Introduction

Organogels are porous materials composed of a three-dimensional (3D) network encapsulating the liquid molecules1,2. Despite their potential in diverse applications such as adhesives1,2,3, signal sensing4,5, impact protection6,7, soft electronics8,9,10, and biomedical engineering11,12, organogels usually exhibit low strength and poor stability, mainly due to their weak loose network and the mismatch between the network and the solvent13,14,15. Moreover, most of the reported organogels require pre-preparation and cannot be regulated via in situ polymerization16,17.

To overcome these obstacles, rationally designing small-sized polymeric network and bulky solvents for the preparation of sterically hindered organogels (SHOs) based on the steric hindrance effect is a promising strategy1,2,6,7. For instance, universal bio-based adhesives and robust underwater adhesives with enhanced toughness and underwater hydrophobicity were developed using a highly crosslinked polyurethane network capturing epoxidized soybean oil or epoxidized linseed oil molecules by virtue of the steric hindrance effect1,2. Furthermore, a bulky solvent with dynamic responsiveness was designed and encapsulated in a polyurethane network, resulting in anti-impact SHOs6. To establish a stable steric hindrance effect for the preparation of SHOs, the mesh size of the polymer network should be smaller than the size of the solvent. Therefore, the previously reported SHOs are basically designed based on bulky solvents1,2,6,7. However, small solvents are more beneficial for practical applications because they allow adjusting the viscosity, curing speed, and toughness18,19. Unfortunately, the development of SHOs with small solvents is challenging because the mesh size of the encapsulating networks should be smaller than that of conventional polymer networks. Actually, the smaller mesh size means the network possesses a higher crosslinked density, leading to stronger limitation to inside solvent, higher mechanical strength and better stability.

Hypercrosslinked polymers (HCPs) are amorphous microporous 3D network materials constructed by covalent connections between the organic components20. Owning to their tunable porous structure, customizable versatility, and high specific surface area, HCPs exhibit a wide range of practical and potential applications, such as gas sorption and separation21,22, heterogeneous organic transformation catalysis23,24,25, pollutant treatment26,27, and element collection28,29. However, most HCPs are brittle solids or powders, and mainly interact with gas molecules or ions, and remaining unexplored for capturing solvent molecules. Compared with conventional polymeric networks, HCPs possess higher crosslinking density and better strength, which render them a promising choice for locking small solvents.

In this study, a hypercrosslinked polymeric network design was proposed for capturing small bio-based molecules via the steric hindrance effect to fabricate highly permeable, instantly forming and mechanically robust polyurea organogels (PUAOs). To this end, acetyl tributyl citrate (ATBC), a low-viscosity, inexpensive, branched small molecule synthesized from renewable citric acid30, was used to regulate the viscosity, curing rate and mechanical strength of the resulting PUAOs. For the network design, we envisioned that using siloxane to prepare hypercrosslinked network through sol–gel chemistry would be a valuable strategy, since the resultant organic–inorganic covalent network exhibits small mesh size, robust strength, and toughness31,32,33,34. Therefore, hyperbranched amine–oligosiloxane (HPSi-NH2) with high functionality and strength was first synthesized via the hydrolysis and condensation reaction of siloxanes. Subsequently, the isocyanate–amine curing reaction between HPSi-NH2 and hexamethylene diisocyanate trimer (HDIT) was conducted to obtain a hypercrosslinked polyurea network, which trapped the ATBC molecules via steric hindrance effect. The reasonable combination of the hypercrosslinked polyurea network with the ATBC molecules endows the resulting PUAOs with low viscosity, high permeability, instant formation, and robust mechanical properties. We determined the sizes of the hypercrosslinked polyurea network and ATBC molecule, and examined the fundamental properties, environmental adaptability, stability, and potential applications of the PUAOs. The results demonstrate that this approach for the design of a hypercrosslinked polymeric network for trapping small solvents efficiently affords multifunctional SHOs with diverse potential applications for rust coating, concrete repair, and low-carbon concrete manufacturing. Considering the above applications utilizing the in situ formation process and the great effects of small solvents, this study could offer perspectives for the application of gel materials.

Results

Design principle of sterically hindered PUAOs

To fabricate highly permeable, mechanically robust and stable organogels, rational solvent selection and network design are critical because the viscosity of the system is largely determined by the solvent, the strength of the gel is mostly determined by the structure of the polymeric network, and the stability depends on the steric hindrance effect between network and solvent13,14,15. Therefore, acetyl tributyl citrate (ATBC), a branched small molecule that can be prepared from biomass citric acid, was selected as the solvent for polyurea organogels (PUAOs) owing to its low viscosity and good plasticizing effect (Supplementary Fig. 1). For the network design, polyurea is commonly employed due to its robust mechanical properties and strong affinity. Nevertheless, the mesh size of conventional polyurea is too large to encapsulate small molecules. To increase the crosslinked density of the polyurea network and decrease its mesh size, the hydrolysis and condensation reactions of amine-functionalized silanes and alkoxysilanes was rationally used to synthesize hyperbranched amine-oligosiloxane (HPSi-NH2) with high functionality and strength (Fig. 1a). The successful synthesis of HPSi-NH2 was confirmed via nuclear magnetic resonance spectroscopy (Supplementary Fig. 2). The HPSi-NH2 exhibited a number-average molar mass (Mn) of ~ 1860 g/mol and its polydispersity index was ~1.06, showing the narrow dispersity feature (Supplementary Fig. 3). The FTIR spectra indicated that residual Si-OH and Si-OEt in HPSi-NH2 were few due to its high degree of hydrolysis and condensation (Supplementary Fig. 4). Then, HPSi-NH2 was reacted with hexamethylene diisocyanate trimer (HDIT) to prepared the polyurea network. Figure 1a shows the preparation of HPSi-NH2 and the chemical structures of HPSi-NH2, HDIT, and ATBC. Figure 1b demonstrates the schematics of formation process and micro structure of PUAOs. When HPSi-NH2, HDIT, and ATBC are mixed together, a hypercrosslinked polyurea network is established based on the isocyanate-amine curing reaction between HPSi-NH2 and HDIT. Meanwhile, the ATBC molecules are encapsulated within the network via the steric hindrance effect.

Fig. 1: Design principle of sterically hindered polyurea organogels (PUAOs).
figure 1

a Preparation of hyperbranched amine-oligosiloxane and chemical structures of HPSi-NH2, HDIT, and ATBC. b Schematic description of the preparation and network structure of the PUAOs. c Optimized ball-and-stick molecular model of the crosslinked unit of the hypercrosslinked polyurea network. d Storage modulus of the rubbery plateau region and the corresponding calculated mesh size of the polyurea network. e Distribution of the mesh sizes calculated using DMA measurement and MD simulations. f Conformations and sizes of ATBC at different moments during the 90-ps MD simulation. g Distribution of the arithmetic and geometric mean radii of ATBC molecules.

Full size image

Obtaining sterically hindered organogels (SHOs) requires the precise size matching between the mesh of the polymeric network and the solvent molecule1,2,6,7. Therefore, we examined the sizes of the mesh and the ATBC molecule. To calculate the mesh size of the polymer network, different numbers of crosslinked units were constructed using HPSi-NH2 and HDIT, which was then optimized via a 150-ps molecular dynamics (MD) simulation to accurately simulate the real network structure (Supplementary Figs. 5–7, and Supplementary Data 1). Following the MD optimization, the edge sizes and average sizes of the crosslinked units were calculated (Supplementary Fig. 8, Supplementary Table 1, and Supplementary Data 1). Figure 1c demonstrates the optimized ball-and-stick molecular model of the crosslinked unit of the hypercrosslinked polyurea network. The real size of the network mesh was also assessed via dynamic mechanical analysis (DMA) using samples synthesized with HPSi-NH2 and HDIT in strict stoichiometric ratios1,2,6,7. The changes in storage modulus (E′), loss modulus (E′′), and loss coefficient (tanδ) as a function of temperature for the polyurea network were measured (Supplementary Fig. 9). According to the DMA results, the length of the chain segments between the crosslinking sites (Dm) were determined to range from 1.13 to 1.18 nm (Fig. 1d). Figure 1d exhibits the distribution of the calculated mesh sizes based on the DMA test and the MD simulations. The mesh size calculated using a single crosslinked unit ranged from 1.00 to 1.05 nm, with the average value being ~1.03 nm. The average mesh sizes calculated using two and four crosslinked units were ~1.07 and ~1.12 nm, respectively, whereas the average DMA-calculated Dm was ~1.17 nm. The DMA-calculated mesh size is larger than the MD-caculated size, indicating that the measured crosslinking density of the network is lower than the theoretical value, which is due to possible network defects and the incomplete reaction of the monomers. Note that the mesh size represents the length of chain segment between crosslinking points, which is determined by the structures of (3-aminopropyl) triethoxysilane and HDIT rather than the structure of HPSi-NH2. A 90-ps MD simulation was also conducted to analyze the conformation and compute the size of the ATBC molecule (Supplementary Fig. 10, Supplementary Data 1). Then, the stereoscopic sizes and molecular radii of gyration (Rg) of the ATBC molecules in different conformations were calculated (Supplementary Fig. 11, Supplementary Table 2, and Supplementary Data 1). As depicted in Fig. 1f, the sizes of the ATBC molecules consistently remained within a certain range despite the movement of the chain segment. Figure 1g displays the approximate distribution of the arithmetic and geometric mean radii of the ATBC molecules. The Rg of ATBC across different conformations varied between 6.03 and 6.47 Å, with a statistical average of 6.25 Å. This indicates that ATBC molecules exhibit molecular dimensions ranging from 1.20 to 1.29 nm, with an average diameter of approximately 1.25 nm, confirming that the molecular size of the ATBC is larger than the mesh size of the hypercrosslinked polyurea network formed by HPSi-NH2 and HDIT. Additionally, the ester groups in the ATBC molecule can form various hydrogen bonds with urea groups in the polyurea network (Supplementary Fig. 12). Therefore, the ATBC molecules can be encapsulated within the network during the curing process via the steric hindrance effect to form SHOs, which were designated as PUAO-x, where the ‘A’ is included to distinguish between polyurea (PUA) and polyurethane (PU), x denotes the weight percentage of ATBC. The mass content of each component in the PUAOs was given in Supplementary Table 3.

Basic characterizations and mechanical properties of PUAOs

Considering that the introduction of solvents can reduce the viscosity and extend the curing time of the system, the curing process of the PUAOs were investigated. Figure 2a shows the viscosity changes of the PUAOs during the curing process. As the ATBC content increased, the viscosity growth rate decreased, taking a longer time to reach a viscosity of 10,000 mPa·s. Figure 2b, c exhibit the initial viscosity and gelation time of the PUAOs. The initial viscosity of PUAO-0 was about 3452 ± 148 mPa·s, and its gelation time was about 10 s due to the extremely fast reaction of amine and isocyanate. The initial viscosities of PUAO−10, PUAO-30, PUAO-50 and PUAO-70 were approximately 1578 ± 58, 487 ± 33, 235 ± 17, and 155 ± 10 mPa·s, respectively. As the ATBC content increased, the initial viscosity gradually decreased, indicating that ATBC effectively reduces the viscosity. The gelation times of PUAO−10, PUAO-30, PUAO-50 and PUAO-70 were approximately 20 ± 3, 105 ± 6, 275 ± 8, and 720 ± 10 s, respectively. As the ATBC content increased, the gelation time gradually increased, indicating that ATBC could slow down the curing rate. Figure 2d demonstrates the change in the state of PUAO-50 during a curing process of 300 s. The evolution of contact angles with time of different PUAOs on concrete indicated the significant differences in the viscosity evolution and wetting ability of different PUAOs (Fig. 2e, Supplementary Fig. 13). With the content of ATBC increasing, the contact angles at the same time of different PUAOs reduced and the change rates of contact angles increased, which can be attributed to the effects of ATBC for reducing viscosity, slowing curing and enhancing wettability. Moreover, the static contact angles (60 s) of the PUAOs on different substrates were also examined (Supplementary Fig. 14). As the ATBC content increased, the contact angles gradually decreased, indicating that ATBC could enhance the wetting and spreading properties. Figure 2f shows the images of a PUAO-50 droplet rapidly penetrating into standard sand and a rusty steel plate within 60 s, indicating that the PUAO-50 fluid exhibits good permeability. Figure 2g displays the image of the PUAOs penetrating into standard sand. The penetration depths gradually increased with the increasing ATBC content, revealing that ATBC could enhance the permeability.

Fig. 2: Characterizations and mechanical properties of the polyurea organogels (PUAOs).
figure 2

a Viscosity changes during the curing processes of the PUAOs. b Initial viscosity of the PUAOs. c Gelation time of the PUAOs. d State change of PUAO-50 within 300 s. e Contact angles of PUAO droplets on concrete surface. f Rapid penetration of a PUAO-50 droplet into standard sand and a rusty steel plate within 60 s. g Penetration depths of the PUAOs in standard sand. h Macroscopic state (1 grid: 1 cm) and SEM images of the PUAOs. i WAXS characteristic curves and j SAXS and WAXS scattering images of PUAO-0 and PUAO-50 samples. k Compressive, l flexural, and (m) tensile curves of the PUAOs. n A tensile sample of PUAO-50 lifting up a weight of 2 kg without failure. The centre of error bars and the error bars respectively correspond to the mean value and standard deviation of 3–6 measurements for each analysis.

Full size image

Figure 2h illustrates the macroscopic states of PUAOs and their microstructure scanning electron microscope (SEM) images. As the ATBC content increased, the milky white samples gradually changed from translucent to opaque, indicating that the introduction of ATBC changes the gel structure. As for the microstructures, upon increasing ATBC content from PUAO-0 to PUAO-30, nanoscale vesicles progressively appeared. Similar fuller nanoscale vesicles were observed on the micro surface of PUAO-50. Meanwhile, the micro surface of PUAO-70 became smooth again, possibly due to phase inversion. The microstructures of PUAO-0 and PUAO-50 were analyzed using SAXS and WAXS to ascertain the presence of microscopic phase separation1,2,7,35,36. The SAXS curves revealed that both PUAO-0 and PUAO-50 are highly homogenous on the 1–10 nm scale, exhibiting no microphase separation (Supplementary Fig. 15). PUAO-0 and PUAO-50 exhibited highly similar WAXS curves, with scattering signals attributed to the butyl citrate and triazine ring structures, respectively (Fig. 2i). The SAXS and WAXS images of PUAO-0 and PUAO-50 were highly consistent, confirming the absence of microphase separation (Fig. 2j). The SAXS and WAXS scattering analyses illustrated that the ATBC molecules were uniformly encapsulated within the hypercrosslinked polyurea network and integrated at molecular level with the polymer.

Figure 2k–m illustrate the mechanical properties of PUAOs after curing for 24 h. The compressive strengths and strains of PUAO-0 (21.8 MPa, 12.7%), PUAO−10 (17.8 MPa, 18.5%), PUAO-30 (15.7 MPa, 27.4%), PUAO-50 (10.7 MPa, 36.8%), and PUAO-70 (2.3 MPa, 58.3%) are presented in Fig. 2k. With increasing ATBC content, the strength of the PUAOs decreased, whereas the strain increased, which can be attributed to the fact that the ATBC molecules do not partake in the crosslinking reaction but rather fill the mesh of the polymeric network, thereby enhancing the toughness. The flexural and tensile properties of the PUAOs (Fig. 2l, m, Supplementary Table 4) exhibited a similar trend to those of their compression properties. Note that the residual Si-OH and Si-OEt of HPSi-NH2 could continue to condense and affect the curing process and final properties of polyurea organogels. To exclude the point, PUAO-50 samples prepared by 15-d HPSi-NH2 and the PUAO-50 samples placed for 15 d after prepared were examined (Supplementary Fig. 16). They demonstrate similar curves to that of control samples, indicating that the residual Si-OH and Si-OEt in HPSi-NH2 slightly affect the curing process and final properties of PUAOs. This is understandable that the residual Si-OH and Si-OEt in HPSi-NH2 were few and the hydrophobic ATBC can block water from entering the system, which would restrict the condensation of residual Si-OH and Si-OEt. Actually, the curing process and final properties of PUAOs are mainly depends on the reaction between HPSi-NH2 and HDIT since the reaction rate of amine and isocyanate is much higher than those of residual Si-OH and Si-OEt. Anyhow, by virtue of the hypercrosslinked polyurea network, the PUAOs exhibit substantially improved mechanical properties compared with those of most conventional gel materials. Figure 2n shows that a tensile sample of PUAO-50 with a thin neck of 5 mm × 2 mm could lift up a weight of 2 kg without fracture, reflecting that the PUAO-50 exhibits good strength and toughness.

Besides, the cyclic compression tests (10 cycles) under several stress thresholds (1, 3, and 5 MPa) have been conducted to investigate the resilience and dimensional stability of PUAOs, taking PUAO-50 as a typical sample (Supplementary Fig. 17). All the compressive strains of PUAO-50 were unable to recover rapidly, and the strains were gradually increased and finally up to 2-3 times of the initial strains as the number of cycles increased. The fact indicated that PUAOs have poor instant resilience and dimensional stability. This is probably due to the rigidity caused by highly crosslinked network structure and the hysteresis caused by ATBC solvent. However, after a longer time, the compressed samples under different stresses finally recovered back to initial state, and the recovered time increased with the stress thresholds (from 5 min to 12 h). In particular, the final mechanical strength of the recovered PUAO-50 samples changed slightly even the stress thresholds and recovered time were different. The result clearly demonstrated that the PUOA-50 had a certain degree of resilience and dimensional stability when the recovered time is enough.

The thermogravimetric analysis (TGA) was conducted to characterize the thermal stability and thermal decomposition behavior of PUAOs (Supplementary Fig. 18). The PUAOs underwent varying degrees of pyrolysis over 200 to 300 °C, and the mass loss and its rate exhibited positive correlation to the contents of ATBC. As the temperature further increased, the mass of PUAOs decreased continuously, and finally reached the minimum values. The temperature of entering the final platform and the residual mass fraction both gradually decreased with the content of ATBC increasing. This is understandable that the thermal decomposition temperature of ATBC is lower than the polyurea network. Note that residual mass fraction exhibited positive correlation to the contents of polyurea network, attributing to inorganic Si-O-Si core from HPSi-NH2.

Considering the initial viscosity, gelation time, mechanical strength and toughness, and thermal stability, PUAO-50 was selected for subsequent experiments.

Environmental adaptability and stability of PUAO−50

Existing gel materials require pre-preparation and commonly suffer from low strength and poor stability, which limit their practical applications13,14,15. Considering the universal adaptability of the isocyanate-amine curing reaction and the high crosslinking density of the hypercrosslinked polyurea network, PUAO-50 may possess environmental adaptability and robust mechanical properties. To confirm this hypothesis, the compressive strength of PUAO-50 after curing for different times under typical temperatures was examined (Fig. 3a). When curing at −20 °C for 3 h, the compressive strength of PUAO-50 was 2.72 ± 0.28 MPa. As the curing time increased to 24 h, the compressive strength increased to 6.03 ± 0.21 MPa. Meanwhile, curing at 0 °C for 3 h, resulted in a compressive strength of 3.30 ± 0.13 MPa, which increased to 8.34 ± 0.24 MPa as the curing time increased to 24 h. When curing at 25 °C for 3 and 24 h, the compressive strength of PUAO-50 was 3.78 ± 0.22 and 10.73 ± 0.26 MPa, respectively. Curing at 40 °C for 3 and 24 h led to a compressive strength of 8.86 ± 0.25 and 10.85 ± 0.22 MPa, respectively. These results indicate the applicability of PUAO-50 over a wide temperature range because of the high reactivity of amine and isocyanate.

Fig. 3: Environmental adaptability and stability of PUAO-50.
figure 3

a Compressive strength of PUAO-50 after curing at several typical temperatures for different times. b Image displaying the underwater injection and rapid solidification of PUAO-50. c Compressive strength of PUAO-50 after curing under different water conditions for different times. d Images of pressure stability test of PUAO-50. e Mass changes of PUAO-50 after being subjected to varying pressures for 1 h. f Mass variations and (g) compressive curves of PUAO-50 samples maintained at different temperatures for 24 h. h Mass variations and i compressive curves of PUAO-50 samples after placing in different media for 24 h. The centre of error bars and the error bars respectively correspond to the mean value and standard deviation of 3–6 measurements for each analysis.

Full size image

The applicability of PUAO-50 under water conditions was also evaluated. Figure 3b displays the underwater injection and solidification of PUAO-50. The PUAO-50 fluid was injected into an underwater Olympic ring mold using a double-barrel syringe. After 10 min, the fluid was solidified and became white. Moreover, the compressive strength of PUAO-50 after curing for different times under different water conditions were measured. As demonstrated in Fig. 3c, the compressive strength of PUAO-50 cured under wet condition, underwater, and in seawater was basically consistent, indicating that PUAO-50 could be used under various water conditions. As the curing time increased from 3 to 24 h, the compressive strength increased from about 3.55 ± 0.25 to 10.47 ± 0.28 MPa, which is similar to the strength achieved when curing at 25 °C in air. These results illustrate that the curing process of PUAO-50 is only slightly affected by water, which could be attributed to the high hydrophobicity of ATBC molecules, as well as the tolerance to water of reaction between amine and isocyanate. These results prove that PUAO-50 performs environmental adaptability, and can be applied under various conditions.

High temperatures and high pressures pose serious challenges to the stability of organogels. Under these conditions, the network may deform and the mesh sizes can change, potentially resulting in solvent leaching from the system13,14,15. As the device and experimental method depicted in Fig. 3d, the pressure stability of PUAO-50 was evaluated by monitoring the mass variations upon continuous compression for 1 h. As given in Fig. 3e, the mass reduction of PUAO-50 after compression at 1 MPa stress for 1 h was 0.02%. As the pressure increased to 2, 3, 4, and 5 MPa, the mass reductions increased to 0.13%, 0.58%, 1.11%, and 1.33%, respectively. These results suggest that the ATBC molecules remained firmly entrapped within the network even under high pressure and considerable deformation (Supplementary Fig. 19). The high strength, modulus, and excellent anti-deformation properties of the hypercrosslinked polyurea network and the silane nanoclusters contribute to maintaining a stable mesh size compatible with ATBC. The temperature stability of PUAO-50 was assessed by measuring the changes in its weight and strength after exposure to different temperatures for 24 h. As depicted in Fig. 3f, the mass losses of PUAO-50 after being subjected to in 50, 75, and 100 °C for 24 h were below 2% (0.88%, 1.41% and 1.82 %, respectively), indicating its exceptional temperature stability. Moreover, the compressive curves of PUAO-50 recorded at 25, 50, 75, and 100 °C for 24 h were consistent (Fig. 3g). These results reveal the good temperature stability of PUAO-50, which can be attributed to the high temperature resistance of the hypercrosslinked polyurea network and the silane nanoclusters and the settled steric hindrance effect.

In addition to high temperatures and high pressures, solvent immersion is a common scenario encountered in practical applications. Small solvent molecules may attack and invade the gels network, adversely affecting their stability2,7,13,14,15. The stability of the polymer gels was investigated by analyzing the changes in their masses, appearance and strength after immersion in water, ethanol, and ATBC as typical solvents for different durations. As illustrated in Fig. 3h, the PUAO-50 samples basically maintained their mass with little increase (less than 1.5%) after being soaked in water, ethanol, and ATBC for 24 h. Figure 3i illustrates that the compressive strength of the PUAO-50 samples immersed in various solvents were remarkably similar, approximately 10.5 MPa, indicating that the migration of solvent molecules was considerably restricted because the ATBC molecules within the network firmly occupied the mesh via the steric hindrance effect1,2. The environmental adaptability and stability test results confirm the properties of PUAO-50 under diverse conditions. Motivated by the comprehensive performance of PUAO-50, we explored its application scenarios.

Application of PUAO-50 as a low-surface-treatment coating for rusty steel

The subsequent corrosion protection of rusty steel is a common industrial challenge. The rust on the surface absorbs water and oxygen from the air, further causing severe corrosion to the inner substrate. This issue is usually addressed by polishing the surface to remove rust, and then painting an anticorrosion protective, which is a complex and time-consuming procedure. Considering the low viscosity, high permeability, and rapid solidification of PUAO-50, it was envisioned to be used as a permeable coating for rusty steel.

Rusty steel plates with different corrosion levels were obtained by conducting salt spray experiments for different times (Supplementary Fig. 20). Subsequently, the rusty steel plates were coated with PUAO-50. As shown in Fig. 4a, the D-level rusty steel plate coated with PUAO-50 exhibits a white surface and the hook adhered on PUAO-50 surface (adhered surface area: ~7.0 cm2) could lift a weight of 20 kg without detachment, indicating that PUAO-50 could be used as a coating to firmly combine with rusty steel. Figure 4b shows the adhesion strength of PUAOs on rusty steel with different levels. The adhesion strength of PUAO-50 was about 2.46 ± 0.19 MPa on the not-rusty steel plate, and about 2.59 ± 0.19, 2.64 ± 0.27, 2.70 ± 0.24, and 2.59 ± 0.37 MPa on the rusty steel plates with A, B, C, and D corrosion levels. The higher adhesion strength on the rusty steel plates can be attributed to that low-viscosity PUAO-50 not only forming interfacial interactions with steel, but also penetrating into the porous rust to form a robust mechanical interlock.

Fig. 4: Application of PUAO-50 as a low-surface-treatment coating for rusty steel.
figure 4

a Images displaying that PUAO-50 could be directly coated on a rusty steel plate and lift up a 20-kg kettle bell. b Adhesion strength of PUAO-50 on steel with different corrosion levels (note that N means not-rusty). c SEM and EDS images of the combining section of PUAO-50 and rust. d Bode plot, e Nyquist plot, f Tafel polarization curves and g fitting circuit diagram of EIS electrochemical impedance spectra of PUAO-50 coating. The centre of error bars and the error bars respectively correspond to the mean value and standard deviation of 3–6 measurements for each analysis.

Full size image

To investigate the permeability of PUAO-50 on rust, the cross-sectional micromorphology of a sample was observed using SEM and energy-dispersive spectroscopy (EDS). As shown in Fig. 4c, the SEM images display that the cross-sectional area can be divided into three parts: steel substrate, the combination of PUAO-50 and rust, and pure PUAO-50 coating, indicating that the PUAO-50 was in close contact with the rust and steel substrate. The SEM images display that the cross-section can be divided into three parts: steel substrate, combination of PUAO-50 and rust, and pure PUAO-50 coating, indicating the PUAO-50 was closely contacted with rust and steel substrate. The EDS spectrograms of Fe and Si elements further prove that PUAO-50 penetrated into the porous rust and formed a mechanical interlock.

To further evaluate the applicability of PUAO-50 as an anticorrosion coating, electrochemical impedance spectroscopy (EIS) tests were conducted to investigate the corrosion resistance properties of the PUAO-50 coating. PUAO-50 was coated on a rusty steel, which was then sealed with paraffin wax and soaked in a 3.5 wt.% NaCl solution for different times. Then, EIS was performed. As illustrated in the Bode plot (Fig. 4d), PUAO-50 exhibits a high impedance modulus of 108 Ω·cm2 in the low frequency range, indicating its good corrosion resistance. As the soaking time increased, the impedance modulus of PUAO-50 in the low-frequency range decreased slightly, implying that its corrosion resistance remained basically stable. In the Nyquist plot (Fig. 4e), the capacitor semicircle of PUAO-50 shows a large diameter, illustrating its exceptional corrosion resistance. The diameter of the capacitor semicircle shows a small decline with increasing soaking time, implying that PUAO-50 can effectively seal the pores of rust and form a dense protective layer. The high anticorrosion performance of PUAO-50 can be attributed to that the high hydrophobic ATBC molecules and the silane nanoclusters of polyurea network can efficiently block corrosive ions in water. The Tafel polarization curves shows the changes of corrosion current density and corrosion potential with soaking time for PUAO-50 coating (Fig. 4f). With the soaking time increased, the corrosion current density decreased and the corrosion potential increased, demonstrating good and stable corrosion protection effect. The fitting circuit diagram of EIS electrochemical impedance spectra of PUAO-50 coating was shown as the equivalent circuit model of a typical electrochemical system consisting of three main parts: electrolyte, coating and substrate (Fig. 4g), indicating that PUAO-50 have passed through the surface rust and closely contact with bottom steel substrate.

Application of PUAO-50 as a strong adhesive to repair concrete

Adhesives have a wide range of applications, from everyday life to industrial uses37,38,39,40,41. Gel materials are emerging as high-performance adhesives in different applications. Considering the high strength and stability of the hypercrosslinked polyurea network, the favorable adhesion effect of the high-density urea bonds, and the hydrophobic effects of the ATBC molecule, the application of PUAO-50 as a versatile adhesive was explored (Supplementary Figs. 21, 22). Both in air and underwater, PUAO-50 exhibits high adhesive strength to various substrates, which can be attributed to its abilities to cure under diverse conditions and to form various interactions with the substrate surface, such as mechanical interlock, chemical crosslinks, hydrogen bonds, hydrophobic interactions, and van der Waals interactions.

Compared with conventional adhesives, PUAO-50 exhibits low viscosity and in-situ solidification, which render it potentially applicable to repair broken concrete. Cracks with a length of 20 cm and different widths were constructed on the surface of concrete using feeler gauges. Then, PUAO-50 fluid was poured over the cracks to explore the permeability of PUAO-50 to the concrete cracks. As shown in Fig. 5a, b, the penetration depths of PUAO-50 gradually increased with the crack width increasing. For cracks with a width of 0.2, 0.4, 0.6, 0.8, and 1.0 mm, the penetration depth of PUAO-50 was about 9.6 ± 0.5, 11.8 ± 0.4, 14.0 ± 0.4, 16.2 ± 0.3, and 17.5 ± 0.4 cm, respectively. These results prove that PUAO-50 possesses a good penetration effect on concrete cracks of different widths.

Fig. 5: Application of PUAO-50 as a strong adhesive to repair concrete.
figure 5

a Image of PUAO-50 permeating into concrete cracks with different widths. b Penetration depth of PUAO-50 into concrete cracks with different widths. c Adhesion strength of PUAO-50 to concrete under different conditions. d Images of repaired broken concrete using PUAO-50 doped with a fluorescent dye. e Images of repaired broken concrete holding a 1 kg weight. f Enlarged images of surface and internal cracks. g 3D CT reconstruction images of concrete, cracks, and PUAO-50 in a repaired sample. h 3D CT tomography images. i SEM image of a section of a repaired crack. The yellow dashed line represents the boundary between PUAO-50 and concrete. j Compressive curves and k compressive and flexural strengths of control and repaired concrete samples. l Compressive strength of concrete repaired after multiple times. The centre of error bars and the error bars respectively correspond to the mean value and standard deviation of 3–6 measurements for each analysis.

Full size image

The adhesive strength of PUAO-50 to concrete was also measured after curing for 24 h under different conditions (Fig. 5c). When curing in air, the adhesive strength of PUAO-50 to concrete was approximately 2.39 ± 0.18 MPa, whereas it was 1.88 ± 0.23 and 1.41 ± 0.27 MPa under wet and underwater conditions, illustrating that PUAO-50 could be firmly bonded with concrete. To visualize the repairing process, PUAO-50 was doped with a fluorescent dye. As shown in Fig. 5d, the green fluorescence indicates that PUAO was distributed throughout the cracks of the broken concrete. The repaired concrete could hold a weight of 1 kg without fracture (Fig. 5e), showcasing the robust repair effect of PUAO-50. Figure 5f presents enlarged photos of surface and internal cracks displaying widths of about 0.3 and 0.1 mm. To investigate the repair effect of PUAO-50 on the interior cracks of concrete, 3D computed tomography (CT) was conducted on a repaired sample of about 1 cm3 to detect the distribution of PUAO-50 in concrete cracks. The 3D refactored models of the restoration body, fractured concrete, and PUAO-50 (Fig. 5g) indicate that PUAO-50 could penetrate into subtle cracks as fine as 0.1 mm and cure for repair. The CT photos at different heights of the sample indicate that the width of the cracks decreased from 0.3 to 0.1 mm from top to bottom (Fig. 5h). The SEM image shows that PUAO-50 has penetrated deeply into and closely adhered to the concrete substrate, further validating the excellent repair effect of PUAO-50 on concrete (Fig. 5i).

In addition to these microscopic characterizations, the repair effect of PUAO-50 on concrete was also evaluated in terms of the mechanical strength of repaired concrete. According to the compressive curves, the strength and strain of a control sample of concrete were 30.2 MPa and 5.37%, respectively, while those of the repaired concrete were 27.5 MPa and 5.98% (Fig. 5j). Moreover, the flexural strength and strain of the control concrete sample were 4.56 MPa and 3.19%, respectively, while those of the repaired concrete were 4.28 MPa and 3.40% (Fig. 5k, Supplementary Fig. 23). These results highlight the exceptional repair efficiency of PUAO-50 on concrete. Moreover, PUAO-50 can be applied for repeated repair when the repaired concrete is damaged again. As the number of repairs increased, the compressive strength of the restoration gradually decreased, but still exhibited a robust strength exceeding 25.0 MPa (Fig. 5l), and the repair efficiency was higher than 80% (Supplementary Fig. 24).

Application of PUAO-50 in low-carbon concrete manufacturing

Emergency engineering problems, such as road emergency maintenance and rapid construction of buildings, often arise in daily life. Using conventional inorganic materials for repairing or building usually requires high material consumption, complex processes, and long curing times, resulting in heavy workload and high carbon emissions. Therefore, the repair materials with the low material consumption and high availability need to be urgently developed. In this context, organic materials are a promising alternative because they can be mixed with various aggregates, thereby reducing the dosage and carbon emissions. Owing to the low viscosity, fast curing, robust strength, and exceptional adhesion performance of PUAO-50, its applicability for constructing low-carbon concrete was evaluated.

Common inorganic aggregates, such as concrete fragments, gravel, brick rubble, soil, clay, coal ash, and river sand, and some special aggregates, such as desert sand, sea sand, and coral sand, could be consolidated into blocks by PUAO-50 at a usage of 20 wt.% (Fig. 6a, Supplementary Figs. 25, 26). The compressive strength of these blocks after curing for 24 h was examined (Fig. 6b, Supplementary Figs. 27, 28), finding values of 6.78 ± 0.16, 6.18 ± 0.10, 7.89 ± 0.25, 6.47 ± 0.14, 5.46 ± 0.17, and 6.13 ± 0.12 MPa for the desert sand, sea sand, gravel, brick rubble, soil, and coal ash samples, respectively. Correspondingly, the elastic moduli were 75.7 ± 2.9, 65.7 ± 2.5, 99.7 ± 4.4, 78.2 ± 5.2, 55.9 ± 2.6, and 60.5 ± 4.6 MPa. These results illustrate that PUAO-50 can be utilized to consolidate various aggregates to manufacture low-carbon concrete without selectivity, and the resulting blocks exhibit excellent compressive strength and high modulus. By virtue of this robust mechanical property, a large-sized desert sand consolidated sample with a size of 30 cm × 20 cm × 4 cm could withstand repeated rolling of a 2-ton car (actual pressure: ~ 196 kPa) for 20 times without deformation or fragmentation (Fig. 6c, Supplementary Fig. 29).

Fig. 6: Application of PUAO-50 for low-carbon concrete manufacturing.
figure 6

a Cube samples obtained by consolidating different aggregates with 20 wt.% PUAO-50. b Compressive strength and elastic modulus of the consolidated samples. c A large desert sand-based consolidated sample (30 cm × 20 cm × 4 cm) withstanding repeated rolling of a 2-ton car. d Compressive strength, elastic modulus, and morphology change of desert sand-based consolidated samples during continuous UV irradiation for 1000 h. e Engineering structures built by mechanically machining and assembling plate, column, and beam samples made of desert sand-based consolidated samples. f Directly formed complex-shaped desert sand-based consolidated samples. g Process of PUAO-50 consolidating standard sand underwater. h Underwater sand consolidated sample could support a 20-kg kettle bell without dispersion. The centre of error bars and the error bars respectively correspond to the mean value and standard deviation of 3–6 measurements for each analysis.

Full size image

To face the various challenges arising from complex environments during practical applications, the consolidated samples should possess various properties such as mechanical stability and ultraviolet (UV) resistance42,43. Therefore, UV-aging experiments were conducted on consolidated samples using desert sand as the typical sample, and the changes in the compressive strength and macroscopic state were monitored to examine the weather resistance. Negligible changes were observed upon UV irradiation for 1000 h (Fig. 6d, Supplementary Figs. 30, 31), demonstrating the excellent UV resistance of the consolidated samples, which can be attributed to the UV tolerance of the chemical groups of PUAO-50.

In practical applications, low-carbon concrete is usually constructed into specific engineering structural shapes. As depicted in Fig. 6e, the desert sand consolidated by PUAO-50 could be shaped into plate, column, and beam samples. Moreover, the samples could be further mechanically machined, such as machining holes on the plate and threads inside the columns, and the machined plate and columns could be assembled together using bolts to build an engineering structure. The desert sand could be also directly consolidated into complex shapes by PUAO-50, such as a pyramid, a bird nest and a Canton Tower (Fig. 6f). Considering its rapid curing property underwater, we also evaluated the ability of PUAO-50 for underwater sand consolidation. Figure 6g displays the process of PUAO-50 consolidating standard sand underwater. The PUAO-50 fluid was injected into water with a double-barrel syringe and then mixed with ssand by stirring. After curing for 10 min, the loose sand was reinforced into a block, which could support a 20 kg kettle bell without dispersion (Fig. 6h), indicating the high strength and excellent reinforcement effect of PUAO-50 underwater. After curing for 24 h, the underwater standard sand consolidation samples exhibited compressive strength and modulus of 7.06 ± 0.25 and 73.3 ± 3.8 MPa, which were similar to those of the samples cured in air for same time (Supplementary Fig. 32), further reflecting the underwater applicability of PUAO-50.

Discussion

In this study, a hypercrosslinked polymeric network design is proposed for capturing small bio-based molecules to fabricate SHOs. The small solvent ATBC considerably reduces the initial viscosity of the system, dramatically slow down the reaction rate, and substantially enhances the toughness. The hypercrosslinked polyurea network (~1.17 nm) formed by reacting HPSi-NH2 and HDIT encapsulates the ATBC molecule (~1.25 nm) via steric hindrance effect. By synergistically combining the hypercrosslinked polyurea network with ATBC molecules, the resultant PUAO-50 SHO demonstrates high permeation, instant formation (~5 min), and mechanical robustness (~10.7 MPa). Moreover, PUAO-50 exhibits good environmental adaptability and stability, resulting in fast curing and robust mechanical strength under various temperatures and water conditions, with only slight changes in quality and strength even under harsh environments of high pressures, high temperatures, and solvent immersion. PUAO-50 can penetrate into porous rust and rapidly cure, exhibiting favorable corrosion resistance performance, and into small concrete cracks, acting as a strong adhesive to efficiently repair concrete. Moreover, PUAO-50 can consolidate diverse aggregates to manufacture low-carbon concrete with mechanical robustness, demonstrating its suitability as a structural material for engineering applications. This study proposes an approach of designing hypercrosslinked network to entrap small solvents, develops a multifunctional SHOs, and provides insights on the application of gel materials.

Methods

Materials

(3-aminopropyl) triethoxysilane (KH550, 99%), methyltrimethoxysilane (MTMS, 98%), acetyl tributyl citrate (ATBC, 98%), and N, N-dimethylformamide (DMF, 99%) were purchased from Energy Chemical and used as received. Anhydrous ethanol (EtOH) and deionized water (H2O) were procured from Sinopharm and used as received. Hexamethylene diisocyanate trimer (HDIT) was purchased from Shandong Wanhua Chemical Group Co., Ltd. and used as received. The ratio of isocyanate group and amine was 1:1 during the preparation experiments. The aggregates such as desert sand, sea sand, gravel, etc. were purchased from Dongguan Calculus New Materials Technology Co., Ltd.

Synthesis of HPSi-NH2

KH550 (22.14 g, 0.10 mol), MTMS (13.62 g, 0.10 mol), H2O (5.40 g, 0.30 mol), and ethanol (80 mL) were mixed in three-neck flask with a magnetic stirrer. The solution was allowed to hydrolyze at 60 °C for 12 h, then it was concentrated under vacuum to remove volatiles. The obtained hyperbranched amine–oligosiloxane (20.40 g) was designated as HPSi-NH2.

Preparation of PUAOs

Series of PUAOs with different ATBC contents were prepared by follow steps: HPSi-NH2, HDIT and ATBC were homogeneously mixed according to the weight contents given in Supplementary Table 3 to obtain a mixture fluid. Subsequently, the fluid was poured into the PTFE mold and cured at room temperature (25 °C).

Nuclear magnetic resonance (29Si-NMR)

The 29Si-NMR spectrum of HPSi-NH2 was recorded on Bruker AV600 NMR spectrometer using CDCl3 as the solvent.

Gel permeation chromatography (GPC) test of HPSi-NH2

The molecular weight and polydispersity index (PDI) of HPSi-NH2 was determined via GPC using DMF as the eluent. The analysis was performed with a refractive index (RI) detector with two identical PL gel MIXED-C columns. The test was operated at a sample concentration of 5 mg/mL and a solvent flow rate of 1.0 mL/min. A series of polystyrene (PS) standards were used for calibration to obtain the apparent number-average molar mass (Mn) and PDI of the HPSi-NH2.

Fourier transform infrared spectroscopy (FTIR)

FTIR spectra were collected from a Bruker Vertex 70 FTIR spectrometer. The HPSi-NH2 was measured with KBr pellet samples in the range of 500 to 4000 cm–1.

Molecular dynamics (MD) simulation and size calculation of polyurea mesh

By using Materials Studio, different numbers of crosslinked units were constructed and then optimized through dynamic running algorithm. The simulation algorithm was Forcite--Calculation--Dynamics. The simulation was conducted at NVT ensemble at 298 K, while the forcefield was COMPASSII and charges were forcefield assigned. For time parameters, the timestep was 1 fs, the number of steps was 150,000, and whole duration was 150 ps. After that, the optimized crosslinked units were disassembled and the stereoscopic size of each edge were calculated by using the coordinates of all atoms and the van der waals radius of each kind of atom. The coordinates of atoms were obtained by opening mol file with ChemBio3D, and the van der Waals radius was taken from the Bondi version44.

Dynamic mechanical analysis (DMA) of polyurea samples

The dynamic mechanical analysis was conducted with a NETZSCH DMA 242 C. The samples were analyzed by tensile mode under a nitrogen atmosphere. The temperature sweep is from 0  to 150 °C with a heating rate of 5 °C/min at a constant frequency of 1 Hz. The test samples are 5 mm in width, 1 mm in thickness, and 15 mm in length.

Calculation of polyurea mesh size based on DMA

For the amorphous cross-linked polymers, the average spacing between cross-linked sites can be estimated from the dynamic mechanical properties. Firstly, the average molecular weight of the chain segments between the crosslinking sites (Mc) can be calculated using the following equation45:

$${Mc}=\frac{3{RT}\rho }{G^{\prime} }$$

Then, the number of the chain segments (Nc) and average occupied volume of chain segments (Vc) can be calculated using the following equation:

$${Nc}=\frac{m\cdot {NA}}{{Mc}},\,{Vc}=\frac{V}{{Nc}},\,\rho=\frac{m}{V}$$

Considering that the cube in space has 12 edges and each edge is shared by 4 cubes, the volume of the cross-linking unit can be taken as three times of \({Vc}\), and the mesh size can be calculated using the following equation:

$${Vm}=3V{{\boldsymbol{c}}},{Dm}=\root{3}\of{{Vm}}$$

By simplifying the above formula, it can be concluded that: \({{Dm}}=\root{3}\of{\frac{9{{\rm{RT}}}}{G{\prime} {{\rm{\cdot }}}NA}}\), where R denotes the gas constant, T indicates the temperature of the material when it is in the rubber plateau region (T is taken between Tg + 30  to Tg + 60 K), G’ is the energy storage modulus of the material at T, NA is Avogadro constant.

MD simulation and size calculation of ATBC

MD simulations were used to analyze the molecular conformations of ATBC molecule. The structural model of ATBC was established using Materials Studio. The condition of this simulation was as same as above MD simulations of mesh, and whole duration was 90 ps. Molecular size of ATBC, including stereoscopic size, arithmetic mean radius and geometric mean radius, was calculated in the same way as the above edges.

Curing tests of PUAOs

The initial viscosity of PUAOs fluid was measured using a HD-C801-2 rotational viscometer at 25 °C. The viscosity changes of PUAOs during curing process were obtained by measuring the viscosity at different times. The time when the viscosity rises to 10,000 mPa·s after mixing is taken as the gelation time.

Surface wettability measurement of PUAOs

The surface wettability of the PUAOs were measured by dropping 3 μl PUAOs fluid on different substrate surface at 25 °C and 55%RH using a contact angle goniometer (Biolin, Model KSV NIMA). Each average value was calculated from 5 different regions of samples.

Permeability measurement of PUAOs

The permeability of the PUAOs fluid were measured by pouring PUAOs fluid on standard sand with depth of 10 cm and measuring infiltration depth. The permeability of PUAO-50 fluid on concrete was measured by pouring PUAO-50 fluid on concrete cracks with different widths and measuring the infiltration depth.

Scanning electron microscopy (SEM) test

The SEM images of PUAO samples were taken with a Zeiss Merlin scanning electron microscope. Before the SEM analysis, the samples were vacuum dried for 1 h and gold sprayed.

Small-angle X-ray scattering (SAXS) and wide-angle X-ray scattering (WAXS) experiments

X-ray scattering techniques, including small-angle X-ray scattering (SAXS) and wide-angle X-ray scattering (WAXS), are extensively used to analyze the microstructures of soft materials such as polymers and gels36. Thus, SAXS and WAXS experiments were conducted with an Xenocs Xeuss 2.0 system. X-Ray wavelength: 1.54189 Å; Detector: Pilatus 3 R 300 K; Single pixel size: 172 μm; Sample detector distance: 1185 mm (SAXS) and 152 mm (WAXS); Exposure time: 300 s. calibrated by Silver Behenate; Experimental environment: room temperature vacuum; Collimation mode: High throughput mode. The data was analyzed in the XSACT software and corrected for background scattering from air and from the sample holder. The lengths of the scattering vector q were 0.007393 to 0.420835 Å−1 for SAXS, and were 0.071473 to 2.349391 Å−1 for WAXS.

Mechanical properties tests of PUAOs

The mechanical strength of PUAOs evaluated using a Sens C51.105 universal mechanical testing machine at a constant temperature of 25 °C and a humidity of 55%RH. The compressive test specimens were cubic in shape, each with a dimension of 20 mm. These tests were performed at a displacement rate of 5 mm/min. For assessing flexural strength, the samples were rectangular bars, each measuring 20 mm × 20 mm × 120 mm. During testing, a span of 100 mm was maintained. The flexural strength tests were conducted at a displacement rate of 2 mm/min. The tensile samples were dumbbell shaped with neck size of 80 mm × 10 mm × 4 mm. The tensile tests were conducted at displacement rate of 5 mm/min. At least 3 samples were measured for each condition.

Cyclic compression tests of PUAO-50

The cyclic compression tests of PUAO-50 were conducted on an Instron 5966 instrument with compression fixture (25 °C, 55%RH). The specimens were cubic with a dimension of 20 mm, and the loading and unloading processes were carried out at a rate of 400 N/min. The three selected stress thresholds were 1, 3, and 5 MPa, and 10 cycles were performed under each stress condition.

Thermogravimetric analysis (TGA) tests of PUAOs

The thermogravimetric analysis was conducted on the TG 209F1 from NETZSCH, Germany. The mass of the thermogravimetric analysis sample was about 5 mg, the testing temperature ranges from 25 to 825 °C with a heating rate of 20 K/min. The atmosphere is nitrogen atmosphere, and the gas flow rate is set to 20 mL/min.

Environmental adaptability tests of PUAO-50

The environmental adaptability of PUAO-50 was measured by testing the compressive strength of PUAO-50 after curing for different times under typical temperatures (−20, 0, 20, 40 °C) or different water condition (in air, 100%RH wet condition, underwater).

Stability tests of PUAO-50

The pressure stability of PUAO-50 was evaluated by monitoring the mass variations upon continuous compression for 1 h. The stability under high temperatures of PUAO-50 was measured by the changes of weight and compressive strength after placed at different temperatures for 24 h. The stability of PUAO-50 under solvent immersion were examined by investigating the changes in appearance, weight and strength of PUAO-50 after soaked in solvents (water, ethanol, ATBC) for 24 h.

Adhesion strength test of PUAO-50 on rusty steel

The adhesion tests were performed by an automatic tester (PosiTest, Model AT-A) according to ASTM D4541-09. The coating surface was polished lightly by 150-grit abrasive paper for rough polishing and 300-grit abrasive paper for fine polishing. The circular aluminum ingot with a diameter of 2 cm were adhered on the surface of the coating using epoxy adhesive. After complete curing, the aluminum ingot was pulled up in rate of 0.2 MPa/s. The average adhesion strength was calculated based on the adhesion strength of 4 different test areas on the coating.

Energy-dispersive spectroscopy (EDS) tests

After SEM test, the energy-dispersive spectroscopy test was conducted out to observe the elemental distribution of cross-sectional of the sample.

Preparation of rusty steel plates

First, rusty steel plates were prepared by conducting a salt spray test on a polished Q235 steel plate. The temperature of salt spray test was 35 ± 2 °C, the concentration of sodium chloride solution was 5.0%, the pH value was 6.5 to 7.2. After the salt spray test, the surface rust on the steel plate was manually removed and the rusty steel plates were cleaned with water and dried at 35 °C. According to the morphology and distribution area of the surface rust, the surface corrosion of steel plates can be divided into four levels (A, B, C, and D).

Electrochemical impedance spectroscopy (EIS) tests

EIS characterization of the anti-corrosion performance of coatings on rusty steel plates was conducted using Zennium electrochemical workstation (Zahner, Germany). When preparing the rusty steel plate sample for testing, the surrounding edges and uncoated part were sealed using adhesive tape, and the mixture of paraffin wax and rosin (mass ratio 1:1). After soaking in a 3.5 wt.% NaCl solution for different time, the samples were taken out and test the electrochemical impedance spectra. The test thickness is 100 μm and the test area is 1 cm2. The electrochemical workstation adopts a three-electrode system for testing, with a carbon rod as the counter electrode, a saturated calomel electrode as the reference electrode, and a rusted coating as the working electrode. The frequency testing range is 10−1 to 105 Hz, and the voltage amplitude is 100 mV. After the EIS test, the electrochemical workstation was set as the POL mode for Tafel polarization curve testing, and the open circuit voltage was range from −250 to 250 mV, with a scan rate of 10 mV/s.

Adhesive Strength Test of PUAO-50

The adhesive strength of PUAO-50 was determined through a lap-shear test, utilizing an Instron 5966 instrument. The adhesive area of test samples was set at 25 mm × 20 mm. The tests were conducted at a displacement rate of 25 mm/min. Samples were prepared in air (25 °C, 55%RH) to test adhesive strength in air. For underwater adhesive strength test, samples were prepared underwater (25 °C), and retrieved from their immersed medium before testing. At least 5 samples were measured for each condition. The lap-shear strength was calculated according to the following equation:

$$\sigma=\frac{F}{A}$$

where σ is lap-shear strength (MPa), F is the maximum loading force to break the joint (N), and A is the adhesive area (500 mm2).

Adhesion strength test of PUAO-50 on concrete

The adhesive strength of PUAO-50 on concrete was determined through a pull-out test, utilizing an Instron 5966 instrument. The adhesive area of test samples was set at 25 mm × 20 mm. The tests were conducted at a displacement rate of 5 mm/min. Samples were prepared in air (25 °C, 55%RH) to test adhesive strength in air. For wet (25 °C, 100%RH) and underwater conditions (25 °C), samples were prepared under corresponding conditions, and retrieved from their immersed medium before testing. At least 5 samples were measured for each condition.

3D computed tomography (CT) test

The 3D computed tomography technique can collect a series of projection data of the sample by tomography scanning, then distinguish different substances by calculating the attenuation coefficient of X-rays of each part, since the coefficient is closely relative to the equivalent atomic coefficient and density of a substance. Finally, the 3D structure reconstruction was completed based on lots of imaging planes46. The 3D CT test was conducted using a GE Vtomex tomography equipment. The sample size was about 1 cm × 1 cm × 1 cm. The distance between scanning faults was 1 μm, and scanning rate was 10 μm/min. After the CT test, the model was rebuilt based on the tomographic images.

Low-carbon concrete manufacturing

The aggregates were mixed with PUAO-50 to obtain mortar, the usage of PUAO-50 was 20 wt.% of the aggregate quality, and the mortar was modeled into blocks or other shapes. After curing in air for 24 h (25 °C, 55%RH), the mortar was transformed into low-carbon concrete. Then the block samples were tested for compressive strength. When the sand was placed underwater, the PUAO-50 was injected to mix with sand underwater. The underwater samples were conducted to compressive strength test after cured underwater (25 °C) for 24 h.

Ultraviolet (UV) aging test

The UV aging test was conducted using the UV accelerated aging testing machine (GT-QUV/spalay) from Q-Lab. The samples were cured in air for 24 h (25 °C, 55%RH), and then placed in a test chamber with a temperature set to 50 °C and a light intensity set to 0.83 (W·m−2)/nm for continuous UV aging test. The macro state and strength of the samples were tracked at regular intervals.