A superstrong, decarbonizing structural material enabled by microbe-assisted cell wall engineering via a biomechanochemical process

Abstract

Lightweight and high-strength structural materials promise exceptional applications in advanced engineering fields. As a productive and sustainable material, wood exhibits exceptional potential to be converted into high-performance structural materials. Inspired by ancient buried wood—a naturally formed material after wood endures in microbial-rich and high-pressure environments for thousands of years—here, we demonstrate a biomechanochemical process to rapidly transform natural wood into artificial ancient buried wood (named Bio-Strong-Wood). Biotreatment depolymerizes the lignin and softens the cell wall. Then, Bio-Strong-Wood components are linked via a strong network of hydrogen and covalent bonds through the mechanochemical treatment. This results in a substantially enhanced mechanical strength (539 ± 21.7 megapascals), which outperforms the SAE 304 stainless steel. In addition, life cycle and technoeconomic assessments reveal that the obtained material achieves negative carbon emissions of 1.17 kilograms of carbon dioxide equivalent per kilogram. Overall, our work provides an economically competitive, environmentally sustainable, and decarbonizing alternative to existing structural materials.

A biomechanochemical process transforms wood into a high-strength, decarbonizing structural material.

INTRODUCTION

Structural materials, as well-established materials since the dawn of human existence, have been continuously evolving along with the development of science and technology (1, 2). Currently, structural materials should fulfill stringent requirements in terms of mechanical strength while also offering simplified processing, lower manufacturing costs, and reduced carbon footprints to satisfy diversified application scenarios (3–5). Against this backdrop, wood, the most sustainable and advanced material for past, present, and future civilizations, along with its derived functional materials, has attracted the attention of scientists (6). Wood is a historical structural material with intrinsic outstanding advantages such as natural abundance, renewability, and biodegradability (7–10). Its mechanical strength, however, is relatively low (with a tensile strength of ~10 to 110 MPa along the longitudinal direction varying among different wood species) compared to conventional structural materials such as steel, alloys, and fiber-reinforced composites that are widely used in advanced engineering fields (11). On the basis of the in-depth understanding of wood structures, a variety of wood-derived composites with impressive properties have been recently fabricated by means of advanced improvements (12–16). Among them, viscoelastic thermal compression is a commonly used processing technique, which leads to a reduction in the porosity of wood and an increase in its density, resulting in a notable increase in the mechanical strength of the material. Because of the poor plasticity of wood, it needs to be softened first to ensure that the integrity of the wood is not destroyed during the compression process. Limited by the hydrothermal treatment that can lead to incomplete densification, scientists have developed a chemical delignification process involving alkali or acid that allows the cells to collapse and close adequately during the hot compression stage (12). This is a subtractive manufacturing technique that generates substantial amounts of waste material and carbon footprint and greatly diminishes the strength of the wood (17–19). Therefore, it is urgent to develop a green treatment technology that can enhance wood properties in a low–carbon-footprint and pollution-free manner.

In this context, we focus our attention on a material known as “oriental sacred wood”—ancient buried wood (20, 21). It is rare and shows multiple excellent characteristics not found in original wood, including high density, corrosion resistance, waterproofing, or mothproofing, among others (22). Ancient buried wood originates from the changes in the Earth’s crust, landslides, and other natural phenomena that bury the ancient forest trees at the bottom of the riverbed. In the lack of oxygen, existence of high pressures, and microbial action, and over thousands of years, the ancient buried wood is formed (Fig. 1A) (23). In such a natural process, the microorganisms decompose the wood without causing environmental pollution. Despite the notable properties of ancient buried wood, it has not yet been possible to implement it in a mass-scalable and time-efficient way due to the lengthy and complex formation process.

Fig. 1. Bioinspired design of Bio-Strong-Wood—an emerging decarbonizing structural material.

(A) Schematic illustration of the formation of ancient buried wood. (B) Design strategy for the preparation of Bio-Strong-Wood via a biomechanochemical process (step I: biotreatment; step II: mechanochemical treatment via hot pressing at 40 to 60 MPa, 130°C). The horizontal axis below the image indicates the time required for forming, and the vertical axis at the right of the image indicates the changes in micromorphology, cellulose microfibril, and lignin molecular structure from natural wood to Bio-Strong-Wood. (C) Photographs of different shapes of heterogeneous Bio-Strong-Wood. Scale bars, 10 cm.

Here, inspired by ancient buried wood, we report a microbe-assisted cell wall engineering strategy that combines biotreatment with a subsequent mechanochemical process (i.e., biomechanochemical process) to rapidly transform natural wood into a superstrong, decarbonizing structural material (named Bio-Strong-Wood) in a pollution-free manner. This strategy is widely applicable to various wood species, such as basswood, oak, pine, maple, birch, and paulownia. Basswood, a common and cost-effective wood species, was chosen as the model wood species for detailed investigation. If not specified, then natural wood and Bio-Strong-Wood refer to basswood species. The most widely investigated and degradative white-rot fungus in the field of lignin decomposition was chosen for the biotreatment process. A “shearing” approach for the lignin instead of its removal was achieved by controlling the biotreatment duration (Fig. 1B). This step softens the wood cell walls and prepares the wood for the subsequent densification step with negligible change in composition. Our obtained Bio-Strong-Wood exhibits a unique combination of high strength and toughness, resulting in a much lower environmental impact and price than other conventional wood-derived composites and commonly used engineering materials. In addition, the Bio-Strong-Wood resembles ancient burial wood in its thermal and humidity resistance properties, high density, and dark color and can be processed into desired shapes and sizes for targeted applications (Fig. 1C).

RESULTS

Biotreatment analyses

Biotreatment is the most important part of our material preparation; thus, determining an optimal biotreatment condition is necessary (for the sake of distinction, the term “biotreated wood” will be used later on to refer to the wood after the biotreatment). White-rot fungi degrade woody biomass through a mechanism involving both ligninolytic enzymes [e.g., lignin peroxidase (LiP) and manganese peroxidase (MnP)] and enzyme-mediated free radical chain reactions (24–26). The elevated activity of these enzymes substantially improves the depolymerization efficiency of lignin (27, 28). On the basis of this mechanistic correlation, we prioritized the determination of the growth of white-rot fungi and the change of enzyme activities over time. Following the preincubation period, the biomass of white-rot fungi exhibited a rapid growth phase, reaching peak levels before stabilizing within a defined range (fig. S1A). Similarly, enzyme activities showed a rapid increase after the preincubation stage, peaking within several days. However, subsequent to peak activity, a marked decline was observed, ultimately plateauing at a lower range (Fig. 2A and fig. S1B). This pattern is likely attributable to nutrient depletion within the culture system. Furthermore, to maximize the efficiency of the biotreatment, experiments were designed to investigate the effect of treatment time on the wood, based on the graphs of enzyme activity versus time (fig. S1B). Because the morphological changes of the wood were not obvious with short-term treatments, the mechanical properties of the biotreated wood with different biotreatment times and the Bio-Strong-Wood prepared accordingly were also determined to obtain a visual comparison of the data (figs. S2 and S3). It can be seen that the mechanical properties of the biotreated wood after 1 day of biotreatment are close to those of the natural wood in both directions, and the mechanical enhancement of the prepared Bio-Strong-Wood is not obvious, probably owing to the incomplete softening of the wood. While the tensile strength of the biotreated wood after 30 days of biotreatment decreased notably, especially in the direction perpendicular to the growth of the wood, the mechanical properties of the prepared Bio-Strong-Wood (30 days of biotreatment) also performed poorly. The wood after 4 days of biotreatment showed the best performance with the prepared Bio-Strong-Wood demonstrating the strongest tensile properties.

Fig. 2. The effect of biotreatment on wood.

(A) Variation of enzyme activity of ligninolytic enzymes with days. To provide a favorable visual impression, LiP and MnP have been set to similar heights. The intensity of MnP has been divided by about 3.6 times. (B) Changes in the content of each component of wood after 4 and 30 days of biotreatment. The abbreviations in the figure represent, respectively: NW: natural wood; BW (4D): biotreated wood (4 days); and BW (30D): biotreated wood (30 days). (C to E) Cross-sectional scanning electron microscopy (SEM) and fluorescence images of natural wood (C) and biotreated wood after 4 days of biotreatment (D) and 30 days of biotreatment (E).

To gain further insights into the impact of biotreatment time on the mechanical strength enhancement of the final Bio-Strong-Wood products, we carried out compositional and morphological analyses of the biotreated wood samples. The lignin content of the wood after 4 days of biotreatment decreased, the cellulose content remained essentially unchanged (Fig. 2B), and a preliminary separation between the wood cell walls was observed versus natural wood (Fig. 2, C and D). In sharp contrast, the cellulose and lignin contents of the biotreated wood after 30 days of treatment decreased notably (Fig. 2B), and the thickness of the cell wall was substantially reduced, indicating that the structure of the wood had been damaged (Fig. 2E). The decrease in lignin content was accompanied by a decrease in the natural fluorescence intensity of lignin in the wood cross section (Fig. 2, C to E). Cell wall perforation and complete separation between adjacent cells caused by the removal of the middle lamella have also been observed in previous studies of wood eroded by white-rot fungi over longer times (26). This implies that the time of the biotreatment is not as long as it should be because white-rot fungi also cause decomposition of cellulose.

In addition to biotreatment time, the proper temperature is also necessary, as too high or too low a temperature can also affect the effectiveness of the biotreatment process (fig. S4). Likewise, hot pressing time also plays a critical role in the strength enhancement effect of the final product—too short hot pressing time was detrimental to the formability of the material, whereas long-term hot pressing was not necessary, as it did not have a notable effect on the strength of the material (fig. S5). Therefore, 4 days of biotreatment at 30°C combined with 5 hours of hot pressing was confirmed as the most efficient treatment.

The homogeneity of the biotreatment was also considered, which is an important guide for the fabrication of large-scale samples. The biotreated wood, with dimensions of 100 mm by 100 mm by 10 mm (longitudinal by tangential by radial), was cut in the manner of fig. S6A and characterized separately. The results showed that after 4 days of biotreatment, the wood remained consistent over a considerable range, which was observed both in the cross section and tangential section of the wood (fig. S6, B to D). This is probably because the enzyme could reach the depths of the wood quickly through the wood rays and the pits in addition to entering the interior of the wood from the vessels and wood fibers on both sides of the wood. Certainly, there were cases where the degree of biotreatment was lower in the center of the wood than at the ends of the wood, but this did not affect the final formability of the material, and the pores of the fabricated Bio-Strong-Wood were tightly closed at all positions (fig. S6E). The biotreatment exhibited excellent homogeneity and demonstrated exceptional potential for the preparation of large samples.

Morphological and structural analyses

While short-term biotreatment did not notably change the morphology of wood, hot pressing transformed the wood from porous to dense. Scanning electron microscopy (SEM) revealed structural distinctions between different states of wood. In natural wood, cylindrical vessels and fiber cells of varying sizes can be observed (fig. S7A), which collectively form a porous cross-sectional structure (Fig. 3A and fig. S7B). Bio-Strong-Wood, on the contrary, presents a different structure, with no pores in the cross section (Fig. 3B and fig. S7D). A three-dimensional (3D) profiler sweep shows no fluctuations in the cross section of Bio-Strong-Wood, indicating the occurrence of a nonporous material (Fig. 3C). Parallel to the wood growth direction, the wood lumina are entirely collapsed, and the cell walls are in tight contact with each other (fig. S7C), resulting in a 75% reduction of the thickness of the Bio-Strong-Wood (fig. S8). As a result, the density of Bio-Strong-Wood increases by 3.1 times compared to natural basswood (fig. S9). Notably, both ancient buried wood and Bio-Strong-Wood display a blackish surface, a sophisticated wood grain appearance, and a highly dense structure (figs. S9 and S10A), but their densification mechanisms are fundamentally different, as evidenced by their contrasting cell wall architectures in fig. S10 (B to D).

Fig. 3. Morphological and structural features of Bio-Strong-Wood.

(A and B) Top-edge SEM micrograph of natural wood (A) and Bio-Strong-Wood (B) showing its 3D structure. (C) 3D profiler sweep of a cross section of natural wood (left) and Bio-Strong-Wood (right). (D to F) 2D wide-angle x-ray scattering (WAXS) diffraction patterns of natural wood (D), biotreated wood (E), and Bio-Strong-Wood (F). (G) Comparison of the crystallinity of wood in different states. (H to J) 2D small-angle x-ray scattering (SAXS) diffraction patterns of natural wood (H), biotreated wood (I), and Bio-Strong-Wood (J). a.u., arbitrary units. (K) Average crystallite size perpendicular to the (200) plane.

Considering the pivotal role of cellulose in the final mechanical properties of wood (29), x-ray scattering was applied to investigate the state of cellulose in natural wood, biotreated wood, and Bio-Strong-Wood. The results of 2D wide-angle x-ray scattering (WAXS) allow the observation that the dominant crystalline polymorph of cellulose I_β persists throughout the Bio-Strong-Wood fabrication process, rather than regenerated cellulose (Fig. 3, D to F). WAXS results also provide valuable information about the arrangement and stacking state of cellulose molecular chains in protofibrils. The degree of cellulose crystallinity calculated by the Segal method (Eq. 1) revealed minor changes in the crystallinity of the wood before and after the biotreatment, whereas it was much higher in Bio-Strong-Wood (63.2% versus 58%; Fig. 3G). The average size of the cellulose microcrystals in the crystal orthogonal directions was calculated according to the Scherrer equation (Eq. 2), and the proportion of cellulose chains inside the microcrystals was estimated and expressed as values of X (Eq. 3). The average crystallite size perpendicular to the (200) plane of Bio-Strong-Wood increased compared to natural wood (fig. S12). This effect, also observed in previous dense wood studies (30), may arise from the cocrystallization of free cellulose chains in the vicinity of the crystalline region of cellulose. Similarly, the Bio-Strong-Wood exhibited higher X values (fig. S12). These results confirmed that the Bio-Strong-Wood contains more cellulose chains that remain highly organized within the cellulose microcrystals. As a result, increased hydrogen bonding interaction among neighboring cellulose chains is expected (31). The absorption peaks of attenuated total reflectance–infrared (ATR-IR) spectra from 3700 to 3000 cm⁻¹ in fig. S13, and the hydrogen bonding model in table S1 [coefficient of determination (R²) > 0.99], further confirm this hypothesis, where the content of O(6)H· · · O(3′) intermolecular hydrogen bonding in Bio-Strong-Wood increased from 27.8 to 43.1%.

The 2D small-angle x-ray scattering (SAXS) graphs reveal that the wood consistently exhibits a strong equatorial streak scattering pattern, indicating that the fibers in the material remain highly oriented after processing (Fig. 3, H to J). The degree of cellulose crystalline orientation was calculated using the Hermans’ orientation function f, according to the azimuthal scan of the scattered intensity at (200) reflection (Eq. 4 and fig. S14). f values of 0.55 and 0.63 have been obtained for natural wood and Bio-Strong-Wood, respectively, indicating an increased orientation of the crystals in the Bio-Strong-Wood (Fig. 3K). Overall, there was little difference in the cellulose state of the biotreated wood compared with the natural wood, whereas Bio-Strong-Wood exhibited a more ordered and denser stacking structure, thus fulfilling the morphological and structural conditions for improved mechanical properties (32).

Mechanism of lignin depolymerization and repolymerization

Beyond the focus on structural changes in cellulose during treatment, the changes occurring in the molecular structure of lignin were also studied. For comparison, the 2D nuclear magnetic resonance (NMR) spectroscopy spectra of natural wood, biotreated wood, and Bio-Strong-Wood were acquired, the cross signals in the heteronuclear single quantum coherence (HSQC) spectra of lignin were assigned by comparison with the published literature (33, 34), and the results are listed in table S2. Furthermore, in reference to previous studies, a quantification method by coupling quantitative ¹³C NMR and 2D HSQC NMR was adopted to quantify the detailed structure of lignin (35, 36). Owing to the degradation of the lignin during the biotreatment, the use of aromatic or methoxy signals as internal standards can yield misleading conclusions, as the samples contain degraded side chains and fragmented aromatic rings (37). Thus, we introduced internal reference compounds for the ¹³C NMR quantitative characterization of lignin samples extracted from each type of wood sample. The relative intensities of the corresponding signals in the quantitative ¹³C NMR spectra, as standardized by the internal reference compounds, are listed in table S3, and the relative intensities of the main lignin cross signals after combining quantitative ¹³C and 2D HSQC NMR are shown in table S4. The side-chain region (δ_C/δ_H: 50 to 90/2.8 to 6.0) of the 2D HSQC NMR spectrum is particularly relevant, as it relates to the interconnectivity between lignin units. Signals for arylglycerol–β–aryl ether (structure A) and pinoresinol (structure B) can be identified in Fig. 4A, revealing that the lignin present in the natural wood is mainly linked via β–O–4, β–β, and α–O–γ bonds. While analogous signals are observed in biotreated wood, reduced intensities are obtained (Fig. 4B and table S4). HSQC contour integration indicated that the biotreatment reduced the intensity of C_α–H_α (δ_C/δ_H: 71.8/4.88) and C_β–H_β (δ_C/δ_H: 83.2/4.33 and 85.8/4.13) of structure A and C_α–H_α (δ_C/δ_H: 84.9/4.66) and C_β–H_β (δ_C/δ_H: 53.6/3.06) of structure B. The weakening of the C–O intensity was also observed by x-ray photoelectron spectroscopy (XPS), where the content decreased from 39.2 to 28.3%, suggesting that ether bond breakage occurred during the biotreatment stage (Fig. 4, H and I). In addition, an intense signal corresponding to C_γ–H_γ (δ_C/δ_H: 61.4/4.10) (sinapyl alcohol; structure I), a typical lignin degradation product, was found in the HSQC spectrum of biotreated wood (Fig. 4B) (38). This result was also observed in the quantitative ¹³C spectrum with a distinct peak at 61.8 parts per million (ppm; fig. S15, inset). A decrease in the signal intensity of the syringyl (S) and guaiacyl (G) lignin units can be observed at the aromatic region of the 2D HSQC spectra, possibly due to the decomposition of lignin, in agreement with the decrease in lignin content that we obtained previously.

Fig. 4. Mechanisms of lignin depolymerization and repolymerization.

Side-chain (first line) and aromatic regions (second line) of lignin in the 2D HSQC NMR spectra: δ_C/δ_H: 52 to 90/2.9 to 5.2 and δ_C/δ_H: 98 to 149/6.0 to 7.6, respectively. (A and D) Natural wood. (B and E) Biotreated wood. (C and F) Bio-Strong-Wood. Symbols are taken from (G). See table S2 for signal assignment. (H to J) High-resolution XPS spectra of C1s peaks at 282 to 290 eV of natural wood (H), biotreated wood (I), and Bio-Strong-Wood (J). (K) Mechanism of lignin change during material preparation. The figure represents only one possible linkage, and the actual bond-breaking and bonding scenarios might be more abundant.

All signals of structure I in biotreated wood disappear in Bio-Strong-Wood after hot pressing as confirmed by 2D HSQC spectra (δ_C/δ_H: 61.4/4.10) and quantitative ¹³C spectra (δ_C: 61 to 62) (Fig. 4C and fig. S15, inset). Besides, integral data acquired by combining the two NMR spectra showed that the signals of C_α–H_α (δ_C/δ_H: 71.8/4.88) and C_β–H_β (δ_C/δ_H: 83.2/4.33 and 85.8/4.13) of structure A exhibited a decreasing trend or a small increase (table S4), implying that hot pressing does not promote the formation of β–O–4 from lignin degradation products. Simultaneously, the number of C–C linkages in the Bio-Strong-Wood increased substantially, confirming that lignin undergoes a self–cross-linking process via the formation of C–C bonds during hot pressing (Fig. 4J). The schematic diagram in Fig. 4K illustrates the complete process of lignin depolymerization and recombination. Briefly, the lignin molecules suffer cleavage by enzymes and free radicals during the biotreatment process, weakening the interfiber strength. Among them, the aryl ether cleavage is the most representative, as we observed the signal change of β–O–4 in the earlier section. After hot pressing, the lignin depolymerization intermediates repolymerize, a process that also involves the repolymerization of the aryl ether bonds, which ultimately creates a strong covalent bonding network.

Functional properties of Bio-Strong-Wood

In the light of structural and morphological analyses, Bio-Strong-Wood fulfills the requirements to exhibit mechanical properties that far surpass those of natural wood and remains even above the benchmark structural materials. Bio-Strong-Wood shows an outstanding high tensile strength of 539.8 MPa, which is not only more than 12 times stronger than natural wood but also substantially better than ancient buried wood (fig. S17), and even remains above the 520 MPa of SAE 304 stainless steel (GB/T 1220-2007, 06Cr19Ni10) (Fig. 5, A and B). The impact resistance according to the split Hopkinson pressure bar (SHPB) dynamic mechanical experiments reveals a 12-fold increase in the energy absorbed by the Bio-Strong-Wood in comparison to natural wood (Fig. 5, C and D).

Fig. 5. Bio-Strong-Wood shows a distinctive combination of high mechanical strength and high toughness.

(A) Tensile stress-strain curves for natural wood and Bio-Strong-Wood. (B) Tensile strength and work for natural wood and Bio-Strong-Wood. (C) Compressive stress-strain curves of natural wood and Bio-Strong-Wood obtained by SHPB testing. The inset shows a schematic of the SHPB test. (D) Impact strength and absorbed energy for natural wood and Bio-Strong-Wood. (E) Ashby plot of specific strength versus fracture toughness of Bio-Strong-Wood compared with benchmark materials. (F and G) Stress-displacement curves of natural wood (F) and Bio-Strong-Wood (G) acquired by in situ tensile testing. The SEM images below represent the changes in wood rays and pits at various stages of fracture occurrence, corresponding to the areas with the same numbers in the stress-displacement graphs.

Attaining materials that combine strength and toughness proves to be elusive during material design (12, 39–41). However, Bio-Strong-Wood fulfills these two often mutually exclusive properties, with a work of fracture that is more than 11 times larger than natural wood (Fig. 5B). Besides, the fracture toughness, K_IC, is more than 10 times larger for Bio-Strong-Wood (~34.2 MPa m^1/2) than the natural wood (2.6 to 3.2 MPa m^1/2) and comparable to aluminum alloys (19 to 41 MPa m^1/2) (Fig. 5E).

The improved mechanical properties can be attributed to a synergistic effect resulting from the reduction of flaws in the original material, together with the tight contact between the interfaces. In particular, the pits and wood rays in natural wood (Fig. 5F) are prone to becoming crack propagation sites (42, 43). Contrarily, Bio-Strong-Wood shows an entirely different morphology where the wood rays and pits appear closed, reducing the susceptibility to crack formation and subsequent propagation (Fig. 5G). In addition, similar to previous findings in densified wood, the collapse of the wood lumina leads to close contact between the cell walls, generating a large amount of friction and dissipating energy when sliding (12, 44). A fracture behavior similar to that of natural wood is also present in Bio-Strong-Wood, where crack propagation between cells tends to be deflected (fig. S18A). This effect can be assigned to the higher mechanical strength of wood fibers in the direction along the cell axis, where a fracture of the intercellular layer is more likely across the cell wall (45). Biotreatment and hot pressing do not alter this characteristic, allowing Bio-Strong-Wood to retain this characteristic mechanical behavior and ultimately to create similar sawtooth-shaped fracture zones as natural wood does (fig. S18B). This deflection increases the energy dissipation to make the material more ductile. To sum up, Bio-Strong-Wood balances high specific strength with fracture toughness, offering notable attributes for future structural applications.

Moreover, the bending and compression properties of the Bio-Strong-Wood are notably strengthened in all the studied directions. The largest enhancement was observed in the parallel growth direction, with values that were 4.2 and 37.1 times the strength of the natural wood, respectively (figs. S19 and S20). The strategy here developed for Bio-Strong-Wood can be extended to a variety of wood species, including hardwood and softwood, suggesting the universality of our biomechanochemical strategy (fig. S21).

In addition to its superior mechanical properties, Bio-Strong-Wood is an attractive material due to a number of other attributes, including stability under moisture attack. Upon exposure to 97% relative humidity for 192 hours, the wood mass exhibited a mere 5.05% increase with a small decrease in tensile strength. Impressively, both the changes in mass and tensile strength of Bio-Strong-Wood were less than those of the previously reported “Super Wood” prepared by alkali treatment, which may be due to the fact that the lignin in Bio-Strong-Wood is not hydrophilically modified (figs. S21 and S22) (12). Moreover, the biological and mechanochemical treatments did not sacrifice the thermal stability of the wood samples, as evidenced by the excellent thermal stability of the Bio-Strong-Wood at high temperature (250°C), which outperforms conventional fossil-carbon–derived polymers. This phenomenon mainly comes from the inherent ignition point of natural wood being higher than the melting point of common plastics (fig. S24). Furthermore, after undergoing rapid thermal shock between extreme temperatures (−196° to 120°C), Bio-Strong-Wood maintains its mechanical properties (fig. S25). Together, these attributes promise the implementation of Bio-Strong-Wood in applications under extreme environments.

Economic and environmental sustainability of the Bio-Strong-Wood

As a key structural material, the economic feasibility of Bio-Strong-Wood is of great concern for its practical implementation. The manufacturing process of Bio-Strong-Wood under industrial production is shown in Fig. 6A. For comparison, the preparation of hand layup glass fiber–reinforced polyester (GFRP), an essential material in construction and automotive manufacturing, is provided. Compared to GFRP, our preparation process is more facile and convenient, enabling the cost of Bio-Strong-Wood (including material acquisition and processing) to be markedly lower (CN ¥2.46 kg⁻¹ versus CN ¥86.61 kg^–1). As illustrated in fig. S26, consumables accounted for most of the cost of Bio-Strong-Wood production with CN ¥1.86 kg⁻¹. In addition, Bio-Strong-Wood holds clear economic advantages over the other five widely used commercial structural materials (Fig. 6B). The estimated production cost of Bio-Strong-Wood remains below that of the previously reported Super Wood, an engineered wood material fabricated by chemical delignification combined with hot pressing, if the assumptions regarding materials, labor, property, and plant are kept (tables S11 and S12) (12). More attractively, from the product value end, the Bio-Strong-Wood can provide similar or even better functionalities to the naturally formed ancient buried wood, showing clear competitiveness in high-value application scenarios. This further highlights the potential of our rapid, mild, and pollution-free biomechanochemical process in transforming low-valued, fast-growing wood into high-valued engineered wood products.

Fig. 6. Economic efficacy and environmental impact of Bio-Strong-Wood production.

(A) Comparison of current manufacturing routes of Bio-Strong-Wood (left) to commercial hand layup GFRP (right). (B) The economic cost is compared to the commercial routes for relevant structural materials based on data from ecoinvent v3.9.1. PVC, polyvinyl chloride. (C and D) Disaggregated cradle-to-gate global warming potential (GWP) for Bio-Strong-Wood according to the GWP-100 by the Intergovernmental Panel on Climate Change (IPCC) 2021 as follows: (C) GWP for on-site processing and (D) GWP for processing considering the credits from embedded biogenic carbon. The diamond represents the total GWP. (E) Cradle-to-gate environmental impacts of Bio-Strong-Wood and its competitors normalized to specific tensile strength.1,4-DCB eq., 1,4-Dichlorobenzene equivalent. PM_2.5 eq., particulate matter 2.5 equivalent.

Furthermore, we also performed a life cycle assessment (LCA) to investigate the environmental sustainability of the Bio-Strong-Wood. First, the greenhouse gas (GHG) emission over a 100-year time horizon was evaluated using the Intergovernmental Panel on Climate Change (IPCC) 2021 method. The cradle-to-gate GHG emissions using a scenario with a 100% renewable electricity production mix (based on current China mix proportion) (46) render a value of 1.18 × 10⁻³ kg·CO₂ equivalent (equiv.) per dm³/MPa. The main contributors are wood acquisition, energy consumption, potato residue treatment, and glucose use (Fig. 6C and fig. S27). As highlighted in Fig. 6D, the biogenic carbon uptake credits (with static accounting) of the used wood exceed the sum of impacts during processing, resulting in a net benefit of 2.95 × 10⁻³ kg·CO₂ equiv. per dm³/MPa for Bio-Strong-Wood. When considering a mass-based functional unit (FU), the GHG emissions are translated into −1.17 kg·CO₂ equiv. kg⁻¹, meaning that 1 kg of Bio-Strong-Wood (end product) stores 1.17 kg of CO₂ and can delay its release into the atmosphere when used in long-lasting load-bearing applications. This ability to remove CO₂ from the atmosphere confirms that our manufacturing method not only produces structural materials with outstanding functional properties but also yields attractive alternatives for decarbonizing materials. Compared to the abovementioned six structural materials, Bio-Strong-Wood demonstrates a notably lower global warming potential (GWP-100) value, regardless of whether biogenic carbon is accounted for or not (Fig. 6E and fig. S28).

To gain a comprehensive understanding, we also consider other environmental impact metrics, and the results are shown in Fig. 6E. In this context, it is important to note the conservation of natural capital, as it cannot be replaced by economic capital (47). Therefore, a careful observation of the “ecological footprint–CO₂ hectares” shows considerably reduced values, indicating that lower amounts of biologically productive terrestrial and aquatic ecosystems would be needed to absorb the emissions produced (2.7 × 10⁻³ m²·annually per dm³/MPa versus 273.5 × 10⁻³ m²·annually per dm³/MPa for hand layup GFRP) (48). This parameter is highly relevant, as humanity’s current demand for ecological resources exceeds what Earth can regenerate (49). As a resource-oriented indicator, the study also accounts for the cumulative energy demand (CED) so that the direct and indirect energy use could be obtained, which includes energy consumed during the extraction, manufacturing, and disposal of the raw and auxiliary material. The results show that our board consumes notably less primary fossil energy (1.4 × 10⁻² MJ-equiv. per dm³/MPa versus 213.6 10⁻² MJ-equiv. per dm³/MPa for hand layup GFRP), which is correlated with a low-energy intensity of the production process. On the contrary, the largest contribution to renewable CED for Bio-Strong-Wood, Super Wood, and plywood originates from the biomass energy stored in the starting biomass (see fig. S28, where materials have a 94.4% share of the renewable CED and the incubation step contributes by 92.9%) (50). The combination of the carbon storage abilities of the Bio-Strong-Wood with the low embodied energy required for its processing provides further evidence of its suitability for material decarbonization (51). Our material not only relies on mostly renewable carbon (wood and potatoes) but also renders reduced environmental impacts in most categories over its competitors (Fig. 6E). Together, these results provide encouraging data that using potato residues to create productive wood is not only in line with circularity goals using renewable materials and upcycling waste streams but also renders value-added products that offer clear environmental and economic benefits over benchmark materials in civil and industrial construction.

DISCUSSION

In summary, this study reports the development of a robust, feasible, and green biomechanochemical approach to processing natural wood into a lightweight, high-strength decarbonizing engineered material—Bio-Strong-Wood. This material mimics the natural formation process of ancient buried wood in a time frame that is five orders of magnitude shorter (thousands of years versus ~5 days). The resulting material combines multiple valuable properties, including ease of processing, high mechanical strength, high toughness, high resistance to humidity and heat, low cost, and good environmental sustainability. By analyzing the material at nano- and molecular scales, we found that Bio-Strong-Wood achieves the enhancement of material mechanics through the concomitant reinforcement of hydrogen bonding between cellulose chains and the formation of a strong covalent bonding network by lignin repolymerization. In addition, failure behavior analysis reveals the initial defect reduction and the mechanism of crack deflection, simultaneously explaining the improvement of both strength and toughness. Under large-scale industrial production, the process is economically feasible, with LCA emphasizing the environmental friendliness of the procedure and Bio-Strong-Wood demonstrating outstanding potential as a decarbonizing material. Overall, this study provides a viable path to use resourceful, renewable, and sustainable biomass to generate engineering materials with notable properties, functionalities, and high value. As a result, we foresee its application in multiple uses related to civil and engineering.

MATERIALS AND METHODS

Materials

The white-rot fungi (Phanerochaete chrysosporium) (BJMCC88160) were purchased from Beijing Microbiological Culture Collection Center. Basswood (Tilia) was purchased from Wood Research Workshop. The samples were cut as 100 mm–by–100 mm–by–10 mm (longitudinal by tangential by radial) blocks. The following types of wood were additionally purchased from this establishment: pine (Pinus sylvestris), candlenut (Paulownia fortunei), oak (Quercus alba), maple (Acer platanoides), and birch (Betula platyphylla). Besides, glucose (>99%), KH₂PO₄ (>99%), and MgSO₄·7H₂O (>99%) were purchased from Aladdin and were used to make the culture medium. The potato was purchased from the Mei Tuan Application. 1,3,5-trioxane (>99.5%) was purchased from Macklin. Dimethyl sulfoxide–d6 (DMSO-d₆) (>99.9%) was purchased from Meryer. Chromium(III) acetylacetonate (>99.99%) and 1,4-dioxane (>99%) were purchased from Aladdin.

Fabrication of Bio-Strong-Wood

Preparation of the potato dextrose broth

Potatoes (300 g) were cubed, added in boiling water, and stirred continuously until the potatoes were fully simmered. The filtered broth was added sequentially with constant stirring and heating accompanied by 20 g of glucose, 3 g of KH₂PO₄, and eventually 1.5 g of MgSO₄·7H₂O to ensure their complete dissolution. The resulting potato dextrose broth (PDB) was cooled down to room temperature in a sterile environment and set aside.

Biotreatment

P. chrysosporium was incorporated into the prepared PDB in a sterile environment and incubated on a shaker with oscillation for several days (preincubation stage). Afterward, the fungus suspension was transferred to a sterile incubator. According to the cultivation protocol in fig. S1b, woods were added at different initiation time points, followed by continuous aeration culture under corresponding durations (1, 2, 4, 7, and 30 days). The wood removed was ultimately soaked in deionized water for several rounds to remove the residual fungus solution. Samples with varying degrees of softening can be obtained by adjusting the incubation time.

Hot pressing

The wood was put into the mold and gradually pressurized to 5 MPa at room temperature and subsequently pressed at 130°C and 40 MPa for several hours to obtain the Bio-Strong-Wood samples. Following the same process as described above, Bio-Strong-Woods were successfully fabricated from various wood species, such as basswood, oak, pine, maple, birch, and paulownia. If not specified, then natural wood and Bio-Strong-Wood refer to basswood species, as it was chosen as the model wood species for detailed investigation. Bio-Strong-Wood made of other wood species is named Bio-Strong-X, where X refers to the wood species name. For example, Bio-Strong-Wood made of oak is named Bio-Strong-Oak.

Characterizations

The morphology of the samples was observed using field emission SEM (FE-SEM; Zeiss, GeminiSEM 500, England) with an acceleration voltage of 5.0 kV and 3D profile measurement (NewView 9000, America). The samples were oven dried at 80°C for 24 hours and then coated with a 4-nm layer of gold before FE-SEM testing. XPS was measured (Thermo Fisher Scientific, K-Alpha, USA) at a pass energy of 50 eV with Al Kα as the x-ray source, and the data were analyzed by Avantage. ATR-IR spectra of bulk samples were recorded by a Fourier transform infrared (FTIR) spectrometer (Thermo Fisher Scientific, Nicolet iS50 FTIR, USA) with an ATR accessory (SMART iTR Diamond). The cellulose content in the wood samples was measured using the nitric acid–ethanol method. Acid-soluble lignin, acid-insoluble lignin, and holocellulose content were measured in accordance with GB/T 10337-1989, GB/T 2677.8-1994, and GB/T 2677.10-1995 standards, respectively.

Isolation of lignin

For a detailed investigation of the changes occurring in lignin during the manufacture of Bio-Strong-Wood, lignin was extracted from the various steps of the preparation (natural wood, biotreated wood, and Bio-Strong-Wood) by the following methods: The samples were cut into small pieces and placed into a ZrO₂ bowl with mixed balls (10 balls of 2 cm diameter and 20 balls of 0.2 cm diameter) and then milled to an 80-mesh powder using a high-energy ball mill (Retsch GmbH, Germany). Organosolv lignin (OSL) was extracted according to reported works (52). The wood powder was dried at 100°C for 24 hours. Subsequently, 20 g of wood powder and 140 ml of 1,4-dioxane were weighed into a 500-ml round-bottomed flask, and 17 ml of hydrochloric acid solution was slowly added. The mixture was refluxed at 110°C for 90 min under a nitrogen atmosphere, then cooled to room temperature, and filtered. The filtration was flushed with 100 ml of 1,4-dioxane in three batches, after which the resulting filtrate was poured into 1 liter of deionized water to precipitate. The sediment was separated on a centrifuge (PingFan, TG16-II, China) at 5000 rpm and dried to obtain crude lignin. The resulting crude birch lignin was dissolved in 60 ml of solvent (acetone:water = 9:1) and precipitated with 1.5 liters of ice-cold water. The liberated lignin was separated from the solution by centrifugation at 11,000 rpm, and the precipitate was adequately washed until the pH was around 7. The dark brown powder obtained by overnight drying at 60°C is the OSL.

Characterization of lignin

Alterations occurring within the lignin during the preparation of the material were investigated using NMR spectroscopy. All NMR spectra were recorded on a Bruker AVANCE NEO 600 MHz spectrometer at 25°C in DMSO-d₆ as the solvent.

For the 2D HSQC NMR, around 50 mg of lignin was dissolved in 0.5 ml of DMSO-d₆. The spectra were collected using the hsqcetgpsi2 pulse sequence with ¹J_(C,H) = 145 Hz. The number of collected complex points was 1024 for the ¹H dimension with a recycle delay of 1.5 s. The number of transients was 8, and 256 time increments were always recorded in the ¹³C dimension. Before the Fourier transform, the data matrices were 0 filled to 1024 and 4096 points in ¹³C and ¹H dimensions, respectively. Data processing was performed using MestReNova software. The central solvent (DMSO) peak was used as an internal chemical shift reference point (δC/δH: 39.5/2.49).

For the quantitative ¹³C NMR, 125 mg of lignin and 45 mg of internal reference compounds (1,3,5-trioxane) were dissolved in 0.5 ml of DMSO-d₆. The inverse-gated decoupling sequence, which allows quantitative analysis and comparison of signal intensities, was used with the following parameters: 30° pulse angle, 1.4-s acquisition time, 2-s relaxation delay, 64,000 data points, and 20,000 scans. Chromium(III) acetylacetonate (0.01 M) was added to the lignin solution to provide complete relaxation of all nuclei.

X-ray scattering analysis

SAXS and WAXS measurements were performed at the School of Materials Science and Engineering, Hubei University, using an x-ray beam with a wavelength = 1.54 Å. The SAXS and WAXS regimes were covered by a Pilatus 300K detector set at a distance of 100 and 10.5 cm from the sample, respectively. All tests were performed in a vacuum environment. The empirical method for the calculation of the crystalline index proposed by Segal et al. (31) is

$$\mathrm{C}·\mathrm{I}=\frac{I_{200}-I_{\text{am}}}{I_{200}}\times 100$$ (1)

where C·I is the crystalline index, I₂₀₀ is the intensity of the scattering peak at 2θ = 18°, and I_am is the intensity of the scattering peak at 2θ = 15.2°. The average crystallite size L was calculated using the Scherrer equation

$$L=\frac{0.9\times \mathrm{\lambda }}{\text{FWH}{\mathrm{M}}_{2\mathrm{\theta }}\times \text{cos}\mathrm{\theta }}$$ (2)

where λ is the x-ray wavelength (0.154 nm), FWHM_2θ is the full width at half maximum of the reflection in the radial direction obtained by the fit of fig. S11, and θ is half of the diffraction angle (2θ). The surface chains occupy a layer ~0.57 nm thick (31), so the proportion of crystallite interior chains (X) is

$$X=\frac{{(L-2h)}^2}{L^2}$$ (3)

where L is the apparent crystallite size for the reflection of the plane (200) and h = 0.57 nm is the layer thickness of the surface chain. The crystalline orientation factor along the fiber axis was calculated from the WAXS pattern by Hermans’ orientation parameter (fc) as

$$f=\frac{3⟨{\text{cos}}^2\mathrm{\Phi }⟩-1}{2}$$ (4)

with

$$⟨{\text{cos}}^2\mathrm{\Phi }⟩=\frac{{\displaystyle \underset{0}{\overset{\frac{\mathrm{\pi }}{2}}{\int }}}I\left(\mathrm{\Phi }\right){\text{cos}}^2\text{Φsin}\mathrm{\Phi }\mathit{d}\mathrm{\Phi }}{{\displaystyle \underset{0}{\overset{\frac{\mathrm{\pi }}{2}}{\int }}}I\left(\mathrm{\Phi }\right)\text{sin}\mathrm{\Phi }\mathit{d}\mathrm{\Phi }}$$

where Φ is the azimuthal angle and I(Φ) is the scattered intensities in the vicinity of the (200) reflection (within the circles of the dashed line) (fig. S14).

Mechanical tests

The tensile, compressive, and flexural properties were tested on two material testing machines (Shimadzu, AGS-X 50 kN and AGS-X 2 kN, Japan) with different ranges as required. Tensile test samples were cut as strips of 1 mm by 5 mm by 100 mm before testing. The samples were clamped at both ends on the fixture, and the experiment was executed at a test speed of 2 mm/min. The dimensions of the bending test specimen were ~5 mm by 10 mm by 100 mm. The specimens were positioned on rollers 80 mm apart, and the upper roller was motioned downward at a speed of 2 mm/min. The samples were cut with a size of about 3 mm by 10 mm by 10 mm for compression strength testing and moved downward along the thickness direction at a speed of 1 mm/min. All tests were operated at room temperature and 25% relative humidity. All the above mechanical tests were replicated at least three times on different wood samples.

The fracture toughness of the samples was determined by a single-edge notched bending experiment on the material as shown in the inset of Fig. 5E (53). The dimensions of the sample are shown in fig. S29. The top roller was lowered at 4 mm/min until the sample failed. On the basis of the force-displacement curves plotted from the tests, K_IC was determined using the criteria given in the Chinese National Standards (GB/T 41932-2022).

Microscopic in situ tensile testing was undertaken on an in situ tensile stage (Mechanical Technology Inc., SEM tester100, USA) under SEM (Zeiss, Sigma 500, England) observation. The samples were fabricated into 0.2 mm–by–5 mm–by–10 mm slices to facilitate the experiments on a sophisticated test bench.

Life cycle assessment

The environmental impacts for the production of Bio-Strong-Wood were quantified according to the International Organization for Standardization 14040/44 international standards using LCA (54). A cradle-to-gate system boundary was established to account for impacts from resource extraction to the factory gate, as in previous reports on lignocellulosic materials (55–57). This includes raw materials and their corresponding upstream processes, on-site production energy, and waste management. The selected provider (market) also considers the transport method for each chemical, which generally occurs via freight train, freight lorry, or freight sea. The impacts of the capital stock, i.e., the emissions caused by manufacturing the installation, were not considered. A 100% renewable energy based on current mix proportions for China has been used for the calculations. Tables S5 to S7 summarize the complete input and output data for Bio-Strong-Wood production together with the energy mix used, as retrieved from primary information sources (own laboratory-scale production) for the life cycle inventory. For comparison, the environmental impacts of Super Wood are also quantified (tables S8 and S9). As the potato used in this study originates from waste streams, it is available burden free according to the cutoff allocation. The potato residues were assumed to be sourced from an average distance of 41 km and are transported by freight using a 16 to 32 metric ton lorry (EURO5).

The assessment used the OpenLCA 2.1.0 software with the ecoinvent v3.9.1 cutoff system model database. Four methodologies have been applied to determine the environmental impacts. First, the IPCC 2021 method was used to calculate the GWP100 and the global temperature change potential (GTP100) using 100-year conversion factors. While the GWP is a measure of the heat absorbed over a given time period due to emissions of a gas, the GTP is a measure of the temperature change at the end of that time period relative to CO₂ (58). The IPCC 2021 was used so the results are compliant with the Kyoto Protocol and the GHG Protocol Standard (solely emissions of carbon dioxide, methane, nitrous oxide, hydrofluorocarbons, perfluorocarbons, sulfur hexafluoride, and nitrogen trifluoride are included). The carbon sequestration of the used wood was estimated on the basis of its 49% carbon content (Scots pine or P. sylvestris) (59) and considering the molecular weight of carbon 12 g·mol⁻¹ and carbon dioxide 44 g·mol⁻¹. As a result, a carbon sequestration value of 1.79 kg CO₂·kg⁻¹ was obtained for P. sylvestris, very close to the generally used value of 1.83 kg CO₂·kg⁻¹ (60). The energy calculations were performed using the CED methodology, following the guidelines provided in the “PCR 2019:14 Wood and wood-based products for use in construction (EN 16485:2014)” Product Category Rule. In addition, the Ecological Footprint methodology was used to determine the global hectares required for CO₂ and land occupation. Last, the ReCiPe 2016 Midpoint (H) methodology was used to obtain the environmental impacts in additional categories (61).

To normalize the impacts, two FUs were used. First, a mass-based FU was applied, allowing for results to be obtained on the basis of 1 kg of material. This provides a common ground for cross-comparison with other materials. However, to enable objective comparisons among different materials serving the same final function, we considered the performance delivered by the Bio-Strong-Wood in its end-use application (62). Therefore, the impacts were normalized (𝐸𝐼_𝑝′) to the specific tensile strength as (57)

$${\mathit{\text{EI}}}_{{\mathit{p}}^{,}}={\mathit{\text{EI}}}_{\mathit{p}}\times \frac{{\mathrm{\rho }}_{\mathit{p}}}{{\mathrm{\sigma }}_{\mathit{p}}}$$ (5)

where ρ_𝑝 accounts for the density (kilograms per cubic decimeter), σ_𝑝 is the tensile strength (megapascals), and 𝐸𝐼_𝑝 is the environmental impact per kg. Details on the used density and tensile strength values are given in table S10. The environmental impacts of both biobased and non-biobased competitor materials were obtained from the ecoinvent v3.9.1 database. The following materials were considered:

1) Sheet rolling, aluminum | sheet rolling, aluminum | cutoff, U – RoW.

2) Steel production, electric, low-alloyed | steel, low-alloyed | cutoff, U – RoW.

3) Glass fiber–reinforced plastic production, polyester resin, hand layup | glass fiber–reinforced plastic, polyester resin, hand layup | cutoff, U – RoW.

4) Plywood production | plywood | cutoff, U – RoW.

5) Polyvinylchloride production, bulk polymerization | polyvinylchloride, bulk polymerization | cutoff, U – RoW + injection molding | injection molding | cutoff, U – RoW.

Economic assessment

The economics for the fabrication of Bio-strong-wood for the year 2024 was conducted for an industrial plant located in Wuhan (China) having a production capacity of 188,887.5 tons·year⁻¹. Table S11 summarizes the plant assumptions, including raw material acquisition, energy consumption, and labor costs (including wages, employee benefits, and payroll taxes) paid by the employer, as well as property, plant, and equipment costs. The plant requires 717 workers to process the material during an 8-hour single work shift, with 2080 working hours per year. A direct salary of CN ¥12.5 hour⁻¹ is set, with an additional 30% for indirect costs (overhead) and 20% for fringe coefficient (benefits and additional costs). An overall equipment effectiveness of 90% accounts for equipment maintenance, raw material supply interruptions, or facility downtime (biomass processing often adopts 90 to 95% values) (63, 64). The equipment is amortized over a period of 10 to 40 years, with 6.7% allocated to capital costs (equipment and factory interests), 5% to maintenance, 1% to insurance, 16% to offsite/outside battery limits (space costs for utility connections and lighting), 20% to engineering, and 10% to possible overruns as contingencies (see table S12 for the results). Financial assessments were conducted using Microsoft Excel (64). The plant requires 307,600 m² of built area at a cost of CN ¥300 m² and a 15-year amortization. Similar assumptions were considered to compare the economics of Bio-Strong-Wood with the Super Wood as previously reported (12). Besides, economic data for comparison were extracted from ecoinvent v3.9.1 using OpenLCA 2.1.0 software. The values are converted from US dollars to Chinese yuan based on the February 2024 currency exchange rate (US $1 = CN ¥7.20).

Supplementary Materials

This PDF file includes:

Figs. S1 to S29

Tables S1 to S12

References