Design Innovation and Thermal Management Applications of Low-Dimensional Carbon-Based Smart Textiles

Abstract

With the rapid development of wearable electronics, traditional rigid thermal management materials face limitations in flexibility, conformability, and multi-physics adaptability. Low-dimensional carbon materials such as graphene and carbon nanotubes combine ultrahigh thermal conductivity with outstanding mechanical compliance, making them promising building blocks for flexible thermal regulation. This review summarizes recent advances in integrating these materials into textile architectures, mapping the evolution of this emerging field. Key topics include phonon-dominated heat transfer mechanisms, strategies for modulating interfacial thermal resistance, and dimensional effects across scales; beyond these intrinsic factors, hierarchical textile configurations further tailor macroscopic performance. We highlight how one-dimensional fiber bundles, two-dimensional woven fabrics, and three-dimensional porous networks construct multi-directional thermal pathways while enhancing porosity and stress tolerance. As for practical applications, the performance of carbon-based textiles in wearable systems, flexible electronic packaging, and thermal coatings is also critically assessed. Current obstacles—namely limited manufacturing scalability, interfacial mismatches, and thermal performance degradation under repeated deformation—are analyzed. To overcome these challenges, future studies should prioritize the co-design of structural and thermo-mechanical properties, the integration of multiple functionalities, and optimization guided by data-driven approaches. This review thus lays a solid foundation for advancing carbon-based smart textiles toward next-generation flexible thermal management technologies.

1. Introduction

With the growing integration of science and daily life, wearable electronic devices [1], flexible sensors, smart textiles, and miniaturized integrated electronic systems are advancing rapidly at an unprecedented pace [2]. These technologies range from smart wristbands for real-time health monitoring to flexible garments with environmental responsiveness [3,4]. They have been widely adopted in both everyday life and professional fields, exerting a profound influence on lifestyles and industrial structures. At the same time, the pursuit of personalized thermal comfort, increased concerns over device thermal safety, and rising demand for adaptive smart garments have jointly driven flexible thermal management materials to the forefront of research [5]. This topic now lies at the intersection of materials science, textile engineering, and electronic technology.

As flexible systems continue to emerge, conventional rigid thermal management approaches have revealed clear shortcomings in deformation adaptability. Materials widely utilized in thermal regulation applications—such as metallic, ceramic, and polymeric systems—exhibit notable efficiency in conducting heat [6,7,8,9]. However, their high modulus and rigidity [10] result in poor mechanical compatibility with flexible substrates. During bending or stretching, these materials are prone to interfacial delamination and crack formation, which significantly compromise thermal transfer efficiency [11]. In contrast, polymer-based composites offer greater flexibility. However, their thermal conductivity is still insufficient to satisfy the heat dissipation requirements of high-power-density electronic systems operating under transient thermal conditions [12,13]. Moreover, most traditional materials conduct heat isotropically and lack directional thermal control. They also offer limited freedom in designing thermal pathways across complex geometries [14,15], making it difficult to simultaneously ensure thermal performance and user comfort.

In this context, textile structural design—owing to its highly tunable geometries—has become a critical component in flexible thermal management systems [16]. Through methods such as weaving, knitting, and three-dimensional (3D) structuring, textile substrates can form multiscale thermal conduction pathways and enable precise control of heat flow directions [17,18,19,20,21]. Additionally, textiles possess a unique porous architecture that enhances convective cooling and overall heat dissipation. These structures also maintain excellent breathability, skin conformity, and mechanical compliance [22]. This makes them well-suited to satisfy the dual demands of structural adaptability and wearing comfort in flexible systems. The porosity and fiber contact arrangement determined by different fabric patterns—including plain, twill, or satin—directly affect thermal conductivity [23,24,25]. The introduction of 3D woven or spacer fabrics further enables the formation of anisotropic heat pathways [26,27], thereby enhancing overall thermal regulation performance.

To overcome the intrinsic trade-off between thermal conductivity and mechanical flexibility, low-dimensional carbon materials have introduced new technological routes for intelligent thermoregulatory textiles [28,29]. The term “low-dimensional carbon materials” refers specifically to one-dimensional (1D) nanostructures such as carbon nanotubes (CNTs) and two-dimensional (2D) materials such as graphene oxide (GO) and reduced graphene oxide (rGO), which exhibit dimensionally confined phonon transport and mechanical flexibility. These are distinct from the three-dimensional porous textile architectures discussed in later sections. These materials exhibit excellent thermal conductivity, low thermal expansion, tunable flexibility, and high thermal stability. They can be integrated into textile substrates through fiber spinning, coating, or hybrid processing techniques [30,31]. Furthermore, these materials lend themselves to biomimetic design. By emulating features such as the porous fur of polar bears [32], the specialized hair structure of tropical animals [33], and the dynamic coloration capability of squid skin [34,35], functional and structural properties can be synergistically optimized. It is worth noting that “active thermal regulation capability” is one of the key performance dimensions for evaluating smart thermal textiles. They are capable of producing controllable and reversible changes in heat flux in response to external stimuli such as temperature, light, or electricity, thereby enabling active thermal management. These textiles typically integrate functional components such as phase-change materials, photothermal converters, or thermoelectric elements, distinguishing them from conventional passive fabrics that rely solely on intrinsic thermal conductivity. To evaluate the effectiveness of such smart textile systems, several assessment methods have been adopted in recent studies. These include stimulus-responsive thermal regulation tests (e.g., under light, electrical, or mechanical input), infrared thermography for real-time surface temperature mapping [36], heat flux measurement under dynamic conditions [37], and durability assessments after repeated deformation or washing [38]. These metrics help differentiate intelligent textile systems from passive counterparts and are crucial for benchmarking performance in practical applications. However, relying solely on thermal conductivity as an indicator is insufficient to comprehensively assess a material’s adaptability and feasibility within textile systems. Therefore,

Table 1. Comparison of key thermal management properties between traditional textile materials and low-dimensional carbon materials.

Material Type	Thermal Conductivity (W/m·K)	Active Thermal Regulation	Flexibility	Wearability	Commercial Availability	Notes
Cotton	~0.0381–0.0546 [45]	None	High	High	High	Natural cellulose fiber; low intrinsic thermal conductivity
Polyester	--	None	High	High	High	Synthetic fiber (PET); widely used functional textile base
Nylon	~0.25 [46]	None	High	High	High	Engineering plastic fiber
Graphene Textile	363 (in-plane); 553 (axial) [47]	Yes	High	Relatively High	Fair	Textile modified with graphene or GO/rGO functional coating layers
CNT Textile	110 (in-plane); 770 (axial) [39]	Yes	High	Relatively High	Fair	High-alignment structure; superior axial thermal conductivity
PBO Fiber	~0.3 [48]	None	Relatively High	Relatively High	Moderate	High-strength, heat-resistant aromatic heterocyclic fiber
PBO/MXene Composite	25.6–42.2 [49]	Yes	Relatively High	Moderate	Fair	Nanocomposite thermal interface film with enhanced flexibility and conductivity

provides a comparative analysis of representative conventional fabrics, PBO materials, and low-dimensional carbon-based materials across multiple dimensions—including thermal conductivity, active thermal regulation capability, flexibility, wearability, and commercial availability—to highlight the overall advantages of carbon-based materials in achieving high-performance flexible thermal management. For example, Thurid S. Gspann [39] synthesized CNTs films and microfibers via floating catalyst chemical vapor deposition. Through an in situ aggregation approach, highly aligned carbon nanotube microfibers were fabricated, exhibiting a record-high axial thermal conductivity of 770 ± 10 W·m⁻¹·K⁻¹ at ambient temperature, thereby setting a new performance benchmark for directionally aligned CNT microfibers under optimized structural conditions. By introducing various reducing agents during the hydrothermal synthesis process, Mu achieved controlled tuning of the pore structure in graphene aerogels [40]. They used pre-constructed small-pore graphene networks to significantly improve the thermal conductivity of the composite. Using an optimized chemical vapor deposition (CVD) technique, Qi et al. [41] fabricated a graphene network with higher density. Compared to composites containing low-density graphene networks produced by conventional CVD, this composite with paraffin wax (PW) exhibited an 87% increase in thermal conductivity. Beyond the carbon matrix itself, functional additives such as phase-change materials (PCMs) [42] and photothermal conversion agents [43] endow textile systems with active thermal responsiveness and dual-mode temperature regulation (heating and cooling). Structural engineering strategies further enhance synergistic effects, including infrared reflection and convective heat exchange. These advances expand the functional range of textile systems under complex thermal conditions. Despite significant progress, challenges remain in developing intelligent thermoregulatory textiles. These include inadequate thermal stability under repeated deformation, poor wash durability, and limited control over thermomechanical coupling in practical scenarios [44]. To address these challenges, a deeper understanding of the interactions among fiber composition, structural design, and interfacial heat transfer is essential. This foundation is critical for developing next-generation thermal management textiles that are high-performing, scalable, and multifunctionally integrated.

This review systematically summarizes recent advances and methodologies for incorporating low-dimensional carbon materials into flexible textile architectures aimed at enhancing thermal management. Section 2 begins by examining the relationship between phonon transport mechanisms and the thermal–mechanical adaptability of carbon nanomaterials. Section 3 focuses on textile architecture strategies, including multidimensional thermal pathways, porosity control, and interfacial coupling. Section 4 evaluates the thermal performance of these textiles in wearable electronics and flexible packaging applications, highlighting key metrics such as thermal conductivity and mechanical durability. Section 5 concludes with a discussion of current bottlenecks and outlines future directions, including multifunctional integration, industrial application, and data-driven design. This review aims to provide both a theoretical framework and practical guidance for researchers working on intelligent textiles for thermal regulation.

2. Thermal Conductivity and Flexibility of Low-Dimensional Carbon Materials

Heat can be transferred via three primary mechanisms: thermal radiation, convection, and conduction. Among them, conduction is the most common heat transfer process in solid materials. It is a microscopic energy exchange that occurs through the vibration of particles and does not require bulk material movement. The carriers responsible for heat conduction include electrons, phonons, and molecules—collectively referred to as thermal carriers (or heat transport agents). Due to variations in the microstructure and vibrational characteristics of different materials, the type and effectiveness of thermal carriers differ, leading to significant differences in thermal conductivity.

2.1. Flexibility and Mechanical Adaptability of Low-Dimensional Carbon Materials

Graphene is a two-dimensional (2D) material consisting of a single atomic layer of carbon atoms organized in a hexagonal lattice structure. Despite its atomic-scale thickness, it exhibits remarkable mechanical flexibility. This exceptional flexibility originates from the intrinsic elasticity of its carbon–carbon covalent bonds. When subjected to bending or stretching forces, carbon atoms in graphene can rearrange within certain limits while maintaining the structural continuity of the lattice, thus avoiding fracture [50]. As a result, graphene can conform to complex surfaces and irregular substrates. In the fabrication of flexible electronic devices, it can be readily deposited or transferred onto flexible substrates such as polymer films or textiles. Even under small bending radii—down to just a few millimeters—graphene maintains stable electrical and thermal properties [51]. Furthermore, the weak van der Waals forces acting between neighboring graphene layers enable relative sliding when the material undergoes deformation. In multilayer graphene or graphene-based composites, this interlayer slippage can further enhance the overall flexibility. For instance, in graphene–polymer composites, when the material bends, the graphene layers can slide within the polymer matrix, acting as stress-buffering elements that preserve structural integrity under mechanical load [52].

CNTs, which can be conceptualized as rolled-up graphene sheets, possess a hollow cylindrical one-dimensional (1D) morphology. Their distinctive structure provides exceptional flexibility, rendering them ideal candidates for applications in flexible electronics and intelligent textile systems. When assembled into aligned bundles or films, their mechanical robustness is significantly enhanced through intertube interactions, solvent-induced densification, and nanobundle formation, resulting in improved resistance to bending and torsion. For example, ethanol- or acetone-treated CNT films have demonstrated high tensile strength (up to 3.19 GPa) and notable toughness, highlighting their potential for mechanically robust and flexible applications [53]. Their mechanical flexibility mainly arises from the continuous axial alignment of carbon–carbon bonds combined with the cylindrical morphology, which facilitates uniform distribution of applied stress and minimizes failure risks caused by localized stress concentrations. The strong covalent bonding also confers mechanical strength, allowing CNTs to resist deformation. When incorporated into fibers or textile matrices, CNTs can significantly enhance the flexibility of the composite material [54]. In CNT-reinforced polymer fibers, the nanotubes act as flexible reinforcement agents, effectively transmitting stress within the matrix and increasing the fiber’s tolerance to bending and stretching. Such composites demonstrate improved softness and tensile strength, making them ideal for applications in flexible sensors and smart textiles that require both mechanical durability and flexibility [55].

In textile applications, the flexibility and structural adaptability of low-dimensional carbon materials are essential for developing next-generation functional fabrics. Incorporating graphene or CNTs into textile fibers not only enhances mechanical flexibility but also preserves—or even improves—other desirable properties [56]. In thermally responsive smart textiles, these materials enable the fabric to conform to body movements including walking, running, or stretching, thereby improving thermal comfort and deformation compliance [57]. Furthermore, the flexibility of low-dimensional carbon materials allows them to support complex textile architectures. They can be integrated into fabrics through patterned weaving, knitting, or functional coating techniques, enabling custom-designed textile systems with multifunctional capabilities. In certain cases, fibers coated with graphene not only maintain excellent flexibility but also exhibit superior electrical conductivity. This makes them suitable for wearable electronic systems that demand simultaneous mechanical flexibility and electrical functionality, such as in sensing or energy storage applications [58]. In summary, low-dimensional carbon materials possess significant advantages in both flexibility and functional performance. Their unique atomic and molecular structures enable them to deform under mechanical stress while maintaining structural and functional integrity. These properties make them ideal candidates for driving innovation in smart, flexible textile systems.

2.2. Mechanisms of Heat Conduction

In solids, heat is primarily conducted through two mechanisms: lattice vibrations (phonons) and free electron transport. Different materials rely on different dominant carriers, resulting in distinct conduction behaviors. Metallic materials exhibit efficient thermal transport through electrons, owing to their high density of free charge carriers—a phenomenon quantitatively governed by the Wiedemann–Franz law [59]. In contrast, ceramics and carbon-based materials exhibit relatively low electronic carrier concentrations, and phonon-mediated transport becomes the primary heat conduction mechanism.

The phonon transport process is critically influenced by the crystal structure. In a crystalline solid, localized lattice vibrations caused by a heat source generate phonons, which propagate through the lattice via atomic interactions. A phonon is a quantized collective vibration of atoms, characterized by its wave vector and group velocity. During propagation, phonons undergo various scattering events, including phonon–phonon (Umklapp) scattering, phonon–impurity scattering, and phonon–boundary scattering. These scattering mechanisms limit the phonon mean free path (MFP), thus reducing the overall thermal conductivity. According to kinetic theory and the Debye model, the lattice thermal conductivity k of a crystalline solid can be expressed as [60]:

$$k={\displaystyle \frac{C_V{V}_{g}l}{\tau }},$$

(1)

where $C_V$ is the specific heat per unit volume, $V_g$ is the average phonon velocity, $l$ is the phonon mean free path, and $\tau$ is the phonon relaxation time.

Low-dimensional carbon materials possess unique advantages in both structural dimensionality and phonon-dominated thermal transport, resulting in outstanding thermal conductivity. The in-plane thermal conductivity of graphene can reach as high as 3500–5300 W·m⁻¹·K⁻¹, which is 2 to 3 orders of magnitude higher than its cross-plane thermal conductivity [61]. Theoretical and experimental studies show that the phonon MFP in nearly defect-free graphene can reach up to 775 nm [62]. The 2D morphology significantly suppresses phonon scattering, allowing for efficient energy transport via ballistic–diffusive hybrid mechanisms [63]. Nevertheless, the thermal conductivity of graphene diminishes significantly with an increasing number of stacked layers. Ghosh et al. [64] demonstrated that increasing the number of graphene layers from a monolayer to four layers results in a pronounced reduction in thermal conductivity, gradually approaching the value characteristic of bulk graphite. This phenomenon is attributed to enhanced interlayer phonon scattering and stronger van der Waals interactions.

Carbon nanotubes exhibit exceptionally high axial thermal conductivity, with values exceeding 3000 W·m⁻¹·K⁻¹ [63,65]. Nonetheless, their thermal conductivity exhibits strong dependence on structural characteristics such as chirality, diameter, and defect density. Smaller diameters reduce phononboundary scattering, and different chiralities (e.g., armchair vs. zigzag) result in distinct phonon dispersion behaviors, which influence thermal transport. Furthermore, inter-tube interactions within CNT bundles further diminish effective conductivity, posing challenges for practical applications.

These unique thermal transport mechanisms impart low-dimensional carbon materials with tremendous potential for application in flexible textile-based thermal management. They not only enable rapid and efficient heat conduction but also maintain performance stability under mechanical deformation—offering a promising design platform for next-generation intelligent thermoregulatory fabrics. Recent advances in hybrid fabrication techniques, such as the integration of aligned CNT sheets with electrospun polymer nanofibers, have further enhanced their mechanical and functional performance, enabling scalable production of conductive, flexible nonwoven films suitable for wearable thermal regulation [66]. However, the actual thermal performance of CNTs and graphene-based composites is often constrained by structural imperfections and interfacial effects, which are discussed in detail in Section 2.3.

2.3. Factors Influencing Interfacial Thermal Conductance

In thermal management systems, interfacial thermal conductance serves as a key physical parameter that directly influences the effectiveness of thermal energy transfer. It remains a central focus in the field. Interfacial heat transfer is strongly influenced by multiple factors, including surface roughness, microstructural incompatibility, the intrinsic properties of the adjoining materials, and the strength of interatomic bonding across the interface. Therefore, a thorough understanding of these influences is of both theoretical and engineering significance for reducing interfacial thermal resistance and enhancing thermal transfer efficiency.

2.3.1. Influence of Interfacial Defects

In flexible textile systems incorporating low-dimensional carbon materials, interfacial defects represent one of the key variables affecting thermal performance. Typical defect types include vacancies, dislocations, impurity atoms, and structural distortions. These imperfections disrupt lattice order and alter local vibrational modes, thereby modulating phonon propagation and interfacial thermal conductance [67].

Vacancies, as the most common point defects, break lattice periodicity and increase phonon scattering. For instance, vacancy defects in graphene can reduce the phonon mean free path by approximately 30–50%, thereby markedly decreasing its thermal conductivity and increasing interfacial thermal resistance [68]. In the case of carbon nanotube systems, the introduction of single, double, and triple vacancy defects has been shown to reduce the thermal conductivity by 8.1%, 10.9%, and 11.7%, respectively [69]. Moreover, the presence of double vacancies increases the interfacial thermal resistance, with the corresponding Kapitza length rising from 0.7752 nm in pristine carbon nanotubes to 0.9476 nm.

Line defects, such as dislocations, induce widespread lattice distortions. In 1D structures like carbon nanotubes, dislocations interfere with interatomic mechanical coupling and hinder heat flow. Studies indicate that a tenfold increase in dislocation density may lead to an estimated 20% reduction in the thermal conductivity of CNTs [70]. Meanwhile, dislocations induce a local structural transition in carbon nanotubes from body-centered cubic (bcc) to body-centered tetragonal (bct) configurations, thereby altering the phonon dispersion relations. As a result, phonon transport is further disrupted, leading to an increase in interfacial thermal resistance.

Impurity atoms disrupt local mass distribution and the electronic environment, thereby enhancing phonon scattering. For example, boron doping in graphene leads to changes in the C–B bond length and bond angles, which disrupt the electron–phonon coupling and raise interfacial thermal resistance [71]. In the case of bilayer graphene systems, an increase in the impurity concentration $n_i$ from 0 to 0.1 leads to an approximate 15% enhancement in thermal conductivity at $v_i/t_{II}$ = 0.6. Conversely, as the chemical potential $\mu /t_{II}$ increases from 0 to 2, the thermal conductivity decreases by nearly 50%, highlighting the complex and non-monotonic influence of impurity atoms on heat transport behavior [72]. Non-equilibrium molecular dynamics simulations have shown that if the impurity mass lies between that of the two interfacing materials, phonon coupling can be enhanced, improving thermal conductance. Conversely, mass mismatch leads to phonon impedance and reduced thermal transfer [73].

Additionally, structural distortions such as wrinkles or folds alter phonon transport paths and increase scattering probability. It has been reported that wrinkles in graphene–polymer interfaces can raise interfacial thermal resistance by up to 80% [74]. Recent advances in machine-learning-driven atomic simulations have revealed that mechanical stress can trigger interfacial structural reconstructions, significantly altering thermal conductance. These findings emphasize the dynamic nature of interfacial phonon mismatch under deformation, which is highly relevant for mechanically active carbon-based fabrics [75]. In graphene–water composite structures, the introduction of 2% Stone–Wales (SW) defects results in a 30% reduction in interfacial thermal resistance. Furthermore, a 3.5% concentration of single vacancies and 4% nitrogen doping reduce the interfacial thermal resistance to 2.1 × 10⁻⁸  m²⋅K W⁻¹ and 2.4 × 10⁻⁸  m²⋅K W⁻¹, respectively, compared to 2.87 × 10⁻⁸  m²⋅K W⁻¹ for pristine graphene [76].

Vacancies and dopants, while common in synthesized carbon nanomaterials, are relatively less dominant in fiber-scale woven structures due to post-synthesis processing. Nevertheless, their presence remains controllable via precursor purity, growth temperature, and post-treatment. Doping strategies such as nitrogen/boron incorporation should balance thermal conductivity and electrical/chemical functionalities depending on target applications. Wrinkles and structural distortions are commonly present during coating, hot-pressing, and lamination processes and exert the most significant impact. Control strategies include applying slow and uniform coating techniques, low-temperature annealing (<150 °C), solvent vapor smoothing, and interfacial functionalization of carbon materials (e.g., polymer grafting or plasma treatment) to enhance interfacial adhesion and reduce residual stress. Dislocations and Stone–Wales defects appear less often in macroscale yarns or films due to high-temperature annealing steps that allow for structural relaxation. However, they may still occur in mechanically deformed or densely packed CNT bundles.

In summary, the formation mechanisms of interfacial thermal resistance are complex. Their effects depend not only on defect type, density, and spatial distribution but also on the intrinsic properties of the materials on either side of the interface. A detailed understanding of how various defects influence heat transport behavior is essential for advancing defect-engineering strategies in the design of flexible thermal management materials. To improve clarity and systematic presentation,

Table 2. The impact of different defect types on interface thermal resistance.

Defect Type	Material System	Thermal Conductivity Change Δκ (%)	Interface Thermal Resistance Influence	Mechanism Description	Reference
Vacancy	CNT	Single vacancies, double vacancies, and triple vacancies cause the thermal conductivity of carbon nanotubes to decrease by 8.1%, 10.9%, and 11.7%, respectively.	Double-vacancy defects increased the Kapitza length corresponding to the interface thermal resistance from 0.7752 nm to 0.9476 nm.	Phonon scattering enhancement.	[69]
Dislocations	CNT	When the dislocation density increases by 10 times, the thermal conductivity of carbon nanotubes decreases by approximately 20%.	Dislocations cause lattice distortion and phase transformation, disrupting the path of phonons and increasing thermal resistance.	Local electronic structure distortion.	[77]
Impurity atoms (dopants)	Double-layer graphene (AA stacking)	Impurity concentration: when ni increases from 0 to 0.1 and $v_i/t_{II}$ = 0.6, the thermal conductivity of bilayer graphene increases by approximately 15%.	Impurity atoms alter the local mass distribution and electronic environment, thereby enhancing phonon scattering. In addition, the application of a bias voltage modulates the interaction strength between impurities and charge carriers, further influencing thermal transport behavior.	Covalent bond interference and interface scattering are severe.	[72]
Structural defects (Stone–Wales)	Graphene–water interface	A 2% Stone–Wales defect (SW) in the material reduces the thermal resistance at the graphene–water interface by 30%.	The SW defect increases the coupling of low-frequency phonons. A single vacancy enhances the aggregation of water molecules and strengthens hydrogen bonds.	Uneven contact, interruption of heat conduction path.	[76]

summarizes the effects of typical interfacial defects—such as vacancies, dopants, and structural distortions—on the thermal conductivity and interfacial resistance of carbon-based composites. These quantitative comparisons not only reveal the practical thermal effects of different defect modulation strategies, but also provide theoretical guidance and data support for structural optimization and performance prediction of flexible textile materials.

2.3.2. Strength of Interfacial Chemical Bonding

In composite systems based on low-dimensional carbon materials (e.g., graphene or CNTs) combined with organic or inorganic fiber substrates, the strength of interfacial chemical bonding directly determines phonon coupling efficiency across the interface. This makes it a core factor influencing interfacial thermal conductance. Compared with interfaces dominated by weak van der Waals forces, covalently bonded interfaces (e.g., C–O, C–N, or Si–C) significantly enhance phonon transmission, leading to reduced interfacial thermal resistance and enhanced overall heat transfer performance [78].

In recent years, researchers have developed strategies to strengthen interfacial bonding by introducing covalent linkages, ionic bridges, or surface functional group modifications. For instance, functionalization of graphene in graphene–epoxy composites has been shown to enhance interfacial thermal conductance [79]. Such bonding reinforcement suppresses phonon scattering and reflection, improving phonon coupling across the interface.

However, excessive bonding strength can be detrimental in flexible systems. In stretchable thermal management textiles, overly strong interfacial bonding may inhibit mechanical compliance, inducing microcracks or even delaminating the interface under strain [80]. Furthermore, if the vibrational modes of atoms on both sides of the interface are mismatched, strong bonding can lead to mode coupling mismatch, resulting in a “strong bond–low conductance” phenomenon [81].

Therefore, in flexible thermal control systems, interfacial bonding strength must be tuned in coordination with material flexibility. Introducing tunable bonding interfaces—such as thermally responsive polymers or ionic liquid interlayers—can offer a balance between mechanical adaptability and efficient thermal conduction, enabling the design of lightweight, high-performance thermal textiles.

2.4. Size Effects

In flexible thermal management systems built upon low-dimensional carbon materials, the physical dimensions of components play a crucial role in dictating heat transport behavior. This influence becomes especially pronounced at the nanoscale, where thermal transport exhibits non-classical and nonlinear characteristics [82]. Size effects refer to the phenomenon whereby thermal conductivity varies significantly when the geometric dimensions of a material or interface approach the phonon MFP [83]. This effect is particularly significant in materials like graphene and carbon nanotubes, as well as at their heterogeneous interfaces with textile substrates, and it is considered a critical physical mechanism that limits thermal transport performance in such systems.

Traditional heat conduction theory, governed by Fourier’s law, assumes a purely diffusive regime and neglects the discrete nature and scattering behavior of phonons [64]. However, when the characteristic dimensions of low-dimensional systems are smaller than the phonon MFP, heat conduction is primarily governed by quasi-ballistic phonon transport mechanisms. In this regime, interfacial scattering increases, leading to highly directional and spatially non-uniform heat flow. Consequently, interfacial thermal resistance rises significantly [84]. For example, at graphene–polyimide interfaces, when the thickness of graphene is reduced below 10 nm, low-frequency phonons fail to fully develop their vibrational modes and are either reflected or localized, resulting in a substantial drop in thermal conductivity [85].

Molecular dynamics simulations have further elucidated the quantitative relationship between structure size and thermal transport. In polymer composites reinforced with carbon fibers, the interfacial thermal conductivity exhibits a length-dependent increase as the fiber length approaches the phonon MFP—typically in the range of tens to hundreds of micrometers—and eventually stabilizes once ballistic transport is saturated [86]. The sensitivity of this size effect is also modulated by the strength of interfacial interactions. At interfaces characterized by covalent bonding and efficient phonon coupling, thermal conductivity exhibits minimal dependence on size variations. In contrast, for weak interfaces dominated by van der Waals forces, increasing the structural length effectively compensates for higher scattering losses, significantly improving thermal transport efficiency [87].

Interestingly, anomalous size effects may also arise in multilayer composite structures. For instance, in three-layer systems, increasing the length of only the outer carbon fibers—while maintaining the middle layer’s dimensions—can paradoxically reduce overall interfacial thermal conductance [88]. This phenomenon is attributed to a combination of decreased phonon transmittance in the central layer, vibrational mode mismatch, and multiple interfacial scattering events [89]. These findings emphasize that size is not the sole determinant of thermal conductivity. A comprehensive design strategy must also account for interfacial phonon matching and coupling strength.

For flexible thermal management textiles, these size effects have direct engineering implications. Such systems typically span three hierarchical scales: nanometer-scale thermal fillers, micron-scale fiber networks, and macroscopic fabric structures. If the multi-scale heat conduction pathways are not well coordinated, even high-conductivity fillers may fail to deliver optimal performance. Studies have shown that when graphene flakes are too short or too densely packed, they induce periodic phonon scattering zones in the fabric, effectively forming a “thermal resistance barrier” that substantially reduces the overall thermal diffusivity [90]. Experimental and modeling studies indicate that for graphene, significant reductions in thermal conductivity are observed when the lateral dimension falls below approximately 1 μm, primarily due to enhanced boundary scattering and defect accumulation [91]. Similarly, the effective thermal conductivity of CNT networks or bundles decreases markedly when the tube length is less than 5–10 μm, particularly in disordered or poorly aligned structures [92].

Therefore, the design of thermal textile systems must integrate structural dimension optimization with interfacial property control. By precisely tuning the size, distribution density, and spatial arrangement of thermally conductive fillers, it is possible to enhance phonon transmission across multiple interfaces, suppress mismatch-induced scattering, and achieve continuous and efficient heat transport pathways. Future advances are likely to benefit from the integration of microstructure modeling, phonon transport simulation, and intelligent material patterning—providing actionable strategies for the fine-tuning of interfacial thermal conductance in flexible thermal control systems.

3. Textile Structural Design Strategies and Their Thermal Management Advantages

3.1. Classification of Textile Structures

Based on the dimensional characteristics of the preform, textile structural composites are classified into three distinct categories: unidirectional-fiber-reinforced composites (UD), two-dimensional structure-reinforced composites (2DSRCs) and three-dimensional structure-reinforced composites (3DSRCs). The structural morphology is depicted in Figure 1a–c [93]. Unidirectional-fiber=reinforced composites usually arrange fiber bundles in a specific direction into a layered structure by laying or winding. In this type of material, the fibers are highly oriented along the axial direction, resulting in exceptional axial properties and markedly enhancing thermal conductivity by providing a directed pathway for heat transfer. Recently, low-dimensional carbon materials—including graphene and CNTs—have been extensively employed in the development of intelligent textile thermal management systems owing to their exceptional thermal conductivity, superior flexibility, and favorable spinnability [94]. Structural configuration design has become the core means of regulating their performance, including 1D, 2D, 3D, and foam/skeleton porous structures. One-dimensional structures are often composed of graphene fibers or CNT yarns, which can be highly axially oriented through wet-spinning, electrospinning, gel spinning, and other processes to form a stable heat conduction path [95,96]. Research by Gao Chao’s team showed that the specific thermal conductivity of graphene/polymer composite fibers prepared by plastic spinning can reach 553 W·m⁻¹·K⁻¹, which is much higher than that of mesophase asphalt-based carbon fiber composites [47]. The axial thermal conductivity of such one-dimensional carbon-based structures is much higher than that of conventional fiber fabrics, and is used to construct linear thermal conductive devices and wearable fiber cooling systems [97]. The thermal conductivity of graphene-reinforced fiber composites measured along the fiber axis can attain values up to 1290 W·m⁻¹·K⁻¹ [98], significantly exceeding their transverse thermal conductivity. This makes them suitable for scenarios that require unidirectional rapid heat conduction, such as linear heat sinks in flexible circuits. However, their out-of-plane thermal conductivity is limited, often necessitating multi-layer stacking or structural composites to enhance the multi-dimensional heat flow regulation performance.

Two-dimensional reinforced textile structures mainly include woven, knitted, and braided types. In 2D woven structures, plain, twill and satin are typical weaving methods [99], which can improve the in-plane heat diffusion capacity by interweaving warp and weft yarns. Two-dimensional knitted structures include warp-knitted and weft-knitted fabrics, and woven structures mainly include triaxial braided fabrics, which further improve the flexibility of the structure and the uniformity of in-plane performance. Compared with 1D materials, 2D fabrics have more balanced performance in all directions, and the in-plane interwoven network helps to control local heat diffusion. Two-dimensional structures mainly include graphene nanopaper, graphene–CNT interwoven fabrics, and ribbon-woven fabrics, which have excellent in-plane thermal conductivity, flexibility, and mechanical strength. Representative examples include graphene yarn/fiber felt fabrics and graphene-based composite natural fiber materials [100,101]. They are often used to construct continuous thermal conductive networks by stacking or interweaving [102]. In flexible electronic thermal management, graphene nanopaper, as a heat dissipation interface material, can effectively reduce the operating temperature of the chip by more than 10 °C [103]. Graphene–CNT interwoven fabrics combine the lateral thermal diffusion performance of graphene with the longitudinal high thermal conductivity of CNTs. By constructing a multi-scale thermal conduction path through an interwoven configuration, the ability to regulate local high heat flux areas is significantly improved. Gao Chao’s team introduced plasticizers to modulate the interlayer spacing of graphene oxide, expanding it from 1.2 nm to 1.8 nm. This adjustment resulted in a plasticity enhancement of up to 580% alongside a thermal conductivity reaching 1480 W·m⁻¹·K⁻¹ [104]. Within the ribbon-woven structure, graphene ribbons demonstrate excellent lateral expansion ability, which enhances the heat flux output per unit area and facilitates the coordinated optimization of both structural stability and high thermal conductivity. Xingfeng Li developed an epoxy resin composite material exhibiting extremely high thermal conductivity through graphitized carbon fiber felt. At 8.5 wt% filler loading, its in-plane thermal conductivity reached 1.88 W·m⁻¹·K⁻¹, representing an increase of approximately 889.47% compared to pure epoxy resin [105]. Jinhong Yu’s research shows that the in-plane thermal conductivity of epoxy/carbon-coated carbon fiber composites treated at 2300 °C is 13.08 W·m⁻¹·K⁻¹, and they had good out-of-plane thermal conductivity [106]. This type of structure effectively combines high-thermal-conductivity carbon materials with flexible fabric substrates, which not only significantly improves the heat transfer efficiency but also retains good mechanical flexibility and wearing comfort [107]. Xinfeng Wu’s team introduced 35.25 wt% graphene into epoxy/carbon-fiber-based composites, achieving an in-plane thermal conductivity of up to 24.09 W·m⁻¹·K⁻¹, as illustrated in Figure 1d [108]. In this ternary system, graphene serves as a bridging component that effectively connects adjacent carbon fibers, increasing both the contact area and filler continuity. As illustrated in the schematic in Figure 1e [108], the introduction of graphene transforms the isolated, point-to-point heat pathways of the epoxy/CF composite into a highly interconnected network, significantly reducing interfacial thermal resistance. This architecture not only facilitates rapid and efficient phonon transport but also preserves the continuity of heat conduction channels under dynamic bending, addressing the challenge of hotspot formation in deformable electronics. It is difficult for traditional metal heat sinks to be compatible with the deformation requirements of flexible carriers.

Figure 1. Optical images of UD (a), 2D (b), and 3D (c) woven preforms [93]. (d) The thermal properties of the epoxy/CF/G composites [108]. (e) The thermal conduction mechanism of epoxy/CF composites (left) and epoxy/CF/G composites (right) [108].

Figure 1. Optical images of UD (a), 2D (b), and 3D (c) woven preforms [93]. (d) The thermal properties of the epoxy/CF/G composites [108]. (e) The thermal conduction mechanism of epoxy/CF composites (left) and epoxy/CF/G composites (right) [108].

Three-dimensional textile structural composites, unlike 2D laminated structural composites, are characterized by bidirectional shedding and picking operations, enabling full interlacing of warp and weft yarns in both horizontal and vertical directions, as schematically illustrated in Figure 2a [109]. Figure 2 demonstrates three typical three-dimensional structures: orthogonal weaving (Figure 2b), angle interlocking (Figure 2c), and fully interlocking structures (Figure 2d) [110]. These configurations enhance mechanical properties in the thickness direction by incorporating vertical yarns. Compared to 1D or 2D reinforcement structures, 3D fabrics establish continuous heat conduction pathways along the thickness direction, thereby significantly reducing the anisotropy of the thermal conductivity of the composite materials [26,111]. Studies have shown that 3D structures enhance thermal conductivity along the through-thickness direction, but their complex internal structure may also introduce local thermal resistance [112]. Varun Joshi’s team constructed a continuous heat conduction path in the thickness direction of a three-dimensional orthogonal woven composite through z-binder yarns. Its thermal conductivity in the thickness direction reached 0.613 W·m⁻¹·K⁻¹m, which is about 10.3% higher than that of the 2D structure (experimental value 0.556 W·m⁻¹·K⁻¹, prediction error 10.3%) [113].

In application scenarios with more stringent thermal management requirements, 3D textile structures are regarded as a promising strategy for enhancing the overall thermal conductivity of composite fabrics due to their excellent interlayer connection ability, designability of the reinforcement direction, and high volume fraction filling potential [114]. Typical structures include three-dimensional angle interlocking woven structures and three-dimensional rectangular braided structures [115]. These structures construct continuous out-of-plane thermal conduction pathways by interlacing yarns in the thickness direction, while significantly enhancing the material’s compressive and thermal fatigue resistance. Based on multi-scale finite element analysis, Yufen Zhao et al. pointed out that 3D fabric structures exhibit substantially enhanced out-of-plane thermal conductivity compared to conventional two-dimensional counterparts [110]. In addition, their multi-directional interwoven spatial network helps to regulate thermal expansion, effectively disperse the stress concentration under high temperature loads, and reduce the risk of thermal fatigue cracking of the materials.

Figure 2. (a) A plain 3D weave, comprising of warp, horizontal (H), and vertical (V) weft, depicted along its three principal directions and in an isometric view, and the sections (A–C) used for evaluating the modeling results are also indicated [109]. (b) A 2.5D angle-interlock woven fabric [110]. (c) A 2.5D angle-interlock (with warp reinforcement) woven fabric [110]. (d) A 3D orthogonal woven fabric [110].

Figure 2. (a) A plain 3D weave, comprising of warp, horizontal (H), and vertical (V) weft, depicted along its three principal directions and in an isometric view, and the sections (A–C) used for evaluating the modeling results are also indicated [109]. (b) A 2.5D angle-interlock woven fabric [110]. (c) A 2.5D angle-interlock (with warp reinforcement) woven fabric [110]. (d) A 3D orthogonal woven fabric [110].

Foam-like and skeleton-like structures are mainly used to construct high-performance porous thermal interface materials. Typical representatives, including graphene foam and CNT sponge, have the characteristics of high porosity and low density, which can significantly reduce the heat capacity while maintaining favorable thermal conductivity. For example, the porosity of graphene foam exceeds 90%, and its thermal conductivity ranges from 50–200 W·m⁻¹·K⁻¹, showing broad application prospects in the fields of high-efficiency heat sinks and thermal conductive fillers [116,117]. Such structures are usually prepared by processes including template-assisted CVD, laser etching, or pyrolysis carbonization. They have excellent 3D interconnectivity and multi-channel heat flow guidance capabilities and are widely used in key thermal management parts, including chip packaging and Light-Emitting Diode (LED) cooling substrates. To enhance clarity and provide a comparison of different textile structural types, Table 3 summarizes the key thermal and mechanical properties of representative 1D, 2D, 3D, and porous carbon-based textile architectures. Parameters such as axial or in-plane thermal conductivity, mechanical flexibility, ease of integration into wearable systems, and compatibility with current industrial processes are listed to aid in evaluating their suitability for practical applications.

Table 3. Performance comparison of textile structures in different dimensions.

Structure Type	Representative Materials	Thermal Conductivity (W·m−1·K−1)	Mechanical Flexibility	Wearable Integration	Industrial Compatibility
1D	CNT yarns, graphene fibers	~500–1920 [47,97,98]	Excellent (axial), moderate in bundles	Easy (fiber spinning, coating)	Limited (lab-scale)
2D	Graphene/CNT woven fabrics, ribbon-woven meshes	~10–1480 [104,105,106,107]	High	Moderate to easy	Compatible with weaving/knitting
3D	Angle-interlock, orthogonal woven composites	~0.6 (Z-direction) [113]	Moderate	Moderate (requires 3D weaving)	Compatible with 3D textile manufacturing
Porous	Graphene foam, CNT sponge	50–200 [116,117]	good	Low (requires template molding)	Low (complex processing)

Table 3. Performance comparison of textile structures in different dimensions.

Structure Type	Representative Materials	Thermal Conductivity (W·m−1·K−1)	Mechanical Flexibility	Wearable Integration	Industrial Compatibility
1D	CNT yarns, graphene fibers	~500–1920 [47,97,98]	Excellent (axial), moderate in bundles	Easy (fiber spinning, coating)	Limited (lab-scale)
2D	Graphene/CNT woven fabrics, ribbon-woven meshes	~10–1480 [104,105,106,107]	High	Moderate to easy	Compatible with weaving/knitting
3D	Angle-interlock, orthogonal woven composites	~0.6 (Z-direction) [113]	Moderate	Moderate (requires 3D weaving)	Compatible with 3D textile manufacturing
Porous	Graphene foam, CNT sponge	50–200 [116,117]	good	Low (requires template molding)	Low (complex processing)

In addition, the hybrid structure of weaving and knitting is a complementary design scheme that combines the dimensional stability of woven fabrics with the elastic advantages of knitted fabrics. This type of composite fabric can maintain structural integrity and continuity of heat conduction paths under high-strain conditions such as repeated bending and stretching, thereby alleviating the phenomenon of thermal stress concentration. Studies have shown that under high-strain-rate conditions, the yarn network in the hybrid fabric can release stress through stretching and slipping processes while maintaining the stability of heat flux transmission, thereby improving the overall thermal shock tolerance performance [118].

In summary, from 1D orientation heat conduction structures to 2D interwoven configurations and then to 3D texture networks and porous skeleton structures, textile thermal management materials based on low-dimensional carbon materials are rapidly evolving towards multi-scale synergy, functional integration, and structural bionics. Combined with the multifunctional properties of high-thermal-conductivity carbon materials such as graphene and CNTs, along with advancements in high-throughput weaving technology and multi-physics field simulation, thermal management fabrics are expected to exert greater potential in high-end fields including smart wearables, flexible electronic packaging, and aerospace thermal control in the future [106,119,120]. Several three-dimensional (3D) textile architectures, such as angle-interlock woven carbon fabrics and sandwich-structured spacer composites, have demonstrated compatibility with current industrial weaving and warp-knitting platforms. This compatibility is enabled by multi-layer loom systems and programmable jacquard technologies, which facilitate scalable manufacturing, as validated by pilot-scale fabrication trials. Such demonstrations underscore the technical feasibility for integrating these structures into functional textile applications, particularly in thermal management systems. Thermal anisotropy plays a pivotal role in governing thermal comfort and energy regulation within wearable systems. Enhanced in-plane thermal conductivity promotes lateral heat spreading, mitigating localized hotspots and improving thermal uniformity across the fabric surface. Concurrently, optimized through-plane conductivity ensures efficient dissipation of body-generated heat to the environment, maintaining thermal balance during dynamic use. These mechanisms highlight the necessity of strategically tailoring directional heat transport through structure-specific design—such as layer orientation and fiber architecture—to reconcile thermal regulation with wearability in practical textile systems. To further accelerate progress in this field, structural design strategies should focus on the following three directions: (1) Precise control of yarn orientation and textile topology to tailor anisotropic thermal flux. (2) Integration of porous structures and interface engineering to achieve a balanced trade-off between thermal conductivity and mechanical flexibility. (3) Establishment of multiscale modeling and experimental validation platforms to bridge microscale heat transfer mechanisms and macroscale thermal performance. With the convergence of material design, textile engineering, and thermal simulation, flexible thermal textiles are poised to advance toward greater efficiency, intelligence, and scalability in future applications.

3.2. Performance Enhancement Mechanisms in Thermal Management

3.2.1. Construction of Multidimensional Heat Conduction Pathways

In flexible thermal management material systems, improving thermal conductivity has always been the core of research. Traditional polymer matrices are limited by inherently low thermal conductivity, porous microstructures, and substantial interfacial thermal resistance, making it difficult to meet the heat dissipation requirements under conditions with high heat flux density. To address these limitations, researchers have proposed the construction of multidimensional heat conduction pathways. This strategy aims to ensure the continuity of thermal transport networks across both macro- and microscales through synergistic material design and structural engineering, thereby enhancing overall heat transfer efficiency, accelerating thermal response, and ensuring long-term operational stability of devices.

The core concept of this approach is to overcome the constraint of in-plane thermal conduction by building integrated thermal pathways along the X (warp), Y (weft), and Z (through-thickness) directions. Wei et al. [121] designed a knitted graphene sheet (KGS) with a staggered woven structure to construct a multidimensional thermal channel in the surface by orthogonally arranged graphene nanoribbons, as shown in Figure 3a. The structure not only realizes the spatial guidance and retention of local heat flow but also has excellent flexibility and structural stability. The multidimensional thermal channel characteristics and the response to the localization of heat flow provide a new structural paradigm for high-precision flexible thermal sensing and thermal management systems. Similar enhancements were observed in vertically aligned carbon nanotube (VACNT) structures designed by Peng et al. [122]. In this design, VACNTs were grown vertically inside the matrix using CVD and integrated with textile substrates, forming a highly anisotropic thermal network. This configuration efficiently guided interfacial heat flow along the CNT channels and dissipated it outward, substantially reducing local heat accumulation—especially beneficial for thermal regulation in flexible electronic devices.

At the microscale, low-dimensional carbon materials such as graphene and CNTs also play a pivotal role in constructing efficient heat pathways. Li et al. [123] developed a spectrally adaptive smart fabric (SSSF) constructed based on a network of silver nanowires (AgNWs), which realized the dynamic regulation of infrared emissivity from 0.39 to 0.83 by wet–dry state-driven microstructural changes (Figure 3b). Although the study mainly focuses on radiative thermal regulation, the continuous metal network of AgNWs constructed on the fabric surface also provides an effective channel for in-plane heat conduction, which enhances the thermal response rate and heat flow regulation. This work demonstrates the potential of the coupled design of interfacial permeability and structural continuity in the construction of two-dimensional thermal channels. Wu et al. [124] introduced a polydimethylsiloxane (PDMS) cross-linked matrix based on the graphene fabric structure and used its curing shrinkage to induce the self-assembly of graphene layers to form wrinkles, thus establishing a multi-level heat conduction channel from the atomic scale to the macroscopic level. The composite material still maintains a strain response coefficient of more than 700 and stable heat output after repeated stretching and bending, demonstrating the synergistic advantages of thermal channel reconstruction ability and stress adaptability. In addition, the weaving angle plays a critical role in determining the spatial distribution and continuity of thermal conduction pathways within the fabric structure. Jang et al. [125] pointed out that a 60° weaving angle in a three-dimensional woven composite material helps to form the most stable fiber interlaced network, which enhances isotropic thermal conductivity, reduces interfacial thermal resistance, and promotes uniform heat flux distribution, as illustrated in Figure 3c. At the same time, this configuration also has excellent mechanical response and interface bonding characteristics, which helps to effectively disperse the heat density. In terms of multidimensional construction methods, layer-by-layer stacking and interlaced graphene–CNT composite networks are also common strategies. Yang et al. [126] induced the orientation of graphene sheets in graphene oxide fibers by stretching combined with high-temperature graphitization treatment to improve fiber crystallinity, making the thermal conductivity of thermally graphitized fiber attain 1480 W·m⁻¹·K⁻¹ and retaining 99% of the stress in the stretched state, which is a model of the synergistic optimization of thermal–mechanical performance. In the process of three-dimensional composite resin molding, Fu et al. [127] clearly pointed out through multi-physics coupling simulation that reasonable control of the fiber weaving angle and resin flow path can significantly reduce the probability of void formation between fiber bundles, thereby reducing the formation of thermal barriers and improving the continuity and stability of multi-channel thermal conduction networks, as illustrated in Figure 3d.

In summary, the construction of multi-dimensional thermal channels is a multi-level collaborative process involving material direction control, structural design, and interface optimization. Its goal is to achieve efficient thermal coupling between micro-fillers and macro-fabrics by constructing a point–line–surface–body heat flow network, laying the foundation for the development of the next generation of high-performance thermal management fabrics.

Figure 3. (a) Location of a loading point on the KGS and a full atomic schematic diagram of the KGS [121]. (b) Thermal management schematic of the SSSF under resting and physical activity status of the human body [123]. (c) Effect of braiding yarn angle on composite interface and resin impregnation [125]. (d) Relationship between position of voids and resin flow time scalar. Subfigures (a,b) illustrate two contrasting flow regimes: (a) Δt_tow > Δt_Btow, where microscopic voids form inside the fiber tows; (b) Δt_tow < Δt_Btow, where mesoscopic voids form between the tows [127].

Figure 3. (a) Location of a loading point on the KGS and a full atomic schematic diagram of the KGS [121]. (b) Thermal management schematic of the SSSF under resting and physical activity status of the human body [123]. (c) Effect of braiding yarn angle on composite interface and resin impregnation [125]. (d) Relationship between position of voids and resin flow time scalar. Subfigures (a,b) illustrate two contrasting flow regimes: (a) Δt_tow > Δt_Btow, where microscopic voids form inside the fiber tows; (b) Δt_tow < Δt_Btow, where mesoscopic voids form between the tows [127].

3.2.2. Pore Structure Regulation and Flexibility Matching

In the design of flexible thermal management materials, achieving a balance between thermal conductivity and mechanical flexibility remains one of the key challenges. Thermal management fabrics need to adapt to complex morphological surfaces—such as human skin, wearable electronic devices, and curved circuits—necessitating not only high thermal conductivity but also excellent flexibility, conformability, and ductility. Consequently, beyond exhibiting high thermal conductivity, these materials are also required to possess superior bendability, conformability, and stretchability. In this context, the regulation of pore structures plays a central role in enabling such multifunctional performance.

On the one hand, an appropriately engineered porous structure helps reduce the material’s density while improving breathability and wearer comfort. On the other hand, by controlling the pore volume fraction, pore size, and spatial distribution, the flexibility and extensibility of the textile can be enhanced without significantly sacrificing thermal conductivity. Zhao et al. [128] engineered a “meta-louver” programmable textile that responds to humidity via spiral yarns, enabling reversible open–close switching of the fabric and thus dynamically regulating both pore configuration and thermal radiation pathways, as shown in Figure 4a. This humidity–thermal coupled response provides a promising strategy for adaptive textile-based thermal regulation. Li et al. [129] designed a graphene nanoribbon–carbon nanotube woven structure that demonstrates a synergistic balance between thermal responsiveness and structural flexibility at the fabric scale, as shown in Figure 4b. The pore morphology within the structure is uniform and stable, with excellent through-thickness thermal stability maintained in the 200–500 K temperature range. Simulations show that when the graphene nanoribbon width exceeds 2.0 nm, wrinkling is virtually absent, effectively preventing disruption of thermal pathways or pore collapse.

Figure 4. (a) Hierarchically structured meta-louver fabric for dynamic thermoregulation [128]. (b) CGWS high-performance nanosensors [129]. (c) Schematic illustration of a wearable wireless musical instrument made from GWF/PDMS sensors [130]. (d) Schematic diagram of the entire procedure for the fabrication of GWF [131]. (e) Writing laser-induced graphene (LIG) on aramid fabric [132]. (f) Snapshots of HWF impacted by a projectile [133].

Figure 4. (a) Hierarchically structured meta-louver fabric for dynamic thermoregulation [128]. (b) CGWS high-performance nanosensors [129]. (c) Schematic illustration of a wearable wireless musical instrument made from GWF/PDMS sensors [130]. (d) Schematic diagram of the entire procedure for the fabrication of GWF [131]. (e) Writing laser-induced graphene (LIG) on aramid fabric [132]. (f) Snapshots of HWF impacted by a projectile [133].

Furthermore, under applied mechanical stress, the structure exhibits stress-induced reconstruction of thermal pathways, exemplifying the dual benefits of pore regulation and flexibility matching. Liu et al. [130] constructed a graphene woven fabric (GWF) using a template-assisted CVD process and embedded it within a PDMS matrix to produce a flexible composite film. This approach effectively combined an ordered porous architecture with matrix flexibility. At the microscale, the GWF features a well-defined woven porous network, while at the macroscale, it maintains excellent in-plane conformity. Experimental results show that the composite retains stable thermal/electrical transport channels during stretching and bending. Moreover, the tight interfacial bonding between GWF and PDMS effectively prevents pore collapse and structural slippage, enabling a highly integrated system where pore control, flexibility, and thermal stability are co-optimized, as depicted in Figure 4c. Yin et al. [131] used flame reduction to fabricate a reduced graphene oxide (rGO) woven fabric, maintaining the integrity of the graphene network at the microscopic level while forming a typical yarn-interlaced structure macroscopically, and the preparation process is shown in Figure 4d. This architecture provides excellent stretchability and continuous thermal conduction. Under multiple strain cycles, the fabric exhibits robust microcrack self-healing capability. When combined with the mechanical constraint of an elastic substrate, this design effectively buffers the disturbances of the porous architecture during deformation and ensures consistent thermal output under high-strain conditions. Naseri et al. [132] proposed a porous textile strategy that integrates aramid fabrics with laser-induced graphene (LIG) (Figure 4e). By directly engraving conductive networks into the woven structure via laser writing, the continuity of through-thickness thermal pathways was significantly enhanced. The LIG structures were uniformly embedded within the yarn interstices and closely matched the morphology of the original fabric. After PDMS encapsulation, the entire composite retained excellent thermal responsiveness even under repeated mechanical loading. This “pore infiltration–network embedding” mechanism notably improved interfacial coupling strength, offering robust experimental evidence for the advancement of thermal management textiles that combine flexibility, thermal conductivity, and structural stability.

In conclusion, the described pore engineering and structural matching strategies enable the integration of thermal conductivity with mechanical flexibility across multiple length scales. From ordered woven fabrics and composite matrices to nanostructured conductive networks, these designs not only optimize heat flow continuity but also enhance resilience to deformation. Such hierarchical and embedded architectures provide foundational insights for building flexible thermal management systems with long-term stability and adaptability under dynamic service conditions.

3.2.3. Stress Buffering Mechanisms

During the long-term service of thermal management fabrics, frequent mechanical deformations caused by human motion, environmental perturbations, and external loads result in repeated stretching, compression, and bending of the material. If these mechanical stresses are not effectively mitigated, they can lead to fracture or severe distortion of thermal conduction pathways, ultimately compromising heat transfer and reducing the system’s cooling efficiency. Therefore, the integration of stress buffering mechanisms in thermal textile design is a key strategy for ensuring stable thermal performance and extending operational lifespan.

Li et al. [134] proposed a braided graphene belt structure that incorporates riveted and sliding zones at yarn intersections. By leveraging geometrically induced energy dissipation mechanisms, this architecture effectively alleviates local stress concentration. Under mechanical loading, the structure redistributes stress across multiple contact points, significantly delaying crack initiation and the interruption of thermal pathways. Remarkably, the composite maintains a stable electrothermal response under high-strain conditions (>20%). Sang et al. [133] developed a helical CNT fabric composed of stretchable, spring-like carbon nanotube networks. These networks exhibit excellent strain dissipation and elastic recovery under deformation (Figure 4f). Even under extreme stretching up to 500%, the material retains its thermoelectric performance, demonstrating outstanding deformation adaptability and cyclic durability. Doan et al. [135] combined molecular dynamics and micromechanical analysis to reveal that stress redistribution zones can form around graphene boundary defects. These zones act as “structural shock absorbers,” enabling localized phonon pathways to remain continuous under mechanical perturbations. This micro-level energy buffering significantly enhances the system’s stress-response stability and thermal reliability.

In summary, the stress buffering mechanism not only helps to improve the mechanical bearing capacity of flexible fabrics, but also is the key support for maintaining the stability of their multi-dimensional thermal channels and continuous thermal function output. In the future, the structural design of thermal management fabrics should focus on introducing graded stress energy absorption configurations, adaptive response units, and interface slip control strategies to achieve 3D coordinated regulation of force, heat, and shape and promote the development of intelligent responsive thermal control fabric systems.

4. Applications of Flexible Thermal Management and Device Performance

As smart wearable technologies, flexible electronics, and mobile integrated devices continue to advance, conventional thermal management materials have increasingly failed to meet the growing demands for thermal regulation in flexible systems. This limitation arises from their rigidity, elevated interfacial thermal resistance, and poor structural adaptability. Intelligent textile structures based on low-dimensional carbon materials, graphene, and CNTs have demonstrated promising potential across various application scenarios. These structures benefit from customizable architectures, high thermal conductivity, and excellent adaptability to mechanical deformation. In this section, the applications of flexible thermal management systems in wearable technologies are reviewed and summarized systematically, focusing on representative scenarios and the performance of key devices.

4.1. Thermal Management Applications in Wearable Devices

Wearable electronic devices have seen extensive application in areas including healthcare, military protection, motion monitoring, and human–computer interaction. These devices are required not only to integrate sensing, communication, and energy modules but also to manage heat generated during operation and to mitigate thermal disturbances arising from environmental fluctuations. On the one hand, continuous operation of electronic components can cause localized temperature increases. Delayed heat dissipation can lead to thermal discomfort or a decline in device performance. On the other hand, dynamic fluctuations in ambient temperature and human physiological temperature can also influence both user comfort and system stability. Therefore, the development of material systems that combine flexibility, high thermal stability, and multifunctional thermal regulation capability is considered a key approach toward high-performance wearable thermal management systems.

In recent years, flexible thermoelectric (TE) materials have attracted increasing interest owing to their capability to convert temperature gradients between the human body and the environment directly into electrical energy. These materials function not only as energy harvesters but also localized heat flow sensors and regulation units. Zhao et al. [136] developed a stretchable and self-healing n-type SWCNT/PEI/PVP composite film. After being cut, the material was able to recover 89.4% of its original power factor. A Seebeck coefficient of −28 μV·K⁻¹ and a power factor of 21.2 μW·m⁻¹·K⁻² have been documented. Under a temperature gradient of 11 K, more than 90% of the output was recovered. The system was demonstrated to be capable of stable power generation even under structural damage, making it a robust and recyclable energy harvesting unit suitable for prolonged dynamic wear. Ryan et al. [137] engineered an all-organic thermoelectric textile by coating multi-walled carbon nanotubes and polyvinylpyrrolidone onto polyester yarns, forming an air-stable n-type conductive yarn. Through pairing with p-type poly(3,4-ethylenedioxythiophene): poly(styrenesulfonate) (PEDOT:PSS) yarn, a thermoelectric array of 38 couples was constructed. Utilizing the natural temperature gradient between human skin and ambient air, the device produced an open-circuit voltage of 143 mV and a power output of 7.1 nW, demonstrating the feasibility and practicality of organic-based flexible thermoelectric energy harvesting for low-grade heat sources. Ding et al. [138] developed a 2D-fabric-based thermoelectric structure by coating cotton strips with a PEDOT:PSS/CNT/Bi₂Te₃ composite layer. Additional thermal pads were introduced at the top and bottom to convert the vertical temperature gradient into an in-plane heat flow. An output voltage of 4.69 mV was achieved at a ΔT of 30 K. The device maintained excellent flexibility and bend resistance, surviving 120 bending cycles without degradation. This result demonstrated the crucial role of structural design in improving the responsiveness and stability of thermoelectric textiles. Yu et al. [139] developed a phase-change-enhanced thermoelectric textile using carbon nanotube fibers modified with polyethyleneimine and gold nanoparticles, as shown in Figure 5a. Microencapsulated phase-change materials (MPCMs) were applied to the hot end, enhancing thermal energy utilization by 25.7%. The maximum power output of the device reached 270 nW under a ΔT of 6.6 K, validating a design concept in which thermal buffering and thermoelectric effects are synergistically optimized. Sun et al. [140] fabricated a “fabric-type” thermoelectric module by interweaving p-type and n-type CNT yarns into a π-type architecture, eliminating the need for rigid substrates. The module achieved a power density of 70 mW·m⁻² under a temperature difference of 44 K. Remarkably, the output showed no significant degradation under 80% strain, highlighting the structure’s high deformation stability. Jang et al. [141] constructed a 3D woven thermoelectric fabric by interlacing polyurethane-coated CNT yarns into vertically aligned p/n couple arrays (Figure 5b). The structure was aligned perpendicular to the skin surface to optimize the exploitation of the temperature gradient between the body and the environment. The fabric achieved a standardized open-circuit voltage of 8.0 mV·K⁻¹ per thermoelectric couple, alongside a unit power output of 1.1 × 10⁻⁴ μW·K⁻². Electrical performance remained stable even under 100% tensile strain, demonstrating excellent mechanical adaptability and self-healing capability. This work represents a benchmark in highly stretchable thermoelectric textile systems for wearable applications.

In the construction of flexible TE systems, the processing methods of TE fibers have a decisive influence on device stability, service lifetime, and performance consistency. The integration efficiency, weavability, and scalability of TE modules are largely dependent on the morphological control of thermoelectric materials and the optimization of interfacial compatibility. Xu et al. [148] fabricated a flexible thermoelectric fiber composed of PEDOT:PSS and SWCNTs via a wet-spinning technique. Through treatment with NaOH, the Seebeck coefficient was markedly enhanced to 36.4 μV/K, accompanied by an increase in the power factor to 280 μW·m⁻¹·K⁻². A module composed of five TE fiber pairs was driven by the temperature gradient between body heat and ambient conditions when worn on the arm, validating its feasibility in lightweight, sewable wearable thermoelectric generators. These devices demonstrate not only high thermoelectric performance but also mechanical and environmental robustness. Notably, the fiber-based TE generator was tested under real-wear conditions, and several textile modules remained functional under large mechanical strains, simulating movement and deformation in actual use scenarios. Khoso et al. [142] successfully constructed a flexible, washable, and reusable TE textile using a pad-dry–cure process. A composite layer of rGO and PEDOT:PSS was applied to cotton fabric (Figure 5c). The device generated a voltage of 25 mV under a ΔT of 16.5 °C while maintaining favorable mechanical stability and air permeability. These results demonstrated the engineering compatibility of conductive polymer-based TE layers for industrial-scale wearable electronics. From the perspective of structural design, the development of flexible TE fibers has evolved from “surface coating” to an “integrated fiber body architecture.” Focus has been placed on the synergistic stretchability of active TE materials and the substrate, as well as their interfacial adhesion and compatibility with fabric functions. At present, techniques such as multimaterial co-drawing, coaxial electrospinning, and solution-assisted coating-and-stretching have provided multiple fabrication pathways for TE textile systems. These methods show great promise in achieving flexibility, weavability, high TE performance, and scalability.

In dynamic wearing environments, the temperature gradient between the human body and the environment, as well as its temporal fluctuation, make it difficult for traditional thermal insulation materials to maintain a stable microclimate. PCMs, which can absorb or release substantial latent heat during solid–liquid phase transitions, have been regarded as key candidates for active thermal buffering and temperature regulation. However, the inherently low thermal conductivity, insufficient mechanical strength, and leakage risks of PCMs have restricted their direct application in flexible and wearable devices. In response, efforts have been made to introduce thermally conductive fillers, construct composite multiphase structures, and adopt innovative processing methods, aiming to improve both thermal performance and mechanical integration. Meng et al. [149] dispersed graphene nanosheets (GNs) within MPCMs using a cellulose-assisted method, thereby creating MPCMs-GNs/textile composites with a stable structure and excellent thermal responsiveness. Within the operating temperature range of 27–35 °C, thermal buffering lag and delayed temperature rise effects were observed. Infrared thermography confirmed their ability to suppress local temperature gradients and enhance thermal comfort, thus improving both wearability and safety. Luo et al. [150] proposed a bilayer composite structure (GRGC) composed of rGO aerogel and n-octadecane PCM, inspired by the thermal gradient structure of the Earth’s atmosphere. Solar absorption and localized heating were achieved via the rGO layer owing to its high absorptivity and low thermal conductivity, while the PCM layer regulated thermal flux through latent heat storage. Under extreme conditions (−5 °C ambient temperature and strong solar irradiation), the GRGC maintained a simulated skin temperature between 32–34 °C, demonstrating excellent environmental adaptability. To improve both breathability and thermal buffering, Li et al. [143] fabricated a fluorine-free composite consisting of silicon-based polyurethane, graphene quantum dots, and stearic acid (SiPU/GQDs/SA) using the electrospinning technique. This composite incorporated paraffin-based PCM and graphene quantum dots (GQDs), achieving heat buffering in cold conditions and heat absorption under high temperatures. After multiple thermal cycles, the membrane maintained stable latent heat performance, along with softness, breathability, and thermal durability—making it suitable for use as a passive thermal management layer in smart clothing and outdoor wearable systems (Figure 5d). Su et al. [151] proposed a flexible composite thermal control material by doping CNTs into a PCM-based gel and integrating it into textile structures via screen printing. This system exhibited a solar absorption efficiency of 96.4% along with a photothermal conversion rate of 94.8%. Under a temperature gradient of ΔT = 30 K, it successfully powered a thermoelectric module, showcasing a fully integrated system capable of thermal regulation, energy storage, and energy conversion. PCM-based thermal buffering systems are evolving from single-function insulation toward multifunctional coupling. Through integration with high-conductivity fillers, infrared-responsive structures, and flexible matrices, their temporal responsiveness in thermal management has been significantly extended. To enhance retention and prevent PCM leakage during use and deformation, strategies such as microencapsulation, confinement in electrospun nanofiber networks, and bilayer architectures have been adopted. These approaches provide mechanical support, enhance thermal reliability, and minimize the risk of phase-change material seepage during stretching or repeated use. This development also provides foundational support for future integrated energy systems.

Apart from phase-change-based heat absorption structures, wearable thermal management systems often require the incorporation of active temperature regulation mechanisms to cope with sudden thermal fluctuations or therapeutic heating demands. Such materials are expected to exhibit low energy consumption, rapid thermal response, mechanical flexibility, and the ability to provide localized and precise heating. Kim et al. [152] created a flexible, transparent heater utilizing large-area graphene films produced through CVD. The device demonstrated an excellent thermal response speed and uniform heat distribution. A surface temperature of 100 °C was achieved under a 12 V input, and the heater maintained performance under repeated bending, making it suitable for applications such as automotive defogging, smart windows, and wearable heating. In terms of composite material design, Li et al. [153] introduced a dry-spun CNT network as a conductive scaffold, combined with graphene sheets to form a transparent heating platform. The device maintained a 71% optical transmittance at 85 °C and remained functional after 10,000 bending cycles, thereby improving the mechanical stability and thermal uniformity of the flexible conductive film. Pillai et al. [144] fabricated a fully printed flexible textile heater using silver/carbon composite conductive ink deposited on a polyester substrate via screen printing (Figure 5e). The heater maintained a stable surface temperature of approximately 50 °C at 5 V, with a power consumption as low as 55 mW·cm⁻². This demonstrated advantages in low power consumption, high design flexibility, and suitability for mass production. Olivieri et al. [154] constructed a bilayer rGO/polyurethane coating system, which endowed cotton textiles with excellent conductivity, thermal responsiveness, and washing durability. Under a low power input (<0.5 W), localized heating with uniform temperature distribution was achieved. Even under 50% strain, the fabric maintained a pressure-sensitive response, with a maximum thermal response time of only 500 ms. This made it suitable for integration into multifunctional wearable platforms with heating, sensing, and monitoring capabilities. Li et al. [155] developed an intelligent aerogel-structured fiber, incorporating graphene aerogel as a scaffold combined with polyethylene-glycol- or paraffin-based phase-change materials. The resulting composite exhibited photo-, electro-, and thermally driven responses, with a thermal activation time below 15 s. It also featured flexibility, self-cleaning behavior, and multimodal environmental responsiveness, providing stable thermal regulation under complex conditions.

Traditional single-sided thermal regulation structures often face limitations under variable environments, including mismatched thermal radiation direction and insufficient heat dissipation. To address this, Luo et al. [145] introduced a wearable Janus textile structure based on a dual-sided design. One side incorporated photothermal materials with low infrared emissivity and high solar absorptivity, while the other side possessed high reflectivity and strong infrared radiation capacity. This enabled dual-mode functionality for both warming in winter and cooling in summer (Figure 5f). Experimental results showed that under sunlight exposure, the skin temperature increased by 8.1 °C, while in high-temperature conditions, the surface temperature was reduced by 6 °C. After integration with a thermoelectric module, the textile system generated electric power densities of 60 mW·m⁻² during the day and 23.5 mW·m⁻² at night, revealing strong potential for coordinated passive thermal control and energy harvesting. Overall, Janus and gradient textiles not only enable bidirectional thermal flux management across front and back surfaces but also offer excellent wearability and system compatibility. This allows for real-time temperature regulation under complex and dynamic environmental conditions.

In extreme-temperature emergencies and industrial/military heat exposure scenarios, traditional wearable thermal management structures often fail to deliver a rapid response or thermal hazard warning. To address this, Sun et al. [146] developed a multifunctional anisotropic aerogel (MGPA) composed of MXene, rGO, and PEDOT:PSS (Figure 5e). This material combines high thermal insulation, photothermal responsiveness, and tunable electrical conductivity, enabling rapid electrothermal triggering. Under an ambient temperature of 0 °C, the device rapidly increased its surface temperature to ~30 °C within seconds through either electrical or photothermal stimulation. Additionally, a thermoelectric voltage was generated within 1 s, enabling thermal warning output. The structure exhibited high configurational stability, thermal aging resistance, and excellent flame retardancy, making it a representative high-integration smart textile platform for deployment in extreme cold zones or emergency protection settings. Meanwhile, Prasanna et al. [147] utilized a microwave-assisted solvothermal technique to grow vertically aligned zinc oxide nanorods in situ on the surface of conductive carbon fabric (CF), creating a 3D thermoelectric heterojunction structure, as shown in Figure 5h. This device retained the flexibility and wearability of the CF while achieving a power factor of up to 777 nW·m⁻¹·K⁻². High charge transport efficiency and strong thermal stability were also reported. In space-constrained or on-skin applications, the structure enabled continuous harvesting of thermal energy from the slight temperature gradient between the human body and its surroundings, which was subsequently converted into electrical signals. These findings demonstrated the potential of low-dimensional structural optimization for flexible nanoscale thermoelectric devices. By embedding thermally sensitive components into aerogel matrices, conductive networks, or micro–nano arrays, precise temperature-triggered responses and active feedback mechanisms can be realized. These features enable the construction of smart wearable systems with early warning capability, suitable for integration into garments, skin patches, or sensor platforms. In summary, wearable thermal management systems have now developed into four major technical pathways: thermoelectric conversion, phase-change buffering, radiative control, and multifunctional coupling. To strengthen the comparison and critical perspective of the described thermoelectric and phase-change textile systems,

Table 4. Durability evaluation data of typical carbon-based flexible thermal regulation structures.

System Type	Representative Material	ΔT (K)	Power Output	Mechanical Durability	Reference
TE (π-type CNT yarn)	Interwoven p/n-CNT yarn array	~44	~70 mW/m2	Stretchable up to 80% without degradation	[140]
Organic TE yarn	PEDOT:PSS/MWCNT and PVP on polyester yarn		~7.1 nW	Stable under ambient wear	[137]
TE + PCM hybrid	CNT/AuNP composite with MPCM	~6.6	~270 nW	Retains integrity after moderate deformation	[139]
PCM-enhanced	Graphene + MPCM in woven cotton matrix	27–35		Maintains phase retention in repeated cycles	[149]
PCM bilayer textile	rGO aerogel with PCM bilayer	Maintains 32–34 °C at −5 °C		Solar-adaptive, stable over outdoor use	[150]

summarizes the key performance metrics of representative devices. Parameters such as the temperature difference (ΔT), power output, and mechanical durability are listed, facilitating a clearer evaluation of their practical potential in flexible wearable thermal regulation.

Thermoelectric textiles are being optimized toward higher figure of merit (ZT) values, organic/inorganic hybrid materials, and π-type architectures for enhanced and stable heat-to-electric energy conversion. Phase-change materials still require further improvements in interfacial compatibility and thermal transport control to enhance engineering stability. Meanwhile, photothermal regulation and Janus textiles exhibit significant potential in advanced material structuring and directional heat flux management.

4.2. Interfacial Thermal Regulation in Flexible Electronic Packaging

As flexible electronic technologies rapidly advance in fields such as wearable devices, flexible displays, biosensors, and artificial intelligence, localized overheating problems have become increasingly prominent. These issues arise from complex packaging structures, high integration density, and concentrated power consumption within the devices. Since heat in flexible electronics is primarily transferred across multiple interfaces, interfacial thermal resistance has emerged as a critical bottleneck limiting heat flow efficiency. Therefore, enhancing interfacial thermal conductance and reducing interfacial thermal resistance have become primary objectives. At the same time, maintaining mechanical flexibility while achieving co-optimization of device thermal stability and functional responsiveness has emerged as a critical direction in the study of thermal regulation for flexible electronic packaging.

Recent studies have demonstrated that constructing multiscale, continuous thermal pathways is an effective strategy for improving interfacial thermal management. Among these approaches, 2D thermal conductive materials such as graphene, MXene, and their derivatives have garnered widespread interest owing to their outstanding in-plane thermal conductivity and superior film-forming capabilities. Hwang et al. [156] proposed a flexible packaging structure by coating CVD graphene onto a silver nanowire network. The resulting structure exhibited both high thermal conductivity and flexibility compatibility. It maintained thermal stability even after 18,000 bending cycles under 15% strain, demonstrating the role of graphene in reducing interfacial contact resistance, alleviating stress concentration, and suppressing local hot spots. Shao et al. [157] designed a Janus-type multifunctional ultra-flexible polytetrafluoroethylene–carbon nanotube–Fe3O4/MXene (FCFe/M) material with a composite interface structure consisting of an MXene layer and a “silk-like” polytetrafluoroethylene (PTFE)/CNT/Fe₃O₄ layer, as shown in Figure 6a. A lateral heat conduction pathway was established, which reduced phonon scattering and improved heat diffusion. An in-plane thermal conductivity of 19.89 W·m⁻¹·K⁻¹ and rapid thermal response were achieved, meeting the requirements of low-voltage flexible packaging devices for fast thermal regulation. In terms of 3D thermal network construction, Wang et al. [158] co-carbonized MXene (CMXene) with CF to fabricate a 3D synergistic thermal network (CMXene/CF), which enhanced interfacial thermal coupling and structural integrity. This network was uniformly embedded in a polyimide (PI) matrix, forming a lightweight PI-based composite. The interfacial design improved electromagnetic shielding stability at high temperatures (400 °C). In elastic polymer composite systems, interfacial construction has also been utilized to enhance thermal conductivity while preserving flexibility. Paleo et al. [159] fabricated a nonwoven nanocomposite membrane composed of SWCNTs and thermoplastic polyurethane (TPU). A stable fibrous interface was formed by coating SWCNTs with TPU, achieving a thermal conductivity of 7.6 W·m⁻¹·K⁻¹ and superior mechanical flexibility. This membrane has potential applications as a thermal interface material for large-area flexible packaging. Kim et al. [160] enhanced the alignment and doping efficiency of wet-spun CNT fibers through solvent optimization and developed a fabric-type thermoelectric generator (FTEG) integrated with high-density p/n thermoelectric pairs (Figure 6b). The highly aligned CNT fibers retained a Seebeck coefficient of −42.4 μV·K⁻¹ after multiple washing cycles. This highly oriented structure effectively controlled interfacial heat flow and distribution, thereby improving thermal stability and integration efficiency in flexible electronic packaging. Luo et al. [161] used electrospun PU microfibers as a scaffold and employed ultrasonic cavitation to anchor multi-walled carbon nanotubes (MWCNTs) firmly onto the fiber surface. In situ polymerization of PEDOT was then conducted to form a multifunctional PEDOT/MWCNT@PU fiber network. This system achieved dual optimization of thermal pathway stabilization and interfacial stress modulation, exhibiting excellent thermal response retention under high-frequency mechanical cycling. Codau et al. [162] fabricated a TPU/MWCNTs/PEDOT:PSS composite nanofiber membrane using a wet-electrospinning process. A continuous conductive chain structure (CNTs/PEDOT:PSS/CNTs) was formed through a nonsolvent-induced phase separation mechanism. This approach not only enhanced thermoelectric performance but also improved interfacial coupling and structural integrity. The membrane demonstrated significant potential for regulating thermal field distribution in regions with temperature non-uniformity. Ka et al. [163] employed plasma treatment and repeated dip-coating techniques to firmly anchor rGO onto the surface of cotton fabrics, forming an e-textile with an integrated thermal conduction pathway. This rGO-based flexible textile exhibited excellent thermal and electrical stability and was used as an interfacial thermal management layer in flexible packaging to mitigate local overheating and enhance thermal distribution uniformity. For functional enhancement, PCMs, owing to their latent heat absorption and temperature regulation capabilities, have been widely adopted to construct dynamic thermal management interfaces with buffering capacity. Zhang et al. [164] developed a paraffin–graphene composite (PGC) composed of high-thermal-conductivity graphene aerogel (HGA) and paraffin, reaching a thermal conductivity of 30.75 W·m⁻¹·K⁻¹. The composite effectively reduced the temperature rise in high-power-density battery systems, demonstrating dual functions in interface temperature stabilization and heat storage. Li et al. [165] developed a densified graphene aerogel–PCM host–guest film (d-GAPHF), as illustrated in Figure 6c, to enhance the mechanical strength of the structure through densification treatment and to construct a continuous in-plane thermal conduction pathway. As a result, high thermal conductivity (2.0 W·m⁻¹·K⁻¹) and significant latent heat (182.5 J/g) were achieved while maintaining processability and mechanical flexibility. This enabled delayed thermal buffering and responsive heat regulation. Sun et al. [166] further introduced a polyvinylidene fluoride–hexafluoropropylene (PVDF-HFP)-enhanced graphene aerogel (GAF) skeleton to construct a GAF–paraffin wax (GAF-PW) composite film. This film maintained flexibility and deformability while achieving a high phase-change enthalpy (154.64 J/g) and outstanding photothermal conversion efficiency (95.98%). Stable thermal performance was retained even after 500 thermal cycles. Zhou et al. [167] proposed a composite PCM by integrating silver-modified multi-walled carbon nanotubes (Ag-MWCNTs) with paraffin wax (PW) and encapsulating the mixture within a carbon nanotube sponge (CNS), forming a Ag-MWCNTs/PW@CNS structure. The 3D network of the CNS provided a robust thermal interface channel, addressing challenges such as PCM leakage and structural delamination. This material represents an excellent interfacial thermal management system under photo-electro-thermal multiphysics coupling. Cui et al. [168] proposed a highly stable phase-change coating composed of cross-linkable polyethylene glycol (RPEG) and MWCNTs. Under the assistance of a self-cross-linking dispersant, a tightly bonded interface network was formed between the RPEG and MWCNTs. The fabricated system demonstrated excellent thermal conductivity along with a substantial heat storage density (142.8 J/g). This structure enhanced both interfacial thermal storage capability and mechanical adhesion, effectively preventing the delamination or thermal failure of the high-temperature packaging layers. To enable intelligent structural response in interfacial thermal regulation, Alehosseini et al. [169] developed a composite material with skin-like softness and self-healing properties. The incorporation of graphene–PEDOT:PSS fillers provided enhanced thermal conductivity and self-repair performance. Even under 600% strain, the material could self-heal rapidly while maintaining excellent thermal responsiveness and a low modulus (<1 MPa), offering a promising strategy for designing flexible self-healing packaging systems. Wu et al. [170] constructed a fabric-type thermoelectric device by uniformly depositing a composite layer of waterborne polyurethane (WPU), MWCNTs, and PEDOT:PSS onto yarn surfaces via dip-coating. The device delivered stable power output and excellent interfacial thermal conductance. Ju et al. [171] designed a flexible mat (TCCNF) by co-doping carbon nanofibers with TiN and Co nanoparticles (Figure 6d). Numerous heterointerfaces (e.g., TiN/C, Co/C) were formed within the carbon nanofiber framework, promoting heat carrier distribution and thermal flux uniformity. Enhanced thermal diffusion and energy conversion capabilities were achieved. The material demonstrated a fast response, high thermal efficiency (1508 °C/W·cm²), and strong electromagnetic interference (EMI) shielding, showing great promise for electro-thermal–magnetic synergistic packaging systems.

Taken together, these studies reveal that interfacial thermal regulation strategies for flexible electronic packaging are evolving toward diversification and integration. By constructing 2D thermal layers, three-dimensional scaffold structures, elastic composites, phase-change components, and multifunctional electro-thermal–magnetic response systems, improvements in thermal conductivity, interfacial stability, and heat transfer efficiency have been achieved across various application scenarios. Future research is encouraged to focus on the mechanisms of thermo-electro-magnetic coupling under multiphysics environments, develop AI-assisted structural design and optimization methods for interfaces, and explore sustainable and recyclable high-performance thermal regulation materials. Under the continuous evolution of high-throughput fabrication, device structural complexity, and system-level intelligence, flexible thermal interface materials are poised to become indispensable in next-generation electronic packaging.

4.3. Thermal Control Coatings and the Expansion of Infrared Shielding Functions

Driven by the rapid progress in infrared imaging technologies, integrating infrared stealth and thermal management has become a key development direction for next-generation smart textiles and electronic thermal control systems. The regulation of infrared radiation not only requires materials to possess low emissivity and excellent thermal insulation properties but also demands integrated performance such as electrothermal responsiveness, electromagnetic shielding, and compatibility with flexible processing. As a result, the development of thermal control coatings with synergistic infrared shielding and thermal regulation functions has become a prominent research focus within the thermal management field.

In response to the growing demand for multifunctional integration, extensive research in recent years has explored the co-assembly strategies of carbon-based nanomaterials and metallic conductive phases to enhance the performance of thermal control coatings. A representative example is the graphene/MXene composite coating fabric (Figure 6e) developed by Hou et al. [172]. Multilayer coatings were constructed via blade coating and spraying methods. The resulting structure exhibited a low infrared emissivity (minimum 0.248), a rapid temperature rise to 91.7 °C under a 4 V drive, and an EMI shielding effectiveness of up to 64.3 dB. This system integrated infrared shielding, electrothermal heating, and electromagnetic interference protection, offering a systematic design approach for multifunctional smart textiles. For the integration of infrared and EMI shielding functions, Navik et al. [173] used a mechanochemical process combined with supercritical CO₂ to prepare few-layer graphene (FLG)-enhanced PA-6 composite films, as shown in Figure 6f. With just 3 wt% FLG, the composite achieved a thermal conductivity of 1.78 W·m⁻¹·K⁻¹ and an EMI shielding effectiveness of 41.8 dB. Superior heat dissipation performance was demonstrated in high-power LED chip testing, offering a reliable thermal–electrical co-management strategy for highly integrated electronic systems. Xie et al. [174] fabricated coated polyamide (PA)–polyethyleneimine (PEI)–rGO composite yarns via chemical reduction. The coating demonstrated high electrical conductivity (4.7 × 10³ S/m), excellent infrared reflectivity, and photothermal/electrothermal conversion properties, achieving an EMI shielding effectiveness of 66.7 dB. This confirmed the adaptability and stealth potential of thermal control coatings in wearable electronics. Beyond thermal and shielding capabilities, dynamic control of infrared emissivity has also become a key design objective. MWCNT-based electrochromic films have been shown to adjust emissivity in the 0.15–0.7 range under voltage modulation. A temperature difference of approximately 8 °C was achieved in vacuum conditions, indicating the tunability required for infrared camouflage under varying backgrounds [175]. From the standpoint of thermal conductivity and heat-spreading structure design, Mandriota et al. [176] applied rGO and silver nanoparticles to the surface of cotton fabrics. A histidine (His)-rGO/Ag network was constructed, doubling the original fabric’s thermal conductivity while reducing surface wettability and preserving flexibility and breathability. This environmentally friendly process avoided toxic reducing agents and is suitable for industrial-scale application. Building on this, Wang et al. [177] developed a graphene oxide (GO)-reinforced superporous hydrogel to stably support phase-change materials. A 3D thermal conduction network and foam scaffold were established. The material achieved a high melting enthalpy (169.0 J/g) and demonstrated strong thermal responsiveness and cooling management performance under infrared thermal imaging. This material system demonstrates significant potential for application in packaging, thermal storage structures, and heat dissipation for electronic devices. In addition, several studies have explored coating strategies combining aerogels and PCMs to achieve both energy storage and stealth. Lyu et al. [178] proposed a Kevlar nanofiber aerogel/PCM (KNA/PCM) composite membrane capable of automatically adjusting thermal radiation under fluctuating sunlight conditions. The emissivity approached 0.94, and the structure exhibited low infrared transmittance in the 3–15 μm range. The phase-change behavior modulated the surface temperature, effectively masking heat sources. Wu et al. [179] developed a dual-functional Janus textile. One side was coated with GO for efficient photothermal conversion, while the opposite side featured a hydrogel to enhance evaporative cooling. Under 1 sun illumination, the GO side produced a 43.7 °C temperature increase, while the hydrogel side achieved a 7.1 °C cooling effect. This bidirectional thermal management system also demonstrated antibacterial properties, offering new ideas for environmentally adaptive and multifunctional thermal control coatings. Nanofillers have also been incorporated into conventional PCM systems to boost both thermal conductivity and structural stability. Hekimoğlu and Sarı [180] used fly ash as a shape-stabilizing support for encapsulating n-octadecane and added carbon-based nanomaterials like graphene and CNTs to improve thermal conductivity. At 8 wt% graphene doping, the thermal conductivity increased by over 200% without compromising latent heat storage, supporting the application of thermal control coatings in building envelopes and thermoresponsive apparel. Sezer Hicyilmaz et al. [181] coated polyester textile substrates with microencapsulated n-octadecane modified by graphene via a coating process. With only 0.1% graphene addition, thermal conductivity increased by 31% while the coating retained flexibility and phase-change functionality.

In conclusion, current research on thermal control coatings is advancing toward multifunctional integration, hierarchical structuring, and intelligent regulation. Through the incorporation of carbon-based and metallic conductive materials, as well as the construction of layered and hybrid functional coatings, synergistic improvements have been achieved in infrared shielding, thermal regulation, energy storage, and EMI protection. These advances provide diversified solutions for thermal management in wearable devices, energy electronics, and military camouflage applications.

5. Challenges and Future Prospects

In recent years, low-dimensional carbon materials have opened new avenues for the design of thermally functional textiles due to their outstanding thermal conductivity, structural tunability, and mechanical flexibility. Continued advances in multiscale structural design and interface engineering have expanded their potential for wearable applications. However, current research remains largely confined to laboratory-scale verification and prototype development. A significant gap still exists between laboratory feasibility and practical application in terms of performance stability, process control, and scalable manufacturing. Bridging the divide between functional performance and engineering viability requires addressing a range of unresolved scientific questions and technical obstacles. The following sections outline core issues and propose future directions for their resolution.

5.1. Fundamental Challenges in Material Design and Multi-Property Integration

Although low-dimensional carbon materials offer promising opportunities for thermal management textiles, major difficulties remain in material selection and structural compatibility. Graphene, carbon nanotubes, and MXenes exhibit significant differences in key performance parameters, including thermal conductivity, flexibility, and stability. However, the absence of a standardized evaluation framework and compatibility criteria has rendered the material selection process largely reliant on empirical judgment. To address this, it is necessary to establish a multi-parameter performance database. Integrating data-driven approaches, such as machine learning, can help construct mapping models linking structure, thermo-mechanical properties, and application scenarios, thereby enabling more systematic and scientific material development.

At the same time, achieving a balance between high thermal conductivity and wearable flexibility remains a central challenge in structural design. Current strategies for enhancing thermal performance primarily depend on the densification or alignment of fillers. However, these methods often compromise mechanical flexibility, which, in turn, affects both wearing comfort and dynamic mechanical response. Although some efforts have been made using composite gradient architectures or rigid–flexible hybrid networks to mitigate this issue, the conflict between thermal performance and flexibility has yet to be fully resolved. Biomimetic gradient interfaces and hierarchical network design may offer promising paths to achieve synergistic optimization of thermal and mechanical performance.

Moreover, in wearable applications, materials are subjected to complex mechanical stresses such as stretching, bending, and twisting. The structural stability of thermal pathways under these conditions is a key factor limiting long-term performance. Current studies have insufficiently addressed the degradation mechanisms of thermal networks under mechanical load. Future work should incorporate in situ multiphysics coupling tests, high-resolution modeling, and microstructural tracking to uncover the interfacial evolution and heat transfer behavior under force–thermal coupling at the micro/nanoscale. These insights will offer a theoretical foundation for the design of stable and efficient thermal conduction structures.

5.2. Critical Bottlenecks in Engineering Construction and Industrial Conversion

The transition of thermal management textiles from “experimental validation” to “practical wearable implementation” requires overcoming challenges related to construction scale, manufacturing processes, and system integration. Although micro-/nano-scale techniques such as in situ growth, laser patterning, and hierarchical deposition have enabled the precise formation of thermal pathways at the fiber level, achieving continuous channels and compatible interfaces at the macroscopic fabric scale remains a major obstacle to overall thermal efficiency. Issues including uneven fiber alignment, unstable interfacial contact, and assembly defects often result in thermal discontinuities and increased local thermal resistance, thereby significantly compromising the thermal conductivity of the entire fabric. Therefore, the development of high-resolution, synergistic fabrication strategies and systematic interfacial thermal optimization approaches is urgently needed. These would enable multidimensional connectivity of thermal pathways and minimize interfacial thermal resistance.

In parallel, facilitating the transition of functional textiles from prototypes to industrial applications requires attention to process stability, cost control, and end-user comfort. Currently, most high-performance devices still rely on small-scale manual fabrication, which fails to meet industrial-scale and standardized manufacturing requirements. Functional integration often leads to increased thickness, reduced flexibility, and impaired breathability, thereby limiting user acceptance in wearable applications. Although fabrication techniques including electrospinning, spray-coating, and printing exhibit certain potential for scalability, a comprehensive processing methodology and evaluation framework tailored to wearable systems is still lacking. In the future, efforts should be directed toward the development of continuous and versatile fabrication platforms. Additionally, the construction of lightweight, low-energy-consumption, and multi-field-tolerant thermal regulation materials will be essential to achieve performance, production, and wearability in a synergistic manner—laying the groundwork for the practical deployment of next-generation thermal management textiles.

5.3. Data-Intelligence-Driven Innovations in Structural Design and Advanced Engineering Applications

The structural design of thermal management textiles requires the simultaneous optimization of thermal conductivity, mechanical flexibility, structural integrity, and interfacial stability. Traditional design strategies, which rely heavily on empirical knowledge and trial-and-error approaches, are inadequate for achieving such multifactorial optimization. With the advancement of artificial intelligence technologies, data-driven methods have provided predictive, quantifiable solutions for textile architecture design. Machine learning and deep learning models may serve to map relationships between material parameters and performance, thereby assisting in structural prediction and functional optimization. By establishing multidimensional performance databases, artificial intelligence (AI) can support target-oriented architecture generation and structure screening, significantly improving design efficiency. Mahmuda’s study compared artificial neural networks (ANNs) with multiple linear regression (MLR) models for predicting the thermal transmittance of plain-woven cotton fabrics based on thread density and fabric thickness [182]. The ANN model demonstrated superior accuracy and robustness, with lower error metrics and higher determination coefficient, effectively capturing complex interactions among fabric parameters. This highlights the advantage of machine learning techniques over traditional statistical methods for thermal property prediction in textiles. A series of studies by Xianyi Zeng demonstrated the effective use of ANNs in modeling and optimizing the performance of knitted and barrier fabrics. The models were trained to predict key comfort-related properties, including global thermal comfort and air permeability, based on structural and processing parameters [183,184,185]. In each case, the ANN models were further applied in reverse to guide the selection of fabric structures, yarn types, and processing conditions, enabling performance-driven textile design without extensive experimental iteration. Chokri et al. applied ANNs to model the complex relationships between air permeability and multiple fabric-related parameters and further utilized the trained models to guide the selection of yarn, fabric, and loom settings for tailoring barrier fabric performance without extensive trial and error [186]. The adoption of generative design strategies has further expanded the accessible structural space, enabling the automatic generation of high-performance designs under defined physical constraints and application requirements. Although preliminary achievements have been reported, several challenges remain unresolved. Due to the high heterogeneity of textile materials and their complex manufacturing processes, publicly available datasets within the industry remain scarce. These include limited model interpretability, inadequate training datasets, and difficulties in integrating multi-scale features. Future progress will depend on the development of more robust multimodal model systems. These must incorporate experimental feedback mechanisms and physically informed optimization frameworks to enable the deep integration of AI technologies into intelligent textile structure design.

Moreover, the potential applications of flexible thermal management materials in high-end and complex environments are becoming increasingly apparent. In extreme conditions such as spacecraft thermal control and high-density chip cooling, thermal systems are required to provide a balance of high thermal conductivity, mechanical flexibility, and environmental stability. Traditional materials often suffer from performance degradation under multi-field coupling and complex deformation conditions, making them unsuitable for practical use. Low-dimensional carbon materials exhibit excellent thermal transport properties and tunable structures. When constructed into textile architectures, they demonstrate outstanding stress adaptability and thermal stability. Their performance stability under high heat flux densities and non-steady-state conditions underscores their suitability for engineering applications. With continued advances in multi-scale interface design and thermal pathway engineering, these materials are expected to be deployed in high-end domains including spacecraft thermal protection layers and thermal interface materials for electronic chips. This would offer crucial support for constructing flexible thermal regulation systems under extreme environments.

5.4. Cost and Process Compatibility of Carbon-Based Materials

Although low-dimensional carbon materials—such as SWCNTs and reduced graphene oxide (rGO)—offer outstanding thermal and mechanical properties, their high cost and limited compatibility with existing textile manufacturing processes remain significant barriers to large-scale industrial application. For example, the cost of single-walled carbon nanotubes typically ranges from 100 to 500 USD per gram, depending on the synthesis method, purity, functionalization, and supplier, making them prohibitively expensive compared to conventional thermally conductive fibers or fabrics. Moreover, although scalable production techniques—such as pad-dry rGO coatings and wet-spinning of CNT/graphene hybrid fibers—have been proposed, integrating these materials into standard textile processes like spinning, weaving, or roll-to-roll printing still poses challenges. This is due to issues in dispersion stability, solvent compatibility, and the need for specialized dispersion aids or post-treatments. As a result, current applications of carbon-based smart textiles are mostly confined to high-value, small-area scenarios—for example, wearable thermal patches, medical sensors, and defense apparel—where the performance benefits justify the cost. This finding further reinforces the importance of microstructural optimization in the scalable processing of carbon-based composites. To bridge the gap between laboratory research and real-world deployment, future development should focus on two synergistic approaches: cost-effective synthesis—including low-temperature, air-based rGO reduction and optimized large-batch production of graphene fibers—and process-compatible composite design—such as printable conductive inks, continuous hybrid filament production, and roll-to-roll assembly—to enable efficient and scalable integration with existing textile infrastructure.

6. Conclusions

Low-dimensional carbon materials, such as graphene and carbon nanotubes, represent a transformative class of thermal management materials due to their extraordinary thermal conductivity, structural tunability, and mechanical compliance. These properties have enabled their integration into flexible electronic systems, where requirements extend beyond mere heat dissipation to encompass conformability, dynamic adaptability, and multifunctional performance. While notable progress has been made in constructing hierarchical architectures and optimizing interfacial coupling to reduce thermal resistance, challenges including scalable fabrication, interfacial stability under mechanical deformation, and integration with electronic components persist. Addressing these issues demands advances in nanoscale design strategies, multi-physics modeling, and process engineering. Future efforts should focus on the development of scalable, structurally engineered materials, the realization of intelligent thermal feedback mechanisms, and cross-disciplinary innovations bridging materials science, electronics, and artificial intelligence. In conclusion, low-dimensional carbon materials are poised to redefine flexible thermal management by enabling adaptive, high-performance thermal control systems for next-generation technologies, including soft robotics, bioelectronics, and wearable devices.