Strategic alkyl chain length matching and covalent bonding ensured compatibility and prevented phase separation. Flexible Azo carbon chains enhanced van der Waals forces and hydrogen bonding with TA, enabling precise supercooling control. Crystallinity could be modulated by adjusting the dopant concentration and the cooling rate, achieving a maximum supercooling degree of 7.67 °C. The charged PCC delivered a latent heat of 239 J g⁻¹ and a power density of 594 W kg⁻¹, enabling photo-triggered thermal storage under ambient conditions.

Ge et al.²³² developed a series of photo-responsive alkoxy-grafted azobenzene-based PCMs (a-g-Azo PCMs). Remarkably, even at an ultralow temperature of −63.92 °C, these materials attained a high energy density of 380.76 J g⁻¹. Based on these materials, they further designed a three-branch light-driven microfluidic control device (LMCD) (Fig. 12d). Leveraging temperature-difference-induced asymmetric thermal expansion, the LMCD enabled remote controllable movement of the a-g-Azo PCMs within microchannels. During movement, the optically triggered heat release of a-g-Azo PCMs achieved a temperature difference of 6.6 °C even at a low temperature of −40 °C. This research integrates photo-responsive PCMs with light-driven microfluidic technology, constructing an intelligent thermal management system capable of integrated energy capture, storage, remote transport and on-demand release. It provides an innovative approach for the application of polymer-based thermal management materials in dynamic temperature control under extreme conditions.

4. Summary and perspectives

Polymer-based thermal management materials have established a diversified technological framework encompassing efficient insulation, thermal conductivity enhancement, radiative cooling and phase change energy storage through bio-inspired modeling, molecular engineering and cross-scale design. Their core strengths manifest in multi-scenario innovations: thermal barriers for construction/personal thermal management via synergistic design of cellular-structured polymer matrices and hollow-structured fillers; exponential thermal conductivity improvement through interfacial phonon regulation between high-thermal conductivity fillers and polymer matrices, meeting demanding electronics cooling requirements; radiative cooling capability with high solar reflectance and infrared emissivity, achieved by periodic micro-nano structures with vibrational modes of polymer molecular groups; and broad-temperature-range thermal buffering through blending modification, leveraging inherent compatibility with organic phase change materials.

These advances fully exploit polymers' intrinsic advantages, including lightweight nature, electrical insulation and processability, enabling a decisive transition from theoretical innovation to practical implementation across diverse fields. Despite significant progress in polymer-based thermal management materials, substantial challenges remain to be addressed in future research.

Extreme environmental adaptability

Polymer-based thermal management materials still face critical challenges in extreme-temperature environments. At low temperatures, reduced molecular chain mobility induces matrix embrittlement, compromising both mechanical properties and filler–matrix interfacial compatibility, thereby degrading thermal management performance. At high temperatures, thermal degradation/oxidation reactions cause material deterioration, while weakened filler–matrix interfacial bonding impairs thermal conduction efficiency.

In the aerospace sector, where components are subjected to cyclic thermal stresses ranging from −100 °C to over 200 °C, materials must maintain exceptional thermal and mechanical stability. Selecting polymer matrices with high-bond-energy structures or rigid aromatic segments (e.g., PI) enhances thermal stability, while introducing ether–oxygen bonds is essential for improving low-temperature toughness. Concurrently, the material's architecture must be meticulously designed. A highly stable cross-linked polymer network can provide robust structural support, thereby strengthening the mechanical stability of the composite. In deep-sea exploration equipment, thermal management systems must operate under harsh conditions such as high pressure and low temperature. The incorporation of ionic liquids into the polymer matrix enables stable heat transfer capability across a broad temperature range. Furthermore, biomimetic gradient porous structures or dynamic covalent networks can offer stress buffering and self-healing capability during thermal cycling, significantly extending the service life of the material. Through the coordinated integration of these strategies, polymer-based thermal management materials are expected to achieve efficient and stable thermal regulation in extreme environments.

Intelligence and adaptive regulation

Conventional polymer-based thermal management materials, which are primarily designed for static performance, lack dynamic responsiveness to real-time changes in multiple factors within complex environments. This limitation causes inefficiency in thermal management and misalignment with practical demands, thereby failing to meet the advanced requirements of smart devices. Consequently, developing intelligent and adaptive control capabilities has become imperative.

Achieving this necessitates constructing multi-stimuli-responsive systems. For example, photo-isomerizable units could trigger conformational changes to modulate thermal conductivity or phase transition temperatures. Electrically/magnetically responsive systems could integrate conductive networks or magnetic nanoparticles for dynamic heat transfer regulation via external fields. Thermally adaptive composites combining shape-memory polymers with PCMs could redirect heat flow directionality. Strain-responsive architectures leverage flexible crosslinked networks to adapt thermal transport through programmable chain deformation. Furthermore, the coupling of multiple stimuli could enhance control precision, enabling closed-loop sensing-response-regulation thermal management in intelligent electronics and soft robotics. Ultimately, this integrated capability will facilitate a paradigm shift from passive adaptation to cognitive thermal autonomy.

Driven by dual trends of device miniaturization and intelligence, polymer-based thermal management materials rapidly transition from single-function static materials to multifunctional intelligent systems. Multifunctional integration is not merely a simple combination. It involves a complex interplay of functionalities that may conflict or constrain one another. It is essential to tailor materials to specific application scenarios and implement a multiscale structural design with precise control, spanning from molecular to macroscopic levels. In particular, data-driven design integrates multidimensional datasets to precisely guide material development and shorten R&D cycles, meeting diverse application demands. Future breakthroughs require deep integration of multidimensional synergistic drivers to overcome critical bottlenecks in extreme-environment adaptability, dynamic responsiveness and cross-scale integration. This will provide core support for thermal management revolution in electronics, energy, and construction sectors, while establishing critical foundations for energy-efficient utilization to achieve carbon neutrality goals and sustainable development of the electronics industry.

Author contributions

Jia Li performed an extensive literature study under the supervision of Mengmeng Qin. Wei Feng supervised, reviewed and edited this review. All authors co-wrote the manuscript.

Conflicts of interest

There are no conflicts to declare.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

The Supplementary Information contains the full names corresponding to some abbreviations used in the article. Supplementary information is available. See DOI: https://doi.org/10.1039/d5mh01396h.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (no. 52173078, 52130303 and 52327802) and the National Key R&D Program of China (no. 2022YFB3805702).

Notes and references

1. W. Zhou, K. Dong, Q. Sun, W. Luo, B. Zhang, S. Guan and G. Wang, Int. J. Energy Res., 2022, 46, 11574–11595.
2. Z. Han, W. Sun, K. Yang, H. Yang, Z. Liu, D. Li, C. Yin, H. Liu, Y. Zhao, Z. Ling, Q. Guan and S. Yu, Angew. Chem., Int. Ed., 2023, 62, e202211099.
3. B. R. Mishra, S. Sundaram and K. Sasihithlu, ACS Omega, 2024, 9, 49494–49502.
4. L. Liu, X. Zhang, X. Xu, X. Lin, Y. Zhao, L. Zou, Y. Wu and H. Zheng, Renew. Energy, 2021, 179, 2348–2358.
5. Y. Lin, P. Li, W. Liu, J. Chen, X. Liu, P. Jiang and X. Huang, ACS Nano, 2024, 18, 3851–3870.
6. Y. Cai, H. Yu, C. Chen, Y. Feng, M. Qin and W. Feng, Carbon, 2022, 196, 902–912.
7. C. Liu, M. Wang, J. Wang, G. Xu, S. Zhang and F. Ding, Small, 2024, 20, 2404104.
8. Y. Zhou, X. Tian, X. Cao, Q. Wang, J. Wang, Y. Xu, M. Luo and Z. Wang, Mater. Today Phys., 2025, 54, 101719.
9. H. Zhang, Q. He, H. Yu, M. Qin, Y. Feng and W. Feng, Adv. Funct. Mater., 2023, 33, 2211985.
10. S. Song, C. Hou, A. Yang, L. Wei, H. Liu, D. Xie and Y. Song, Mater. Horiz., 2025, 12, 2535–2544.
11. S. Liu, C. Sui, M. Harbinson, M. Pudlo, H. Perera, Z. Zhang, R. Liu, Z. Ku, M. D. Islam, Y. Liu, R. Wu, Y. Zhu, J. Genzer, S. A. Khan, P.-C. Hsu and J. E. Ryu, Nano Lett., 2023, 23, 7767–7774.
12. F. Zhao, W. Yuan, H. Chen, H. Fu, Z. Li, J. Xiao and Y. Feng, Carbon Energy, 2025, 7, e719.
13. K. Wang, H. Yu, J. Gao, Y. Feng and W. Feng, J. Mater. Chem. C, 2024, 12, 3811–3837.
14. Y. Wang, K. Wang and C. Zhang, Compos. Part B Eng., 2024, 285, 111740.
15. H. Zhang, Q. He, F. Zhang, Y. Duan, M. Qin and W. Feng, Adv. Mater., 2025, 2503140.
16. S. V. Thummalapalli, D. Patil, A. Ramanathan, D. Ravichandran, Y. Zhu, V. Thippanna, M. T. Sobczak, A. Sajikumar, L. B. Chambers, S. Guo, A. M. Kannan and K. Song, Energy Storage Mater., 2024, 71, 103670.
17. M. Wu, Z. Shao, N. Zhao, R. Zhang, G. Yuan, L. Tian, Z. Zhang, W. Gao and H. Bai, Science, 2023, 382, 1379–1383.
18. Y. Cui, H. Gong, Y. Wang, D. Li and H. Bai, Adv. Mater., 2018, 30, 1706807.
19. Y. Wu, J. Chen, C. Xu and C. Chen, Chin. J. Chem., 2025, 43, 1691–1706.
20. J. Luo, Z. Lv, Q. Wang, L. Zhang, Y. Zhong, H. Xu and Z. Mao, Chem. Eng. J., 2024, 495, 153202.
21. N. N. Shi, C. Tsai, F. Camino, G. D. Bernard, N. Yu and R. Wehner, Science, 2015, 349, 298–301.
22. H. Guo, B. Ma, J. Yu, X. Wang and Y. Si, Adv. Fiber Mater., 2025, 7, 106–116.
23. H. Zhang, Z. Xu, Z. Wei, T. Zhang, X. Wang and Y. Zhao, ACS Appl. Mater. Interfaces, 2025, 17, 17580–17589.
24. J. Chen, Y. Zhou, X. Huang, C. Yu, D. Han, A. Wang, Y. Zhu, K. Shi, Q. Kang, P. Li, P. Jiang, X. Qian, H. Bao, S. Li, G. Wu, X. Zhu and Q. Wang, Nature, 2023, 615, 62–66.
25. B. Liu, Y. Zhou, L. Dong, Q. Lu and X. Xu, iScience, 2022, 25, 105451.
26. Y. Kou, K. Sun, J. Luo, F. Zhou, H. Huang, Z. Wu and Q. Shi, Energy Storage Mater., 2021, 34, 508–514.
27. H. Zhang, Y. Guo, Y. Zhao, Q. Zhu, M. He, H. Guo, X. Shi, K. Ruan, J. Kong and J. Gu, Angew. Chem., Int. Ed., 2025, 64, e202500173.
28. Z. Fu, D. Yu, T. Xue, X. Zhang and W. Fan, Compos. Commun., 2025, 56, 102335.
29. L. Wang, J. Feng, S. Zhang, Q. Sun, Y. Luo, J. Men, W. He, Y. Jiang, L. Li and J. Feng, Addit. Manuf., 2023, 71, 103583.
30. Y. Duan, H. Yu, F. Zhang, M. Qin and W. Feng, Nano Res., 2024, 17, 9796–9814.
31. T. Ma, Y. Zhang, K. Ruan, H. Guo, M. He, X. Shi, Y. Guo, J. Kong and J. Gu, InfoMat, 2024, 6, e12568.
32. Z. Yang, S. Jia, Y. Niu, X. Lv, H. Fu, Y. Zhang, D. Liu, B. Wang and Q. Li, Small, 2021, 17, 2101093.
33. Y. Liu, J. Zhou, Z. Ning, H. Huang, Z. Cheng, L. Duan, Y. Wang, X. Tao, P. Liu, Y. Ma and Z. Yao, Adv. Funct. Mater., 2024, 34, 2411573.
34. B. Wei, W. Luo, J. Du, Y. Ding, Y. Guo, G. Zhu, Y. Zhu and B. Li, SusMat, 2024, 4, e239.
35. Y. Yao, H. Chen, Z. Ding, W. Xiao, N. Luo, J. Zeng, L. Tang and K. Chen, J. Phys.:Condens. Matter, 2024, 37, 6.
36. J. Kim, Y. Liu, T. Luo and Z. Tian, ASME J. Heat Mass Transf., 2025, 147, 030801.
37. A. Vasilev, T. Lorenz and C. Breitkopf, Polymers, 2022, 14, 2046.
38. F. Jabbari, A. Rajabpour and S. Saedodin, Chem. Eng. Sci., 2017, 174, 67–81.
39. H. Yu, C. Chen, J. Sun, H. Zhang, Y. Feng, M. Qin and W. Feng, Nano-Micro Lett., 2022, 14, 135.
40. H. Dong, Z. Fan, L. Shi, A. Harju and T. Ala-Nissila, Phys. Rev. B, 2018, 97, 094305.
41. F. Zhang, Y. Sun, L. Guo, Y. Zhang, D. Liu, W. Feng, X. Shen and Q. Zheng, Adv. Funct. Mater., 2024, 34, 2311906.
42. M. Puligheddu, Y. Xia, M. Chan and G. Galli, Phys. Rev. Mater., 2019, 3, 085401.
43. C. Breitkopf, Thermo, 2024, 4, 31–47.
44. A. Chernatynskiy and S. R. Phillpot, Curr. Opin. Solid State Mater. Sci., 2013, 17, 1–9.
45. Ł. Łach and D. Svyetlichnyy, Energies, 2025, 18, 1302.
46. M. N. Z. Abidin and M. Y. Misro, J. Adv. Res. Fluid Mech. Therm. Sci., 2022, 92, 104–115.
47. Q. Wen, Appl. Comput. Eng., 2024, 78, 115–122.
48. M. Qin, Y. Huo, G. Han, J. Yue, X. Zhou, Y. Feng and W. Feng, J. Mater. Sci. Technol., 2022, 127, 183–191.
49. Y. Tomizawa, K. Sasaki, A. Kuroda, R. Takeda and Y. Kaito, Appl. Therm. Eng., 2016, 98, 320–329.
50. Q. He, M. Qin, H. Zhang, J. Yue, L. Peng, G. Liu, Y. Feng and W. Feng, Mater. Horiz., 2024, 11, 531–544.
51. M. Karuppusamy, R. Thirumalaisamy, S. Palanisamy, S. Nagamalai, E. E. S. Massoud and N. Ayrilmis, J. Mater. Chem. A, 2025, 13, 16290–16308.
52. C. Shi, J. Zheng, Y. Wang, C. Gan, L. Zhang and B. W. Sheldon, Atmosphere, 2025, 16, 95.
53. R. Ma, H. Zhang and T. Luo, ACS Appl. Mater. Interfaces, 2022, 14, 15587–15598.
54. S. Wu, Y. Kondo, M. Kakimoto, B. Yang, H. Yamada, I. Kuwajima, G. Lambard, K. Hongo, Y. Xu, J. Shiomi, C. Schick, J. Morikawa and R. Yoshida, Npj Comput. Mater., 2019, 5, 1–11.
55. X. Huang, S. Ma, Y. Wu, C. Wan, C. Y. Zhao, H. Wang and S. Ju, J. Mater. Chem. A, 2023, 11, 20539–20548.
56. C. Yuan, Y. Shi, Z. Ba, D. Liang, J. Wang, X. Liu, Y. Xu, J. Liu and H. Xu, Gels, 2025, 11, 70.
57. R. I. Rabady and A. Ababneh, Optik, 2014, 125, 548–553.
58. A. Sakurai, K. Yada, T. Simomura, S. Ju, M. Kashiwagi, H. Okada, T. Nagao, K. Tsuda and J. Shiomi, ACS Cent. Sci., 2019, 5, 319–326.
59. T. Ma, M. Ma and L. J. Guo, iScience, 2025, 28, 112222.
60. X. Huang, C. Y. Zhao, H. Wang and S. Ju, Mater. Today Phys., 2024, 44, 101438.
61. S. Kim, W. Shang, S. Moon, T. Pastega, E. Lee and T. Luo, ACS Energy Lett., 2022, 7, 4134–4141.
62. C. Zhu, E. A. Bamidele, X. Shen, G. Zhu and B. Li, Chem. Rev., 2024, 124, 4258–4331.
63. M. Li, L. Dai and Y. Hu, ACS Energy Lett., 2022, 7, 3204–3226.
64. Z. Wang, G. Wang, Z. Xu, C. Ma and G. Zhao, Adv. Funct. Mater., 2025, 2504560.
65. J. Yue, Y. Feng, M. Qin and W. Feng, Nano Res., 2024, 17, 883–903.
66. J. Yue, M. Qin, H. Yu, Q. He and W. Feng, Adv. Funct. Mater., 2025, 2508319.
67. F. Du, W. Zhu, R. Yang, Y. Zhang, J. Wang, W. Li, W. Zuo, L. Zhang, L. Chen, W. She and T. Li, Adv. Sci., 2023, 10, 2300340.
68. F. Hu, S. Wu and Y. Sun, Adv. Mater., 2019, 31, 1801001.
69. J. Feng, Z. Ma, J. Wu, Z. Zhou, Z. Liu, B. Hou, W. Zheng, S. Huo, Y. Pan, M. Hong, Q. Gao, Z. Sun, H. Wang and P. Song, Adv. Mater., 2025, 37, 2411856.
70. Q. Xue, N. Dong, P. Fan, D. Lin, G. Tian, J. Liu, J. Wei, S. Qi and D. Wu, Chem. Eng. J., 2024, 493, 152424.
71. J. Men, C. He, L. Wang, J. Feng, Y. Jiang, L. Li, Y. Hu, G. Tang and J. Feng, Chem. Eng. J., 2025, 503, 157989.
72. L. Ge, S. Shang, Y. Ma, T. D. Koudama, K. Yuan, W. Liu and S. Cui, ACS Appl. Mater. Interfaces, 2025, 17, 26091–26116.
73. Y. Shi, L. Li, J. Zhao, X. Zheng, Z. Feng, J. Huang, Z. Liu and Y. Lai, Energy Environ. Mater., 2025, e70046.
74. X. Chen, Z. Zhang, C. Hu, Y. Zhao, S. Wang and C. Huang, Small, 2025, 21, 2502291.
75. L. Guo, H. Deng, Y. Chen, S. Zhang, C. Tian, X. Qiang, S. Liang, J. Wang and N. Song, Chem. Eng. J., 2025, 520, 165671.
76. Z. Liu, T. Liu, H. Dong, B. Yang, X. Li, X. Li, Y. Wu and K. Xu, ACS Nano, 2025, 19, 6826–6839.
77. L. Dou, X. Zhang, X. Cheng, Z. Ma, X. Wang, Y. Si, J. Yu and B. Ding, ACS Appl. Mater. Interfaces, 2019, 11, 29056–29064.
78. R. Zheng, Y. Cheng, X. Jiang, T. Lin, W. Chen, G. Deng, H. N. Miras and Y. Song, ACS Appl. Mater. Interfaces, 2022, 14, 27214–27221.
79. A. Varamesh, Y. Zhu, G. Hu, H. Wang, H. Rezania, Y. Li, Q. Lu, X. Ren, F. Jiang, S. L. Bryant and J. Hu, Chem. Eng. J., 2024, 495, 153587.
80. H. Huang, Y. Lv, X. Jin, H. Wang, C. Wu, Y. Pan, X. Yan, C. Hong, W. Han and X. Zhang, Chem. Eng. J., 2023, 470, 144413.
81. S. Sui, H. Quan, J. Wang, Y. Lu, Y. Yang, Y. Sheng, Z. Sun and Y. Zhang, Adv. Sci., 2025, 12, 2410938.
82. Y. Xu, C. Yan, C. Du, K. Xu, Y. Li, M. Xu, S. Bourbigot, G. Fontaine, B. Li and L. Liu, Compos. Part B Eng., 2023, 261, 110809.
83. Q. Zhang, Q. Xu, D. Yang, X. Wang, M. Zheng, Z. Liu, Z. Zhu, Z. Li, H. Wei, X. Cao and J. Wang, Chem. Eng. J., 2024, 500, 157355.
84. K. Wu, Z. Ye, J. Cheng, Y. Zeng, R. Wang, W. Sun, Y. Kuang, F. Jiang, S. Chen and X. Zhao, Ind. Crops Prod., 2023, 205, 117495.
85. X. An, C. Ma, L. Gong, C. Liu, N. Li, Z. Liu and X. Li, J. Colloid Interface Sci., 2024, 668, 678–690.
86. J. Li, H. Li, J. Lin, Y. Lu, J. Qin, J. Gao, X. Zhang and Q. Zhao, Chem. Eng. J., 2025, 507, 160807.
87. W. Zhang, G. Liang, S. Wang, F. Yang, X. Liu, J. Yu, S. Zhang and B. Ding, Adv. Funct. Mater., 2025, 35, 2412424.
88. R. Li, Y. Bao, R. Guo, L. Gao, W. Zhang, C. Liu and H. Zhao, Small Struct., 2025, 6, 2400439.
89. J. Zhou, E. Wu, Y. Hu, M. Jiang, C. Liu, L. Xue, H. Li, Y. Liu and X. Zhuang, Chem. Eng. J., 2025, 505, 159507.
90. H. Zhan, K. Wu, Y. Hu, J. Liu, H. Li, X. Guo, J. Xu, Y. Yang, Z. Yu, H. Gao, X. Luo, J. Chen, Y. Ni and S. Yu, Chem, 2019, 5, 1871–1882.
91. G. Wang, L. Wang, J. You, Y. Yang, Y. Wang, W. Wang, L. Liu, J. Huang, Y. Li, W. Yu, H. Chen, Y. Lai and W. Cai, Appl. Therm. Eng., 2024, 255, 123959.
92. M. Fuji, C. Takai, H. Watanabe and K. Fujimoto, Adv. Powder Technol., 2015, 26, 857–860.
93. A. Jeong, S. Mo Son, S. Lee and D. Gi Seong, Compos. Part Appl. Sci. Manuf., 2024, 179, 108041.
94. J. Wang, M. Pan, J. Yuan, G. Liu and L. Zhu, ACS Appl. Mater. Interfaces, 2021, 13, 14669–14678.
95. Y. Zhang, Y. Chen, S. Li and W. Bian, J. Alloys Compd., 2025, 1021, 179781.
96. H. Yang, X. Liu, A. Liu, X. Yao, B. Liu and M. Zhang, Chem. Eng. J., 2025, 506, 160205.
97. G. Sun, Z. Guan, Z. Shen, J. Luo and R. Liu, ACS Appl. Eng. Mater., 2024, 2, 593–603.
98. H. Chen, V. V. Ginzburg, J. Yang, Y. Yang, W. Liu, Y. Huang, L. Du and B. Chen, Prog. Polym. Sci., 2016, 59, 41–85.
99. B. Wang, H. Zhang, Q. He, H. Yu, M. Qin and W. Feng, Chin. J. Polym. Sci., 2024, 42, 1002–1014.
100. Y. Dong, H. Yu, Y. Feng and W. Feng, J. Mater. Sci. Technol., 2024, 200, 141–161.
101. J. Zha, F. Wang and B. Wan, Prog. Mater. Sci., 2025, 148, 101362.
102. Y. Zhang, Y. Heo, Y. Son, I. In, K. An, B. Kim and S. Park, Carbon, 2019, 142, 445–460.
103. Y. Guo, K. Ruan, X. Shi, X. Yang and J. Gu, Compos. Sci. Technol., 2020, 193, 108134.
104. G. Lv and J. Zhu, Adv. Funct. Mater., 2025, 2420708.
105. R. Li, X. Yang, J. Li, Y. Shen, L. Zhang, R. Lu, C. Wang, X. Zheng, H. Chen and T. Zhang, Mater. Today Phys., 2022, 22, 100594.
106. C. Wong, S. Wang, L. Ren, J. Xu, X. Zeng and R. Sun, Adv. Funct. Mater., 2024, 34, 2401157.
107. S. Shen, A. Henry, J. Tong, R. Zheng and G. Chen, Nat. Nanotechnol., 2010, 5, 251–255.
108. A. Henry and G. Chen, Phys. Rev. Lett., 2008, 101, 235502.
109. S. Ronca, T. Igarashi, G. Forte and S. Rastogi, Polymer, 2017, 123, 203–210.
110. C. M. Hartquist, B. Li, J. H. Zhang, Z. Yu, G. Lv, J. Shin, S. V. Boriskina, G. Chen, X. Zhao and S. Lin, Nat. Commun., 2024, 15, 5590.
111. J. Gu and K. Ruan, Nano-Micro Lett., 2021, 13, 110.
112. H. Yu, Y. Feng, C. Chen, H. Zhang, L. Peng, M. Qin and W. Feng, Adv. Sci., 2022, 9, 2201331.
113. R. Roy, K. C. Stevens, K. A. Treaster, B. S. Sumerlin, A. J. H. McGaughey, J. A. Malen and A. M. Evans, Mater. Horiz., 2024, 11, 3267–3286.
114. S. Wang, K. Ruan, Y. Guo, J. Kong and J. Gu, Angew. Chem., Int. Ed., 2025, 64, e202501459.
115. J. Qin, Z. Sun, M. Wang, J. Luo, J. Zhang, Q. Liu, W. Liu, H. Zhang and J. Yu, Angew. Chem., Int. Ed., 2025, 64, e202504637.
116. K. Zhang, J. Zhang, L. Dang, Y. Wu, M. He, H. Guo, X. Shi, H. Qiu and J. Gu, Sci. China Chem., 2025, 68, 2615–2627.
117. J. Tan and Y. Zhang, Molecules, 2024, 29, 3572.
118. D. Zhang, J. Zha, C. Li, W. Li, S. Wang, Y. Wen and Z. Dang, Compos. Sci. Technol., 2017, 144, 36–42.
119. X. Zhang, X. Yang, Y. Fan, R. Hu, B. Xie and X. Luo, Int. J. Extreme Manuf., 2024, 7, 025504.
120. X. Lu, J. Liu, C. Shu, S. Zhang, H. Zhao, Y. Zhang, Q. Wang, Z. Yu and X. Li, Adv. Funct. Mater., 2025, 35, 2417324.
121. H. Yu, P. Guo, M. Qin, G. Han, L. Chen, Y. Feng and W. Feng, Compos. Sci. Technol., 2022, 222, 109406.
122. M. Qin, Y. Xu, R. Cao, W. Feng and L. Chen, Adv. Funct. Mater., 2018, 28, 1805053.
123. X. Dong, B. Wan, L. Huang, Q. Zhao, R. Yao, J. Gao, C. Ding, X. Wang, Z.-M. Dang, G. Chen and J. Zha, Adv. Funct. Mater., 2025, 35, 2417843.
124. H. Jiang, Y. Xie, M. He, J. Li, F. Wu, H. Guo, Y. Guo, D. Xie, Y. Mei and J. Gu, Nano-Micro Lett., 2025, 17, 138.
125. X. Wu, H. Yu, L. Wang, X. Gong, D. Chen, Y. Hong and Y. Zhang, Compos. Sci. Technol., 2023, 243, 110268.
126. F. Yao, W. Xie, C. Ma, D. Wang, Z. M. El-Bahy, M. H. Helal, H. Liu, A. Du, Z. Guo and H. Gu, Compos. Part B Eng., 2022, 245, 110236.
127. H. Yu, L. Peng, C. Chen, M. Qin and W. Feng, Nano-Micro Lett., 2024, 16, 198.
128. L. Peng, H. Yu, C. Chen, Q. He, H. Zhang, F. Zhao, M. Qin, Y. Feng and W. Feng, Adv. Sci., 2023, 10, 2205962.
129. W. Dai, T. Ma, Q. Yan, J. Gao, X. Tan, L. Lv, H. Hou, Q. Wei, J. Yu, J. Wu, Y. Yao, S. Du, R. Sun, N. Jiang, Y. Wang, J. Kong, C. Wong, S. Maruyama and C. Lin, ACS Nano, 2019, 13, 11561–11571.
130. J. Lu, X. Ming, M. Cao, Y. Liu, B. Wang, H. Shi, Y. Hao, P. Zhang, K. Li, L. Wang, P. Li, W. Gao, S. Cai, B. Sun, Z. Yu, Z. Xu and C. Gao, ACS Nano, 2024, 18, 18560–18571.
131. H. Niu, H. Guo, L. Kang, L. Ren, R. Lv, L. Liu, A. Bashir and S. Bai, Compos. Part B Eng., 2024, 272, 111219.
132. K. Xu, Y. Wang, Z. Zhang, M. Li, R. Yang, Y. Guo, J. Zhang, B. Zhu, Y. Zhou, X. Wang, Y. Qin, L. Li, T. Cai, W. Dai, C. Lin, K. Nishimura, X. Wu, N. Jiang and J. Yu, Adv. Funct. Mater., 2025, 2505225.
133. W. Dai, X. Ren, Q. Yan, S. Wang, M. Yang, L. Lv, J. Ying, L. Chen, P. Tao, L. Sun, C. Xue, J. Yu, C. Song, K. Nishimura, N. Jiang and C. Lin, Nano-Micro Lett., 2022, 15, 9.
134. Y. Zhou, Y. Wang, M. Li, Y. Qin, R. Yang, K. Xu, Y. Guo, L. Li, Z. Zhang, J. Zhang, B. Zhu, C. Lin, Y. Xu, K. Nishimura, N. Jiang and J. Yu, Adv. Compos. Hybrid Mater., 2025, 8, 237.
135. F. Peng, Y. Song, T. Xu, M. Wang, N. Song, S. Sun and P. Ding, Compos. Part B Eng., 2024, 283, 111617.
136. H. Lin, M. Ma, Q. Chu, L. Xu, S. Chen, Y. Shi, H. He and X. Wang, Compos. Struct., 2023, 312, 116896.
137. R. Chen, X. Li, J. Ma, L. Cheng, F. Wen, L. Li, Y. Bai, Y. He and J. Mu, Compos. Part Appl. Sci. Manuf., 2023, 173, 107633.
138. J. Li, Y. Tao, S. Rao, W. Ma, M. Li, Y. Luo, X. Li and J. Yang, Adv. Funct. Mater., 2025, e08494.
139. L. An, Y. Yu, Q. Cai, S. Mateti, L. H. Li and Y. I. Chen, Prog. Mater. Sci., 2023, 138, 101154.
140. W. Lu, Q. Deng, M. Liu, B. Ding, Z. Xiong and L. Qiu, Nano-Micro Lett., 2023, 16, 25.
141. W. Dai, Y. Wang, M. Li, L. Chen, Q. Yan, J. Yu, N. Jiang and C. Lin, Adv. Mater., 2024, 36, 2311335.
142. A. Bashir, M. Maqbool, A. Usman, U. Younis, A. Z. Khan, Z. Li, C. Liu, D. Chen and S. Bai, Mater. Sci. Eng. R Rep., 2025, 164, 100968.
143. Y. Du, B. Wu, R. Xia and J. Qian, Compos. Commun., 2023, 41, 101636.
144. X. Meng, H. Yu, L. Wang, X. Wu and N. Wang, Compos. Sci. Technol., 2023, 234, 109931.
145. S. Zhou, T. Xu, N. Song, J. Dai, G. Qian and P. Ding, Chem. Eng. J., 2024, 481, 148653.
146. L. Guo, L. Feng, C. Huang, Q. Song, P. Wei, D. Xu, M. Hou and H. Song, Compos. Sci. Technol., 2025, 259, 110954.
147. X. Zhang, Z. Zhang, H. Wang and B. Cao, ACS Appl. Mater. Interfaces, 2023, 15, 3534–3542.
148. W. Luo, B. Wei, T. Luo, B. Li and G. Zhu, Small, 2024, 20, 2406574.
149. J. Xie, G. Zhou, Y. Sun, F. Zhang, F. Kang, B. Li, Y. Zhao, Y. Zhang, W. Feng and Q. Zheng, Small, 2024, 20, 2305163.
150. X. Sun, C. Huang, L. Wang, L. Liang, Y. Cheng, W. Fei and Y. Li, Adv. Mater., 2021, 33, 2001105.
151. H. Yu, C. Chen, Y. Duan, L. Wang, M. Qin and W. Feng, SmartMat, 2025, 6, e70006.
152. X. Li, B. Wu, Y. Lv, R. Xia and J. Qian, J. Mater. Chem. A, 2024, 12, 864–875.
153. Y. Song, N. Jiang, S. Li, L. Wang, L. Bai, J. Yang and W. Yang, Mater. Horiz., 2025, 12, 4260–4273.
154. J. Lu, Y. Fu, Y. He, K. Zheng, F. Sun, J. Zhang, X. Cao and Y. Ma, Carbon, 2024, 226, 119236.
155. H. Ma, B. Gao, M. Wang, Z. Yuan, J. Shen, J. Zhao and Y. Feng, J. Mater. Sci., 2021, 56, 1064–1086.
156. T. Yao, C. Zhang, K. Chen, T. Niu, J. Wang and Y. Yang, Compos. Sci. Technol., 2023, 234, 109934.
157. H. Yu, Y. Feng, C. Chen, Z. Zhang, Y. Cai, M. Qin and W. Feng, Carbon, 2021, 179, 348–357.
158. J. Zhang, H. Wang, T. Zhang, X. Sun, Y. Meng, C. Ma, T. Zhang, N. Lu, C. Liu and Y. Zeng, Compos. Sci. Technol., 2023, 233, 109915.
159. W. Lin, Y. Li, X. Liu, R. Xu, J. Huang, Z. Jiang, Z. Qu, K. Xi and Y. Lin, Mater. Horiz., 2025, 12, 6765–6773.
160. Y. Wei, M. Wen, Y. Wen, Q. Wen, J. Wen, Q. Fu and H. Deng, Chem. Eng. J., 2024, 499, 156069.
161. X. Li, B. Wu, Y. Lv, R. Xia and J. Qian, Chem. Eng. J., 2025, 503, 158246.
162. X. Xie and D. Yang, Ceram. Int., 2023, 49, 17650–17659.
163. B. Wan, X. Li, X. Zeng and J. Zha, Compos. Commun., 2024, 45, 101803.
164. S. Liu, Y. Chen, J. Wu, T. Zhang, L. Tong and X. Liu, Chem. Eng. J., 2025, 518, 164719.
165. F. Wang, X. Dong, G. Liu, B. Wan, J. Gao, B. Liu and J. Zha, Compos. Part Appl. Sci. Manuf., 2024, 185, 108348.
166. F. Wang, X. Dong, G. Liu, J. Gao, X. Wang and J. Zha, Compos. Sci. Technol., 2025, 259, 110963.
167. X. Zeng, J. Fan, X. Zeng, C. Yang and X. Xia, Chem. Eng. J., 2024, 495, 153161.
168. T. Liu, Z. Liu, W. Gu, J. Zhang, S. Gong and J. Li, Compos. Part B Eng., 2025, 291, 111998.
169. J. Liang, J. Wu, J. Guo, H. Li, X. Zhou, S. Liang, C. Qiu and G. Tao, Natl. Sci. Rev., 2023, 10, nwac208.
170. C. Lin, K. Li, M. Li, B. Dopphoopha, J. Zheng, J. Wang, S. Du, Y. Li and B. Huang, Adv. Mater., 2025, 37, 2409738.
171. K. Zhang and B. Wu, Mater. Today Phys., 2025, 51, 101643.
172. M. Lee, G. Kim, Y. Jung, K. R. Pyun, J. Lee, B. Kim and S. H. Ko, Light Sci. Appl., 2023, 12, 134.
173. H. Ma, K. Yao, S. Dou, M. Xiao, M. Dai, L. Wang, H. Zhao, J. Zhao, Y. Li and Y. Zhan, Sol. Energy Mater. Sol. Cells, 2020, 212, 110584.
174. A. P. Raman, M. A. Anoma, L. Zhu, E. Rephaeli and S. Fan, Nature, 2014, 515, 540–544.
175. Z. Chen, L. Zhu, A. Raman and S. Fan, Nat. Commun., 2016, 7, 13729.
176. S. Jiang, M. Li, Z. Hu, F. Zhang, X. Zhang, W. Liu and A. A. A. Omer, ACS Appl. Mater. Interfaces, 2024, 16, 69940–69950.
177. B. Li, Z. Luo, H. Sun, Q. Quan, S. Zhou, W. Yang, Z. Yu and D. Yang, Adv. Funct. Mater., 2024, 34, 2407644.
178. H. Zhang, K. C. S. Ly, X. Liu, Z. Chen, M. Yan, Z. Wu, X. Wang, Y. Zheng, H. Zhou and T. Fan, Proc. Natl. Acad. Sci. U. S. A., 2020, 117, 14657–14666.
179. Q. Pan, M. Wang, Z. Huang, X. Qiu, Y. Wang, F. Zhao, M. Zhu, X. Guo, C. Chen, S. Zhang, J. Wang, Z. He and S. Yu, Adv. Sci., 2025, 12, 2413559.
180. Y. Fu, L. Chen, Y. Guo, Y. Shi, Y. Liu, Y. Zeng, Y. Lin and D. Luo, Adv. Sci., 2024, 11, 2404900.
181. L. Xing, Y. Chen, X. Liu, W. Shao, Y. Li, C. You, X. Huang and B. Sun, Chem. Eng. J., 2025, 515, 163748.
182. F. Xie, W. Jin, J. R. Nolen, H. Pan, N. Yi, Y. An, Z. Zhang, X. Kong, F. Zhu, K. Jiang, S. Tian, T. Liu, X. Sun, L. Li, D. Li, Y. Xiao, A. Alu, S. Fan and W. Li, Science, 2024, 386, 788–794.
183. J. Zhou, T. G. Chen, Y. Tsurimaki, A. Hajj-Ahmad, L. Fan, Y. Peng, R. Xu, Y. Wu, S. Assawaworrarit, S. Fan, M. R. Cutkosky and Y. Cui, Joule, 2023, 7, 2830–2844.
184. W. Su, Z. Guo, Z. Chang, Y. E. W. Li, J. Li, S. Yuan, X. Yao, S. Yan, M. Ma, K. Wang and J. Jiang, Chem. Eng. J., 2025, 503, 158220.
185. Y. Wang, T. Wang, J. Liang, J. Wu, M. Yang, Y. Pan, C. Hou, C. Liu, C. Shen, G. Tao and X. Liu, Mater. Horiz., 2023, 10, 5060–5070.
186. H. Li, C. Liu, X. Yuan, Y. Ma, C. Zhi, H. Li, Y. Hu, L. Xue, G. Yang, X. Zhuang and B. Cheng, Carbohydr. Polym., 2025, 359, 123583.
187. H. Yu, J. Lu, J. Yan, T. Bai, Z. Niu, B. Ye, W. Cheng, D. Wang, S. Huan and G. Han, Nano-Micro Lett., 2025, 17, 192.
188. X. Guo, S. Cheng, W. Cai, Y. Zhang and X. Zhang, Mater. Des., 2021, 209, 109936.
189. C. Ren, Z. Wei, J. Wang, C. Cai, B. Cai, Z. Wang and H. Lei, Nano Energy, 2025, 141, 111076.
190. M. Wu, Y. Li, G. Huang, R. Xu, X. Yin and G. Zhang, Mater. Horiz., 2025, 12, 3024–3035.
191. J. Park, D. Chae, H. Lim, J. Ha, S. Park, H. Sung, C. Park and H. Lee, Adv. Sci., 2025, 12, 2417584.
192. S. Liang, Y. Lu, Q. Zhang, J. Miao, H. Liao, Y. Wu, X. Shi and Y. Zhou, ACS Appl. Mater. Interfaces, 2025, 17, 35902–35911.
193. Y. Zhai, Y. Ma, S. N. David, D. Zhao, R. Lou, G. Tan, R. Yang and X. Yin, Science, 2017, 355, 1062–1066.
194. L. Zhang, Q. Zhu, Y. Zhou, B. Xu, M. Zhou, P. Wang, Q. Wang and Y. Yu, Chem. Eng. J., 2025, 513, 162980.
195. X. Cai, L. Gao, J. Wang and D. Li, ACS Appl. Mater. Interfaces, 2023, 15, 8537–8545.
196. D. V. Lam, D. T. Dung, U. N. T. Nguyen, H. S. Kang, B. Bae, H. Kim, M. Lim, D. Kim, J. Kim and S. Lee, Small Methods, 2025, 9, 2401141.
197. L. Qi, W. Cai, T. Cui, L. Chen, J. Gao, W. Wang, M. Z. Rahman, Z. Gui, B. Fei, Y. Hu and W. Xing, Chem. Eng. J., 2025, 507, 160469.
198. R. Wu, C. Sui, T. Chen, Z. Zhou, Q. Li, G. Yan, Y. Han, J. Liang, P. Hung, E. Luo, D. V. Talapin and P. Hsu, Science, 2024, 384, 1203–1212.
199. Z. Guo, P. Xiong, H. Nan, D. Yan, G. Zhong, J. Lei and Z. Li, Matter, 2025, 102057.
200. N. Xu, S. Ren, J. Chen, Q. Gou, G. Xue, H. Mei, Z. Chen, K. Wang, J. Wang, J. Cui and M. Li, Chem. Eng. J., 2025, 512, 162301.
201. L. Li, Q. Zhang, G. Liu, R. Shi, H. Zhao, K. Huang, Y. Zang, Y. Xu, C. Li, L. Wu and Y. Zhang, Adv. Compos. Hybrid Mater., 2024, 8, 97.
202. X. Liu, P. Li, Y. Liu, C. Zhang, M. He, Z. Pei, J. Chen, K. Shi, F. Liu, W. Wang, W. Zhang, P. Jiang and X. Huang, Adv. Mater., 2024, 36, 2409473.
203. X. Zhang, F. Wang, H. Guo, F. Sun, X. Li, C. Zhang, C. Yu and X. Qin, Adv. Sci., 2024, 11, 2305228.
204. R. Liu, S. Zhao, X. Wu, Z. Zhao, K. Wang, Z. Guo, A. Xi, F. Lan, R. Li, R. Liu, M. Ma, Y. Zang, F. Wang, Q. Jiang, Y. Li, P. Lyu and R. Zhang, ACS Nano, 2025, 19, 826–836.
205. X. Zhang, W. Yang, Z. Shao, Y. Li, Y. Su, Q. Zhang, C. Hou and H. Wang, ACS Nano, 2022, 16, 2188–2197.
206. X. Shao, S. Yang, L. Fan and Y. Yuan, J. Energy Storage, 2023, 68, 107848.
207. B. M. Tripathi, S. K. Shukla and P. K. S. Rathore, J. Energy Storage, 2023, 72, 108280.
208. Y. Zhang, M. M. Umair, S. Zhang and B. Tang, J. Mater. Chem. A, 2019, 7, 22218–22228.
209. Y. Jiang, J. Liu, W. Luo, X. Quan, H. Li, J. Huang and W. Feng, Surf. Interfaces, 2021, 24, 101071.
210. L. Sun, Y. Wang, L. Chen, J. Ying, Q. Li, L. Fu, Q. Yan, K. Wu, C. Xue, J. Yu, N. Jiang, K. Nishimura, C. Lin and W. Dai, Mater. Horiz., 2024, 11, 5031–5044.
211. B. Kalidasan and A. K. Pandey, Prog. Mater. Sci., 2025, 148, 101380.
212. C. Wu, L. Zeng, G. Chang, Y. Zhou, K. Yan, L. Xie, B. Xue and Q. Zheng, Adv. Compos. Hybrid Mater., 2023, 6, 31.
213. S. Xue, G. Zhang, Y. Zhang, K. Wu and Q. Fu, Chem. Eng. J., 2024, 497, 154562.
214. X. Zhang, K. Sun, H. Liu, J. Chen, X. Yan, Y. Kou and Q. Shi, Nano Energy, 2024, 121, 109256.
215. L. Jin, W. Cao, P. Wang, N. Song and P. Ding, Nano-Micro Lett., 2022, 14, 133.
216. C. Zhang, Y. Mao, K. Li, Y. Liu, Z. Xu, K. Pang, S. Cai, L. Fan and C. Gao, Energy Storage Mater., 2025, 75, 104003.
217. B. Hu, H. Guo, T. Li, J. Li, M. Cao, W. Qi, X. Li and B. Li, Chem. Eng. J., 2024, 483, 149240.
218. J. Tong, Z. Liu, Y. Liu, X. Huang and G. Wang, Adv. Funct. Mater., 2025, 35, 2411744.
219. X. Chen, J. Xu, Y. Li, Y. Gao and G. Wang, SusMat, 2023, 3, 510–521.
220. H. Zhao, C. Shu, X. Wang, P. Min, C. Li, F. Gao, X. Li and Z. Yu, Adv. Funct. Mater., 2023, 33, 2302527.
221. M. M. Umair, Y. Zhang, K. Iqbal, S. Zhang and B. Tang, Appl. Energy, 2019, 235, 846–873.
222. C. Shu, H. Zhao, S. Zhao, W. Deng, P. Min, X. Lu, X. Li and Z. Yu, Compos. Part B Eng., 2023, 248, 110367.
223. T. Shi, X. Gao, H. Liu and X. Wang, Nano-Micro Lett., 2025, 17, 236.
224. W. Ding, Y. Liu, Y. Meng, H. Yang, X. Jiao, P. Hou and C. Liu, Carbon, 2025, 234, 120027.
225. E. M. Shchukina, M. Graham, Z. Zheng and D. G. Shchukin, Chem. Soc. Rev., 2018, 47, 4156–4175.
226. X. Lu, R. Qian, X. Xu, M. Liu, Y. Liu and D. Zou, Nano Energy, 2024, 124, 109520.
227. C. Wang, S. Wang, J. Zhang, W. Song and J. Wang, Chem. Eng. J., 2025, 519, 165539.
228. R. Al-Shannaq, M. M. Farid, M. Wasi Ahmad, S. A. Al-Muhtaseb and J. Kurdi, Chem. Eng. J., 2024, 493, 152807.
229. M. Pan, D. Wang, L. Wang, N. Dong, H. Xie and W. Yu, Chem. Eng. J., 2024, 499, 156058.
230. X. Xu and G. Wang, Small, 2022, 18, 2107473.
231. G. G. D. Han, H. Li and J. C. Grossman, Nat. Commun., 2017, 8, 1446.
232. J. Ge, M. Qin, X. Zhang, X. Yang, P. Yang, H. Wang, G. Liu, X. Zhou, B. Zhang, Z. Qu, Y. Feng and W. Feng, SmartMat, 2024, 5, e1300.

---