Chemical O₂-generated aerogel-structured wearable microreactors

Abstract

Wearable chemical microreactors are highly desirable for life-sustaining processes such as on-demand oxygen-generation, but their development has been hindered by the lack of materials that combine dynamic robustness, textile integrability, and ultra-efficient fluid transport. Herein, we show a wearable microreactor based on armored aramid aerogel hollow fibers. Reinforcing these fibers with heat-shrink tubing yields textile-compliant elasticity while preserving their nanoporous structure. Subsequently, infusing a low-surface-energy liquid into the mesopores of aerogel walls creates a nanoconfined lubricant layer that reduces microfluidic drag by up to 68.1% compared to solid-walled channels. Finally, a O₂-generated wearable microreactor prototype (268 g) achieves high space-time yield, 100% oxygen purity, and maintaining oxygen-supply under physical stress. In vivo trials at 4998 meters altitude demonstrate rapid blood oxygen saturation recovery, effectively alleviating acute hypoxia. This work establishes a materials paradigm that overcomes the trade-off between wearability and continuous-flow microreactor in on-body chemical synthesis.

Wearable chemical microreactors show applications for life-sustaining processes but their development suffers from the lack of materials with dynamic robustness, textile integrability, and ultra-efficient fluid transport. Here, the authors develop a wearable microreactor based on armored aramid aerogel hollow fibers for oxygen generation with high space-time yield and maintaining oxygen-supply under physical stress.

Introduction

Microreactor technology has fundamentally advanced chemical processing by leveraging micro-scale confinement to achieve exceptional mass transfer efficiency, precise fluid control, and enhanced safety, enabling a significant shift from bulk reactors to compact, integrated systems¹. Extending these transformative capabilities to the realm of wearable technology holds immense promise for groundbreaking applications, such as on-demand, precision medicine (e.g., in vivo synthesis of therapeutics²) and continuous physiological monitoring (e.g., respiratory analysis³ and dynamic biomarker tracking⁴). However, the transition to body-compliant, wearable systems is fundamentally hampered by an irreconcilable materials dilemma. On the one hand, the exceptional efficiency of conventional microreactors relies on precisely engineered, rigid microchannels (glass, metals) that maintain stable geometries for predictable fluid dynamics⁵. Wearable operation, however, demands continuous mechanical compliance to withstand dynamic stresses—bending, stretching, compression, and impact⁶—encountered during human movement. Rigid substrates fracture under such conditions, while compliant polymeric alternatives deform unpredictably, disrupting critical microfluidic patterns and reaction kinetics. On the other hand, seamless integration into textiles or garments necessitates lightweight, flexible, and processable materials compatible with textile manufacturing. Current attempts⁷ to adapt microreactors for wearability often result in bulky external housings⁸ or necessitate significant performance sacrifices⁹, thereby forfeiting the very advantages of miniaturization and integration that define microreactor technology. Therefore, the foremost challenge is to create a microreactor substrate that is simultaneously mechanically robust under dynamic wear and efficient for mass transfer.

To overcome the predicament of this wearable microreactor, aerogel fibers have emerged as a unique and promising material. They represent an innovative class of materials that transcend conventional textile frameworks by combining the nanoporosity, ultralow density, and exceptional surface activity of aerogels with the mechanical flexibility and processability of fibrous architectures^10–13. Over the past decade, rapid progress has been made in material diversification (e.g., towards polyimide¹⁴, aramid^15–17, graphene^18–20, silica²¹, and titanium carbide²² systems), structural engineering (including hollow^23–25, coaxial^26,27, and porous‑cortex‑dense‑core structure²⁸), and advanced processing techniques (such as dynamic sol-gel wet spinning combined with supercritical drying^22,29, and wet-freeze spinning coupled with freeze-drying^30,31). Unlike traditional fibers, the nanoporous architecture and tunable composition of aerogel fibers confer several critical functionalities for microreactors: efficient mass transport via interconnected structures, tailorable surface chemistry/thermal stability for diverse reactions, and the unique ability to host and stabilize functional fluids via nanoconfinement. Their quasi-1D geometry further enables seamless incorporation into textile-based devices—a critical advantage over brittle monolithic aerogels. However, a persistent technological paradox continues to constrain progress: the very nanoporous framework that enables their function is inherently mechanically fragile. While recent encapsulation strategies³⁰ have partially addressed mechanical fragility, they inadvertently sacrificed porosity—the very feature that underpins aerogels’ performance. Thus, the specific and unresolved challenge for wearable microreactors is: how to radically enhance the mechanical durability of aerogel fibers without compromising their nanoporous, functional interior? This question defines the key knowledge gap our work addresses.

To address the aforementioned challenges, we introduce a two-fold, synergistic strategy that transforms fragile aerogel fibers into high-performance wearable microreactors. First, we address the mechanical fragility externally. We develop a non-destructive armoring strategy using heat-shrink tubing to create an armored aramid aerogel hollow fiber (AAAHF). This design concurrently delivers traditionally mutually exclusive properties: mechanical robustness (17.6 MPa tensile strength), preserved chemical functionality (99% mesoporosity with 2–50 nm pore distribution), 770% elongation, and multifunctional protection (hydrophobicity, oleophobicity, flame retardancy, stain resistance, and textile compatibility). However, armor alone is insufficient for a high-performance microreactor. The preserved nanoporous network, while intact, still presents high fluidic resistance.

Therefore, the second, critical step addresses internal transport efficiency. Functionalization with perfluoropolyether (PFPE) oil creates a solid-liquid composite microreactor (AAAHF@K101, K101 represents the brand name of the PFPE oil used), enabling dual breakthroughs: (1) mesoscale stabilization of functional fluids through mesoporous confinement effects, and (2) enhanced fluid transport efficiency surpassing conventional microtubules, achieving a drag reduction rate of up to 68.1%. The synergistic strategies between armoring walls and nanoconfined lubricants pave the way for assembling wearable microreactors. We integrated meter-long AAAHF@K101 into a textile-based O₂-generated wearable microreactor weighing ~268 g with a space-time yield (STY) of 26498 g L⁻¹ h⁻¹ (three orders of magnitude beyond traditional reactors) (Fig. 1). The system consists of a ~68 g core unit (containing AAAHF@K101, two miniature peristaltic pump, a custom gas-liquid separator, a commercial gas purification module, and oxygen supply medium) plus ~200 g of working substance. The system demonstrates sustained oxygen production (100 h) with replaceable oxygen sources, delivering 100% pure oxygen at adjustable flow rates. Remarkably, it maintains stable oxygen generation under extreme conditions, including high-intensity activities (e.g., sporting and mountaineering) and high-altitude (up to 4998 m) environments, while combining portability, wearability, and rechargeability. The demonstrated methodology provides both technical feasibility and theoretical foundation for developing advanced wearable microreactors, while offering insights into multifunctional textile-integrated devices.

Fig. 1. Schematic diagram on a chemical O₂-generated wearable microreactor based on armored aramid aerogel hollow fibers.

This design addresses internal transport efficiency through two synergistic mechanisms: (1) mesoscale stabilization of the slippery fluid via confinement within the mesoporous aerogel wall, and (2) enhanced fluid transport that surpasses conventional microtubules, achieving a substantial drag reduction effect. Meter-long AAAHF@K101 is integrated into a textile-based wearable microreactor. The system exhibits a high space-time yield for oxygen generation, which is orders of magnitude higher than that of traditional reactors. aq indicates aqueous solution.

Results

Fabrication of AAAHF and the popularity of general strategies

We employed commercial heat-shrink tubing to construct armored aerogel fibers through a solvent-free fabrication method, achieving a nanoporous core/dense-tough shell architecture. The comprehensive preparation protocol is illustrated in Fig. 2a. The process commenced with the synthesis of aramid aerogel hollow fibers via dynamic bidirectional sol-gel wet spinning technology. In this technique, aramid nanofiber dispersion as the spinning solution and deionized water as the coagulation bath were respectively injected through the outer and inner channels of a coaxial needle, followed by co-extrusion into the deionized water coagulation bath. The sol-gel transition of aramid nanofiber dispersion was triggered by proton acquisition from the aqueous coagulation medium under dual-phase bath interaction³², resulting in the formation of aramid hydrogel hollow fibers. Subsequent solvent exchange and freeze-drying procedures yielded aramid aerogel hollow fibers. The fiber samples underwent thermal annealing at 400 °C for 30 min under an argon atmosphere to optimize network crosslinking, thereby yielding heat-treated aramid aerogel hollow fibers³³. The final armoring step involved applying polyethylene (PE) heat-shrink tubing to heat-treated aramid aerogel hollow fibers, generating AAAHF with preserved nanoporous core and integrated dense-tough shell. PE was selected as the primary armor material due to its advantageous properties: low shrinkage temperature, mechanical flexibility, additional flame retardancy, and cost-effectiveness³⁴. All subsequent characterizations refer to PE-armored systems unless otherwise stated. The thermal shrinkage mechanism, depicted in Fig. 2b and Fig. S1, originates from the polymer’s shape memory behavior³⁵. During manufacturing, heat-shrink tubing undergoes radial expansion under applied mechanical load when heated to its high-elastic state (above glass transition temperature, T_g), followed by rapid cooling to fix the expanded configuration. Upon reheating during application, thermal activation above T_g releases stored elastic energy, enabling polymer chain segments to recover their pre-expansion conformation through entropy-driven relaxation. This intrinsic shape memory property facilitates non-destructive aerogel encapsulation. Briefly, the heat-treated aramid aerogel hollow fiber was inserted into pre-expanded PE heat-shrink tubing (Movie S1) and heated at 110 °C for 15 s to induce complete shrinkage. Thermal crosslinking significantly improved elastic recovery while maintaining structural integrity (Fig. S2). As evidenced in Figs. S3 and S4, the heat treatment preserved original network structure, chemical composition, nanoporous architecture, and tensile property, while substantially enhancing compression resilience—a critical attribute for elastic microchannels in wearable microreactors. The dimensions of both the nanoporous aerogel core and the PE armor shell can be independently and precisely tuned. The thickness of the aerogel wall was controlled in the range of 150–350 µm by adjusting coaxial spinning parameters (needle gauge combinations and extrusion rates of the inner/outer spinning solutions) (Fig. S5 and Table S1). Meanwhile, the thickness of the PE armor layer was tuned between 180 and 250 µm by selecting commercial heat-shrink tubing with specific initial inner diameters and shrinkage ratios (Fig. S6). This dimensional control ensured intimate interfacial contact between the PE armor and the aerogel fiber, as confirmed by scanning electron microscopy (SEM) (Fig. 2c) and summarized in Table S2. High-magnification SEM imaging revealed undisturbed integration between nanoporous aerogel core and dense outer shell (Fig. 2d). Critically, the non-destructive armoring process preserves the intrinsic nanoporous network of the aerogel core. Comparative SEM images of the cross-sections of unarmored aerogel fibers and AAAHF reveal identical, continuous porous networks without collapse or blockage (Fig. S3). Brunauer–Emmett–Teller (BET) specific surface area analysis of the aerogel core (Fig. 2e) demonstrated preserved mesoporosity with 183.8 m²/g specific surface area, 0.7 cm³/g mesopore volume, and 18.4 nm average pore diameter, which was almost consistent with the measurements before armoring (197.1 m²/g specific surface area, 0.6 cm³/g mesopore volume, and 13.7 nm average pore diameter) (Fig. S4a, b). The combination of SEM and BET analyses conclusively demonstrates that the inner morphology and porosity of the aerogel fiber remain unaltered after armoring, effectively preventing the capillary-induced nanopore collapse associated with conventional coating-drying approaches. The mechanical properties of the AAAHF exhibited a remarkable enhancement, with tensile strength doubling from 6.0 MPa (before being armored) to 17.6 MPa, while elongation at break demonstrated an extraordinary 38.5-fold increase from 20% (before being armored) to 770% (Fig. 2f). The AAAHF (500 μm wall thickness) demonstrated greatly enhanced mechanical robustness, sustaining 9.8 N loads (equivalent to 1 kg mass under standard gravity) without fiber fracture—a 100% improvement over state-of-the-art thermoplastic polyurethane-encapsulated aerogel fibers (maximum 4.9 N/500 g capacity)³⁰ (Fig. S7). This load-bearing performance fundamentally surpasses the capabilities of aerogel fibers, which failed catastrophically under identical loading conditions. The proposed solvent-free and non-destructive armoring strategy significantly enhanced the mechanical properties of aerogel fibers. By using a core-shell structure, the brittle aerogel fibers were transformed into a tough composite material. The continuous PE sheath acts both as a crack-arresting barrier, confining the brittle core to suppress catastrophic crack propagation, and as a load-bearing, energy-dissipating matrix. Thus, the role of aerogel shifts from a load-bearing skeleton to a functional core protected by a tough, energy-dissipating shell.

Fig. 2. Preparation and characterization of armored aramid aerogel hollow fibers.

a Schematic diagram on the preparation process of AAAHF. b Upon thermal stimulation, the heat-shrink tubing radially contracts to structurally reconfigure the heat-treated aramid aerogel hollow fibers into AAAHF, accomplishing a seamless interface through morphological evolution from a seamed to a self-adaptive fit together. The experiment was repeated independently three times with similar results. Scale bar was 6 mm. c Cross-sectional view of the AAAHF demonstrates fits tightly between the heat-shrink tubing’s inner wall and the aramid aerogel hollow fiber’s outer surface. The experiment was repeated independently three times with similar results. Scale bar was 500 μm. d High-magnification imaging reveals intimate bonding between the aerogel’s mesoporous architecture and the heat-shrink tubing’s dense shell. The experiment was repeated independently three times, and a representative image is shown. Scale bar was 5 μm. e Nitrogen adsorption-desorption curve of the inner aerogel wall of AAAHF (the inset is the pore size distribution diagram). f Tensile stress-strain curves of aramid aerogel hollow fibers, heat-treated aramid aerogel hollow fibers, heat-shrink tubing (PE stands for polyethylene), and AAAHF. g Summary of the adaptation between various aerogel fibers and various heat-shrink tubing (aerogel fibers including aramid aerogel fibers (AAF), cellulose aerogel fibers (CAF), sodium alginate aerogel fibers (SAAF), graphene aerogel fibers (GAF), MXene aerogel fibers (MAF), silica aerogel fibers (SAF), graphene aerogel hollow fibers (GAHF), cellulose aerogel hollow fibers (CAHF), and aramid@MXene aerogel coaxial fibers (AMACF). Alternative armor materials including ethylene-vinyl acetate copolymer (EVA), polyvinyl chloride (PVC), polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), fluorinated ethylene propylene (FEP), silicone rubber (SR), fluororubber (FR), and ethylene propylene diene rubber (EPDM). The five-pointed star and cross symbols denote compatibility and incompatibility, respectively, between the aerogel fiber and heat-shrink tubing.

To systematically validate the versatility of our methodology, diverse aerogel fibers were effectively armored through PE heat-shrink tubing processing, yielding armored aerogel fibers including armored aramid aerogel fibers, armored cellulose aerogel fibers, armored sodium alginate aerogel fibers, armored graphene aerogel fibers, armored MXene aerogel fibers, armored silica aerogel fibers, armored graphene aerogel hollow fibers, armored cellulose aerogel hollow fibers, and armored aramid@MXene aerogel coaxial fibers (Fig. S8). All systems maintained intact nanoporous cores with tightly bonded dense shells. Alternative armor materials, including ethylene-vinyl acetate copolymer, polyvinyl chloride, polytetrafluoroethylene, polyvinylidene fluoride, fluorinated ethylene propylene, silicone rubber, fluororubber, and ethylene propylene diene rubber, demonstrated compatibility contingent upon thermal stability matching between aerogel matrices and polymer shrinkage temperatures (Fig. 2g). Observed incompatibilities primarily occurred when aerogel decomposition temperatures overlapped with polymer shrinkage activation ranges.

Endowed versatility of AAAHF

The AAAHF exhibits significant improvements in mechanical properties, while the dense polymer shell endows the aerogel fiber with exceptional characteristics, including water and oil resistance, washability, multi-color capability, and flame retardancy. Notably, tens of meters of AAAHF can be produced via this simple, solvent-free, and non-destructive armoring process (Fig. 3a). And AAAHF demonstrates flame-retardant properties. heat-treated aramid aerogel hollow fiber is readily carbonized by the outer flame of an alcohol lamp (Fig. 3b), while AAAHF withstands exposure to the same flame for 1–2 s and self-extinguishes upon flame removal (Fig. 3c and Movie S2). This marked enhancement in flame resistance is primarily attributed to the flame-retardant additives within the commercial heat-shrink tubing armor. Furthermore, multi-color functionality is easily achieved by armoring AAAHF with PE heat-shrink tubing with varying colors (Fig. 3d). By applying different-colored PE heat-shrink tubing to a single continuous aerogel fiber, multicolored segments can be fabricated on a single AAAHF strand (Fig. S9), offering a dramatically simplified alternative to conventional complex fiber-dyeing processes. Owing to its high flexibility, AAAHF can be woven into intricate shapes such as traditional Chinese knots (Fig. 3e). Combined with its high strength, AAAHF is compatible with wearable devices through hybrid weaving techniques (Fig. 3f). Unarmored aerogel fibers respectively exhibit super-amphiphicity, which are rapidly wetted by water and oil (Fig. S10) accordingly, while armored aerogel fibers show water-/oil-repellency. Unlike unarmored aerogel fibers, which is susceptible to contamination by water or organic solvents (Fig. S11), AAAHF simultaneously resists both aqueous and oily contaminants (Fig. 3g). Immersion tests further confirm AAAHF’s superior water and oil resistance: after 5 h in water or organic solvent (K101), its water absorption decreases from 1011% (before being armored) to 0.07% (5 orders of magnitude reduction), and oil absorption drops from 2004% (before being armored) to 1.2% (1670-fold improvement) (Fig. 3h). The robust mechanical properties and stable hydrophobicity of AAAHF also ensure excellent wash durability (Fig. S12). To further elucidate the fluid transport properties of our designed fibers, we evaluated their capillary-driven anti-gravity water transport behavior. This test primarily aimed to demonstrate the sealing effect of the armor layer and the significant change in wettability induced by lubricant injection. As shown in Fig. 3i, AAAHF was able to continuously supply water for over 10 days, reaching a final height of 990 mm (Fig. S13a), due to the effective suppression of evaporation by its dense armor layer. In contrast, the unarmored aerogel fibers reached equilibrium within 1 h due to the inability to suppress evaporation. Crucially, after injecting a hydrophobic PFPE lubricant into the aerogel core layer of AAAHF to form AAAHF@Oil (Fig. 4a), the wetting property of the inner surface of AAAHF changed from superhydrophilic (Fig. S10) to hydrophobic (Fig. 4b). Consequently, its anti-gravity water transport height was drastically reduced to only about 20 mm (Fig. S13b). This decisive result confirms that the lubricating layer uniformly and stably alters the chemical properties of the pore surface, which is essential for forming the fluid slip interface responsible for drag reduction. Therefore, the anti-gravity transport study clearly highlights the function of each component in our design: the armor layer provides a long-term seal, while the injection of lubricant determines the interfacial transport characteristics.

Fig. 3. Flexibility, weavability, and multifunctional performance of armored aramid aerogel hollow fibers.

a AAAHF wound on a wire roll, indicating that aerogel fibers can be continuously armored up to 12 m. Scale bar was 2 cm. b, c Ignition comparison tests between unarmored aerogel fibers and AAAHF, indicating the good flame retardancy of AAAHF. d AAAHF can be armored with heat shrink tubing of different colors, indicating its multi-color nature. Scale bar was 1 cm. e Chinese knot woven with AAAHF, indicating its flexibility with a radius of curvature of 1 mm. Scale bar was 1 cm. f AAAHF combined with wool textiles through knitting, indicating the wearability of AAAHF. Scale bar was 1 cm. g Water stains or oil stains on the surface of AAAHF can be easily removed by simple rinsing, indicating its anti-pollution property. Scale bar was 1 cm. h Comparison of oil and water absorption rates between unarmored aerogel fibers and AAAHF immersed in water or oil (PFPE oil), indicating the excellent water and oil repellency of the AAAHF surface (the ends of unarmored aerogel fibers and AAAHF were knotted and sealed during immersion). Values in (h) represent their means ± SD (standard deviations) from n = 3 independent samples. Source data are provided as a Source data file. i Anti-gravity water transport height comparison between unarmored aerogel fibers and AAAHF, indicating that AAAHF has good sealing properties against water vapor.

Fig. 4. Drag reduction performance of microchannel using armored aramid aerogel hollow fibers.

a Schematic diagram on the preparation of a microchannel with drag reduction effect based on AAAHF, where perfluoropolyether oil is injected into the aerogel mesoporous wall of the inner layer of AAAHF. b Water contact angles on the inner surfaces of AAAHF@K101, AAAHF@GPL103, and AAAHF@K105. Values in (b) represent their means ± SD from n = 3 independent samples. Source data are provided as a Source data file. c The interfacial tension of different types of PFPE oils (brand name: K101, GPL103, K105) with water. Values in (c) represent their means ± SD from n = 3 independent samples. Source data are provided as a Source data file. d Schematic diagram on the pressure-drop test device for microchannel pipelines and the schematic diagram on the drag reduction mechanism of the mesoporous wall fluid injected with perfluoropolyether oil in AAAHF. e Comparison of pressure drop between PE heat-shrink tubing, AAAHF, and AAAHF@K101 for transporting water. f Apparent Darcy friction factor comparison between PE heat-shrink tubing, AAAHF, and AAAHF@K101. g Pressure drop comparison between PE heat-shrink tubing and AAAHF@K101 transporting wastewater, seawater, and hydrogen peroxide. Values in (e–g) represent their means ± SD from n = 3 independent experiments. In each experiment, 70 data points were collected over 70 s (raw data in Fig. S17), and the average was used as a single measurement. Bar heights represent the mean of the three experiment means; error bars indicate SD. Source data are provided in the Source data file. h Drag reduction rate comparison among AAAHF@K101 and other methods (hybrid Euler-Lagrange⁴², machine learning⁴³, pipe shape modification^44,45, and surface configuration^46–48 of pipe inner) in the literature.

Drag reduction properties of AAAHF@K101

Given the critical role of the lubricant in determining interfacial transport properties, we sought to optimize the lubricant for the microreactor application. Inspired by the Nepenthes pitcher plant³⁶, which features a slippery liquid-infused porous surface to create uniform, stable, and smooth interfaces for friction reduction, we implemented a strategy where lubricating liquids infused and anchor within the nanoporous structure of aerogel rather than a hollow one (Fig. S14). This drag reduction (DR) mechanism fundamentally resembles that of superhydrophobic surfaces, utilizing a liquid lubrication layer to minimize interfacial friction between solid surfaces and transported fluids³⁷, aiming to optimize its performance as a microfluidic pipeline in wearable microreactors. PFPE was selected as the lubricant due to its non-volatility, chemical stability, immiscibility with most working fluids, and low surface energy³⁸. Through systematic interfacial characterization of three PFPE variants (K101, GPL103, K105), K101 was identified as the optimal lubricant, based on its hydrophobicity (contact angles of water on the inner surface of AAAHF@K101: 121.7°) (Fig. 4b) and the highest interfacial tension with water (57.2 mN/m) (Fig. 4c). These properties synergistically contribute to drag reduction through two mechanisms: (i) enhanced hydrophobic slip at the interface^39,40 and (ii) improved stability during fluid transport cycles. To quantitatively demonstrate the reduced transport resistance afforded by the liquid wall of the AAAHF@K101 compared to conventional solid walls, we measured the pressure drop of laminar water flow using a custom-built apparatus (Figs. 4d and S15). The experiment compared the AAAHF@K101 against original PE heat-shrink tubing, with both substrates featuring comparable inner diameters (620 vs. 630 μm, respectively, Fig. S16). This minimal dimensional variance ensures that the observed drag reduction can be unequivocally attributed to the slip effect at the liquid-liquid interface, rather than to geometric differences. As shown in Fig. 4e, at a water flow rate of 350 μL/min, the pressure drop of the PE heat-shrink tubing is 285 Pa, while those of the AAAHF and AAAHF@K101 are 173 and 129 Pa, respectively. To quantify the flow characteristics of the microchannels, we employed an apparent Darcy friction factor (f ’) based on the Darcy-Weisbach equation⁴¹ (1):

$$f’=\frac{2D\mathrm{\Delta }P}{L{\rho V}^2}$$ 1

Where f ’ is the apparent Darcy friction factor, ΔP is the pressure drop, L is the pipe length, ρ is the density, V is the flow velocity, and D is the diameter of the pipe. It is important to note that the tested microchannel is not a classical circular pipe but a composite structure consisting of a nanoporous aerogel wall and a lubricant-infused interface. The f ’ defined here is therefore a lumped parameter that collectively represents the overall flow resistance of the complex channel system. According to the above formula, the apparent Darcy friction factors of PE tube, AAAHF and AAAHF@K101 are 10.90, 6.09 and 4.54, respectively (Fig. 4f). And the accurate drag reduction can be further estimated by defining the DR parameter in terms of the apparent Darcy friction factor⁴²,

$${DR}_1=\frac{{f’}_{PE}-{f’}_{AAAHF\mathrm{@}K101}}{f_{PE}}\cdot 100\%$$ 2

$${DR}_2=\frac{{f’}_{AAAHF}-{f’}_{AAAHF\mathrm{@}K101}}{{f’}_{AAAHF}}\cdot 100\%$$ 3

Where DR₁ is the drag reduction rate compared to PE pipe, DR₂ is the drag reduction rate compared to AAAHF, f ’_PE is the apparent Darcy friction factor of PE pipe, f ’_AAAHF@K101 is the apparent Darcy friction factor of AAAHF@K101, and f ’_AAAHF is the apparent Darcy friction factor of AAAHF. Compared with bare PE tubing, the DR₁ of AAAHF@K101 can be up to 58.2% when transporting water. When transporting other liquids such as wastewater, seawater, and even hydrogen peroxide, AAAHF@K101 still demonstrated superior transport efficiency, with DR₁ values ranging from 47.3 to 68.1%. These results underscore the pronounced slip advantage conferred by the lubricant-infused aerogel inner wall. Even when compared with uninfused AAAHF, AAAHF@K101 maintained a clear drag reduction benefit, achieving DR₂ values between 4.4 and 31.0%. This additional improvement highlights the effective lubricating role of the immiscible, low-surface-energy PFPE oil (Figs. 4g, S17 and 18; Table S3). According to the Darcy-Weisbach equation⁴¹, the pressure drop for flow in a pipe is directly proportional to the fluid’s density and viscosity. Therefore, under identical flow conditions, the higher density (~1.11 g/cm³) and viscosity (~1.1 mPa s) of 30% H₂O₂ intrinsically lead to a greater pressure drop than that of water. Notably, AAAHF@K101 achieved a superior drag reduction efficiency in fluid transport, surpassing conventional approaches^42–48 such as hybrid Euler-Lagrange⁴² (26%) pulsating pipe flow⁴³ (38.6%), shape-modified pipes^44,45 (4–25%), and surface-configured pipe^46–48 (23%) (Fig. 4h and Table S4). The performance superiority originates from the liquid wall of AAAHF@K101, in which the aerogel mesopores confine the K101 to create a smooth lubricating layer. The slip at this liquid-liquid interface during transport effectively minimizes viscous dissipation. To quantify energy efficiency, we calculated the pumping power per unit length (P_unit = ΔP × V/L), a widely used metric for microfluidic transport⁴⁹. At 350 μL/min, the P_unit values for PE tubing, AAAHF, and AAAHF@K101 are 16.6, 10.1, and 7.5 μW/m, respectively (Fig. 4e). Compared with uninfused AAAHF, AAAHF@K101 achieves a ~25% reduction in pumping power, confirming that the nanoconfined lubricant layer effectively lowers flow resistance. Notably, its energy consumption is ~55% lower than that of smooth PE tubing, highlighting the superior energy-saving potential of the liquid-infused wall design—a critical advantage for power-limited wearable applications.

While lubricant leaching is a recognized challenge for conventional liquid-infused surfaces^50,51, the nanoconfined architecture of our system effectively enhances lubricant retention. In contrast to simply coating micro-scale roughness⁵¹ (Fig. S19a) or porous surfaces⁵⁰ (Fig. S19b), this study achieves lubricant infusion into a three-dimensional nanoporous network of interconnected aramid nanofibers (Fig. S19c). This nanoconfinement promotes stronger capillary forces and greater solid-liquid contact area, thereby intrinsically improving lubricant lock-in capability. To quantitatively evaluate the leaching resistance, we continuously pumped water through the AAAHF@K101 for 21 h and measured the pressure drop (at a flow rate of 300 μL/min) at 0, 12, and 21 h. The overall pressure drop remained stable throughout the test (Fig. S20a). The associated mass loss plateaued after 18 h, with a total loss of less than 1.5% (total mass loss ~2.0 mg, Fig. S20b). This suggests that the initial minimal mass loss originates from the removal of excess lubricant adhering to the rough outer surface, while the lubricant confined within the nanopores of aerogel exhibits strong retention. Furthermore, we continuously monitored the pressure drop of the AAAHF@K101 fiber in an aqueous fluid across varying flow rates (300, 500, and 700 μL/min). The ΔP remained generally stable during the 30-min testing period (Fig. S20c), with no sharp increase that would indicate lubricant depletion. Gravimetric analysis confirmed a negligible mass loss of less than 1.5% (Fig. S20d). These combined results demonstrate that the nanoconfinement effect provided by the aerogel matrix ensures exceptional long-term stability of the lubricant-infused interface. This robustness establishes a solid foundation for utilizing AAAHF@K101 as a reliable and efficient fluid transport channel in wearable microreactors. To assess stability under static conditions, we immersed AAAHF@K101 samples in an aqueous medium for 30 days. The total mass loss was less than 1.7% (total mass loss ~1.9 mg). And after being immersed in water for 30 days, the unarmored aerogel fiber@K101 remained hydrophobic and thus did not get contaminated even after being immersed in the methyl orange solution subsequently (Fig. S21). Integrating the results from both dynamic and static tests, these findings compellingly demonstrate that the inherent nanoconfinement effect of the aerogel matrix ensures exceptional long-term stability of the lubricant layer. This durability is a key foundation for the application of AAAHF@K101 as a reliable and efficient fluid transport channel in wearable microreactors.

Design of O₂-generated wearable microreactors (OWM)

The armored aerogel fiber design successfully retained the intrinsic nanoporous structural characteristics of pristine aerogels while achieving mechanical robustness compatible with wearable devices (tensile strength: 17.6 MPa, elongation at break: 770%). Without the armoring of heat-shrink tubing, unarmored aerogel fibers exhibited fragility, susceptibility to liquid infiltration (water/oil contact angle = 0°), and leakage risks under dynamic deformation, highlighting the critical role of the armor layer in addressing practical challenges. Leveraging the confined lubricant drag-reduction effect enabled by the nanoporous aerogel structure within AAAHF inner walls, we achieved a significant enhancement in microchannel fluid transport efficiency (DR up to 68.1% improvement). To systematically validate its potential for wearable microreactors, we successfully developed an O₂-generated wearable microreactor prototype utilizing the microchannel properties of AAAHF@K101. Notably, efficient oxygen supply technology holds critical strategic importance across medical emergencies, high-altitude survival support, confined-space operations, and oxygen-enriched engineering systems⁵². The World Health Organization has included medical oxygen in its Essential Medicines List since 2017, emphasizing its foundational role. Current mainstream oxygen technologies face notable limitations: (1) bulky high-pressure tanks with poor portability; (2) pressure swing adsorption (PSA) systems constrained by low oxygen concentration (up to 93%) and excessive weight⁵³; (3) water electrolysis requiring high energy consumption while generating hazardous hydrogen byproducts⁵⁴; (4) solid chemical methods posing fire/explosion risks from exothermic reactions⁵⁵. Therefore, developing OWM integrating lightweight design, cost-effectiveness, and operational safety presents crucial research value and engineering prospects for enhancing emergency response capabilities and safeguarding personnel in extreme environments.

The OWM assembly schematic (Fig. 5a) comprises two liquid reservoirs, two micro peristaltic pumps, 1.5 V batteries (power source), a T-junction (reactant/catalyst mixing), 1.5 m AAAHF@K101 (reaction zone), a gas-liquid separation bag, a gas filter (calcium hydroxide particles for moisture/acid gas removal), and a nasal cannula (Fig. S22). The total system mass remains 67.8 g (Fig. S23). The integration of AAAHF@K101 with knitted textiles serves multiple purposes. Firstly, the AAAHF@K101 is woven into the fabric using a plain stitch (Fig. S24a, b), which demonstrates its mechanical compliance—a critical attribute for a functional wearable microreactor. The use of an approximately meter-long AAAHF@K101 fiber is essential for three reasons: it ensures sufficient residence time for complete H₂O₂ decomposition, demonstrates the effectiveness of our drag-reduction strategy at a practically relevant scale, and provides ample material for textile integration. Secondly, organizing such meter-long fibers within the textile matrix not only ensures structural regularity and esthetic appeal but also leverages their launderability—after detaching external components (e.g., pumps, batteries, etc.)—allowing the textile component to be washed. Other components are securely affixed via stitching, maintaining a cohesive garment esthetic (Fig. 5b) and ensuring wearing comfort (Fig. S24c). Furthermore, this integration strategy is versatile; as an alternative to direct weaving, the system can be housed in a backpack or integrated into footwear, highlighting its adaptability for diverse wearable applications. Dual 100 mL reservoirs (30 wt% H₂O₂ and 0.3 mol/L FeCl₃ solutions) feed reactants into the AAAHF@K101 microchannel through controlled pumping (Fig. S25). The primary reaction governing this process is the Fe³⁺-catalyzed decomposition of H₂O₂. The mechanism involves two key steps⁵⁶:

Fig. 5. Assembly and performance of an O₂-generated wearable microreactor.

a Schematic diagram on the assembly of the O₂-generated wearable microreactor. The main accessories include knitted fabric, a raw material bag (hydrogen peroxide as oxygen source, ferric chloride solution as catalyst), a peristaltic pump, a three-way valve, AAAHF@K101, a gas-liquid separation device, a gas purification device, and a nasal aspirator. b Photo of the O₂-generated wearable microreactor worn on a person. From the outside, the clothes look the same as ordinary knitted clothes. The white dotted boxes circle AAAHF@K101 and the oxygen-liquid separation device, respectively. AAAHF@K101 is combined with ready-made clothes by weaving, and the oxygen-liquid separation device is fixed to the clothes by bundling. c Infrared photo of the O₂-generated wearable microreactor worn on a person. d The effect of different raw material feed rates on oxygen generation rate and oxygen concentration. e The effect of different raw material feed rates on liquid temperature in the oxygen-liquid separation device. Values in (e) represent their means ± SD from n = 3 independent samples. Source data are provided as a Source data file. f Four types of pipelines (PE pipe, PE@K101, AAAHF, and AAAHF@K101) were selected as microreactor channels for OWM reaction to compare their pressure monitoring curves over a 25-min period. The results demonstrate that while the internal pressure of AAAHF@K101 remained consistently stable, the other three pipelines exhibited significant pressure fluctuations with sustained high-pressure levels. g Stability test of the O₂-generated wearable microreactor. The O₂-generated wearable microreactor was operated continuously for 100 h, and the oxygen generation rate and oxygen generation amount were monitored in real time. h Comparison of blood oxygen saturation in subjects wearing OWM when they were sitting still with and without oxygen inhalation. i Comparison of blood oxygen saturation in subjects wearing OWM at different altitudes when they inhaled oxygen for 200 ± 3 s and when they did not inhale oxygen.

1. Reduction of Fe³⁺: ${\mathrm{H}}_2{\mathrm{O}}_2\left(\mathrm{aq}\right)+2{\mathrm{FeCl}}_3\left(\mathrm{aq}\right)\to 2{\mathrm{FeCl}}_2\left(\mathrm{aq}\right)+2\mathrm{HCl}\left(\mathrm{aq}\right)+{\mathrm{O}}_2\left(\mathrm{g}\right)\uparrow$

2. Re-oxidation of Fe²⁺: ${\mathrm{H}}_2{\mathrm{O}}_2\left(\mathrm{aq}\right)+2{\mathrm{FeCl}}_2\left(\mathrm{aq}\right)+2\mathrm{HCl}\left(\mathrm{aq}\right)\to 2{\mathrm{FeCl}}_3\left(\mathrm{aq}\right)+2{\mathrm{H}}_2\mathrm{O}\left(\mathrm{l}\right)$

Net reaction: ${2\mathrm{H}}_2{\mathrm{O}}_2\left(\mathrm{aq}\right)\to 2{\mathrm{H}}_2\mathrm{O}\left(\mathrm{l}\right)+{\mathrm{O}}_2\left(\mathrm{g}\right)\uparrow$

In this cycle, FeCl₃ acts as a catalyst. Under the strongly acidic conditions maintained, potential side reactions such as Fe³⁺ hydrolysis are suppressed. Carrying 100 g of 30 wt% H₂O₂ generates 10 L pure oxygen at normal pressure, equivalent to commercial portable tanks (1 L, 1.0 MPa). Crucially, while conventional tanks are single-use, our system enables continuous oxygen supply through reagent replacement. The core component of the OWM, designated AAAHF@K101, acted as a microreactor, achieving an oxygen generation STY of 26498 g L⁻¹ h⁻¹. This value surpassed those of conventional stirred batch reactors and continuous flow systems by three orders of magnitude⁵⁷, a result attributed to the intensified mass transfer enabled by the unique design of AAAHF@K101 microreactor, as quantified by the STY formula below⁵⁸:

$$STY=\frac{M_P}{V_e\cdot T}$$ 4

where M_p is the mass of product (g), V_e is the effective volume of the reactor (L), T is the reaction time (h). The length of the wearable microreactor is 1.5 m, and its diameter is 620 μm (Fig. S12). The production rate of oxygen is 140 mL/min, as shown in Fig. S21 (The injection rate of the reactants is 1.1 mL/min).

Infrared thermography confirms safe operation: AAAHF@K101 surface temperature (~25 °C) and gas-liquid interface temperature (37 °C) match body conditions (Figs. 5c and S26). Systematic analysis reveals: (1) oxygen generation rate increases proportionally with injection rate (controlled via pump regulation, Fig. 5d); (2) AAAHF@K101’s drag-reducing effect ensures stable flow uniformity (Fig. S27); (3) the oxygen concentration can reach 100% in stable operation. (Figs. S28 and 5d). Parametric studies demonstrate that temperature elevation at the gas-liquid separator (Te) correlates with injection rate (enhanced reaction kinetics, Fig. 5e), while remaining unaffected by controlling catalyst concentration or AAAHF@K101 length (Fig. S29). To demonstrate the unique advantages of AAAHF@K101 as an irreplaceable OWM microreactor, four pipelines (PE, PE@K101, AAAHF, and AAAHF@K101) were engineered into microfluidic reaction channels for comparative analysis of real-time pressure profiles during 25-min continuous-flow. Real-time pressure monitoring revealed that the AAAHF@K101 maintained exceptional pressure stability with ultralow fluctuations (σ = 159 Pa, 0.11 MPa peak)—representing 10.1-fold, 30.8-fold, and 113.8-fold reductions in standard deviation compared to PE@K101 (σ = 1606 Pa, 0.12 MPa peak), PE (σ = 4898 Pa, 0.13 MPa peak), and pristine AAAHF (σ = 18,085 Pa, 0.22 MPa peak), respectively (Fig. 5f). The liquid-infused surface establishes a liquid-liquid interface, modifying the conventional no-slip boundary condition as described by the Navier slip model³⁷:

$$\mu =\lambda {\frac{d\mu }{dy}}_{y=0}$$ 5

where u is the slip velocity, λ is the slip length, and du/dy is the velocity gradient. The liquid-liquid interface (AAAHF@K101) exhibits a slip length of λ ≈ 10 ~ 100 μm, whereas solid interfaces (PE and AAAHF) show negligible slip (λ ≈ 0 μm)³⁷. This boundary slip stabilizes fluid flow by suppressing perturbation growth⁵⁹, enabling AAAHF@K101 to maintain stable flow patterns with lower pressure fluctuation amplitude (σ = 159 Pa). Conversely, PE@K101 lacks mesoporous confinement, leading to rapid lubricant depletion and consequent flow destabilization. The long-term operational stability of the single-cell OWM was further evaluated under an injection rate of 800 mL/min. As demonstrated in Fig. 5g, the oxygen generation rate remained unchanged during total operation for 100 h, while the cumulative oxygen quantity progressively increased. Comparative analysis of the AAAHF@K101 inner walls using 3D confocal microscopy before and after the 100-h test revealed no discernible surface changes, with surface roughness increasing marginally from 0.8 μm (Fig. S30a) to 1.6 μm (Fig. S30b)—a negligible variation that confirms the structural stability of the OWM. These findings collectively underscore the robust and stable oxygen production capability of the OWM system. The integrated OWM established benchmarks for oxygen generation technologies, achieving a high mass-specific oxygen evolution rate (OER) of 519 mL min⁻¹ kg⁻¹ at 0.27 kg total mass—representing 9.8-fold, 8.1-fold, and 1.6-fold improvements over portable oxygen generators (53 mL min⁻¹ kg⁻¹ @ 0.11 kg)⁶⁰, PSA oxygen generators (64 mL min⁻¹ kg⁻¹ @ 15.6 kg)⁶¹, and electrochemical oxygen generators (319 mL min⁻¹ kg⁻¹ @ 4.7 kg)⁶² respectively (Fig. S31, Table S5). This performance matrix (oxygen purity: 100%; weight: 0.27 kg; OER: 519 mL min⁻¹ kg⁻¹; wearability; rechargeability) redefines wearable oxygen generation technology, outperforming commercial oxygen generation limited by bulkiness (>15.6 kg), low purity (93%), and static operation modes (Fig. S32).

The field validation demonstrated OWM’s unimpeded functionality during mountaineering and running activities (Movie S3), positioning it as a significant approach in next-generation oxygen delivery systems. To quantitatively assess its hypoxia mitigation capability, we conducted systematic trials in high-altitude (>3000 m above Sea Level) environments using nasal cannula-mediated oxygen delivery. Initial static condition testing at 3650 m elevation revealed rapid blood oxygen saturation (SpO₂) normalization: Within 60 ± 3 s of oxygen supplementation, subjects’ blood oxygen saturation elevated from baseline 84% to 96%, maintaining ≥95% with continuous administration (Fig. 5h). Subsequent ambulatory testing under identical altitude conditions showed comparable efficacy, with SpO₂ increasing from 86% to 94.6% after 150 ± 3 s of oxygen inhalation (Fig. S33a). Repeatability was confirmed through cyclic oxygen administration protocols (Fig. S33b), demonstrating OWM’s prompt response to acute hypoxic episodes. Multivariate altitude analysis further validated system performance across 3650 m, 4280 m, and 4998 m elevations. Notably, 200 ± 3 s oxygen supplementation consistently restored SpO₂ to physiologically normal ranges regardless of elevation gradients (Fig. 5i). These findings collectively establish OWM’s capacity for immediate hypoxia correction and sustained oxygen maintenance, even under extreme altitude conditions. The results substantiate two critical advantages: (1) rapid-response oxygen replenishment within 60–200 s across activity states, and (2) altitude-independent stabilization of blood oxygenation levels. This performance profile suggests OWM’s potential as a wearable solution for altitude-related hypoxemia in both stationary and ambulatory scenarios. This successful implementation of AAAHF@K101-based OWM not only validates aerogel fibers’ potential in wearable microreactors but also provides strategic insights for developing advanced wearable life-support systems.

Discussion

This work demonstrates feasibility and realizes fabrication of wearable chemical microreactors by using lab-made armored aerogel architecture as the key component. First of all, commercial heat-shrink tubing AAAHFs were successfully fabricated, resolving the fundamental incompatibility between mechanical robustness (tensile strength 17.6 MPa) and ultrahigh porosity while conferring textile-compliant elasticity (770% elongation), flame retardancy, and environmental resistance. Subsequent confinement of perfluoropolyether oil within the mesoporous aerogel wall, exhibiting exceptional microchannel drag reduction (up to 68.1%) through nanoconfined lubricant-mediated fluid-slip interfaces. Leveraging these advances, a wearable microreactor for chemical oxygen generation was constructed. This system integrates a meter-long microchannel with aerogel-confined liquid as the inner wall, which serves as the core reaction zone, with commercially available medical oxygenation components. The resulting device demonstrated particularly performance, achieving 100% oxygen purity, total running time exceeding 100 h, a high STY of 26498 g L⁻¹ h⁻¹, and robust operational stability. The efficacy of this system in rapidly alleviating acute hypoxia was definitively confirmed through in vivo trials at 4998-meter altitude, where blood oxygen saturation recovery to 95% within 200 ± 3 s was observed. This study also establishes a foundational materials paradigm that enables the realization of wearable chemical microreactors, providing a versatile strategy for the integration of high-performance aerogel fibers into advanced wearable technologies with significant implications for emergency medicine and portable chemical synthesis systems.

Methods

Materials

Kevlar fiber was purchased from DuPont. Dimethyl sulfoxide (DMSO), potassium tert-butoxide, methanol, sodium alginate, lithium fluoride, iron(III) chloride hexahydrate (FeCl₃·6H₂O, AR), calcium hydroxide (Ca(OH)₂, AR), and tetraethyl orthosilicate (TEOS, AR) were purchased from Shanghai Aladdin Biochemical Technology Co. Ltd. Ethanol, tert-butanol, hydrochloric acid (AR, 36.0–38.0%), hydroiodic acid (AR, ≥45.0%), hydrogen peroxide solution (AR, 30 wt. % in H₂O), ammonia aqueous solution (NH₃·H₂O, 25–28 wt %, AR), and anhydrous calcium chloride (CaCl₂) were purchased from Sinopharm Chemical Reagent Co. Ltd. Titanium aluminum carbide (Ti₃AlC₂) was provided by Jilin Yiyi Technology Co., Ltd. Graphene oxide was provided by Hangzhou Gaoxi Technology Co. Ltd. Cotton balls were kindly provided by Winner Medical Supplies Co., Ltd. 1-Allyl-3-methylimidazolium chloride was purchased from Energy Chemical Co. Deionized water was obtained from a Millipore-Q system. All other reagents were used as received without further purification.

Fabrication of heat-treated aramid aerogel hollow fiber

First, 50.0 g of potassium tert-butoxide was added to 850 g of DMSO and stirred until it was completely dissolved. Then, 50.0 g Kevlar fibers and 50.0 g of anhydrous methanol were added and stirred until a stable and uniform dispersion was formed to obtain a 5.0 wt.% Kevlar dispersion as a spinning solution. The aramid aerogel hollow fibers were prepared by coaxial wet spinning and freeze-drying. The specifications of the coaxial needle were 21 G + 16 G (inner needle 21 G, inner diameter, 0.5 mm; outer needle 16 G, inner diameter, 1.15 mm). 5.0 wt.% Kevlar dispersion and deionized water were injected into the outer and inner needles of the coaxial needles through separate injection pumps, with injection speeds of 400 μL min⁻¹ (inner needle)/400 μL min⁻¹ (outer needle). No draft ratio has been added. Subsequently, the extruded Kevlar dispersion underwent sol-gel transition in a deionized water coagulation bath to form aramid hydrogel hollow fibers, which were replaced three times in tert-butyl alcohol/water (1:4, m/m) solvent and freeze-dried for 12 h to obtain aramid aerogel hollow fibers. Finally, aramid aerogel hollow fibers were heat-treated at 400 °C in an argon atmosphere for 30 min to obtain heat-treated aramid aerogel hollow fibers.

Fabrication of armored aramid aerogel hollow fiber

Heat-treated aramid aerogel hollow fibers were placed into polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the heat-treated aramid aerogel hollow fibers to obtain AAAHF. The armor length of the AAAHF can reach several meters. The resulting armor layer exhibited a thickness of 180–250 μm, and the heat-treated aramid aerogel hollow fiber-to-tubing mass ratio between 0.0362 and 0.0820 (Table S1). Notably, the tight adhesion between the polyethylene armor and the heat-treated aramid aerogel hollow fiber was achieved through precise coordination of three key parameters: the initial inner diameter of the heat-shrink tubing, its shrinkage ratio, and the outer diameter of the heat-treated aramid aerogel hollow fiber. Specifically, the initial inner diameter of the tubing must be larger than the outer diameter of the heat-treated aramid aerogel hollow fiber to facilitate smooth insertion during assembly, while the fully shrunk inner diameter must be smaller than that of the heat-treated aramid aerogel hollow fiber to ensure a constant compressive force and thus intimate interfacial contact. For instance, with a heat-treated aramid aerogel hollow fiber outer diameter of 0.95 mm, we selected polyethylene heat-shrink tubing with a 2:1 shrinkage ratio, an initial inner diameter of 1.5 mm (allowing easy insertion), and a fully shrunk inner diameter of ≤0.6 mm (smaller than the heat-treated aramid aerogel hollow fiber diameter). Upon heating, the tubing radially shrinks, resulting in a robust and mechanically tight interface between the polyethylene armor and the heat-treated aramid aerogel hollow fiber surface.

Control of the AAAHF core and shell dimensions

The thickness of the nanoporous aerogel core (i.e., the hollow fiber wall) and the outer PE armor layer were independently controlled as follows. Aerogel Core Thickness: The wall thickness of the aramid aerogel hollow fiber precursor was tuned within 150–350 µm by varying the coaxial wet-spinning parameters. Key variables included the gauge combination of the coaxial needles and the extrusion rates of the inner (coagulation bath) and outer (aramid nanofiber dispersion) solutions. Representative parameter sets, and the resulting thicknesses are provided in Table S1 (Supplementary Information). PE Armor Shell Thickness: The thickness of the final PE armor layer was controlled between 180 and 250 µm by selecting polyethylene heat-shrink tubing with different initial inner diameters (e.g., 1.0, 1.5, 2.0 mm) and a fixed shrinkage ratio (2:1). The tubing was shrunk uniformly onto the heat-treated aramid aerogel hollow fiber by heating at 100 °C for 15 s.

Fabrication of AAAHF with drag reduction

Perfluoropolyether oil (K101, a perfluoroalkyl substance) was injected into AAAHF through an injection pump at a speed of 50 μL/min, and both ends were sealed and left to stand for 6 h to allow the perfluoropolyether oil to fully infiltrate and infused into the aerogel skeleton of AAAHF to form an aerogel-confined solid-liquid composite hollow fiber, named AAAHF@K101. It is important to note that while Perfluoropolyether oil offers exceptional chemical inertness and stability required for this proof-of-concept study, its environmental persistence and bioaccumulation potential raise serious concerns. We fully acknowledge this critical limitation and are actively pursuing more sustainable and environmentally benign alternatives for future applications.

Design of O₂-generated wearable microreactor

30% H₂O₂ was used as the oxygen source, and a 0.3 mol/L FeCl₃ solution was used as the catalyst. AAAHF@K101 was mixed with knitted fabrics to obtain the residence area of the OWM, that is, the mixing and reaction area. 100 g of FeCl₃ solution and H₂O₂ solution were taken and placed in plastic bags and connected to a micro peristaltic pump (powered by a 3 V battery). The other end of the micro peristaltic pump was connected to a Y-type 3-way interface; the outlet of the Y-type 3-way was connected to AAAHF@K101. The other end of AAAHF@K101 was connected to a self-made gas-liquid separation device, and the outlet of the gas-liquid separation device was connected to a gas filter (with built-in calcium hydroxide particles as an absorbent), and the outlet of the filter was connected to a nasal inhaler for oxygen supply. All accessories except AAAHF@K101 were fixed inside the knitted clothing using a needle and thread.

Fabrication of armored aramid aerogel fiber

Kevlar dispersion (5 wt%) was used as the spinning solution. The spinning solution was injected into a rotating crystallization disk containing a coagulation bath (i.e., water) through an injection needle (needle diameter 21 G, needle diameter 500 μm; injection speed, 200 μL/min). The obtained aramid hydrogel fiber was solvent-displaced at least four times with a water/tert-butyl alcohol solution (mass ratio 4:1) and then freeze-dried (−55 °C, 5 Pa, 12 h) to obtain aramid aerogel fibers. The aramid aerogel fibers were sheathed in a PE heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the aramid aerogel fibers to obtain the armored aramid aerogel fibers.

Fabrication of armored cellulose aerogel fiber

A 4 wt% cellulose solution was prepared by dissolving 4 g of cotton balls in 96 g of 1-Allyl-3-methylimidazolium chloride. Cellulose dispersion (4 wt%) was used as the spinning solution. The spinning solution was injected into a rotating crystallization disk containing a coagulation bath (ethanol) through an injection needle (needle diameter 21 G, needle diameter 500 μm; injection speed, 200 μL/min). The obtained cellulose hydrogel fibers were replaced with ethanol solvent at least four times, and then supercritical CO₂ dried (40 °C, 10 MPa, 12 h) to obtain cellulose aerogel fibers. The cellulose aerogel fibers were sheathed in polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the cellulose aerogel fibers to obtain armored cellulose aerogel fibers.

Fabrication of armored sodium alginate aerogel fiber

A 4 wt% sodium alginate solution was prepared by dissolving 4 g of sodium alginate in 96 g of water, and this solution was used as the spinning dope. The spinning solution was injected into a rotating crystallization disk containing a coagulation bath (CaCl₂ solution) through an injection needle (needle diameter 21 G, needle diameter 500 μm; injection speed, 200 μL/min). The obtained sodium alginate hydrogel fibers were replaced with ethanol solvent at least four times, and then supercritical CO₂ dried (40 °C, 10 MPa, 12 h) to obtain sodium alginate aerogel fibers. The sodium alginate aerogel fibers were placed into polyethylene heat-shrink tubing and heated to 90 °C for approximately 30 s to shrink the heat-shrink tubing and fit the outer wall of the sodium alginate aerogel fibers to obtain armored sodium alginate aerogel fibers.

Fabrication of armored MXene aerogel fiber

2 g of Ti₃AlC₂ powder was slowly added to an etching solution containing 3.2 g of LiF and 40 mL of 9 M HCl. The mixture was kept at 50 °C for 30 h. The resulting dispersion was centrifuged at 3500 rpm for 5 min repeatedly until the supernatant reached pH ~ 6, whereupon self‑separation occurred. The dispersion was then manually shaken for 20 min and centrifuged at 3500 rpm for 20 min. The Ti₃C₂T_x MXene precipitate (large‑lamellar MXene) was collected, redispersed in water, and centrifuged at 3500 rpm for 2 min to remove the unetched MAX phase in the pellet. The supernatant was further centrifuged at 12,000 rpm for 1 h to obtain a highly concentrated MXene dispersion (140 mg/mL).

MXene dispersion (140 mg/mL) was used as the spinning solution²². The spinning solution was injected into a rotating crystallization disk containing a coagulation bath (CaCl₂ solution) through an injection needle (needle diameter 21 G, needle diameter 500 μm; injection speed, 200 μL/min). The obtained MXene hydrogel fibers were replaced with ethanol at least four times, followed by supercritical CO₂ drying (40 °C, 10 MPa, 12 h) to obtain the MXene aerogel fibers. The MXene aerogel fibers were sheathed in polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the MXene aerogel fibers to obtain armored MXene aerogel fibers.

Fabrication of armored graphene aerogel fiber

Graphene oxide dispersion (15 mg/mL) was used as the spinning solution. The spinning solution was injected into a rotating crystallization disk containing a coagulation bath (i.e., CaCl₂ solution) through an injection needle (needle diameter 21 G, needle diameter 500 μm; injection speed, 200 μL/min). The obtained graphene oxide hydrogel fibers were immersed in a hydroiodic acid solution (10 wt%, 60 °C, 4 h), and after removal, they were replaced with ethanol solvent at least eight times, and then supercritical CO₂ dried (40 °C, 10 MPa, 12 h) was performed to obtain the graphene aerogel fibers. The graphene aerogel fibers were sheathed in polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the graphene aerogel fibers to obtain armored graphene aerogel fibers.

Fabrication of armored silica aerogel fiber

A silica spinning solution was prepared by mixing TEOS, ethanol, and 0.01 M HCl at a molar ratio of 1:4:2.8, followed by refluxing at 130 °C for 12 h. The spinning solution was injected into a rotating crystallization disk containing a coagulation bath (ammonia water) through an injection needle (needle diameter 21 G, needle diameter 500 μm; injection speed, 200 μL/min). After aging, the obtained silica hydrogel fibers were replaced with ethanol solvent at least four times, and then supercritical CO₂ dried (40 °C, 10 MPa, 12 h) to obtain silica aerogel fibers. The silica aerogel fibers were sheathed in polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the silica aerogel fibers to obtain armored silica aerogel fibers.

Fabrication of armored graphene aerogel hollow fiber

Graphene oxide dispersion (15 mg/mL) was used as the spinning solution. Subsequently, graphene aerogel hollow fibers were prepared via coaxial wet spinning and supercritical drying. The specifications of the coaxial needle were 21 G + 16 G (inner needle 21 G, inner diameter ≈ 0.5 mm; outer needle 16 G, inner diameter ≈ 1.15 mm). The graphene oxide dispersion and CaCl₂ solution were injected into the outer and inner needles of the coaxial needle, respectively, through separate injection pumps at an injection speed of 400 μL min⁻¹ (inner needle)/400 μL min⁻¹ (outer needle). No draft ratio has been added. Subsequently, the extruded graphene oxide dispersion underwent a sol-gel transition in a CaCl₂ solution coagulation bath to form graphene oxide hydrogel hollow fibers. The obtained graphene oxide hydrogel hollow fibers were immersed in a hydroiodic acid solution (10 wt%, 60 °C, 4 h), and after removal, they were replaced with ethanol solvent at least eight times, and then supercritical CO₂ dried (40 °C, 10 MPa, 12 h) to obtain graphene aerogel hollow fibers. Finally, the graphene aerogel hollow fibers were sheathed in polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the graphene aerogel hollow fibers to obtain armored graphene aerogel hollow fibers.

Fabrication of armored cellulose aerogel hollow fiber

Cellulose dispersion (4 wt%) was used as the spinning solution. Subsequently, cellulose aerogel hollow fibers were prepared by coaxial wet spinning and supercritical drying. The specifications of the coaxial needle were 21 G + 16 G (inner needle 21 G, inner diameter, 0.5 mm; outer needle 16 G, inner diameter, 1.15 mm). The cellulose dispersion and ethanol were injected into the outer needle and inner needle of the coaxial needle, respectively, through separate injection pumps at an injection speed of 400 μL min⁻¹ (inner needle)/400 μL min⁻¹ (outer needle). No draft ratio has been added. Subsequently, the extruded cellulose dispersion underwent sol-gel transition in an ethanol coagulation bath to form cellulose hydrogel hollow fibers. The obtained cellulose hydrogel hollow fibers were replaced with an ethanol solvent at least four times, and then supercritical CO₂ drying (40 °C, 10 MPa, 12 h) was performed to obtain cellulose aerogel hollow fibers. The cellulose aerogel hollow fibers were sheathed in polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the cellulose aerogel hollow fibers to obtain armored cellulose aerogel hollow fibers.

Fabrication of armored aramid@MXene aerogel coaxial fiber

Aramid dispersion (5 wt%) and MXene dispersion (140 mg/mL) were used as the spinning solutions. Aramid@MXene aerogel coaxial fibers were prepared via coaxial wet-spinning and supercritical drying. The specifications of the coaxial needle were 21 G + 16 G (inner needle 21 G, inner diameter, 0.5 mm; outer needle 16 G, inner diameter, 1.15 mm). The aramid dispersion and MXene dispersion were injected into the outer and inner needle of the coaxial needle, respectively, through separate injection pumps at an injection speed of 400 μL min⁻¹ (inner needle)/400 μL min⁻¹ (outer needle). No draft ratio has been added. Subsequently, the extruded dispersion underwent sol-gel transition in a CaCl₂ solution coagulation bath to form aramid@MXene hydrogel coaxial fibers, which were replaced with ethanol solvent at least four times and then dried with supercritical CO₂ (40 °C, 10 MPa, 12 h) to obtain aramid@MXene aerogel coaxial fibers. The aramid@MXene aerogel coaxial fibers were sheathed in polyethylene heat-shrink tubing and heated to 100 °C for approximately 15 s to shrink the heat-shrink tubing and fit the outer wall of the aramid@MXene aerogel coaxial fibers to obtain armored aramid @ MXene aerogel coaxial fibers.

Characterization

Scanning electron microscopy (SEM, 400 FEG) was used to observe the microstructure of the aerogel fibers. To improve the conductivity, gold was deposited on the samples before SEM testing. The pore structures of the aramid aerogel hollow fibers were analyzed using a nitrogen adsorption desorption instrument (ASAP 2020, Micromeritics, USA) and the Barrett-Joyner-Halenda (BJH) nitrogen adsorption desorption method. The specific surface area of aerogel fibers at 77 K was determined by the BET method. The tensile stress-strain of a single fiber was tested using a universal material testing machine (INSTRON 3365). Interfacial characterization was performed under controlled conditions (25 ± 0.5 °C): (i) water contact angles were measured with 2 μL droplets deposited via automated syringe (Model OCA 15EC, DataPhysics Instruments), (ii) oil wettability was quantified using perfluoropolyether as the probe liquid (5 μL), and (iii) water-perfluoropolyether (PFPE) interfacial tension was determined via pendant drop technique (ASTM D971-12) with axisymmetric drop shape analysis (ADSA) calibration. Infrared thermal images were captured using an infrared thermal imager (Fluke TiX580) and analyzed using Smart View. The surface roughness and morphology of the samples were measured using a three-dimensional laser confocal microscope (VK-X200, Keyence).

Measurement of liquid anti-gravity transport characteristics

The AAAHF sample was vertically fixed on the wall, and the length of the sample under the water surface was 1 cm. The height and time of the anti-gravity rise of the pure water in the sample were recorded. There were 3 groups of samples were used for each test to evaluate the reliability of the results.

Measurement of AAAHF@K101 drag reduction characteristics

The drag-reduction performance of AAAHF@K101 was evaluated using self-assembly equipment based on pressure sensors. The system comprises a computerized data acquisition module and pipeline transport assembly with controlled flow parameters. Prior to testing, each fiber module (length: 10 cm) was pre-conditioned by flushing with deionized water at 500 μL/min for 30 min to remove any excess lubricant not confined within the nanopores, until no free oil droplets were observed in the effluent. The test sample (10 cm AAAHF@K101-coated pipe) was connected via a three-way valve to a syringe pump (LSP01-1BH, Baoding Lange Constant Current Pump Co., linear thrust exceeding 294 N) and the high-pressure port of the differential pressure sensor (PX2300-25BDI, Omega Engineering; low-pressure port exposed to ambient air). The liquid was transported through an injection pump at a speed of 515 μm/min, and then flowed out of the test section through the three-way valve. During the entire transport process, the pressure sensor transmits the pressure drop of the pipe segment to the digital source table (HWCL-3, KEITHLEY) and monitors the change in the pressure drop when transporting different types of liquids and pipes of different materials. The acquired pressure drop data for different fluid/pipe combinations can enable a quantitative comparison of drag reduction performance. For meaningful comparison, traditional solid-wall microchannels with nearly identical inner diameters (630 μm for conventional vs. 620 μm for the AAHHF@K101) and equivalent lengths were used.

Evaluation of lubricant leaching resistance in AAAHF@K101

Sample Preparation: AAAHF@K101 fibers were prepared following the protocol described in the main text. Prior to testing, each fiber module (length: 10 cm) was pre-conditioned by flushing with deionized water at 500 μL/min for 30 min to remove any excess lubricant not confined within the nanopores, until no free oil droplets were observed in the effluent.

Long-term stability test (21 h): Water was continuously circulated at a flow rate of 300 μL/min for 21 h. Pressure drop values were recorded at 0, 12, and 21 h. After testing, the fiber was first vigorously shaken to remove residual water from its inner channel, blotted dry with absorbent paper to eliminate surface moisture, dried in an oven at 60 °C for 1 h, and weighed using a microbalance (precision: 0.1 mg) before and after the test to determine mass loss.

Flow rate-dependent stability test: Separate fiber samples were subjected to continuous water flow at different flow rates (300, 500, and 700 μL/min) for 3 h. Pressure drop was continuously recorded throughout the testing period. Mass loss was determined gravimetrically before and after each test as described above.

Static Leaching Test under Immersion: To evaluate long-term stability under quiescent conditions, pre-conditioned AAAHF@K101 fibers and unarmored aerogel fiber@K101 were fully immersed in sealed vials containing deionized water at room temperature (25 °C) for 30 days. After immersion, the fiber was first vigorously shaken to remove residual water from its inner channel, blotted dry with absorbent paper to eliminate surface moisture, dried in an oven at 60 °C for 1 h, and weighed using a microbalance (precision: 0.1 mg) to determine mass loss.

Measurement of O₂-generated wearable microreactor characteristics

Prior to oxygen generation, the AAAHF@K101 module was thoroughly pre-flushed with water (500 μL/min, 30 min) to remove any excess PFPE oil not confined within the mesopores, continuing until no floating oil was observed. The generated oxygen first passed through a gas-liquid separator to capture liquid droplets, followed by a multi-stage filtration membrane to remove particulates and ensure high-purity output. The oxygen concentration was measured using an oxygen sensor (model ST8100A, SIGMA). The oxygen flow rate was monitored through a combined system consisting of a gas mass flow meter and digital flow monitoring equipment (model AMS2016R07, Guangzhou Aosong Electronics Co., Ltd.). Temperature measurements were acquired using a thermocouple connected to a data logging system.

Ethics statement

The study involving human participants was approved by the Ethics Committee of the Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Sciences (Approval No.: SINANO/EC/2025-087). Written informed consent was obtained from the participant shown in Fig. 5b, c and Movie S3 for the publication of the images and video.

Animals

Male C57BL/6 mice (species: Mus musculus, substrain: C57BL/6, genetic background: inbred, age: 9–12 weeks) were obtained from Suzhou Jingweiyu Biotechnology Co., Ltd. (Suzhou, China). All animals were housed under a 12 h light/12 h dark cycle at 22 ± 2 °C and 50 ± 10% relative humidity with free access to food and water. All experimental procedures were approved by the Ethics Committee of the Suzhou Institute of Nano-Tech and Nano-Bionics, Chinese Academy of Sciences (Approval No.: SINANO/EC/2025-087).

High-altitude oxygen inhalation study

Study design

The safety of the O₂-generated wearable microreactor was first validated in animal models. Subsequently, a high-altitude study was conducted with three human volunteers (two females, A and B, and one male, C) in Lhasa, Tibet, China. All volunteers were healthy adults with a long-term residence altitude of 3–4 m. The experimental protocols were initiated on the second day after arrival in Lhasa to allow for initial acclimatization.

Volunteer profile

Female A: 31 years, 165 cm, 60 kg; Female B: 26 years, 172 cm, 55 kg; Male C: 36 years, 173 cm, 65 kg.

Experimental protocols (all performed at 3650 m unless specified. The oxygen generation rate was around 200 mL/min):

Static oxygen inhalation

Volunteers remained seated at rest. Blood oxygen saturation (SpO₂) was monitored simultaneously using three pulse oximeters. The protocol involved measuring SpO₂ under two conditions: without oxygen supplementation and while inhaling oxygen supplied by the wearable microreactor via a nasal cannula. This test was repeated at least three times.

Dynamic Oxygen inhalation during ambulation

Volunteers walked at a steady pace. SpO₂ was monitored as in Protocol 1, comparing the same two conditions (no oxygen vs. microreactor oxygen). This test was repeated at least three times.

Cyclic oxygen on/off during ambulation

Volunteers walked while undergoing a cyclic intervention to assess the system’s transient response. SpO₂ was continuously monitored. The cycle consisted of: oxygen inhalation for 120 s, followed by cessation of oxygen (device off, cannula removed) for 120 s (until 240 s), then oxygen resumption for 120 s (until 360 s), and finally another cessation for 120 s. This cycle was performed twice.

Altitude-dependent response

SpO₂ was monitored under both conditions (no oxygen vs. microreactor oxygen) at three different altitudes: 3650 m, 4280 m, and 4998 m, to evaluate performance across varying hypoxic intensities.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

Xuetong Zhang conceived the project and acquired funding, and was responsible for project administration and supervision. Yuzhen Li performed the methodology, investigation, visualization, and co-wrote the original draft and the review and editing. Yafei Ding contributed to the investigation, visualization, and co-wrote the original draft and the review and editing. Yongji Li contributed to the investigation and visualization. All authors reviewed and approved the final manuscript. This study was financially supported by the National Natural Science Foundation of China (22572216) (X.Z.).

Peer review

Peer review information

Nature Communications thanks Wanli Cheng, who co-reviewed with Qingxiang Wang; Uttam Manna and the other, anonymous, reviewer for their contribution to the peer review of this work. A peer review file is available.

Data availability

All data are available in the main text or the Supplementary Materials, and all data are also available from the corresponding authors upon request. Source data are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-72772-3.