1 Introduction

Space exploration offers an opportunity to gain insights into our origins and discover other planetary systems [1]. A remarkable example is Voyager 1, launched in 1977, which continues to transmit valuable data back to Earth after more than 45 years in space [2]. Space-based radiation, such as galactic cosmic rays, solar flares, and neutrons, poses a substantial threat to space missions [3]. Exposure to space radiation increases astronauts’ risks of both cancerous and non-cancerous ill-effects, which may manifest years after returning to Earth [4]. Moreover, radiation-induced lattice rearrangement, component ionization, and trapped-charge formation degrade electronics, causing malfunctions that range from minor data loss to catastrophic failure [5].

To enable further long-distance space exploration and colonization, effective radiation shielding material that integrates a thermal-management system is needed [6, 7]. In the case of the Solar Orbiter probe, which approaches approximately 42 million kilometres from the Sun and experiences temperatures of nearly 600 °C, robust thermal control measures are imperative [8]. Ensuring that spacesuits provide adequate thermal management for astronauts is vital for their safety and well-being in space’s harsh and fluctuating temperature conditions [9]. Simultaneously, maintaining appropriate temperatures inside electronic equipment is crucial to prevent overheating, malfunctions, and potential mission failure [10].

This study aimed to fabricate 2D textiles specifically engineered for neutron protection and thermal management. Our approach was based on combining AAP and BNNT. AAP belongs to a category of synthetic engineering materials characterized by collective hydrogen bonding and π-stacking between molecular building blocks [11]. This robust material demonstrates superior mechanical stability and heat resistance compared to common polymers [12]. While considering various boron-functionalized materials, such as 0D boron carbide nanoparticles or 2D boron nitride nanosheets, we posited that the 1D cylindrical geometry in 1D BNNT would be advantageous for constructing 1D composite fibers [13]. Furthermore, BNNT’s distinctive properties, such as neutron absorption and high thermal conductivity, render it suitable for space applications [14].

However, a significant challenge in continuously and reliably producing AAP and BNNT composite (ABC) fibers lies in the absence of a nano-to-macroscale co-assembly. This challenge arises from the fact that neither AAP nor BNNT soften or melt under accumulated heat; instead, they degrade into gaseous products [15]. Consequently, melt-spinning is unviable for forming composite fibers [16]. To address this challenge, we employed the LLC self-assembly method, creating a homogeneous solution with liquid-like fluidity and crystal-like order [17]. The resulting ABC solution facilitates industrially viable wet-spinning processes [18]. The long, flexible 1D ABC fibers could be fabricated into 2D ABC textiles using techniques such as braiding, knotting, and weaving. The tailorable and washable ABC textile demonstrated simultaneous mechanical strength, thermal stability, neutron attenuation, and thermal conductivity. The utilization of neutron shielding and thermal managing ABC textiles is expected to effectively protect astronauts and electronics against space radiation and high temperatures.

2 Experimental Section

2.1 Materials

AAP (964HP, moisture regain = 7%) was purchased from DuPont. BNNTs (NAIEEL, purity > 90%) were used as received. Their diameter and length were 30–50 nm and 5–10 µm, respectively, as specified by the manufacturer. Acetone (99.5%), chlorosulfonic acid (CSA, 99.0%), and water (ACS reagent) were purchased from Sigma-Aldrich.

2.2 Preparations

AAP was washed with acetone multiple times and dried under reduced pressure at 60 °C for 12 h. The AAP and BNNT were combined in a glass vial filled with CSA. A series of AAP/BNNT solutions with different BNNT contents (100/0, 90/10, 80/20, 70/30, 60/40, 50/50, 40/60, and 30/70) were prepared. The mixtures of ABC solution were stirred magnetically for 48 h at room temperature and planetary mixed for 10 min to ensure homogeneity. The resulting solution of ABnC, where n ranged from 0 to 7, with varying BNNT contents of 0, 10, 20, 30, 40, 50, 60, and 70 wt%, respectively, was carefully transferred to a glass syringe to prevent air bubble formation. It was then extruded through a spinneret, comprising a needle and nozzle, into a coagulation bath filled with acetone at room temperature. During winding, the ABnC fibers were collected. A consistent draw ratio (1.22) was maintained across all ABnC fiber production. This ratio was calculated by dividing the draw speed of the winding roller (0.77 m min−1), by the spinning rate of the dope solution (0.5 ml min−1). The ABnC fibers were then rinsed in a water bath to remove any residual solvents. The ABnC fibers were placed on a bobbin and at 80 °C for 12 h in a vacuum oven. Lastly, the ABnC textile was formed by crafting the continuous ABnC fibers.

2.3 Characterizations

Macroscopic photographs were captured by digital camera (Canon EOS 5D). Microphotographs were obtained using polarized optical microscopy (POM, Nikon E600POL). The thermal degradation temperature was measured by thermogravimetric analysis (TGA, TA Q50). The heat capacities (Cp, J g−1 K−1) of the ABnC fibers were obtained by differential scanning calorimetry (DSC, Perkin-Elmer Pyris Diamond DSC). Scanning electron microscopy (SEM, FEI Nova NanoSEM) was used to investigate the microscopic morphology. The chemical state was analyzed by X-ray photoelectron spectroscopy (XPS, Thermo Scientific K-Alpha) and Energy Dispersive X-Ray Spectrometer (EDS, FEI Helios NanoLab650) mapping. The characteristic Raman shift of ABnC fibers were measured using Fourier Transform Raman (FT-Raman, Renishaw inVia Reflex) spectroscopy. The ABC fiber samples were placed onto a glass substrate and analyzed within the Raman shift range of 700–1800 cm−1. An excitation laser with a wavelength intensity of 785 nm was focused on the surface of the ABnC fibers. To determine the degree of BNNT alignment within the ABnC fibers, we monitored changes in the intensity of the Raman peak as the polarization angle was rotated from 0° (in vertical/vertical (VV) mode) to 90° (in horizontal/horizontal (HH) mode) relative to the fiber’s long axis. This analysis was conducted using an optical microscope with a × 50 objective lens, providing a spatial resolution of approximately 1 µm. Scattering light from the ABnC fibers was dispersed by a monochromator using 1200 grooves mm−1 grating. The fibers’ structures were investigated by wide-angle X-ray diffraction (WAXD, D8 Discover, Bruker) patterns. Their density (ρ, g cm−3) was obtained using a gradient density column (Ray-Ran POLYTEST). The fiber tester (Textechno FAVIMAT) was used to measure mechanical properties. Knot efficiency was calculated by dividing the tensile force needed to break a knotted fiber by that of a flat fiber to evaluate the flexibility. A scanning laser heating thermal diffusivity meter (LaserPIT, ULVAC-Riko) and a high temperature thermal diffusivity system (LFA467, Netzsch) were employed to measure the textiles’ in-plane and out-of-plane thermal diffusivity (α, mm2 s−1) along the longitudinal direction (LD) and transverse direction (TD) of the fibers, respectively. To conduct these measurements, a bundle comprising ABnC fiber multifilaments was loaded after plain weaving onto the sample holder. Thermal conductivity (k, W m−1 K−1) was subsequently calculated according to the following formula:

$$k = \alpha \times \rho \times C_{{\text{p}}} .$$
(1)

An infrared thermal imaging instrument (FLIR C3) recorded the heat dissipation from the ABnC textiles. A thermal neutron shieldability test was performed using an SP9 3He proportional counter (Centronic) and 241Am-9Be source with a neutron emission rate of 1.23 × 107 s−1. The thermal neutron field followed the Maxwell–Boltzmann distribution with a neutron energy of 0.6 eV.

3 Results and Discussion

Ensuring homogeneity is imperative for attaining high-performance fibers, as uneven dispersion can lead to defects within the resulting composites [19]. As shown in Fig. 1a, AAPs exploit an extensive collective hydrogen bonding network and π–π interaction between the molecular building blocks [20]. Meanwhile, BNNT exhibits a unique hexagonal lattice structure with a 1D tubular shape and highly inert surface properties (Fig. 1b) [21]. The high cohesive energy of AAP and BNNT due to the close molecular stacking and strong inter-tube interaction, respectively, is a major challenge to co-assembly. To overcome this challenge, CSA was selected as the true solvent because of its effective dispersion of both AAP and BNNT [22]. The necessary amounts of AAP and BNNT were combined in a glass vial filled with CSA to produce the ABC fiber (Fig. 1c). This process yielded a series of solutions containing ABnC, where n ranged from 0 to 7, with varying BNNT contents of 0, 10, 20, 30, 40, 50, 60, and 70 wt%, respectively. A solution of AB5C at 40.0 mg mL−1 in CSA solution exhibited the LLC phase as confirmed by POM images (Fig. S1). Schlieren textures, thread-like topologic defects, were used to identify nematic phases of ABnC solutions [23].

Fig. 1
figure 1

Material properties of AAP (a) and BNNT (b) attributed to their chemical structures. Formation of homogeneous ABnC solution through the LLC self-assembly method and its mesomorphic phase behaviors (c). Schematic illustration of constructing the uniaxially oriented composite fibers via wet-spinning (d). Continuous ABnC fibers wound on the bobbin (e). Replica spacesuit stitched with ABnC fibers (f)

Full size image

LLC phase occurs when a subtle balance of specific interactions exists between the materials. Initially, AAP dissolved in CSA as the acid disrupts the strong interchain hydrogen bonds, leading to the loss of aromatic stacking, and finally resulting in the dissociation of crystallite aggregates [24]. Simultaneously, BNNTs should be individually dispersed through the protonation of nitrogen atoms on the outer BNNT wall, causing positively charged BNNTs to repel each other in CSA [25]. Subsequently, the rod-like AAP and intrinsically rigid rod-shaped BNNT tend to align to minimize the excluded volume [26]. This increases the anisotropic volumes, forming the director (n) in the nematic domain [27]. Here, electrostatic interactions between the electron lone pair of the amide functions’ nitrogen atom and the boron atom in the partially charged hexagonal lattice enhance the intermolecular interaction between AAP and BNNT. Consequently, the LLC mesophase forms, indicating a homogeneous dispersion of AAP and BNNT with a long-range order [28]. The absence of undissolved solids was not only exclusive to AB5C but also extended to the AB0C, AB1C, AB2C, AB3C, and AB4C solutions, resulting in the depolarization of plane-polarized light and opalescence at room temperature. At high concentrations, significant BNNT aggregation was observed, coalescing to form a large number of macroscopic particles through coalescence.

The ABnC solutions, serving as the spinning dope, were wet-spun through a fine needle with an inner diameter of 159 μm into a coagulation bath using a custom-made spinning rig (Fig. 1d). During wet-spinning, the applied shear flow and extensional flow along the fiber direction (FD) lead to the uniaxial orientation of the nematic domains. Simultaneously, the fast exchange of CSA with acetone resulted in the immediate loss of fluidity of the long-range order in 1D. The ABC fiber was removed from the coagulant bath. The coagulant evaporated, and the fiber, composed entirely of solid content, was wound onto a bobbin (Fig. 1e). The AB0C fibers exhibited a yellowish color, while the fibers progressively whitened with increasing BNNT content (Fig. S2). SEM images revealed that the ABnC fibers had an approximate thickness of 10 µm (Fig. S3). As shown in Fig. 1f, the resultant AB5C fibers were used to stitch “KIST” onto a spacesuit, demonstrating that the ABnC fibers possess sufficient continuity and flexibility for their designed purpose.

The LLC facilitated AAP and BNNT’s dispersion in the ABnC fibers, enabling the tuning of target physical properties, like mechanical strength, thermal stability, neutron attenuation, and thermal conductivity. As shown in Fig. 2a, TGA thermogram revealed that the residual weights of the ABnC fibers at 700 °C increased from 1.85% to 69.72% as the BNNT content rose relative to AAP. Under an O2 atmosphere, the neat AAP used in this study underwent thermal degradation at 647 °C, whereas BNNT did not exhibit thermal decomposition up to 1,000 °C (Fig. S4). The residual weight ratio of the ABnC fibers matched well with the BNNT content added to the spinning dope. WAXD was utilized to further investigate the dispersion of BNNT within the AAP of the composite fiber, as shown in Fig. 2b. The AB0C fiber showed characteristic diffraction peak values at 20.8° and 23.2°, which respectively correspond to the (110) and (200) crystal planes of AAP. In the absence of BNNT, the diffraction peaks of the AB0C fibers were almost identical to those observed in commercial AAP-based fibers, such as Acepara (Taekwang, Korea), Taparan (Tayho, China), and Technora (Teijin, Japan) [29]. Upon incorporating BNNT, broad diffraction peaks emerged at 2θ = 25.53°, indicative of the (002) characteristic peak associated with basal plane spacing and curvature [30].

Fig. 2
figure 2

TGA (a), WAXD (b), XPS (c), and FT-Raman (d) spectra of ABnC fibers. SEM images and corresponding EDS mapping of AB5C fibers along the TD and LD (e). Polarized Raman spectra of AB5C fibers with V0° and V90° modes (f)

Full size image

XPS was used to determine the elemental composition. The N1s peak observed at 397.1 eV is attributed to BNNT’s hexagonal lattice, and its development is associated with an increased BNNT concentration (Fig. 2c). This peak is distinct from the amide function peak, which is observed at 399.0 eV. The appearance of B1s at 189.9 eV indicated an increase in the relative atomic content of B with the addition of BNNT (Fig. S5). The FT-Raman spectra of the AB0C fibers were consistent with AAP-based materials (Figs. 2d and S6). The peaks at 1182 cm−1, 1278 cm−1, 1515 cm−1, and 1611 cm−1 are indicative of C–C ring stretching. In addition, the peak at 1328 cm−1 results from C-H in-plane bending. The spectral features at 1579 cm−1 and 1649 cm−1 correspond to N–H bending and C–N stretching, respectively, with the 1649 cm−1 peak also predominantly associated with C=O stretching [31]. The spectra of the ABnC fibers reveal an increased relative intensity of the peaks at 1367 cm−1, attributed to E2g vibration of B-N, highlighted by a rectangular box [32]. The spectra across all compositions appeared nearly homogeneous, indicating no chemical reactions or new compound formations between AAP and BNNT during the wet-spinning of the LLC dope solution. The EDS mapping of the ABnC fibers, both in the TD and LD relative to the FD, revealed an even distribution of BNNT within the AAP, evidenced by the green, red, and yellow areas corresponding to B, C, and N atoms, respectively (Figs. 2e, S7, and S8). That all elements coexisted within the same region of the composite fibers suggests that AAP and BNNT were homogeneously dispersed in the CSA, without undergoing decomposition, facilitating the fabrication of continuous and reliable ABnC fibers.

Besides the uniform dispersion of BNNT, polarized Raman and POM analyses confirm its uniaxial orientation within the composite fibers. Characteristic Raman peaks observed around 1367 cm−1 show polarization-dependent intensity variations under parallel and perpendicular polarization modes relative to the BNNT long axis (Fig. 2f) [33]. Therefore, the degree of BNNT alignment in the ABC fibers was evaluated by tracking the intensity of the 1367 cm−1 Raman peak as the incident beam’s direction rotated from the VV (V0°) mode to the HH (V90°) mode, with the FD of ABC fibers aligned along the V direction of the incident beam (inset of Fig. 2f). The highest intensity of this peak occurred under parallel polarization at V0°, while the lowest intensity was observed under perpendicular polarization at V90° (Fig. S9). This pattern is consistent with the alignment of BNNTs. For AB5C, the intensity ratio of the characteristic peak (IV90°:IV0°) was about 1:6.41. At higher BNNT mass fractions, this ratio for AB6C and AB7C fibers is significantly lower compared to other ABnC fibers. This reduced ratio indicates that BNNT aggregation is occurring, which in turn makes their Raman peak intensities less sensitive to changes in the incident beam direction.

From the POM analysis, the ABC fibers were rotated in 45° increments under cross-polarized light, and images were captured at each rotational angle (Fig. S10). The most pronounced birefringence for the AB1C, AB2C, AB3C, AB4C, and AB5C fibers was observed when the fibers were at a 45° angle relative to the polarizer, indicating that the BNNTs in these composite fibers were well-aligned along the FD. In contrast, for composite fibers spun from the AB6C and AB7C dope, the variation in birefringence intensity with respect to the polarizer angle was minimal, suggesting that the LLC domains were not uniformly aligned along the FD. The poor alignment of the BNNTs in these composite fibers could be attributed to the inefficient debundling of BNNTs at high mass fractions. This birefringence, along with the polarized Raman spectra, indicated that the long axes of the inherently rigid BNNT aligned parallel to the FD.

The mechanical properties were evaluated using both tensile tests and morphologic distortions. The maximum tensile load increased up to 6.84 cN, representing the actual loading force applied to the AB5C fibers (Fig. 3a). The incorporation of BNNT enhanced the strength while reducing the strain on the ABnC fibers, because BNNT exhibits exceptional modulus and stiffness [34]. The electrostatic interaction between the AAP and BNNT may also have contributed to the enhanced mechanical strength, facilitating efficient load transfer across the matrix and nanofiller interfaces. The tensile strength and modulus of AB5C fibers increased to 188.57 MPa and 18.16 GPa, respectively (Fig. S11). However, when the BNNT loading increased to 60% and 70%, the tensile strength of the composite fibers declined sharply, decreasing to 3.71 cN for the AB6C fiber and 2.32 cN for the AB7C fiber. This significant reduction in tensile force suggests that there was insufficient AAP to effectively bind the BNNT. Moreover, imperfect AAP and BNNT dispersion results in defects that cause concentrations of stress, ultimately leading to the premature failure of the composite fibers under load. The AB5C fibers displayed increased stiffness compared to the AB0C fibers, yet remained capable of being knotted and braided (Figs. 3b and S12). The AB5C fiber’s knot efficiency was calculated to be 16.6%. When the AB5C fibers were twisted with two and three strands of monofilaments, the resulting tensile load of 2-ply and 3-ply was 1.66 and 2.43 times higher than that of the original monofilament strand, respectively. SEM images illustrated various morphological distortions of the ABnC fibers, including flat, knotted, and twisted shapes (Figs. 3c and S13). Due to their high stiffness, the AB6C and AB7C fibers could not form knots or twists. It is important to note that pliability is a critical attribute for textile applications. Consequently, the composite fiber with a BNNT content of 50% offers the most balanced physical properties for practical textile applications.

Fig. 3
figure 3

Tensile force and strain spectra of ABnC fibers (a). Mechanical behaviors (b) and SEM images (c) of the flat, knot, and twist shape of ABnC fibers. Wet-spinning ABnC fibers using a 40-hole nozzle (d). Macroscopic photographs of ABnC textiles (e). Demonstration of tailorability (f), flexibility (g), and washability (h) of ABnC textiles sewn onto a black cloth glove

Full size image

For practical applications, the 1D fibers needed processing into 2D textiles. The ABnC fibers were sufficiently long and pliable to be woven into a textile’s complex structure; however, weaving monofilament ABnC fibers, each only a few micrometers thick, proved challenging due to their thinness. Thus, we attempted fabricating multifilament ABnC fibers (Fig. 3d). The ABnC solution prepared by LLC self-assembly method was spun through fine nozzles with 40 circular holes and an inner diameter of 159 μm. SEM analysis revealed a substantial increase in fiber thickness, reaching nearly 135 µm (Fig. S14). The tensile load of these multifilament ABnC fibers increased, approximately 30-fold compared to the monofilament (Fig. S15). Wet-spinning with multi-hole nozzles effectively thickened and strengthened the ABnC fiber tufts, facilitating weaving. These ABnC fibers were successfully plain woven into a formation of ABnC textile (Fig. 3e).

The tailorability, flexibility, and washability directly impact whether they can be used in practical life outside the laboratory; therefore, we then evaluated the ABnC fibers’ real-world applicability. The resulting textiles were easily cut into various forms, such as hearts or stars, using simple tools like scissors (Fig. S16). A piece of ABnC textile was sewn onto a black cloth glove, as demonstrated in Fig. 3f. This tailorability demonstrates the ABnC textiles’ adaptability to a wide range of applications, including patterning of future protective clothing and coverings. The ABnC textile integrated with the glove withstood repeated bending, up to 100 times, to accommodate hand gestures, as shown in Fig. 3g. Due to the inherent flexibility of ABnC fibers, the textile could be bent to various angles (Fig. S17). This flexibility is essential for comfort and mobility, particularly in the case of spacesuits where astronauts need to move freely. The ABnC textiles proved to be washable, enduring exposure to tap water (500 ml) and commercial detergents (1 ml) under continuous stirring (600 rpm) (Fig. S18). After subsequent rinsing in running water, the ABnC textile sewed onto the glove was dried at 70 °C for 12 h. The ABnC textile retained its original shape (Fig. 3h). In addition, the tensile force of the fibers within the ABnC textiles were almost identical before and after the simulated washing, rinsing, and subsequent drying process (Fig. S19). The textile’s washability ensures that it can maintain its properties over time by removing contaminants, contributing to its durability and long-term functionality.

As shown in Fig. 4a, spacesuits are intricately designed with complex and multiple layers of functional fabrics serving various purposes, including stabilizing the internal pressure, supplying oxygen, and protecting against micrometeoroid shock [35]. Our ABnC textile could be utilized within spacesuit linings to mitigate radiation even at high temperatures. The thermal neutron-shielding capability of the ABnC textiles was evaluated according to Beer Lambert’s law [36]. The attenuation of thermal neutron through the ABnC textiles can be obtained based on the following equation:

$$\mu = {\text{ ln}}\left( {I_{0} I^{{ - {1}}} } \right)t^{{ - {1}}} ,$$
(2)

where μ is the linear attenuation coefficient; t is the thickness of the textile; and I and I0 denote the number of transmitted neutron fluxes with and without the shielding materials, respectively (Fig. 4b). The mass attenuation coefficients (μρ, cm2 g−1) were calculated by dividing the µ of ABnC textiles by their measured ρ.

Fig. 4
figure 4

Schematic illustration of spacesuit fabricated with functional fabrics (a). Schematic illustration of measuring the ABnC textiles’ thermal neutron-shielding performance (b). Thermal neutron-shielding properties of ABnC textiles (c). Radiation shieldability and mechanical behavior of the AB5C textile before and after thermal annealing (d)

Full size image

The thermal neutron-shielding properties of ABnC textiles, influenced by BNNT concentration, are summarized in Fig. 4c. The primary absorption property of the B nucleus in BNNT is facilitated through nuclear reactions with thermal neutrons, transforming them into charged alpha particles accompanied by gamma emissions [37]. However, it was observed that the µ of ABnC textiles improved marginally, from 0.105 to 0.228 mm−1, as the BNNT content increased from 0 to 30 wt%. Considering that the interaction probability between thermal neutrons and B elements at the macroscopic level is critical in determining the absorption cross-section, a BNNT content below 30 wt% in ABnC textiles proved insufficient for effective thermal neutron shielding. The µ of ABnC textiles markedly improved when the BNNT concentration surpassed 40 wt%, with the AB5C textile achieving a µ of 0.73 mm−1.

When ρ of the AB5C fiber was 1.80 g cm−3, the μρ of the AB5C textile reached 4.06 cm2 g−1 (Fig. S20). This offers a distinct advantage for space applications, where minimizing volume and mass is crucial. Concrete and water, both rich in hydrogen atoms, possess significant neutron scattering capabilities that effectively decelerate neutrons [38]. However, their application in aerospace is limited by their considerable volume [39]. In contrast, metallic and polymeric materials provide effective yet compact neutron-shielding solutions [40]. Specifically, hard metal-based neutron-shielding materials are preferred for structural components due to their durability, whereas soft polymer-based materials are ideally suited for personal radioprotective wear, because of their flexibility [41]. For example, aluminum alloy and polyethylene film exhibit μρ values of approximately 0.24 cm2 g−1 and 0.89 cm2 g−1, respectively [42]. Compared to these materials, the ABnC textile demonstrates superior neutron shielding, meaning that ABnC textiles achieve the equivalent thermal neutron-shielding efficiency with significantly lower volume and mass.

As expected, increasing the thickness by simply stacking the AB5C textiles boosted the thermal neutron-shielding performance. The 1-, 3-, and 5-layer configurations of AB5C textile with thicknesses of nearly 130 µm, 410 µm, and 680 µm prevented thermal neutron penetration by 9.44%, 27.65%, 41.97%, respectively (Fig. S21). The thermal neutron permeability of AB0C textiles only decreased very slightly by increasing the thickness through vertical stacking. This result supports that the thermal neutron-shielding performance is mainly attributable to the concentration of BNNT. The thermal neutron shielding of ABnC textiles could be further enhanced by reducing the inherent open spaces between adjacent warp and weft fibers. As demonstrated in Fig. S22, tightening the AB5C textile weave directly correlated with improvement in the µ. This reduction in space increases the probability of thermal neutron interactions with B atoms within the textile.

Beyond the neutron-shielding performance, the thermal stability of ABnC textile expands its potential space applications [43]. As shown in Fig. 4d, whether subjected to thermal annealing for 1 h under an O2 atmosphere or not, the AB5C textile’s neutron attenuation coefficient remained almost constant compared to its initial value. This is attributed to the high thermal stability; even at 479 °C, the AB5C textile had only lost approximately 5% of its weight. The AB5C textile annealed at 150 °C, 300 °C, and 450 °C remained chemically and structurally unchanged, reflected in the lack of significant binding energy shift (Fig. S23) or morphological change (Fig. S24). Similarly, the AB5C fiber’s tensile load (1.88 N) was almost identical before and after heat treatment (Fig. S25). Due to its thermal and mechanical stability, the macroscopic 2D structures of ABnC textiles did not collapse, rendering the fiber effective in providing thermal neutron protection during exposure to high heat.

The extreme radiation in space causes physical damage and impairs the functionality of electronic devices within spacesuits, which are vital for life-sustaining equipment. Electronic devices include sensitive components, like CPUs and memory units, which were not designed for such harsh conditions [44]. ABnC textiles, characterized by their tailorability and flexibility, offer protection. Their shape, size, and dimensions can be customized to conform to the diverse forms of electronic components, as illustrated in Fig. 5a. By strategically tailoring ABnC textiles to closely envelop CPUs, we can significantly lower the likelihood of thermal neutron interactions with the electronics, as shown in Fig. 5b, thereby reducing the risk of malfunctions. Concurrently, when designing an ABnC textile shield, it is important to incorporate thermally conductive pathways. This aspect becomes increasingly critical considering the ongoing trend toward smaller, more integrated, and higher-power electronic systems [45]. Maintaining an optimal temperature within these devices is essential to prevent overheating and consequent performance decline.

Fig. 5
figure 5

Macroscopic photograph of the tailored ABnC textile in square and circular shapes designed to shield various target electronics (a). CPU protected by simple coverage with the 2D ABnC textile (b). Thermal conductivity of ABnC fibers along the LD and TD (c). Infrared images of ABnC textiles on an CPU (d). Schematic illustration of thermal conduction pathway of ABnC fibers (e). Temperature of ABnC textiles onto the CPU according to the operation time (f)

Full size image

We measured the thermal conductivity of ABnC fibers in both the LD and TD relative to the FD (Figs. 5c and S26). The thermal conductivity in the LD significantly increased with the addition of BNNT. When the BNNT loading reached 50 wt%, the thermal conductivity in the LD peaked at 7.88 W m−1 K−1, about 4.12 times that of the AB0C fiber. This is because BNNT, has a higher intrinsic thermal conductivity than AAP and, positive influences the thermal conductivity of the composite fibers [46]. However, the AB5C fiber demonstrated only a slightly improved the thermal conductivity in the TD, measuring approximately 0.31 W m−1 K−1. BNNT is recognized for its high thermal conductivity along the LD of the tube axis, with experimental values up to 200 W m−1 K−1 and theoretical estimates as high as 6600 W m−1 K−1, attributed to the long phonon mean free path [47]. In contrast, thermal conductivity in the TD of the tube axis was significantly lower, approximately 20 W m−1 K−1, owing to weak van der Waals interactions between the tubes, which impede phonon transport [48]. Thus, the arrangement of the rod-like AAP and intrinsically rigid BNNT parallel to the FD results in anisotropic thermal conductivity between the LD and TD of the ABnC fibers [49].

Figure 5d compared the thermal-management capabilities of the AB0C textile versus the AB5C. The higher thermal responses of the AB5C textile relative to the AB0C textiles were achieved by continuously monitoring surface temperature shifts [50]. The two textiles, both with the same dimensions, were placed onto the uppermost layer of a CPU. The thermal compound was evenly coated onto the CPU to avoid the low thermal emissivity of metallic surface of integrated heat spreader, and to ensure the transfer of heat generated by the core chipset processor inside the CPU to the ABnC textile. Then, the surface temperature was recorded using an infrared thermal imaging instrument [51]. The AB5C textile’s equilibrium temperature was about approximately 9 °C lower than the AB0C textile’s. Due to the higher thermal conductivity of ABnC fibers in the LD compared to the TD, heat flow was predominantly directed along the textile’s in-plane, as illustrated in Fig. 5e [52]. Consequently, the textile mitigated surface heat accumulation and enhanced heat dissipation along the FD of the AB5C fiber [53]. Conversely, the AB0C struggled to dissipate heat in both the textile’s in-plane and out-of-plane, owing to its lower thermal conductivity in both the LD and TD [54]. This difference became apparent as the intermediate temperature of the CPU consistently increased over time. The AB5C textile’s heating rate reached 1.08 °C s−1, about 1.21 times lower than that of the AB0C textile, as shown in Fig. 5f. The AB5C textiles exhibited slower changes and lower steady temperature, attributable to their uniaxially aligned structure of fibers facilitated by the LLC phase, which effectively promotes in-plane heat flow.

4 Conclusions

The discovery of LLC enabled the high loading and uniform dispersion of functional nanofiller (here, BNNT) in a robust matrix (in this case, AAP). The 1D oriented and densely packed ABC fibers were prepared by continuous wet-spinning process. These fibers were woven into tailorable, flexible, and washable 2D textile. We conducted a suite of experiments to demonstrate that the ABC textile containing 50 wt% BNNT has the optimum balance of physical properties, namely, neutron shieldability (0.73 mm−1), thermal conductivity (7.88 W m−1 K−1), mechanical modulus (18.16 GPa), and thermal stability (479 °C). The ABC textile’s thermal neutron attenuation and heat dissipation offer protection to both astronauts and electronics against intense radiation and high temperatures in space.