Scalable phase-change fibres with record thermo-mechanical properties for cooling textiles Authors: Xiaolong Li, Qianyi Li, Xiaoyu Zhang, Tural Khudiyev Published: 2026 ABSTRACT AbstractEfficient thermal regulation in clothing is critical for comfort, safety, and energy conservation, yet achieving it without active cooling remains a major challenge. Phase-change materials (PCMs), which store and release heat during solid–liquid transitions, offer a promising route toward passive and durable temperature control in everyday apparel. Embedding liquid-state PCMs into fibres (PCFs) is inherently challenging due to the instability of micro-scale liquid cores and the inability of conventional microfluidics to sustain continuous injection beyond centimetre lengths, limiting scalable fibre production. We overcome these barriers through a size-reduction and speed-matching strategy during thermal and cold fibre drawing, yielding uniform 35 m-long, 200 µm-thin core–shell PCFs. The fibres exhibit low supercooling (4.2 ± 0.3 °C), high latent enthalpy (122.6 ± 1.5 J g⁻¹), and record-level mechanical performance (32.1 ± 0.8 MPa strength, 671 ± 10.2% elongation). Fabrics incorporating 50% PCFs achieve sustained temperature reductions of 3–6 °C under realistic conditions, passively dissipating heat from sunlight and battery-generated hotspots while outperforming conventional cotton. These findings pave the way for scalable, high-performance textiles that provide passive thermal regulation, offering a practical strategy to reduce energy demand and mitigate environmental impact.Subject terms: Design, synthesis and processing; Polymers; Polymers Introduction Extreme heat is an escalating global health threat. The World Meteorological Organization reports that every year from 2015 to 2024 ranks among the hottest on record, with ~489,000 annual deaths attributed to heat exposure1. Heatwaves of increasing frequency, duration, and intensity highlight the urgent demand for passive, always-available thermal protection that integrates seamlessly into daily life, without the bulk, power requirements, or discomfort of conventional cooling devices2–8. Textiles, ubiquitous and in direct contact with the human body, offer a compelling platform for embedding unobtrusive thermoregulation functionalities9–13. Among various passive cooling strategies—such as radiative cooling, evaporative cooling, biomimetic and structural designs, thermal insulation, and reflective coatings—the incorporation of phase-change materials (PCMs) into fabrics is particularly promising and practical. PCMs store and release heat through reversible phase transitions, buffering temperature fluctuations without external energy input14–19. Yet incorporating sufficient PCM loading into textiles while preserving flexibility, breathability, and mechanical robustness remains a central challenge. Conventional approaches such as electrospinning are unsuitable for wearable long PCM fibres (PCFs), as PCMs lack polymeric chain entanglement and form nanofiber mats rather than continuous filaments20–22. Core-shell PCFs have predominantly been produced via wet spinning, wherein room-temperature liquid PCMs are encapsulated through solvent coagulation, restricting material choice suitable for thermal regulation and environmental compatibility23–25. Solvent-induced coagulation severely limits the usable shell and core chemistries, excluding many functional and eco-benign polymers that are incompatible with the solvent system. More recently, microfluidic extrusion combined with wet spinning has been explored for liquid-metal (LM) fibres in flexible electronics, but these methods target room-temperature LMs with ultrahigh surface tension and oxide skins, relying on external baths or interfacial chemistry, and do not prioritize sealing reliability under thermal cycling26–28. Post-injection of molten PCMs into prefabricated hollow fibres also faces inherent interfacial and fluidic limitations: adhesion to a solidified inner wall is weak, leading to debonding upon cycling, while Hagen-Poiseuille law dictates that filling meter-long hollow fibres (100–500 µm) requires tens of bar even after heating to lower viscosity, undermining scalability, uniformity, and sealing integrity29–33. These limitations underscore the need for an in-process core delivery strategy that achieves both strong interfacial bonding and continuous, high-loading fibre production. Here, we introduce a thermal drawing strategy that integrates liquid PCM injection into the hot cone region of the preform, where size reduction occurs. Guided by rheological and interfacial dynamics, this approach synchronizes the injection flow with the drawing speed, ensuring process stability and strong core-shell bonding. The interfacial stability and scalability of the microfluidic–thermal drawing process are governed by coupled rheological and interfacial dynamics (Supplementary Note 1). This enables the continuous fabrication of long core-shell PCFs with high-melting-temperature liquid cores that suit harsh environmental conditions. Unlike conventional PCM textile systems that mainly rely on PCM-dispersed composite structures or microencapsulated coatings34, the present strategy enables programmable continuous liquid-core architectures with enhanced structural stability and thermal durability. Molten-molten contact promotes dynamic wetting and interfacial adhesion, yielding strong bonding, molecular interpenetration, and mechanical interlocking upon cooling35–38. Controlled thermal processing with microfluidic core delivery and cold drawing in the high-elasticity regime yields mechanically robust fibres with enhanced thermal storage and diameters as small as ~195.3 ± 7.7 µm. The resulting phase-change textiles buffer rapid temperature surges and deliver >5 °C cooling in practical scenarios. Proposed approach supports continuous and scalable production and leakage resistance under repeated thermal cycling conditions over phase-change cycles. The core–shell design enables high PCM loading with structural stability and tunable functionalities, advancing thermal management textiles toward seamless integration of comfort and heat protection. Results Design and fabrication of PCF Integrating liquid PCM into microscale fibres at scale presents a fundamental challenge for conventional fabrication methods, including thermal fibre drawing. During conventional thermal drawing process, in the cone region of the preform (the hot zone), the macroscale liquid reservoir generates hydrostatic pressure that exceeds the mechanical resistance of the softened shell39,40, causing bulging during size reduction (Supplementary Fig. 1). To address this challenge, we developed a hybrid thermal-drawing and microfluidic strategy that enables precise control of the liquid-core flow during fibre draw. We found that tuning the flow rate of the PCM to a critical ratio relative to the preform feed speed allows the fabrication of continuous fibres with long and continuous liquid core (Supplementary Figs. 2–5). The coordination among the thermal-drawing speed, paraffin wax (PW) injection rate, and the cold-drawing process governs the formation of continuous liquid-core fibres (Supplementary Note 2, Supplementary Figs. 5 and 6). Fibre formation proceeds through three sequential processing zones: the thermal drawing–injection coupling region, the structural stabilization region, and the cold-drawing refinement region. These zones impose distinct thermo-mechanical conditions that govern interfacial formation, structural stability, and mechanical reinforcement during fibre fabrication (Supplementary Note 3). The stability of the liquid core is governed by the balance between the internal pressure of the injected PW and the viscoelastic resistance of the polymer shell during thermal drawing (Supplementary Note 4). This behaviour results from a dynamic equilibrium between the internal pressure of the liquid core and the mechanical resistance of the softened cone region during size reduction. Additionally, when the liquid-core fibre is reduced to sub-millimetre diameters, Rayleigh–Plateau instabilities dominate, fragmenting the core into discrete, non-uniform domains and preventing continuous fibre formation41. To achieve fibre diameters suitable for textile applications, we further implemented a post-processing cold-drawing step. This bypasses the need to remelt the PCM and enter the Rayleigh–Plateau instability regime, relying instead on viscous stretching of the polymer shell. Using a semicrystalline elastomer shell, we achieved stable cold drawing with a draw ratio of ~2×, producing flexible fibres with weaving-compatible diameters while maintaining uniform PW cores. Designing PCFs involves optimizing the fluidic core, which provides maximal latent heat and thermal regulation, while maintaining the structural integrity afforded by the solid shell (Supplementary Fig. 7 and Supplementary Note 5). Our multimaterial preform comprised a PW functional core, an elastomeric cyclic olefin copolymer (ECOC) shell for mechanical support, and a sacrificial, peelable styrene-ethylene-butylene-styrene (SEBS)-polymethyl methacrylate (PMMA) cladding (Fig. 1a, Stage 1, Supplementary Figs. 8 and 9). The SEBS/PMMA/ECOC multilayer architecture was designed to provide ductile mechanical support during thermal drawing, controllable sacrificial-layer removal, and robust long-term encapsulation of the PCM core. The thickness of each layer was designed to provide sufficient viscoelastic confinement to stabilize the injected PW core during thermal drawing while maintaining a high PCM loading fraction. In addition, the multilayer preform follows the geometric similarity scaling law of thermal-drawing processes, enabling proportional scaling of the preform without altering the stability of the core–shell configuration or the flow-matching relationship between injection and drawing (Supplementary Fig. 8). In this process, the PW core was heated above its melting point and maintained at constant temperature to ensure flowability and continuously injected into the multilayer preform during drawing via a microfluidic system (Supplementary Fig. 10), achieving real-time coaxial compounding with the ECOC shell. This coupling strategy ensured that the liquid core remained encapsulated within the neck-down region, suppressing gravity-induced pooling in the tapered melt flow and thereby mitigating bulge formation (Supplementary Fig. 1). The external SEBS-PMMA cladding provided structural support under thermos-mechanical stress, buffering perturbations from the fluid core while maintaining shell-core concentricity through interfacial tension and elasticity42,43. By transforming the inherently unstable liquid core into a processable long fibre, we enabled continuous fabrication of core–shell structures. Driven by interfacial energy, the PW spontaneously filled the interior of the fibre. In the subsequent removal stage (Fig. 1a, Stage 2), the cladding was peeled off to yield ECOC/PW fibres. Cold-drawing step (Fig. 1a, Stage 3) exploited the elastic shrinkage of ECOC in its rubbery state in concert with the solid-liquid transition of PW. This dual action diminished the fibre diameter further and improved mechanical robustness. As a result, ECOC/PW core-shell fibres could be continuously spooled in indefinitely lengths from metres to kilometres while retaining core integrity and thermo-mechanical stability. Programmable temperature control of paraffin across the thermal- and cold-drawing processes ensures stable phase-state regulation during fibre fabrication (Supplementary Fig. 10). Thermal drawing of PCFs and the mechanism for wearable thermal management. a Three-stage scalable process enables continuous PCF: during coaxial thermal drawing (Stage 1), VFeed = a × VFlow, where a ≈ 5 in this work. The ECOC shell provides viscoelastic confinement to stabilize the PW core; removal of protective layers (Stage 2) yields ECOC/PW fibres; subsequent cold drawing (Stage 3) reduces diameter and enhances mechanical strength. b Schematics of PCFs woven into fabric, the fibres both attenuate external heat flux and regulate body heat through latent-heat absorption and release, achieving peak-shaving and valley-filling effects that maintain thermal comfort under dual thermal loads. When woven into textiles, the hierarchical fibre-yarn porosity introduces extended heat transfer pathways, thereby attenuating and delaying instantaneous radiative, convective, and conductive fluxes to the skin (Fig. 1b). Under external (solar) or internal (metabolic/locomotion/battery generated heat) thermal loads, the PW undergoes a solid-liquid transition within its phase-change window, preferentially absorbing latent heat to form a quasi-isothermal plateau that suppresses thermal peaks and minimizes skin-temperature fluctuations. Upon cooling, the stored heat is gradually released, moderating the temperature drop and producing damped thermal oscillations under diverse conditions, including direct solar exposure during rest, shaded walking, and smartphone charging via a power bank in the pocket. Collectively, this work establishes a scalable route for fabricating wearable PCFs capable of sustaining thermal comfort under environmental, physiological, and device-induced thermal stress. Morphological, thermal, and mechanical properties of PCF The formation of ultrafine PCFs follows a multistage thermal-cold drawing strategy (Fig. 2a). In stage 1, thermal drawing produced fibres of ~1274.6 ± 51.6 µm, in which the SEBS-PMMA cladding provided viscoelastic support, maintaining interfacial stability and preventing molten PW from drifting or leaking during flow. Removing the protective layer yielded an ECOC/PW core-shell system (Stage 2), where interfacial energy redistribution drove the structure toward a bilayer morphology and reduced the diameter to ~389.5 ± 15.8 µm. In stage 3, cold drawing below the ECOC Vicat softening temperature induced axial stress on the PW core, and together with capillary contraction refined the diameter to ~195.3 ± 7.7 µm. As a result, we produced uniform and mechanically robust PCFs with microscale dimensions and high-throughput compatibility. Optical and scanning electron microscopy (SEM) imaging confirmed that, despite more than an order of magnitude reduction in diameter, the concentric core-shell architecture was preserved. This demonstrates that diameter reduction is a stable multistage evolution process, wherein viscoelastic confinement of the shell consistently suppresses interfacial instabilities and core displacement, thereby preventing leakage and geometric drift during repeated phase transitions44,45. The sequential process of thermal drawing-sacrificial removal-cold drawing enables controllable fibre diameter scaling from millimetre to hundred-micrometre dimensions. Structural design, mechanical robustness, and thermal functionality of PCFs. a PCFs with diameter of ~195.3 ± 7.7 µm, with corresponding optical and SEM images (scale bar, step Ⅰ: 400 µm, step Ⅱ and Ⅲ: 100 µm). b Example of PCFs threading into a sewing needle (scale bar, 2 cm). c Demonstration of PCFs manually twisted into yarn-like bundles (~10 turns cm⁻¹) (scale bar, 1 mm). d Tensile stress-strain curves of a representative PCF (R1) compared with reference fibres. e Scalable production of 35-metre-long PCFs shown on a spool (scale bar, 2 cm). f A single microscale PCF filament supporting 1 kg weight, highlighting its mechanical robustness. g DSC curves of pure PW and PCF, showing melting and crystallization transitions. h PCFs integrated into fabric by machine sewing and manual weaving, with thermal images confirming reversible heat regulation (scale bar, 1 cm). Fibre sizes reduced to hundreds-micrometre dimensions readily passed through a sewing needle, evidencing their ultrafine nature and flexibility (Fig. 2b). Multiple PCFs could be twisted into yarns of varying thickness and configurations (Fig. 2c and Supplementary Fig. 11), forming yarn-like bundles that demonstrate their structural adaptability for textile integration. The resulting bundles also exhibited enhanced mechanical strength (Supplementary Fig. 12). Distinct mechanical behaviour was observed before and after cold drawing (Fig. 2d): As-drawn fibres showed extraordinary ductility (2047 ± 28.4%) but low strength (15.3 ± 1.4 MPa). Cold drawing reduced elongation to 671 ± 10.2% while doubling tensile strength to 32.1 ± 0.8 MPa and further twisting yielded yarns with balanced ductility (603 ± 16.4%) and markedly higher strength (46.7 ± 1.2 MPa). The tensile strength of pristine PCFs shows a clear negative correlation with paraffin content yet remains at a relatively high level when the loading is kept at below 55% (Supplementary Fig. 7 and Supplementary Note 5). These results highlight how multistage processing and hierarchical structuring endow PCFs with record tensile performance while retaining flexibility (Supplementary Fig. 13 and Supplementary Table 1). This indicates that cold drawing is associated with enhanced molecular/crystalline ordering, which correlates with the observed strength increase. We continuously spooled 35 m of cold-drawn fibre (Fig. 2e), demonstrating meter- to tens-of-metres-scale production compatible with textile processing. Diameter measurements along metre-scale fibres, together with cross-sectional SEM images at multiple positions, further confirm the structural stability and uniformity of the fibres during continuous drawing (Supplementary Figs. 14 and 15). A 0.025 g single fibre of 280 µm diameter supported a 1 kg load without fracture (Fig. 2f)—over 104 times its own weight. This observed specific strength arises from chain-alignment reinforcement and synergistic load sharing in the core-shell architecture, where the shell bears tensile loads while the core dissipates strain energy. The thermal response of the fibres was further elucidated by differential scanning calorimetry (DSC) analysis (Fig. 2g). Pure PW displayed melting and crystallization enthalpies of 227 ± 0.8 / 226.9 ± 0.5 J g⁻¹ with a melting point of 37.8 ± 0.3 °C and a supercooling of 5.4 ± 0.2 °C (Supplementary Table 2). ECOC-encapsulated fibres maintained high enthalpies of 122.6 ± 1.5 / 119.8 ± 1.3 J g⁻¹ with a melting point of 37.1 ± 0.3 °C, exhibiting sharpened and concentrated endothermic/exothermic peaks in the 32–38 °C window, falling squarely within the human comfort range. Notably, the ECOC provided not only physical encapsulation but also kinetic modulation of the phase change via interfacial heterogeneous nucleation and confinement46,47, thereby reducing supercooling to 4.2 ± 0.3 °C and improving crystallization kinetics for rapid, reversible cycling. The crystallization temperature of PW confined in the ECOC shell shifts to a higher temperature compared with pure PW, suggesting that the ECOC–PW interface promotes heterogeneous nucleation. The crystallization behaviour and cycling stability of the fibres were further analysed using DSC and X-ray diffraction (XRD) (Supplementary Figs. 16–19). The reduced supercooling did not compromise fibre thermal stability (Supplementary Fig. 20), and the enthalpy efficiency closely matched the paraffin loading in PCF. This indicates that the thermal drawing process preserves the latent-heat capacity of PW while ensuring interfacial stability and record high energy-utilization efficiency. The ECOC/PW system thus combined high energy density, body-relevant transition window, structural stability, and leak-free operation over repeated cycles. These phase-change characteristics were vividly demonstrated at the fabric level (Fig. 2h and Supplementary Fig. 21), the fibres integrated seamlessly—whether machine-sewn in tracks or manual-woven into free patterns (e.g., ‘NUS’). Infrared imaging showed localized cooling zones during heating (latent-heat absorption, Fig. 2h) and delayed warming during cooling (heat release, Supplementary Fig. 21), with thermal maps matching the fibre geometry. Thus, once embedded, the fibres act not as passive fillers but as programmable thermal modulators that spatialize latent heat into controlled temperature distributions for localized and dynamic regulation. Textile integration of PCFs for breathable, anisotropic, and scalable wearable fabrics PCFs can be continuously unwound from spools and directly woven without binders (Fig. 3a), yielding fabrics with programmable architectures and intrinsic thermal regulation. Near the glass transition temperature, the molecular segments retain high reversible mobility, ensuring structural integrity and preventing fracture during weaving. Their flexibility and linear density (Supplementary Fig. 22) fall within the spinnable window for standard textile manufacturing, enabling stable fabric formation (Fig. 3b). The coupled thermal-mechanical characteristics of PCFs permit homogeneous interweaving with common fibres such as cotton (Fig. 3b and Supplementary Fig. 23), wool (Fig. 3c), and nylon (Fig. 3d). Even after tight weaving, the fibre cross-sections and surfaces remain intact, evidencing interfacial compatibility and processing robustness. The polar and compliant surface promotes mechanical interlocking and physical affinity at contacts, mitigating delamination or slippage and preserving mechanical uniformity. Fabrication and characterization of PCFF. a Schematic of the fabrication process from PCFs to woven fabrics, highlighting both structural and functional features. b PCFF composed of PCFs and cotton fibres (scale bar, 10 cm). Optical micrographs of hybrid fabrics blended with 50% wool (c) and 50% nylon (d), showing uniform interlacing and preserved fibre morphology after weaving (scale bars, 500 µm). e Moisture permeability remains comparable to conventional textiles, indicating that PCFF maintains comfort. f Mechanical test of PCFF, demonstrating enhanced strength and extensibility along the fibre axis while retaining softness in the orthogonal direction, evidencing programmable anisotropy. g Large-area PCFF (50% cotton blend), underscoring scalability toward practical textile manufacturing. h A baby garment produced from PCF/cotton fabric (scale bar, 7 cm). PCF fabrics (PCFF) textiles retain polyester-like water absorption (Supplementary Fig. 24) and moisture permeability (Fig. 3e), polyester 85.8 ± 0.9, PCF/cotton 74.1 ± 0.5, and PCF/nylon 81.1 ± 1.1 g m⁻² h⁻¹, indicating that fibre diameters match fabric pore scales and do not occlude pathways, thus preserving breathability. Notably, PCF/cotton shows ~7.5% higher moisture permeability than neat cotton, which arises from the latent-heat-driven regulation of local thermal and humidity gradients: PCFs absorb and release heat during environmental fluctuations, actively facilitating vapor diffusion as dynamic transport modulators. At the fabric scale, stress-strain tests quantify the programmable mechanical anisotropy introduced by PCFs. Neat wool exhibits a tensile strength of 4.1 ± 0.3 MPa and a strain at break of 50 ± 3.6% (Fig. 3f). Along the wool direction, PCF/wool fabrics reach 5.1 ± 0.6 MPa and 73.2 ± 5.3%. Loading along the PCF direction yields 24.5 ± 0.7 MPa and 215.3 ± 9.3%, corresponding to 6.0× higher strength and 4.3× higher extensibility than neat wool. Relative to the wool-direction loading, the PCF-direction response remains strongly enhanced (4.8× in strength and 2.9× in extensibility), confirming that placing PCFs along the principal load path imparts programmable, direction-specific reinforcement while preserving fabric deformability. Beyond laboratory-scale demonstrations, PCFs can be woven into continuous 1.6 m textiles (Fig. 3g) and tailored into garments (Fig. 3h), underscoring their compatibility with industrial textile integration workflows. Full-scale prototypes retaining breathability, anisotropic mechanics, and defect tolerance underscore the scalability of PCFs toward practical wearables. In addition, the practical durability of the PCM-integrated textile was evaluated through machine washing, cyclic bending, and abrasion tests. The fabric exhibits negligible mass loss, preserved fibre morphology, and unchanged phase-transition behaviour after washing, while the woven structure remains intact after repeated mechanical deformation (Supplementary Figs. 25–27). Bidirectional and durable thermoregulation of PCFF under practical heat flux conditions Building on the spinnability, interfacial stability, and scalability demonstrated in Fig. 3, we evaluated the thermoregulation of PCFF under realistic heat fluxes (Fig. 4a; detailed setup in Supplementary Fig. 28). Using a dual-thermocouple arrangement (TC-I near the fabric and TC-II near the skin), the surface heat absorption associated with phase-change was decoupled from the microclimate response on the skin, revealing a bidirectional thermal management mechanism. As PCF content increased, the surface temperature curves exhibited clear plateaus within the transition window (Fig. 4b), indicating preferential absorption of the input heat by latent enthalpy. Compared with the non-PCF control (PCF content: 0%, 58.5/48.1 °C, Supplementary Fig. 29), the equilibrium temperatures were markedly reduced in PCFF (Fig. 4c): 15% (55.8/39.7 °C), 30% (53.6/39.4 °C), and 50% (53.1/38.6 °C). These results show that even under steady heating conditions, PCFF continuously stores part of the incoming heat and suppresses the microclimate temperature. Acting as an enthalpy sink, the phase-change layer effectively attenuates the net heat flux toward the skin. The consistently higher thermal shielding difference further confirms enhanced cooling and stabilization effects. Thermoregulation performance of PCFF. a Schematic of the experimental setup with a data acquisition system to evaluate cooling under heating and heat retention during rapid cooling; Thermocouples I and II are located close to the fabric and skin, respectively. Temperature profiles recorded by thermocouples I (b) and II (c) for white fabrics with varying PCF contents. d Rapid temperature-drop tests using thermocouple III, illustrating the enhanced cooling-buffering capability of PCFF during transition from hot to cold environments compared with cotton textiles. Infrared image of a 50% PCF fabric versus cotton (e) and the corresponding temperature–time curves (f), showing delayed heating and cooling in the PCF region due to latent heat effects, scale bars, 10 mm. g Long-term stability tests of a 50% PCF fabric versus cotton, revealing substantially enhanced thermal buffering. h Cyclic heating tests of the 50% PCFF, confirming stable and repeatable thermoregulation over 11 on/off cycles. White PCFF most effectively suppressed surface heating and enhanced cooling perception. Black PCFF with 50% PCF also reduced surface and microclimate temperatures to 57.3 °C and 47.8 °C, respectively, substantially outperforming black non-PCF textiles (60.3/38.6 °C, Supplementary Fig. 30). This indicates that latent-heat buffering, rather than optical contrast (colour-dependent optical absorption), dominates the thermal response. Under abrupt hot-to-cold transitions, PCFF charged with latent heat markedly delayed cooling by releasing stored enthalpy, yielding prolonged half-life (t₁/₂) and smoother thermal trajectories compared with their pre-transition state and polyester controls (Fig. 4d). Acting as localized heat sources, the PCFF buffers cold shocks and reduces physiological stress. Overall, PCFF redistributes heat flux in both time and space: suppressing peak rises on the hot side and smoothing cooling on the cold side. Infrared thermography (Fig. 4e) reveals that PCFF regions divert part of the input energy into phase-change near melting temperature, elevating the effective heat capacity and suppressing temperature rise. This leads to isothermal retention and reduces thermal peaks. During cooling, latent-heat release prolongs heat dissipation, yielding extended half-lives and smoother cooling trajectories. Temperature-time curves (Fig. 4f) quantify these effects: plateaus and delayed peaks during heating, and shoulder-like segments with prolonged t₁/₂ and reduced the maximum cooling rate during cooling. PCFs thus function as dynamic enthalpy reservoirs, shaving thermal peaks on the hot side and filling valleys on the cold side. Long-term tests confirm robust stability: PCFF consistently maintains lower equilibrium temperatures than cotton (Fig. 4g and Supplementary Fig. 31) and preserves nearly constant heating delays and cooling half-lives across 11 on/off cycles with negligible drift (Fig. 4h). These results establish PCFF as stable and repeatable thermal buffers, delivering substantially stronger and more durable regulation than conventional textiles. Practical demonstrations of versatile thermal management using PCFF Building on controlled laboratory demonstrations, we next evaluated PCF fabrics under outdoor conditions to assess their practical cooling performance. In field tests, garments were designed with one half made of PCFF and the other of cotton (Supplementary Fig. 32), while an IR camera tracked surface temperatures under direct sunlight (Fig. 5a). Even under ambient air temperatures of ~33 °C, ground surface temperatures in direct sunlight could exceed 70 °C (as measured on the testbed), creating an extreme thermal environment for evaluating PCFF performance. Under these conditions, the PCFF side retained distinct cold zones and sustained lower temperatures, whereas cotton rapidly heated into high-temperature regions (Fig. 5b and Supplementary Movie 1). Corresponding profiles showed that the skin temperature beneath cotton rose to 42 °C, while PCFF stabilized near 39 °C, yielding a ~ 3 °C offset and slower heating rates (Fig. 5c). The cooling advantage stems from latent-heat absorption in PCFF, which delays skin temperature rise and stabilizes the microclimate. Cotton, lacking this enthalpy reservoir, cannot resist rapid heat influx—explaining the sharp contrast in infrared images. We further simulated garment conditions by integrating PCFF into the inner layer of clothing and applying external heating from the outside (Supplementary Fig. 33). Both laboratory fabric-sample experiments and wearable-clothing tests (Supplementary Fig. 34) demonstrated that, even without direct solar irradiation, PCFF segments substantially suppressed local temperature rise and delayed the onset of thermal peaks. To demonstrate the versatility and practical relevance of PCFFs in garments, we also designed a PCFF newborn onesie to stabilize body temperature and reduce the likelihood of illness. Infrared images show superior cooling of onesie made from PCFF compared with commercial cotton onesie (Supplementary Fig. 35). This result demonstrates that PCFF provides effective thermal buffering not only under sunlight but also in enclosed configurations, underscoring their versatility and practical relevance for wearable thermal management. Real-world validation of thermoregulation in PCFF. a Outdoor test setup under high-temperature conditions, with PCFF sewn on the right side of a garment and cotton on the left for direct comparison. Infrared images (b) and corresponding surface-temperature profiles (c), showing that PCFF substantially suppressed temperature rise relative to cotton. Surface-temperature responses under moderate outdoor conditions during two representative activities—lying (d) and walking (e)—where PCFF lowered equilibrium body temperatures. f Schematic of a pocket and smartphone case fabricated from PCFF, with monitored device heating during wireless charging. g Comparative illustration of temperature control for a smartphone using PCFF versus cotton fabrics, with an inset photograph showing a practical smartphone case fabricated from PCFF. h Maximum surface temperatures across outdoor and device scenarios, demonstrating consistent cooling performance of PCFF (3–6 °C reduction) compared with cotton. Building on short-term validations, we performed long-duration field trials under two representative conditions: direct sunlight during sitting/lying (Fig. 5d and Supplementary Fig. 36a) and walking (Fig. 5e and Supplementary Fig. 37a) with convective cooling. Across both scenarios, PCFF consistently produced lower and more stable skin-temperature trajectories than cotton, with peak suppression of 3–5 °C, plateau levels reduced by 2-3 °C, and marked temporal delays. Importantly, transient irradiance spikes further amplified this advantage; under intense sunlight at 12:57, the differential reached 5.4 °C (Fig. 5d), substantially improving thermal comfort in hot outdoor environments. Infrared thermography confirmed that PCFF-covered regions sustained cold zones over tens of minutes, effectively shaving peaks and filling valleys in the thermal response (Supplementary Fig. 36d). Even during walking, where convection reduced absolute skin temperatures, PCFF retained its relative advantage, yielding smaller oscillations and faster recovery to the comfort range (Supplementary Fig. 37d). Under combined strong irradiance and convection, vertically suspended garments still exhibited fabric-level temperature differences of ~4.2 °C (at ~11:14, Fig. 5e). These observations point to a unified mechanism: PCFF functions as latent-heat buffers, absorbing thermal loads during irradiance or conduction surges to delay heating, and releasing stored heat during relaxation to smooth cooling. Across diverse perturbations—direct and indirect solar exposure, static and dynamic states—PCFF consistently maintains lower peaks, reduced fluctuations, and a more stable thermal microclimate (Supplementary Fig.36b,c and 37b,c). Additionally, the effective cooling duration was further quantified by analysing the temperature difference between PCFF and cotton fabrics extracted from these outdoor measurements (Supplementary Note 6). The PCFF maintains a sustained cooling advantage (ΔT ≥ 1–3 °C) for tens of minutes under realistic outdoor heat flux (Supplementary Fig. 38 and Supplementary Table 3). Repeated cooling peaks in the outdoor temperature curves indicate that the PCM naturally resolidifies under fluctuating solar irradiation, sustaining multiple phase-change cycles (Supplementary Note 7). To further demonstrate the versatility of PCFF, we integrated them into pocket liners and phone cases and monitored device heating during wireless phone charging (Fig. 5f–h). Without PCFF, phone temperatures rose sharply to 43-44 °C within 600–900 s, while the power bank reached 42-43 °C (Fig. 5g and Supplementary Fig. 39). With PCFF protection, thermal responses were markedly suppressed and delayed: peak temperatures dropped to 37-38 °C on the power bank (5–7 °C) and 40-41 °C on the phone (3-4 °C). The initial heating rate was reduced, and the duration of elevated temperature shortened. Compared with cotton-based pockets, PCF-based cases or pockets provide more stable fitting during phone charging, ensuring reliable buffering even under charging-while-using conditions. Overall, this demonstration confirms that PCFF enables bidirectional thermal management for body-adjacent electronics: reducing peak temperatures and thermal dose without sacrificing portability. Moreover, a direct comparison of maximum temperature differences across multiple conditions, including outdoor lying (5.3 °C), walking (4.3 °C), and operation of heat-generating mobile (4.6 °C) electronic devices (Fig. 5h), confirmed the robust cooling effect of PCFF. Discussion In this work, we established a scalable strategy to integrate liquid phase-change systems into continuous, micron-scale fibres with lengths extending to tens of metres. By combining microfluidic–thermal drawing with post-cold-drawing, we achieved uniform and 200 µm-thin liquid-core PCFs that overcome the fundamental instabilities of conventional fibre processing. The resulting fibres with ultra-high strength and flexibility (671 ± 10.2% of elongation, 32.1 ± 0.8 MPa of tensile strength) exhibit record supercooling stability (4.2 ± 0.3 °C) compared with raw PW and can be woven directly into full fabrics without compromising textile mechanics. When incorporated into garments, PCFF consistently delivered 3–6 °C reductions in surface temperature, suppressing peak heating and delaying thermal rise through sustained latent-heat cycling. This capability was validated across both controlled laboratory settings and outdoor conditions, as well as mitigating heat accumulation in electronic devices. Beyond demonstrating continuous fibre fabrication and functional fabrics, our results establish PCFs as a platform for scalable, application-relevant thermal management. The core-shell design principle is versatile, offering opportunities to integrate additional functions into the shell—which serves as mechanical support in this study—such as radiative cooling, photonic reflection, moisture regulation, or sensing. The convergence of scalable manufacturing, tunable functionality, and demonstrated cooling efficacy positions PCFF as a transformative material platform for sustainable thermal regulation in apparel, infrastructure, and next-generation electronics. Methods Fabrication of SEBS-PMMA/ECOC preform Prior to thermal drawing, hollow SEBS-PMMA/ECOC preform was assembled through a three-stage process. The preform architecture comprised two U-shaped SEBS-PMMA cladding layers and an intermediate rectangular ECOC tube containing a central hollow channel reserved for the injection of PW (PCM-A-37, Shengbang Plastics, Dongguan, China) injection. SEBS (G1657, KRATON) and ECOC pellets (E-140, TOPAS) were dried at 50 °C for 12 h and hot-pressed into films with thicknesses of 4 mm and 2 mm, respectively (5 t, 160 °C for SEBS; 5 t, 135 °C for ECOC). ECOC films were cut into strips of 25 cm × 4 mm × 2 mm and 25 cm × 8 mm × 2 mm, while SEBS films were cut into two 30 cm × 24 mm × 4 mm strips and a U-shaped strip with a central cut-out of 25 cm × 8 mm × 4 mm (Supplementary Fig. 40). Using a PTFE support rod, the films were precisely aligned and consolidated at 110 °C under 0.3 t for 5 min. Subsequently, PMMA films (LORIKA, thickness: 45 μm) were laminated onto the inner walls of the U-shaped SEBS layers, into which the ECOC hollow column was embedded. A second U-shaped SEBS-PMMA strip was placed on top to form a sandwich-like architecture (Supplementary Fig. 8). Before consolidation, a compression-resistant PTFE rod (3.9 mm × 3.9 mm× 25 cm) was inserted at the centre to preserve the hollow channel under heat and pressure, ensuring a stable cavity for subsequent PW infusion. Final consolidation at 110 °C and 0.3 t for 5 min yielded intact SEBS-PMMA/ECOC preforms. After rod removal, the preforms measured 30 cm × 24 mm × 24 mm with a central channel of 25 cm × 4 mm × 4 mm, as shown in Supplementary Fig. 9. Thermal drawing of SEBS-PMMA/ECOC/PW preform The SEBS-PMMA/ECOC preform was mounted on a hollow support and connected via tubing to a heated syringe loaded with molten PW, integrated into a microfluidic injection system (Fig. 1a). The drawing tower was operated with furnace zones set at 110 °C (top), 175 °C (middle), and 70 °C (bottom). The preform was fed into the furnace at 1 mm min⁻¹, and drawn at 0.25 m min⁻¹, corresponding to a draw ratio of 17. Once the fibre diameter stabilized, PW was continuously delivered at 0.2 mm min⁻¹ (corresponding to ~69 μL min⁻¹ for a 20 mL syringe) during thermal drawing, enabling uniform filling of the central channel during elongation. The resulting SEBS-PMMA/ECOC/PW PCFs were continuously collected on a take-up roller with consistent diameters. Following thermal drawing, the sacrificial SEBS-PMMA cladding was removed to yield ECOC/PW PCFs. A subsequent cold-drawing step, performed with the second take-up roller at 0.5 m min⁻¹, further reduced the fibre diameter and enhanced mechanical robustness, yielding ultrafine PCFs. Characterization The surface morphology of the fibres was examined using an optical microscope (CLuosiOptics, SZM), while microscopic structural features were analysed by SEM (Zeiss Sigma 300). Prior to SEM analysis, fibre samples were sputter-coated with gold, and measurements were performed at an accelerating voltage of 5 kV. The phase-change behaviour of PCFs was investigated by DSC (TA, DSC25). The testing protocol involved heating the sample to 80 °C and holding for 3 min, followed by cooling to −20 °C at a rate of 5 °C min⁻¹ and holding for 3 min, and subsequently reheating to 80 °C at the same rate. Besides, the thermal cycling stability of the fibres was evaluated using a dry bath (DB100C, Zhejiang Oumaike Experimental Instrument Co., Ltd., China). The samples were subjected to 200 consecutive thermal cycles following the temperature program 0 °C (3 min) → 70 °C (3 min) → 0 °C (3 min). After cycling, the phase-change behaviour of the fibres was analysed by DSC, and the melting/crystallization temperatures and phase-change enthalpies were compared with those measured before cycling. XRD measurements were carried out using a Bruker D8 Advance diffractometer with Cu Kα radiation (λ = 1.5406 Å). The diffraction patterns were recorded in θ–2θ mode over a 2θ range of 5°–60°, with a step size of 0.0197° and a counting time of 0.2 s per step. Thermogravimetric analysis (TGA, TA, TGA50) was performed under nitrogen flow (50 mL min⁻¹). Samples of approximately 5–10 mg were heated from 25 to 600 °C at a rate of 10 °C min⁻¹. TGA and DTG curves were recorded to determine the onset of decomposition, peak degradation temperature, and residual mass. Water absorption was measured following ASTM D570. Heat-sealed specimens (≈50 × 50 mm) were conditioned at 23 °C and 50% relative humidity for 24 h, then immersed in deionized water at 23 °C. At 10 min and 5 h, samples were drained for 30 s, blotted, suspended for 60 s, and weighed. Water absorption (%) was calculated as (mt–m₀)/m₀ × 100, where m₀ and mt represent the sample mass before immersion and after immersion for time t, respectively. The thermal response of fibres and fabrics was recorded using an infrared thermal imaging camera (FOTRIC 618 C, FOTRIC Inc.) simulated solar irradiation provided by a xenon lamp light source. Infrared thermal images of Fig. 2h and Supplementary Fig. 21 were recorded under ambient laboratory conditions (23 ± 1 °C). Heating was provided by an infrared lamp with a maximum power of 375 W, operated at 60% output during the experiments. The lamp emits broadband infrared radiation with a characteristic wavelength in the infrared region (~4 μm) and was positioned 20 cm above the sample. The samples were sewn into textile fabrics and placed on a wooden table during the measurements. Cooling occurred naturally in air without forced convection. Infrared images were captured at specific time points during the heating and cooling processes for comparison between different patterns. Mechanical properties of the fibres were evaluated using a universal testing machine (Instron 34SC-05) at a crosshead speed of 5 mm min⁻¹. At least five independent specimens were tested for each condition. In addition, fabric-level cyclic bending durability tests were performed using a universal mechanical testing machine (PY680, PUYAN, China). Fabric specimens were mounted between two clamps and subjected to repeated vertical motion at a speed of 100 mm min⁻¹. Each cycle consisted of a 0.2 min displacement followed by a holding period of 0.01 min at the maximum displacement. The test was conducted for 500 cycles to evaluate the mechanical durability of the textile under repeated deformation. Fabric-level abrasion durability tests were performed using a reciprocating sliding configuration. The textile specimen was placed on a cotton fabric substrate, and a 500 g weight was applied to the central region to introduce a constant normal load. The sample was subjected to reciprocating sliding motion with a stroke length of 20 cm per cycle (1 cm s−1) for a total of 500 cycles48. Optical microscopy was used to examine the fabric morphology before and after the abrasion test. Moisture permeability was assessed using the cup method (ASTM E96/E96M, water procedure). Deionized water was placed in the cup, and the specimen was sealed over the orifice. The assemblies were equilibrated at 38 °C and 90% relative humidity. Cup mass was recorded at fixed intervals, and the steady-state slope of mass change versus time was used to calculate the water vapour transmission rate (WVTR, g m⁻² h⁻¹) based on the calibrated test area. The washing durability of PCM-integrated textile was evaluated using a household washing machine (Yangzi, China) operated in quick-wash mode. The washing program consisted of the following sequence: water filling (2 min), washing (5 min), draining and spin-drying (5 min), followed by a second water filling (2 min) and washing step (5 min), and a final draining and spin-drying process (5 min). During the washing process, the drum agitation cycle was approximately 2 s, generating repeated mechanical bending, friction, and water flow similar to practical laundering conditions. The garment was weighed before and after washing using an electronic balance to evaluate potential mass loss associated with PCM leakage. After washing, the garment was air-dried at room temperature before reweighing and further characterization. Supplementary information Supplementary Information Description of Additional Supplementary Files Supplementary Movie 1 Peer Review file Author contributions X.L. and T.K. conceived and planned this work. X.L. fabricated the preforms. X.L. and Q.L. performed fibre drawing. X.L., X.Z., and Q.L. carried out characterization of the fabric. X.L. and Q.L. built custom experimental apparatuses. X.L., Q.L., and T.K. analysed and interpreted the data and prepared the figures. X.L. and T.K. drafted the manuscript. T.K. supervised and coordinated the study. All authors contributed to the discussion of results and critically reviewed the manuscript. X.L. and Q.L. contributed equally to this work. Peer review Peer review information Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available. Funding This research was supported by the National Research Foundation, Singapore, under its Competitive Research Programme (NRF-CRP30-2023-0001); the Ministry of Education, Singapore, under its Academic Research Fund Tier 1 (FY2022); and the National University of Singapore, under its Start-up Grant (FY2022). Data availability The data generated and analysed during this study are provided in the main manuscript and the Supplementary Information. All data are available from the corresponding author upon request. Source data are also available in GitHub at https://github.com/xlli63563-ui/Source-data-PCF. Competing interests The authors declare no competing interests. Human participant consent Written and informed consent was received from all volunteers in this study. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Xiaolong Li, Qianyi Li. Supplementary information The online version contains supplementary material available at 10.1038/s41467-026-75800-4. References Associated Data Supplementary Materials Supplementary Information Description of Additional Supplementary Files Supplementary Movie 1 Peer Review file Data Availability Statement The data generated and analysed during this study are provided in the main manuscript and the Supplementary Information. All data are available from the corresponding author upon request. Source data are also available in GitHub at https://github.com/xlli63563-ui/Source-data-PCF.