Super-robust and sustainable bamboo structural material enabled by bonding network reconstruction

Abstract

Lightweight, homogeneously high-strength, and sustainable structural materials exert a crucial role in advanced engineering fields. Bamboo can be processed into high-performance structural material if its mechanical strength and homogeneity are remarkably enhanced. Herein, we report a strategy to process natural bamboo into a lightweight, robust, and durable structural material (HoRo-Bamboo) with homogenously high strengths in both longitudinal (L) and transversal (T) directions through bonding network reconstruction technology. The process requires the dispersion of fibers through H₂O₂/O₃ treatment, followed by the introduction of Al³⁺ and rapid pressing to reconstruct the ionic-hydrogen bonding network (IHBN). The as-prepared HoRo-Bamboo exhibits a low density of 0.83 g·cm⁻³ with superb high tensile strengths in both L (631 MPa) and T (628 MPa) directions, as well as the excellent environmental resistance and flame retardancy. Life cycle assessment shows that the carbon emissions from producing 1 kg of HoRo-Bamboo are 1.67 kg of CO₂ eq. Our approach enables bamboo to be an excellent alternative to existing structural materials, offering lightweight, super-robust, decarbonized, and sustainable features.

Bamboo can be processed into high performance structural material if its mechanical strength and homogeneity are enhanced. Here the authors report a strategy to process natural bamboo into a lightweight, robust, and durable structural material with high strengths in longitudinal and transversal directions through bonding network reconstruction.

Introduction

The robust progression of modern industry is intrinsically intertwined with the extensive employment of advanced materials^1,2. Steel, aluminum, as well as a series of alloys, have been ubiquitously adopted in domains, such as aerospace, construction, and transportation, primarily owing to their high-strength attributes^3–5. Nevertheless, these materials are fraught with multiple challenges, encompassing excessive deadweight, exorbitant costs, and substantial carbon emissions^6–9. Hence, it is of paramount importance to pioneer novel advanced materials that are lightweight, robust, durable, low-cost and consonant with sustainability requisites^10–12. Against this backdrop, natural and sustainable resources like woody biomass are viable alternatives for prospective development and utilization^13,14. Multiple wood-based robust materials have been reported including bulk natural wood, healed wood, microwave bamboo, etc.^15–22.

Bamboo is a typical lignocellulosic biomass resource. Globally, bamboo is abundant and widely distributed, with a total planting area of approximately 36 million hectares (Fig. 1b)²³. The feature of rapid growth (Fig. 1c) enables bamboo the short harvesting cycle ( ~ 3 years)²⁴, which is in stark contrast to the long maturation period of wood (50-60 years)²⁵. Additionally, bamboo harbors remarkable carbon sequestration capabilities with an annual carbon sequestration potential of ~5.1 t/hectare, approximately 1.46 times that of Chinese fir and 1.33 times that of tropical rainforests (Supplementary Fig. 1)^26–28. These extraordinary natural traits endow bamboo with a significant edge and an auspicious application perspective in the context of carbon neutrality development²⁹. The outstanding mechanical properties exhibited by bamboo lay a firm foundation for the exploration and development of novel bamboo composites³⁰. However, the mechanical properties of natural bamboo have significant limitations, with relatively favorable longitudinal (L) strength and poor transversal (T) strength^31,32. Existing work mainly relies on adding resin-based polymers or other natural polymers to the pores of bamboo or using high-strength mechanical action to enhance hydrogen bonding^33–35. These measures have significantly improved the mechanical properties of bamboo, achieving high strength (770-1098 MPa); however, they have not effectively resolved the issue of mechanical properties disparity in the L and T directions, and the density of the material has also been increased^17,20,36. In addition, natural bamboo has limited flame-retardant properties and releases large amounts of heat and smoke when exposed to situations, such as fire³⁷. These defects seriously affect the durability and reliability of bamboo-based materials^37,38.

Fig. 1. The manufacturing process of lightweight and robust HoRo-Bamboo.

a Photograph of a natural bamboo forest. b Global distribution of bamboo-producing areas. c Statistics and forecasts for the global bamboo industry from 2020 to 2028, as well as a comparison of growth rate comparison of bamboo with some engineered wood. d Feature advantages of HoRo-Bamboo. e The manufacturing process of HoRo-Bamboo and reconstruction of IHBN. f Photographs of bamboo raw materials and HoRo-Bamboo.

Herein, we report an economically feasible and sustainable method for the construction of lightweight (ρ = 0.83, Fig. 1d) and strong bamboo materials (HoRo-Bamboo) that are isotropic, with homogeneous high mechanical strength in both the L (631 MPa) and T (628 MPa) directions (Fig. 1d, Supplementary Figs. 2, 3 and 4), as well as excellent flame retardancy and durability. The process applies H₂O₂/O₃ fiber dispersing followed by ionic-hydrogen bonding network (IHBN) reconstruction, including Al³⁺ cross-linking, orthogonal paving (Supplementary Figs. 5 and 6) and rapid pressing (Fig. 1e). We have developed the fabrication process of HoRo-Bamboo across a range of scales (Fig. 1f, Supplementary Fig. 7), thereby demonstrating its industrially scalable prospects. The by-products generated during the fiber dispersion process can be used as bio-based chrome-free tanning agents for sustainable eco-leather manufacturing³⁹, achieving full-component utilization, which tremendously improves the sustainability and environmental merits (Supplementary Fig. 8). This HoRo-Bamboo manufacture process not only helps to advance the development of sustainable structural materials, but also helps to establish the circular bamboo-based bioeconomy.

Results

Fabrication and characterization of HoRo-Bamboo

We transform natural bamboo into HoRo-Bamboo for the ultra-high mechanical strength through a combination of ultra-dispersal (H₂O₂/O₃ fiber dispersing), and IHBN reconstruction steps (Fig. 2a). It is the process of the bonding network reconstruction and end enhancement of bamboo cellulose fibers. Natural bamboo is a typical biomass resource consisting of cellulose, lignin and hemicellulose, and the continuous cellulose bundles are bonded by low-strength polymer matrix of lignin and hemicellulose through hydrogen bonds, leading to high mechanical properties in L direction with weak properties in T direction (Supplementary Fig. 9). Thus, enhancing the bonding between high strength cellulose bundles and rearranging fiber direction with enhanced bonding system are critical to fabricate a robust bamboo-based structural material with homogeneous ultra-high mechanical strength. Breaking the lignin adhesion and reconstructing cellulose bundles need to be conducted. The current chemical delignification processes (alkali, sulfite, soda, etc.) carry a risk of damaging cellulose bundles, and the high temperatures involved could raise concerns regarding energy consumption, potentially offsetting the advantage of bamboo’s carbon neutrality⁴⁰. Ultra-dispersal is a mild oxidative process with consecutive ozone flow and hydrogen peroxide, which can efficiently remove lignin and hemicellulose from bamboo by continuous free radical oxidation (80 °C, 2 h). Free radicals penetrated into the interior of natural bamboo through the vessels and pores to degrade lignin and hemicellulose, and cellulose was extensively retained in the process (Fig. 2b). A loosen fibril structure of the treated sample was obtained after ultra-dispersal (Supplementary Fig. 10). The bamboo fiber was then obtained after NaOH impregnation, rinsing and slight rolling (Fig. 1f). NaOH impregnation removed residual H₂O₂ by promoting its decomposition under mild alkaline conditions, which prevented cellulose from being overoxidized. This process avoided the potential mechanical strength reduction and also adjusted the pH of the dispersed bamboo. Furthermore, these dispersed fibers maintained excellent tensile property and flexibility (Supplementary Figs.11 and 12).

Fig. 2. Changes in chemical composition and reconstruction of IHBN.

a Schematic diagram of HoRo-Bamboo fabrication. b The content analysis of cellulose, lignin, and hemicellulose. Values in b represent their mean ± SD from n  =  3 independent samples. c FTIR spectra of natural bamboo, dispersed bamboo fibers and HoRo-Bamboo. d Relative content and cross-linking efficiency of Al³⁺, the group of samples indicate the number of the small cut of HoRo-Bamboo. e MD simulation of natural bonding network, hydrogen bonding network and IHBN. Strain represents the degree of deformation experienced by the molecules during stretching, and stress (in GPa) reflects the internal stress generated by different interaction networks to resist this deformation. f Strengthening mechanism and comparison of natural bonding network, hydrogen bonding network and IHBN. The three rightmost images are snapshots of the bonding networks after bond breakage.

Fourier transform infrared spectroscopy (FTIR) analysis shows that the intensities of characteristic peaks of lignin (1635, 1509, and 1441 cm⁻¹) in dispersed bamboo are significantly decreased (Fig. 2c). Meanwhile, an obvious increase of the peak at 1730 cm⁻¹ can be observed in the spectra of dispersed bamboo, which corresponds to the C = O stretching vibration of carboxyl groups, revealing the formation of carboxyl and carbonyl groups after the ultra-dispersal process (Fig. 2c). No significant changes of the characteristic peak of cellulose β−1,4-glycosidic bond (1160 cm⁻¹) can be found, indicating surface modification without disruption of the cellulose backbone. This is consistent with the degree of polymerization (DP) results of cellulose, which showed that the reduction in DP (945 for natural bamboo and 849 for dispersed bamboo) was inconspicuous (Supplementary Table 1). Raman analysis also confirmed that cellulose was extensively retained (Supplementary Fig. 13). X-ray diffraction (XRD) showed similar results that the pattern corresponding to the crystalline natural cellulose I (2θ = 22.8°) intensified, indicating the retention of cellulose and the removal of lignin and hemicellulose (Supplementary Fig. 14). Through scanning electron microscope (SEM) and X-ray microscopy (XRM), we can clearly observe that the bamboo cell walls have become thinner, mainly due to the removal of components, such as lignin and hemicellulose (Figs. 3a and 3d). Cellulose is a linear polymer formed by glucose via β−1, 4 glycosidic bonds with a massive number of hydroxyl groups at C2, C3 and C6 positions of D-glucose units. C1s and O1s spectra of XPS analysis of dispersed bamboo (Supplementary Fig. 15) reveal that the hydroxyl groups of cellulose is partially oxidized into carboxyl and carbonyl by the process, which is consistent with FTIR analysis. Ultra-dispersal removes the components coated on the outer layer of cellulose fibrils and widens the spacing among the fibrils (Supplementary Fig. 16). Partial surface oxidation of cellulose is also carried out by ultra-dispersal, which exposes carboxyl, carbonyl, and hydroxyl groups on the cellulose fibrils (Fig. 2c, Supplementary Figs. 15 and 17). The resulting pores and channels, primarily composed of partially surface-oxidized cellulose fibers, enable metal ions to enter and coordinate (Supplementary Fig. 18). Specific surface area measurements and pore size distribution analysis demonstrated that the dispersed bamboo exhibited enlarged pores and increased specific surface area and porosity after ultra-dispersal as compared with natural bamboo (Supplementary Fig. 19), which primarily due to the removal of lignin and hemicellulose, providing pathways for Al³⁺ penetration. Meanwhile, the oxidative solutes can be collected, which can be further used in eco-leather manufacture (Supplementary Figs. 20–24 and Supplementary Tables 2 and 3)³⁹.

Fig. 3. Structural changes in bamboo fibers.

a SEM images of natural bamboo, dispersed bamboo fibers and HoRo-Bamboo. b Laser Raman microscope spectra of the cross-section of natural, dispersed and HoRo-Bamboo. c XRM 3D scanning analysis of natural bamboo, dispersed bamboo fibers and HoRo-Bamboo. d XRM internal 2D imaging analysis of natural bamboo, dispersed bamboo fibers and HoRo-Bamboo. e XRM numerical modeling of changes in bamboo cell wall thickness.

IHBN reconstruction of bamboo fibers was then carried out to fabricate HoRo-Bamboo. Metal ions, such as Ca²⁺, Fe³⁺, and Al³⁺ could effectively promote the cross-linking between celluloses (Fig. 2b, d, Supplementary Figs. 15, 25–27). Among ions of the same valence, those with a higher ionic potential (charge/radius) exhibit stronger cross-linking effect⁴¹. Al³⁺ is a promising choice due to its small ionic radius (0.535 Å) and high valence. Thus, Al³⁺ was used as the crosslinker by vacuum infiltration, and the following orthogonal assembly and hot-pressing were conducted. The bonding between Al³⁺ and bamboo accompanied by the entry of Al³⁺ into ionic channels through diffusion and electrostatic interactions, promoting the in-depth coordination between Al³⁺ and the functional groups of bamboo cellulose (Supplementary Figs. 28 and 29). We tested Al³⁺ content through inductively coupled plasma-mass spectrometry (ICP-MS), and further calculated its absorption rate, which indicated Al³⁺ content accounted for ~1% of the total mass, and the effective combination rate was ~60% (Fig. 2d). Moreover, the ICP-MS results indicated the uniform distribution of Al³⁺ (Fig. 2d, Supplementary Figs. 30 and 31), in which the Al³⁺ contents of each part of HoRo-Bamboo were stable. It further proved that effective and stable ion cross-linking between cellulose molecular chains was formed (Fig. 2d), which laid a foundation for the subsequent fabrication of HoRo-Bamboo.

Compared with HoRo-Bamboo, the internal structure of natural bamboo is primarily composed of rigid fiber bundles and parenchyma cells, exhibiting a porous cross-sectional architecture (Fig. 3a). Fiber bundles are highly organized and aligned parallel to the growth direction, while parenchyma cells orient either parallel or perpendicular to the growth axis²⁰. These components are bonded by a low-strength polymer matrix of lignin and hemicellulose⁴², where cellulose fibers are enveloped by a lignin-hemicellulose network connected via hydrogen bonds⁴³. However, the inherent brittleness of this lignin-hydrogen bond network leads to weak interfacial interactions between sclerenchyma and parenchyma cells^20,44, resulting in low load transfer efficiency and numerous defects. This hinders natural bamboo from achieving the theoretical mechanical strength of cellulose fibers⁴⁵. Our IHBN reconstruction can effectively address this issue. SEM and XRM analysis directly visualize changes in bamboo cell wall thickness, fiber bonding, and internal compactness of the bamboo fiber block (Fig. 3c–e), demonstrating a structural transition from dense-porous to a more robust configuration. XRM analysis reveals that natural bamboo has a regular internal structure with solid vessel walls and intact chambers (Fig. 3c, Supplementary Fig. 32). Bamboo fiber bundles prepared by ultra-dispersal show loose pore structure, and the chambers expand without destruction. The pore wall thickness of dispersed bamboo fibers declines remarkably, as low as 3.5 μm, which confirms the removal of lignin and hemicellulose, providing ionic channels that can facilitate Al³⁺ penetration (Fig. 3d). The average pore size of dispersed bamboo is 872.9 nm after ultra-dispersal, which enables Al³⁺ (hydrated radius of ~0.535 Å) to diffuse freely (Supplementary Fig. 19). The oxygenated groups of cellulose on the inner surface of the channels attract and immobilize Al³⁺ through electrostatic interactions and coordination bonds^46,47. The IHBN reconstruction is accomplished after the hot-pressing of the reordered Al³⁺ loaded bamboo fiber bundles, and the loose chambers are compressed into a tight structure (Fig. 3c). The average thickness of bamboo chamber wall rose from 5.15 μm (dispersed bamboo fibers) to 10.15 μm (HoRo-Bamboo), owing to the multi-wall fusion (Fig. 3e). The collapse of inter-cell pores and vessels enhances the hydrogen bond between cellulose molecular chains, which can be confirmed by small angle X-ray scattering (SAXS) that the cellulose crystallinity and orderliness is improved by hot-pressing (Supplementary Fig. 33). More importantly, the collapsed ion channels capture Al³⁺, reinforcing the ionic bonds (ionic bonding between Al³⁺ and carboxyl groups, and coordination of Al³⁺ with hydroxyl groups facilitated by carboxyl groups of partially oxidized cellulose) that act as anchors. The ionic bonds ( ~ 500 kJ·mol⁻¹) are far stronger than the hydrogen bonds (17–30 kJ·mol⁻¹), which can enhance the bonding energy without extensive density increasing^48,49. We simulated the stress of natural bonding network, hydrogen bonding network (densified wood/bamboo) and IHBN (HoRo-Bamboo) by molecular dynamics (MD), which indicated the bonding force of IHBN is significantly stronger than that of natural bonding network and hydrogen bonding network (Fig. 2e, f).

The formation mechanism of cross-linking between cellulose and Al³⁺ is clarified as follows: Al³⁺ in aqueous solution attracts the negatively charged oxygen atoms of water molecules, forming hydrated aluminum ions [Al(H₂O)₆]³⁺ and hydrated hydrogen ion, which generate a weakly acidic environment. Hydrated [Al(H₂O)₆]³⁺ undergoes deprotonation (hydrolysis) of coordinated water molecules, generating partially hydrolyzed species including [Al(OH)₂(H₂O)₄]⁺. Elevating the temperature effectively lowers the activation energy barrier for the deprotonation of coordinated water molecules in hydrated Al³⁺ complexes, thereby accelerating the hydrolysis kinetics^50–52. These hydrated aluminum species form coordination bridges with oxygenated groups (carboxyl, hydroxyl, and carbonyl) of dispersed cellulose fibers via lone-pair electron donation, establishing a stable bridging three-dimensional (3D) network structure, which endows the cross-linking with high strength and stability. The uniformly distributed Al³⁺ locks the fiber microdomains by forming ionic bridges and coordination bonds, enabling a robust binding of IHBN upon hot-pressing. IHBN consists of integrated interactions, including coordination bonds, ionic bonds, hydrogen bonds, and van der Waals forces. The crossing of cellulose fiber bundles with these interactions between cellulose fibers will remarkably strengthen the mechanical properties and durability of HoRo-Bamboo in L and T directions after the hot-pressing process.

Strength of HoRo-Bamboo

HoRo-Bamboo, enabled by the innovative IHBN internal hierarchical network, exhibits superior performance in strength, lightweight characteristics, sustainability, energy efficiency, and cost compared to bamboo-steel, bamboo plywood, stainless steel, and concrete (Fig. 4a, Supplementary Tables 4-13)^{10,30,35,53,54}. Its processing conditions also outperform other sustainable structural materials (Supplementary Tables 14 and 15). The IHBN generates a uniform internal adhesive matrix within HoRo-Bamboo, tightly bonding the multi-layered bamboo fiber matrix, enhancing its mechanical properties in both L and T directions and overall toughness (Fig. 4b, Supplementary Fig. 34). The tensile strength of HoRo-Bamboo increased about 485 MPa in L direction (631 MPa) and 618 MPa in T direction (628 MPa) compared with natural bamboo (Supplementary Fig. 4), which were approximately 1.5 and 14 folds in comparison to the densified wood, respectively¹⁵. Meanwhile, a conspicuous enhancement in bending can be observe, and HoRo-Bamboo reached about 350 MPa in both L and T directions (Fig. 4c). The bending strength in the T direction increases by 32-fold (Fig. 4c) compared with natural bamboo, which also significantly higher than the reported densified wood ( ~ 220 MPa) and healed wood ( ~ 180 MPa)^15,18. The mechanical tests about the bamboo control sample without Al³⁺ demonstrated IHBN plays critical role in enhancing the mechanical strengths of HoRo-Bamboo (Supplementary Figs. 35). The bamboo control sample relies on hydrogen bond interactions between cellulose, resulting in lower bonding strength. The fiber bonding between the staggered L and T layers is weak, and delamination is prone to occur (Supplementary Fig. 36). The mechanical performance of HoRo-Bamboo was close to the reported literature about bamboo-based materials, with a superior mechanical homogeneity^17,20,36. The fracture mechanism of HoRo-Bamboo was also investigated by the cross-sectional SEM observation (Supplementary Fig. 37). The cross-section of natural bamboo shows a smooth linear shape, with neat fractures along the direction of vascular bundles, exhibiting typical brittle characteristics (Supplementary Fig. 37a–c). In contrast, the cross-section of HoRo-Bamboo demonstrates the pull-out of a small amount of longitudinal fiber and the local crack propagation of transversal fiber (Supplementary Fig. 37d–f). The stress buffer zone formed by matrix deformation effectively inhibits delamination, and the overall structure is uniformly and intactly maintained. The robust interfacial bonding by IHBN between the fibers in orthogonal directions plays a critical role in preventing the fracture, which can effectively transfer stress and further dissipates energy through interfacial friction. Furthermore, a stress buffer zone is formed in T direction due to matrix deformation, preventing the complete penetration of local crack propagation. The synergistic effect of the strong fiber bearing in L direction and the matrix stress buffering in T direction enables the uniform stress distribution in both directions, avoiding the stress concentration. We had also prepared and tested reconstructed bamboo samples with identical fiber alignment (Supplementary Fig. 38), demonstrating that orthogonal paving is critical for the homogeneously high mechanical strengths of HoRo-Bamboo. In addition, the mechanical strengths of HoRo-Bamboo were higher than those of the reconstructed bamboo materials by Ca²⁺ and Fe³⁺ (Fig. 4c and Supplementary Fig. 39), demonstrating that the reconstruction of IHBN by Al³⁺ was more robust than other ions. HoRo-Bamboo not only successfully addressed the weaknesses of the uneven mechanical strength in L and T directions of natural bamboo, densified wood and healed wood, but also maintained the lightweight nature of bamboo (Fig. 4d). This improvement is attributed to the replacement of the natural lignin brittle network by IHBN. Compared with traditional metal alloys and recently reported bamboo structural materials (Fig. 4e)^19,35, its specific strength is 760 MPa·g·cm⁻³, indicating a significant advancement in lightweight biomass-based structural materials. The mechanical property evaluations of the bamboo/wood-based structural materials revealed that Horo-Bamboo had higher tensile strength than bamboo-based polymer composites ( ~ 190 MPa) at a lower density (0.83 g·cm⁻³, Fig. 4f, g), and the mechanical performance was similar with other bamboo structural materials^20,34.

Fig. 4. The mechanical advantages of HoRo-Bamboo.

a Comprehensive comparison of HoRo-Bamboo and other materials in terms of strength, weight, sustainability, energy consumption and cost. The scoring criteria can be found in Supplementary Methods. b Mechanical improvements in L and T directions by IHBN. c Tensile and bending mechanical homogeneities. d Density comparison of natural bamboo, dispersed bamboo and HoRo-Bamboo. e Specific strength comparison among HoRo-Bamboo and other engineering materials, such as stainless steel and alloys, etc^14–20. f Comparison of tensile strength between HoRo-Bamboo, heated bamboo¹⁹, bamboo steel³⁵, and bamboo-based composites^34,35,37,38. g Comparison of tensile strength and density between HoRo-Bamboo and other composite materials^34,35,37,38. h Diagram of the increase in compressive strength of HoRo-Bamboo compared with natural bamboo under a loading pressure of 30 kN. i Response of HoRo-Bamboo to puncture force with the change of puncture depth. j Impact toughness of HoRo-Bamboo and natural bamboo. k Flat surface of HoRo-Bamboo. Values in (d–f) represent their mean ± SD from n  =  3 independent samples.

The incorporation of Al³⁺ and a 3D bridging structure compacts the cellulose chains, which enhances the toughness of HoRo-Bamboo. In compressive tests, HoRo-Bamboo withstood 30 kN force with an outstanding compressive strength of 64 MPa (Fig. 4h, Supplementary Fig. 40), while natural bamboo easily cracked or crushed. The impact toughness (Fig. 4i) of HoRo-Bamboo achieved 8.3 J/cm², which is about three times that of natural bamboo (2.9 J/cm²). A 10 mm thick HoRo-Bamboo resisted a maximum penetration force of 538 N at 30% penetration depth (Fig. 4j), meeting the standard of high-strength engineering materials⁵⁵, suitable for applications like wind turbine blades and automotive components. The stable IHBN and the coordination-dominated scaffold structure ensure the structural stability of HoRo-Bamboo even after localized damage. The above results demonstrate that HoRo-Bamboo—a multi-level bamboo fiber scaffold structural material, provides robust mechanical strength, hardness and toughness, and can be considered as a promising candidate for the future advanced sustainable structural materials.

Improvement in durability

Durability is one of the critical properties of advanced structural materials. The manufacturing process of HoRo-Bamboo not only provides significant mechanical advantages but also greatly enhances its durability in many aspects. We provide comprehensive evaluations on the stability and durability of HoRo-Bamboo, demonstrating significant improvements in thermal, flame, moisture, and ultraviolet (UV) resistance.

HoRo-Bamboo exhibits remarkable flame retardancy: the limiting oxygen index (LOI) increased from 25% to 35%, and ignition time (TTI) extended from 49 to 63 s (Fig. 5b). Thermogravimetric (TG) analysis revealed that after IHBN reconstruction, HoRo-Bamboo has higher thermal stability (Supplementary Fig. 41). The presence of Al³⁺ in HoRo-Bamboo facilitated the rapid dehydration of bamboo cellulose into charcoal and the formation of dense charcoal layers (Supplementary Fig. 42). These dense charcoal layers and Al³⁺, along with the inert substances formed during combustion, collectively provide HoRo-Bamboo with a protective layer from the outside in (Supplementary Fig. 42). This protective layer helps dilute oxygen concentration and block heat transfer. Additionally, the high-energy ionic bonds in IHBN significantly increase HoRo-Bamboo’s thermal decomposition temperature, delaying its thermal degradation process. Combustion performance tests revealed HoRo-Bamboo reduced the heat release rate (HRR), total heat release (THR), effective heat of combustion (EHC), and total smoke production (TSP) compared with natural bamboo. HRR of HoRo-Bamboo was 3.5 times lower than that of natural bamboo at 300 s (Supplementary Fig. 43), with THR (Fig. 5b) decreasing from 31.1 (natural bamboo) to 15.2 MJ/m² (HoRo-Bamboo). HRR of HoRo-Bamboo remained stable with the prolonging of time between 300 and 600 s, while a significant HRR rise of natural bamboo could be observed at the same time, leading to 2.4 times higher THR value compared with HoRo-Bamboo at 600 s. The smoke release of HoRo-Bamboo was also tremendously suppressed, because the presence of Al³⁺ in HoRo-Bamboo facilitated cellulose carbonization that could isolate heat. The total smoke production (TSP, Fig. 5b) of HoRo-Bamboo decreased ~3.4 folds (0.75 to 0.22 m²) at 300 s compared with natural bamboo, with the ~3.3 folds decline in total smoke rate (TSR, Supplementary Fig. 44). EHC is another critical safety metric, and HoRo-Bamboo shows a 3.1-fold decline compared with natural bamboo at 300 s (Supplementary Fig. 45). It was mainly attributed to the lignin removal during the processing. The high ionicity and IHBN hierarchical structure of HoRo-Bamboo confer it exceptional flame-retardant properties, which provides valuable insights for the flame-retardant design for bamboo/wood-based materials.

Fig. 5. Durability, stability and life cycle assessment of HoRo-Bamboo.

a Improved stability and durability of HoRo-Bamboo. b Flame retardancy of HoRo-Bamboo and natural bamboo, including LOI, TTI, THR and TSP. c Environmental stability and durability of materials, including the moisture content impact of different samples, the bending strength change of HoRo-Bamboo after UV aging, and the bending strength change of HoRo-Bamboo after vapor aging. Values represent their mean ± SD from n  =  3 independent samples. d Life cycle assessment of HoRo-Bamboo, including environmental impacts per kg for HoRo-Bamboo production process and life cycle emissions of HoRo-Bamboo compared with normalized data (polyethylene and alloy). (Note: Particulate matter—RI, Abiotic depletion potential—ADP, Ecotoxicity freshwater—ET, Eutrophication—EP, Climate change—GWP, Human toxicity-cancer effects—HT-cancer, Ozone depletion potential—ODP, Ionizing radiation, human health effects—IRP, Acidification—AP, Resource depletion water—WU).

The weather resistance of HoRo-Bamboo was also significantly improved compared with natural bamboo, which was mainly because IHBN reconstruction enhances the molecular chain stability. The removal of lignin reduces HoRo-Bamboo’s sensitivity to ultraviolet light to some extent, helping to inhibit UV-induced free radical degradation (such as hydroxyl oxidation and bond breakage). Additionally, the presence of Al³⁺ also absorbs and/or scatters ultraviolet light. These properties enable HoRo-Bamboo to maintain favorable stability under solar radiation. UV aging (30 days) was conducted to simulate the solar radiation in natural environment, and no significant surface discoloration (Supplementary Fig. 46) or mechanical properties crash (Fig. 5c) could be found. On the other side, as is well known, cellulose is hydrophilic with numerous hydroxyl groups on its surface. Thus, the moisture resistance is also important for the application of HoRo-Bamboo in daily environment. The long-term durability of HoRo-Bamboo was also evaluated by using the natural aging at 50 °C with a humidity of 85% for 60 d. The results demonstrated HoRo-Bamboo maintained excellent mechanical strength with the tensile strength of 622 MPa after 60 d aging (Supplementary Figs. 47 and 48), recommending HoRo-Bamboo was durable and stable. Although the overall integrity of IHBN and the high bonding energy of ionic bonding can effectively mitigate swelling and strength reduction caused by water absorption, the lack of a physical barrier layer for moisture resistance of HoRo-Bamboo still needs to be resolved. We further addressed this issue by applying an industry-standard coating onto the surface of HoRo-Bamboo, which led to the remarkable improvement in moisture resistance (Fig. 5c, Supplementary Fig. 49). We had exposed the coated HoRo-Bamboo in a 95% relative humidity environment for 120 h, and it remained stable with the flexural strength decline of <3% (Fig. 5c). The above test results indicate that through the IHBN reconstruction and surface treatment, the obtained HoRo-Bamboo is durable and stable in daily climates.

Environmental benefits and economic feasibility of HoRo-Bamboo

We quantified HoRo-Bamboo’s environmental performance for carbon neutrality and emission reduction using a cradle-to-factory-gate life cycle assessment (LCA), and also assessed conventional metals and plastics for comparison. We primarily assessed the carbon intensity, and compared it through process standardization. The result indicates that producing 1 kg of HoRo-Bamboo emits 1.67 kg CO₂ eq. Compared to the common structural materials including polyethylene (PE) and alloy, the processing of HoRo-Bamboo reduces energy/resource use and environmental burdens (Fig. 5d), which demonstrates it exhibits lower environmental impacts in almost all categories, contributing to the achievement of global carbon peak and carbon neutrality goals. The bio-based renewability of HoRo-Bamboo is the core driver of these advantages, achieving such benefits by reducing the dependence on fossil resources and lowering the ecosystem impact.

Techno-economic analysis (TEA) was conducted to evaluate the economic feasibility of HoRo-Bamboo. The material cost and material cost per unit strength of HoRo-Bamboo are 1497 USD/m³ and 3.76 USD/average strength, respectively, which are lower than the commonly used structural materials, including bamboo steel, bamboo plywood, polyethylene sheet and aluminum alloy sheet (Supplementary Figs. 50), indicating a favorable economic feasibility^{15,17,35,43,56}. The cost consists of total investment cost, materials cost and operating cost, in which labor, oxygen and hydroperoxide cost exhibit the greatest share (Supplementary Fig. 51), which can be further reduced in subsequent stages through enhanced automation and process optimization measures.

Discussion

A novel, sustainable, lightweight, super-robust, and durable biomass-based structural material—HoRo-Bamboo is ingeniously produced from natural bamboo through a facile process. The process includes the ultra-dispersal to partially remove non-cellulosic components and the reconstruction of IHBN, ultimately resulting in the successful production of HoRo-Bamboo, which addresses the issues of mechanical inhomogeneity in L and T directions and the density increase caused by extensive pressing. HoRo-Bamboo maintains a low density (0.83 g·cm⁻³) with high mechanical properties. Notably, its tensile strength reaches 631 MPa and 628 MPa in the L and T directions, respectively, and its flexural strength reaches 353 MPa and 351 MPa in the L and T directions, respectively. Its mechanical performances are superior to those of bamboo plywood and certain metal alloys, with excellent durability and carbon neutrality. Therefore, HoRo-Bamboo holds great potential to replace traditional non-renewable structural materials, such as metal alloys, glass fiber, and plastic in many fields. The proposed reconstruction process is highly scalable, paving the way for efficient utilization of other wood-based cellulosic materials and promoting the development of related industries.

Methods

Materials and chemicals

Natural bamboo was procured from Sichuan, China. The samples were cut as 500 × 30 × 5 mm (longitudinal × tangential × radial) stripes. A roller press (BP-8170, BaoPin Technology), an ozone generator (WH-H-Y10, Nanjing WoHuan Technology Company) and a hot-pressing machine (BY70×60, Qingdao GuoSen) were used. Chemicals, including H₂O₂ (30%), NaOH ( ≥ 96%) and Al₂(SO₄)₃ ( ≥ 97%) were purchased from Shanghai Aladdin Reagent, and high-purity O₂ (99.999%) was also used to generate ozone.

Reconstruction

HoRo-Bamboo processing entails three consecutive stages: ultra-dispersal, deep cross-linking, and reconstruction. The cut bamboo strips were completely immersed in a 30 wt% H₂O₂ solution at a solid-to-liquid ratio of 1:10 (w/v), and then heated to 80 °C with ozone gas (flow rate: 1 L/min; ozone production: 10 g/h) for 2 h to partially remove lignin and hemicellulose. The softened bamboo strips were mechanically rolled, and then transferred into deionized water for stirring and dispersion. The dispersed fibers were then neutralized in 0.1 mol/L NaOH at a solid-to-liquid ratio of 1:10 (w/v) for 30 min to adjust the pH to neutral. The neutralized fibers were re-washed and transferred into a 1 wt% Al₂(SO₄)₃ solution (pH ~4) at a solid-to-liquid ratio of 1:10 (w/v), which was heated to 70 °C for 1 h. Subsequently, the dispersed fibers were sealed within custom dual-valve bags and evacuated for 10 min to achieve a vacuum of 0.08 MPa. Then the fibers were taken out for drying, and the dried fibers were orthogonally arranged in a mold and hot-pressed at 5 MPa and 70 °C for 1 h to complete reconstruction. Besides, we had also used CaCl₂ and FeCl₃ to replace Al₂(SO₄)₃ to evaluate the cross-linking. The reconstructed bamboo by IHBN with identical fiber alignment was also prepared to evaluate the effect of orthogonal paving.

In the orthogonal paving process, precise orientation control was achieved by regulating the length of dispersed fibers and implementing dimensional constraints within custom steel molds. Specifically, the first layer of dispersed fibers was laid flat within the mold, with adjacent fibers tightly arranged such that their fiber axes aligned along the mold’s X-direction. The second layer of bamboo fibers was placed atop the first layer, with their fiber axes perpendicular to the first layer, aligned along the mold’s Y-direction. Each layer maintained a consistent bamboo fiber mass, and subsequent layers were stacked according to an alternating X-Y-X-Y rule until the target thickness was achieved. The entire hot-pressing process occurred within the mold’s constraints, ensuring the orthogonally oriented layers remain intact.

Characterization

Samples were characterized by SEM (Apreo 2, Thermo Scientific) to observe the morphology change. Laser Raman microscope (Dxr 2xi, Thermo Scientific) and XRM (Xradia 615, Zeiss) were also employed to characterize natural bamboo and HoRo-Bamboo. Lignin, cellulose, and hemicellulose contents were measured by the NREL standard methods¹⁷. FTIR (Nicolet IS50, Thermo Scientific), XPS (K-alpha, Thermo Scientific), Nanoparticle size and zeta potential analyzer (Nanobrook Omni, Brookhaven), XRD (D8 Advance, Bruker) and SAXS (Xeuss 3.0, Xenocs) were also used to analyze the bamboo samples.

ICP-MS was used to test the relative content of Al³⁺ in HoRo-Bamboo and calculate its cross-linking efficiency (the content of each part was averaged to analyze the overall relative content). 0.5 g HoRo-Bamboo was microwave digested in a mixed solution of 6 mL of nitric acid and 2 mL of hydrogen peroxide. The digested solution was transferred to a 50 mL volumetric flask and diluted to the marked volume with deionized water to obtain the test solution. To ensure data reliability, each sample was tested three times and the average value of the three measurements was taken as the final result. The cross-linking efficiency was obtained by calculating the ratio of the Al³⁺ content in HoRo-Bamboo to the initial Al³⁺ content added in the cross-linking process, which referred to the method described in Supplementary Methods.

The density of HoRo-Bamboo was measured by dividing the material into 20 pieces with a size of approximately 20 mm × 25 mm × 10 mm. The mass and volume of all samples were measured, and the average value of the samples was calculated.

Mechanical and durable testing

Mechanical testing system (68TM-30, Instron), TG analyzer (TG 209F1, Netzsch), oxygen meter (T2406-1, TTech) and cone calorimeter (i-CONE, FTT) were used to determine the mechanical property, thermal stability and flame retardancy of bamboo samples, respectively^15–20,30. Oxygen meters and cone calorimeters shall be tested in accordance with UL-94 and ISO 5660-1 standards, respectively. Surface roughness was characterized by a white light 3D optical microscope (GT-K1, Bruker). Moisture resistance and ultraviolet resistance were conducted in a water vapor aging chamber (GSU-24V, Espec) and a xenon lamp tester (Q-sun Xe 8, Q-Lab), respectively.

The comprehensive comparison of HoRo-Bamboo and other materials was conducted by evaluating specific strength, light-weight, sustainability, energy efficiency and cost by combing the LCA and TEA with the assessment of the materials reported in literatures^{10,30,35,53,54}.

Molecular dynamics simulation

All simulations were performed using the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS)^57,58. The Velocity-Verlet scheme was employed, with a fixed integration time step of 1.0 fs. Temperature and pressure were controlled by a Nosé-Hoover isothermal-isobaric coupler. The cutoff radius for van der Waals interactions was set to 1.2 nm, and long-range electrostatic interactions were handled using the Particle-Particle Particle-Mesh (PPPM) method with a relative accuracy of 10⁻⁵. Isothermal-Isobaric Ensemble (NPT) molecular dynamics simulations were conducted for 200 ps at 300 K and 1 atm. Under 300 K conditions, these equilibrated models were uniaxially stretched along the X direction for 600 ps at a constant engineering strain rate of 1 ns⁻¹.

Cradle-to-gate life-cycle carbon analysis

A cradle-to-gate life-cycle carbon analysis was conducted using OpenLCA 2.5.0 software with the Ecoinvent 3.1.0 cutoff system model database to quantify the carbon footprints of the HoRo-Bamboo. The analysis followed the ISO standard 14040 and 14044 for carbon footprint analysis²².

Techno-economic analysis

The TEA analysis was used to evaluate the cost in the production process of HoRo-Bamboo. The calculation boundary includes total investment cost, operation cost, and material cost. The total investment cost consists of equipment purchase cost and other investment costs.

Author contributions

T.D. designed the experiments, following the directions of X.X. and X.P.L. The experiments were carried out by T.D. The assistance of J.L.C., Y.L.Z., Z.J.S., and Y.W. was instrumental in creating the illustrations. T.D., G.Y.H., J.K.X., and L.J. performed material characterizations. T.D., X.X., X.P.L., and B.S. collectively wrote the paper. All authors commented on the final manuscript.

Peer review

Peer review information

Nature Communications thanks Blaise Tardy, who co-reviewed with Qian Liu; and the other, anonymous, reviewers for their contribution to the peer review of this work. A peer review file is available.

Funding

X.X. acknowledges the support by the Tianfu Yongxing Laboratory Organized Research Project Funding (No. 2024KJGG20). B.S. disclosed support for publication by the Tianfu Yongxing Laboratory Organized Research Project Funding (No. 2024KJGG20). Y.L.Z. thanks the support by the Tianfu Yongxing Laboratory Organized Research Project Funding (No. 2023CXXM16).

Data availability

The data supporting the findings of this study are available within the article and source data files. Any other information needs, please contact the corresponding author. Source data are provided with this paper. 10.6084/m9.figshare.30091039. 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/s41467-026-73719-4.