Towards negative carbon footprint: carbon sequestration enabled manufacturing of coral-inspired tough structural composites

Introduction

In recent years, the impacts of global warming have become increasingly evident, manifesting in phenomena such as rising sea levels^1,2, intensified storms^3,4,5, insect infestations^6,7,8, and wildfires^6,9,10. To address global warming, carbon sequestration provides a mid-term solution to mitigate environmental impacts and facilitate the continued use of fossil fuels until renewable energy technologies reach maturity¹¹. Several existing methods of carbon sequestration have been explored (Table S1). For example, direct CO₂ sequestration involves injecting CO₂ into geological formations and oceans^11,12,13; calcium loop processes immobilize carbon through chemical reactions with calcium salts or oxides^14,15,16; hydrate-based separations offer a means to isolate CO₂ from industrial emissions for storage in a liquid phase^17,18; strategies such as afforestation^19,20,21 and oyster²² farming enhance natural carbon sequestration by absorbing CO₂; and artificial photosynthesis transforms CO₂ into chemicals, such as energy fuels or other liquid molecules^23,24,25. Despite the diversity of approaches, most existing efforts have been focused on storing CO₂ or converting CO₂ into liquid substances (Table S1), with minimal attention devoted to transforming CO₂ into load-bearing solid materials²⁶.

In contrast to most man-made carbon sequestration technologies, natural carbon sequestration processes exhibit a remarkable ability to convert CO₂ into load-bearing solids with exceptional mechanical properties. Take coral as an example (Fig. 1a): Coral reefs demonstrate remarkable resilience against various environmental stressors in the ocean, including attacks by marine organisms and strong ocean currents²⁷. The resilience is attributed to the formation of their aragonite skeletal corallite structure within a semi-enclosed extracellular compartment, termed the extracellular calcifying medium, which can precipitate calcium minerals around organic templates to form a micro-structured skeleton by harnessing CO₂ sequestered from the atmosphere via photosynthesis and calcium ions from the seawater (Fig. 1a)^28,29,30. Corallites contain a circle wall and several septa as primary vertical microstructures (Fig. 1a)³¹. The septum is a biocomposite with a microstructure consisting of alternating organic and mineral phases³², which deviates from a layer-by-layer biochemical growing process (Fig. 1b)³³. Essentially, the biochemical growing process sequesters atmospheric CO₂ and transforms it into a mechanically strong mineral-organic composite septum. This natural carbon sequestration process, if successfully implemented in engineering systems, may lead to a new paradigm that transforms undesirable atmospheric CO₂ into robust engineering solids for diverse structural applications. However, the translation of this process from natural coral ecosystems to engineering practice remains a formidable challenge.

Towards negative carbon footprint: carbon sequestration enabled manufacturing of coral-inspired tough structural composites — **Fig. 1: The overall concept of coral-inspired carbon sequestration enabled manufacturing.**

Inspired by coral, we present a paradigm that effectively captures and immobilizes carbon dioxide, thereby enabling the manufacturing of robust solid composites with exceptional mechanical strength and toughness. The technology harnesses an electrochemical process which efficiently sequesters CO₂ and converts it into calcium carbonate minerals³⁴ around 3D-printed polymer scaffolds, thus yielding mineral-polymer composites with pre-designed microstructures (Fig. 1c). Essentially, the electrochemical manufacturing process gives birth to structured mineral-polymer composites by sequestrating atmospheric CO₂ with a high uptaking rate, thus leading to a negative carbon footprint (see “Methods and Materials” for the estimation method), which strikes a drastic distinction from the manufacturing of most existing engineering materials where positive carbon footprints may be induced (Fig. 1d). In terms of mechanical performance, the manufactured mineral-polymer composites demonstrate a remarkable resistance against shear forces with their specific stiffness comparable to existing engineering composites and their specific strength comparable to technical ceramics (Fig. 1d). Besides mechanical properties, the mineralized composites also exhibit promising functionalities, including structure-programmability, composition-reversibility, fire-resistance, and crack-repairability. Furthermore, through the integration of a modular assembly strategy, the proposed manufacturing can be scaled up to fabricate load-bearing meso-structures with dimensions significantly exceeding those of individual modules. Through leveraging carbon sequestration to manufacture multifunctional composites, the paradigm presented in this study paves the way for negative carbon manufacturing with a more sustainable and promising future.

Results

Mechanism and process of electrochemical manufacturing

To elaborate on the mechanism of the electrochemical manufacturing process, we first take a rectangular structure as an example. The rectangular polymer lattice structure is first 3D-printed and later coated with a thin conductive layer (i.e., ~100 nm Palladium) (Figs. 2a, S1). As the conductive layer assists the electrochemical water splitting process, cobalt sulfide²⁰, titanium²¹ and nickle²² may be alternative options for the conductive layer. The coated structure is then connected to electrochemical circuits to serve as a cathode and immersed in a CaCl₂ solution (1 mol/L) (Figs. 2a, S2). The mechanism of CO₂ conversion to calcium carbonate (CaCO₃) minerals can be divided into 4 steps. Step 1: CO₂ gas is pumped into the solution to enable a hydrolysis process to form bicarbonate and hydrogen ions; Step 2: The electrochemical water splitting around the cathode structure obtains hydroxide ions^35,36; Step 3: The bicarbonate ions from Step 1 and the hydroxide ions from Step 2 react with calcium ions in the solution to generate CaCO₃ solute³⁴; Step 4: When the concentration of CaCO₃ solute is high enough, it is precipitated out through heterogeneous nucleation on the porous surface of the cathode structure, thus gradually filling up the 3D-printed pores to form a condensed mineral-polymer composite³⁷. A low voltage (~6 V) is applied to ensure slow precipitation of CaCO₃ minerals layer by layer, thus enabling a strong bonding between mineral layers and a high mechanical strength of the precipitated mineral phase. Otherwise, a high voltage may trigger a high speed of precipitation of sand-like CaCO₃ mineral particles with an extremely low mechanical strength of the mineral phase³⁸.

To depict the electrochemical mineralization process, we examine samples with reaction durations of 0, 1, 2, 4, and 6 days (Fig. 2b–d). Figure 2b provides a schematic representation of the mineralization process for corresponding various days, and Fig. 2c and 2d show the images obtained using an optical microscope and a scanning electron microscope (SEM), respectively. The images reveal that CaCO₃ minerals grow starting from the inner surface of the lattice structure and gradually extend out to fill the lattice pores (Fig. 2e). After a 4-day reaction, the lattice pores are approximately filled up and the overall structure becomes mineralized polymer composite (Fig. 2b). The thickness of the growing minerals on the inner surface of the lattice structure increase with an accelerating rate until reaching a plateau on day 4 (Fig. 2e). Figure 2e shows the estimated mass of uptaken CO₂ per unit mass of the hosting structure for various reaction durations. The results reveal that a reaction for four days leads to uptaking 2720 kg CO₂ per ton of 3D-printed structure (see “Materials and Methods” for the calculation method). This is significantly higher than the CO₂ uptaking of carbon-negative concrete (150 kg per ton)²⁶ and CO₂-absorbing functional wood (30–40 kg per ton)³⁹, not to mention CO₂ emission of 201 to 222 kg/ton associated with Portland cement⁴⁰. Next, we employ a compressive test to measure the stiffness of the structure over the electrochemical manufacturing process. We find that the structural stiffness increases by over 500 times over the electrochemical mineralization process (Figs. 2f and S3, Table S2). We also measure the flexural strength of the mineralized samples with different reaction durations via three-point bending tests (Figs. 2g, S4, Table S3). Upon mineralization after a 4-day reaction, the flexural strength of the structure increases by around 50 times.

Coral-inspired mineralized structural composites

Next, we harness the proposed electrochemical manufacturing method to fabricate the coral-inspired structural composites. We first construct a computer-aided design (CAD) structure of the organic phase of coral septa, which is then 3D-printed, followed by a coating with a thin conductive layer (Fig. 3a, b). With the electrochemical process, CaCO₃ minerals (Fig. 3c, d) are gradually precipitated out within the pre-designed pores, resulting in the creation of mineral-polymer composites (Fig. 3e, f). Computed tomography (CT) scanning is employed to reveal the internal microstructure of the mineralized structural composite (Figs. S5ab, 3d, f). CT scanning reveals that the electrochemical mineralization process can fill the lattice pores over 85% before the outer surface of the structure is mineralized.

**Fig. 3: Coral-inspired structural composites.**

The mineralized structural composite with alternating mineral-polymer phases is expected to feature outstanding mechanical performance in resisting the shear loads. To experimentally prove this, we employ unnotched two-section compact shear tests to measure the shear strengths (Fig. 3g, S6a) and notched single-section compact shear tests to measure the shear fracture toughness (Fig. 3h, S6b) of the fabricated mineralized composites over various reaction durations. After complete mineralization for 6 days, the shear strength increases by 17.5 times and the shear fracture toughness enhances by 74 times (Fig. 3i). The shear fracture toughness of the mineralized composite is also 102 times higher than that of the CaCO₃ mineral phase (Fig. S7a) and 81 times higher than that of the polymer phase (Fig. S7b). To reveal the mechanism for the high shear fracture toughness, we employ SEM to image the cracking surface after the notched single-section compact shear tests and find a clear zig-zag crack propagation pathway (Fig. 3j). The observed curved crack propagation pathway assists the mineralized composite to dissipate a large amount of energy during the fracture process, thus exhibiting a high fracture toughness of the composite^41,42.

To compare the mechanical performance and CO₂ uptaking capability of the proposed mineralized composites with those of the state-of-the-art carbon-reducing composites, we present the material property charts for specific shear strength (i.e., shear strength per unit density), CO₂ uptaking rate (see Materials and Methods for calculation method), and specific shear fracture toughness (i.e., shear fracture toughness per unit density) in Fig. 3k, l. Compared to the state-of-the-art carbon-reducing composites, including lightweight concrete with reduced carbon emissions⁴³, carbon-negative concrete²⁶, and CO₂-absorbing functional wood³⁹, both specific shear strength and specific shear fracture toughness of the mineralized composites are comparable to the highest existing performances (i.e., of CO₂-absorbing functional wood³⁹), while the CO₂ uptaking rate is an order of magnitude higher than that of highest existing performance (i.e., of carbon-negative concrete²⁶). Impressively, the specific shear strength of the mineralized composite is 100% higher and the specific shear fracture toughness is 600% higher than those of the natural corallite, respectively^{26,39,40,44,45,46,47,48,49,50}.

Multifunctionalities of the proposed paradigm

Besides exceptional mechanical performance and CO₂ uptaking capability, the proposed electrochemical manufacturing method and the fabricated mineralized composites feature several interesting functionalities (Fig. 4). First, the mineralized composite can be facilely programmed by freely designating the mineralization region of the 3D-printed lattice structure via selectively coating conductive layers (Fig. 4a–c). For example, a rectangular polymer lattice structure is coated only at both ends, ensuring that mineralization occurs exclusively at these two locations during the electrochemical process (Fig. 4a, b). Upon compressing this structure, a distinct variance in the deformation is observed, affirming the notable difference in structure stiffness (Fig. 4c). In addition, in contrast to most existing manufacturing methods that are typically not reversible, the electrochemical manufacturing process proposed in this work is fully reversible (Fig. 4d–g). When a mineralized composite sample is submerged in a hydrochloric acid solution (1 mol/L), the mineral phase can be dissolved within 10 min (Fig. 4d, e) and the resulting demineralized lattice can be re-mineralized with the electrochemical process (Fig. 4f, g), which does not require the redeposition of a conductive layer. Moreover, though the 3D-printed polymer is not fire-retardant, the mineralized composite is found to feature an exceptional fire-retardant property (Fig. 4h–k). As shown in Fig. 4h–j, the shape of a mineralized sample does not evidently change after being placed on fire for 30 min. The mechanical performance of the mineralized sample after burning for 30 min is also mostly equivalent to that of the sample before burning. We hypothesize that fire-retardant performance is probably attributed to the fire-quenching effect of the released CO₂ when decomposing small amounts of CaCO₃ minerals under high temperatures (Fig. S8)^51,52. Furthermore, the mineralized composites, if damaged, can be repaired by triggering the electrochemical process around the damage interface (Fig. 4l–m). We take a mineralized composite and break it into two segments. These two segments are later fastened together and finally fused together through the mineralization upon application of our electrochemical process on the interface (Figs. 4l, S9). The repaired composite is loaded with three-pointing bending and its flexural strength can attain 60.5% of that of the unbroken original sample (Fig. 4m).

**Fig. 4: Multifunctionalities of mineral-polymer composites.**

Modular assembly enabled scaling-up of the mineralized composites

One promising vision is to extend the proposed electrochemical manufacturing paradigm to construct future prefabricated carbon-negative buildings. The prefabrication approach is appealing because it can drastically reduce CO₂ emissions compared to on-site construction^53,54,55,56. Besides, the buildings are expected to continuously sequester CO₂ from the atmosphere to strengthen the buildings’ mechanical strength and resistance (Fig. 5a). Despite the promise, one key challenge is how to scale up the fabricated mineralized composite structure.

To address the scaling-up challenge, we propose to employ a modular assembly method that large-scale structures can be fabricated by assembling small-scale puzzle modules of the mineralized composite (Fig. 5b, c). To demonstrate the concept, we fabricate exemplary puzzle modules by first 3D-printing polymer structure units (Fig. 5d), then coating the units with a thin conductive Pd layer (Fig. 5e), and finally mineralizing the units with the proposed electrochemical manufacturing (Fig. 5f). To examine the modular connectivity, we measure the tensile and shear strength of the assembled two units (Fig. 5g–j). Following a 6-day reaction duration, the tensile and shear strength of the connection increase by 27 and 11 times, respectively (Fig. 5j). The tensile strength of the assembled units is in the same order as flexural strength obtained in Fig. 2g and the shear strength is in the same order of the shear strength obtained in Fig. 3g. Finite element simulations reveal that maximum stress is located at the neck region of the connection (Fig. S10). The mineralized neck regions, with high mechanical resistance, ensure a safe connection between puzzle modules. To further evaluate the robustness of the assembled structure, we measure the out-of-plane resistance of a sixteen-assembled-unit plane and find it could resist an out-of-plane force equivalent to 50 times its own weight (Fig. 5k). To demonstrate the feasibility of the modular assembly method in fabricating 3D structures with curved surfaces, we construct a dome model by assembling 16 curved units (Fig. 5l). To further evaluate the molecular assembly method in fabricating future buildings, we build up a house model by assembling 31 units of 4 categories (Fig. 5m). The house model can withstand a load equivalent to 176 times its own weight (Fig. 5n).

Discussion

In summary, we present an electrochemical manufacturing method to sequester CO₂ to fabricate coral-inspired structural composites with exceptional mechanical properties. Different from most carbon sequestration methods that only store CO₂ or convert CO₂ into liquid substances, the coral-inspired manufacturing method turns undesirable atmospheric CO₂ into load-bearing strong solid composites (Table S1). Compared to an existing method that utilizes CO₂ to fabricate low/negative carbon concrete²⁶, our method not only enables structures with designed microarchitectures but also features a 25-times higher carbon sequestration efficiency (i.e., captured CO₂ mass per unit volume over carbon sequestration timescale shown in Table S1). This paradigm may lead to a research area where diverse new manufacturing methods are proposed to sequester CO₂ to fabricate functional solids for various engineering applications. Besides, our paradigm in this work not only enables the fabricated structural composites with exceptional mechanical strength and toughness but also opens opportunities to fabricate hybrid structural composites like sophisticated bouligand structures⁵⁷ and nacre-like structures⁵⁸. Moreover, our fabricated structural composites also exhibit unusual functions such as fire-resistance and crack-repairability. Along with the proposed modular assembly method to scale up the proposed carbon-reducing composites, the proposed paradigm may find broad applications in future engineering structures where both structural reliability and carbon footprint should be synergistically considered.

Methods

Materials

Clear V4, a transparent 3D-printing resin, was sourced from Formlabs. A CO₂ generator from WuyouChy generated gaseous CO₂ for the project. The acetone and calcium chloride were purchased from Sigma-Aldrich. The platinum electrode was purchased from TizZo.

Preparation of 3D-printed structure

To fabricate the 3D-printed structure, we employed Clear V4 resin in the stereolithography process using a Form 3 + 3D printer. A computer-aided design (CAD) model was sliced into sequential images projected onto a liquid basin, with the exposed resin solidified by light exposure to form a bonded solid layer on the printing stage. By repeatedly lifting the printing stage to a prescribed height and illuminating the resin with another slice image, additional layers were printed and bonded onto the previous layers until the entire structure was complete. Each layer was set at a thickness of 25 µm and took 83 s to solidify. Following the 3D-printing process, the resulting polymer was washed with acetone and air-dried for 24 h at 25 °C to remove any residual solvent.

Sputter Coating of 3D-printed structure

The 3D-printed polymer structures were coated with palladium using a Cressington 108 Manual Sputter Coater. After loading the sample into the coater and securing the top plate, the power was set to 30 mA. The chamber was evacuated to a pressure below 0.05 mbar. To eliminate residual oxygen and water vapor, the chamber was flushed with argon at 0.4 mbar for 20 s. Subsequently, the argon pressure was adjusted to 0.08 mbar, and the sputtering process was carried out for 20 s for one side. To achieve a uniform palladium coating, the sample was rotated to the other 3 sides and coated for another 20 s for each side.

Electrochemical manufacturing of mineralized composites

The coated 3D-printed structure served as the cathode, while a platinum electrode acted as the anode, with both immersed in the calcium chloride solution (1 mol/L) (Fig. S2). Two electrodes were then connected to a 6 V power supply, while the gaseous CO₂ was pumped through the calcium chloride solution. During the electrochemical reaction, the cathode generated hydroxide ions by electrolyzing water. The hydroxide ions reacted with the bicarbonate ions produced from the hydrolysis of CO₂ to form carbonate ions, which further reacted with calcium ions to produce calcium carbonate. Over 6 days, the calcium carbonate minerals were gradually precipitated out and firmly adhered to the internal surface of the cathode lattice.

Calculation of CO₂ uptaking

The CO₂ uptaking was calculated using the following equation²⁶,

$${{CO}}_{2}{Uptake}( % )=frac{{M}_{{rm{mineralized; composite}}}-{M}_{3{rm{D}}-{rm{printed; polymer}}}}{{M}_{3{rm{D}}-{rm{printed; polymer}}}}times frac{{m}_{{{rm{CO}}}_{2}}}{{m}_{{{rm{CaCO}}}_{3}}}times 100$$

where ({M}_{{rm{mineralized; composite}}}) and ({M}_{3{rm{D}}-{rm{printed; polymer}}}) were the measured mass of mineralized composites and 3D-printed polymer, respectively; and ({m}_{{{rm{CO}}}_{2}}) and ({m}_{{{rm{CaCO}}}_{3}}) were the molar mass of CO₂ and CaCO₃, respectively.

Calculation of carbon footprint of the mineralized composite

The carbon footprint of the mineral phase was calculated as:

$$begin{array}{l}{C{O}_{2}cdot rho }_{{rm{mineral}}}left({rm{kg}}/{{rm{m}}}^{3}right)=frac{{M}_{{rm{mineralized}}; {rm{composite}}}-{M}_{3{rm{D}}-{rm{printed}}; {rm{polymer}}}}{{M}_{3{rm{D}}-{rm{printed}}; {rm{polymer}}}}\qquadqquadqquadqquadqquad,,,times ,{rho }_{3{rm{D}}-{rm{printed}}; {rm{polymer}}}times ({C{O}_{2}}_{{{rm{CaCl}}}_{2}}times frac{{m}_{{{rm{CaCl}}}_{2}}}{{m}_{{{rm{CaCO}}}_{3}}}-frac{{m}_{{{rm{CO}}}_{2}}}{{m}_{{{rm{CaCO}}}_{3}}})end{array}$$

where ({M}_{{rm{mineralized; composite}}}) and ({M}_{3{rm{D}}-{rm{printed; polymer}}}) are the mass of the mineralized composite and 3D-printed polymer, ({C{O}_{2}}_{{{rm{CaCl}}}_{2}},)(around 38.6 g CO₂eq/kg⁵⁹) is the carbon footprint of calcium chloride per kilogram, ({m}_{{{rm{CO}}}_{2}}), ({m}_{{{rm{CaCl}}}_{2}}) and ({m}_{{{rm{CaCO}}}_{3}}) are the molar mass of carbon dioxide, calcium chloride, and calcium carbonate, ({rho }_{{{rm{CaCl}}}_{2}}) is the density of calcium chloride and ({rho }_{3{rm{D}}-{rm{printed; polymer}}}) is the bulk density of 3D-printed structural polymer (see Supplementary methods for the calculation method). Plugging the carbon footprint of the polymer phase⁶⁰ and the consumed electricity in California, the total carbon footprint of the mineralized composite can be written as:

$$begin{array}{ll}{C{O}_{2}cdot rho }_{{rm{total}}},left({rm{kg}}/{{rm{m}}}^{3}right)={C{O}_{2}}_{{rm{electricity}}}{times E+C{O}_{2}cdot rho }_{{rm{polymer}}}times frac{{rho }_{3{rm{D}}-{rm{printed}}; {rm{polymer}}}}{{rho }_{{rm{polymer}}}}\qquadqquadqquadqquadquad+,frac{{M}_{{rm{mineralized}}; {rm{composite}}}-{M}_{3{rm{D}}-{rm{printed}}; {rm{polymer}}}}{{M}_{3{rm{D}}-{rm{printed}}; {rm{polymer}}}}times {rho }_{3{rm{D}}-{rm{printed}}; {rm{polymer}}}\qquadqquadqquadqquadquadtimes ,({C{O}_{2}}_{{{rm{CaCl}}}_{2}}times frac{{m}_{{{rm{CaCl}}}_{2}}}{{m}_{{{rm{CaCO}}}_{3}}}-frac{{m}_{{{rm{CO}}}_{2}}}{{m}_{{{rm{CaCO}}}_{3}}})end{array}$$

where ({C{O}_{2}}_{{rm{electricity}}}) denotes the carbon footprint of electricity per kilowatt-hour in California (0.22 kg CO₂eq/kWh⁶¹), (E) denotes the electricity quantity consumed per unit volume of mineralized composites in kilowatt-hours (60-100 ({{rm{kWh}}/{rm{m}}}^{3}) for our experiment), ({C{O}_{2}cdot rho }_{{rm{polymer}}}) denotes the carbon footprint of the polymer phase and ({rho }_{{rm{polymer}}}) denotes the density of polymer phases.

Characterization of mineralized composites

The mineralized composites were imaged with a Nikon camera, a Nikon microscope, a scanning electron microscope (Nova NanoSEM 450), and a Micro-CT instrument (Nikon/XT H 225 ST), respectively.

Mechanical testing of mineralized composites

Young’s modulus of the mineralized composite was measured via the compression test using an Instron mechanical tester (Fig. 2f) The flexural strength of the mineralized beam was measured via the three-point bending test using the Instron mechanical tester (Fig. 2g). The flexural strength was determined by using σ_max = 3SF_max/(2BW²)⁶², where S denotes the span length, B represents the width of the sample, and W represents the depth of the sample. The shear strength of the coral-like mineralized composites was measured via unnotched two-section compact shear tests (Fig. 3g, S6a). The shear strength was calculated as τ_max = F_max/(S₁ + S₂), where F_max was the maximum load under the compact shear test, and S₁ and S₂ were the areas of two sections under shear loads. The shear fracture toughness was measured via notched single-section compact shear tests with an initial crack depth of 1 mm (Fig. 3h, +Fig. S6b). The shear toughness was calculated by ({G}_{{rm{II}}}=,frac{{F}_{max }^{2}}{2B}frac{partial C}{partial a})^63,64, where B was the thickness, a was the crack length, F_max was the critical load, and C was the compliance. Compliance values were found from the inverse slope of the load versus displacement curves of the shear fracture toughness tests. Tensile and shear tests on two assembled units (Fig. 5g) and concentrated loading on the 16-unit plate (Fig. 5k) were carried with the Instron mechanical tester.

Selected mineralization

A rectangular polymer lattice structure was initially designed using computer-aided design (CAD) software and subsequently 3D-printed with Clear-V4 resin. During the palladium coating process using a Cressington 108 Manual Sputter Coater, the central part of the structure was masked to ensure that only the ends were coated. These two ends were then connected to a voltage source to function as a cathode, resulting in mineralization occurring solely at these sites during the electrochemical process (Fig. 4a, b).

Reversibility of electrochemical mineralization

A mineralized composite sample was submerged in a hydrochloric acid solution (1 mol/L) for 10 min (Fig. 4d, e), and the resulting demineralized lattice was then re-mineralized via the same electrochemical methods (see Electrochemical manufacturing of mineralized composites for the manufacturing method) (Fig. 4f, g).

Fire resistance test

A mineralized sample was placed on fire for 30 min and then tested via the unnotched two-section compact shear test with the Instron mechanical tester (Fig. 4h–j).

Crack-healing

We took a mineralized composite and broke it into two segments. These two segments were later fastened together while being connected to a voltage of 6 V to function as a cathode, resulting in mineralization occurring at the interface over 6 days (Fig. 4l, S9).

Finite element simulation of two assembled units

The finite element simulation of the two assembled units was carried out using ABAQUS. The simulations utilized 3D deformable element types. The total number of elements used was 172,332. The material properties were specified as follows: Young’s modulus of 400 MPa and Poisson’s ratio of 0.3. The result accuracy were verified by refining the mesh.