Atomic Insight into Durability and Interfacial Stability of Novel Hydrophobic Composite in Concrete Alkaline Environment for Marine Engineering

Ao Zhou , Kexuan Li , Zechuan Yu , Guangzhao Yang , Tiejun Liu

Engineering ›› 2026, Vol. 61 ›› Issue (6) : 196 -210.

PDF (4878KB)
Engineering ›› 2026, Vol. 61 ›› Issue (6) :196 -210. DOI: 10.1016/j.eng.2025.12.020
Research
research-article
Atomic Insight into Durability and Interfacial Stability of Novel Hydrophobic Composite in Concrete Alkaline Environment for Marine Engineering
Author information +
History +
PDF (4878KB)

Abstract

The integrity of organic-inorganic interface determines the performance of composite material systems, such as concrete reinforced with basalt fiber reinforced polymer. The integrity of the interface, which depends on the epoxy resin, may be degraded in harsh environments such as in seawater and concrete alkaline environments. In this study, a novel resin cross-linked with polydimethylsiloxane (PDMS) was developed to enhance the performance of composite material in harsh environments. The long-term mechanical strength of the composite (after modification to enhance its hydrophobicity) increased by 20% in concrete alkaline environments, based on micro- and macro-experiments. This improvement is attributed to cross-linking between PDMS and epoxy molecules and the formation of PDMS phase-separated circular domains, which achieve dynamic equilibrium and simultaneously enhance the densification and hydrophobicity. Molecular dynamics simulations revealed that PDMS reinforces interface adhesion and significantly improves the corrosion resistance by facilitating covalent bond formation at the resin-fiber and even resin-concrete interfaces. This study provides a feasible strategy and atomic insights for durability enhancement of composite with similar organic-inorganic interfaces in concrete structures, thus advancing the safety and service life in marine engineering.

Graphical abstract

Keywords

Polydimethylsiloxane / Hydrophobic modification / Concrete alkaline environment / Degradation mechanism

Cite this article

Download citation ▾
Ao Zhou, Kexuan Li, Zechuan Yu, Guangzhao Yang, Tiejun Liu. Atomic Insight into Durability and Interfacial Stability of Novel Hydrophobic Composite in Concrete Alkaline Environment for Marine Engineering. Engineering, 2026, 61 (6) : 196-210 DOI:10.1016/j.eng.2025.12.020

登录浏览全文

4963

注册一个新账户 忘记密码

1. Introduction

The performance of steel-reinforced concrete structures in marine environments usually degrades within 10 years because of corrosion by chloride‒ ions [1,2]. To mitigate these issues, fiber reinforced polymer (FRP) is employed as a substitute for steel [3], [4], [5]. It has demonstrated significant potential for application in infrastructure construction due to its lightweight nature, high strength, excellent corrosion resistance, and cost-effectiveness. Basalt FRPs have been widely promoted and applied because of their environmental friendliness and low cost. However, the resin matrix is susceptible to swelling and cracking because it absorbs water in alkaline environments [6]. Particularly, moisture and corrosive ions rapidly penetrate the fiber-matrix interface through microcracks [7], [8], [9] to degrade both the interface and matrix itself, with debonding of the basalt fiber and resin matrix considered the key factor responsible for the deterioration of basalt FRP composites. The diffusion and corrosion rates of alkali ions increase with alkalinity [10] and basalt FRP undergoes the fastest degradation in alkaline solution [11,12]. The mechanical properties of composites utilized in marine engineering deteriorate by over 50% during long-term service in harsh environments [13,14]. For example, after immersion in seawater for 12 months, the tensile strength of basalt FRP bars in beams decreased by over 50%, causing the failure mode to change from concrete crushing failure to shear failure [15,16]. Under seawater condition, the ultimate load capacity of FRP-reinforced concrete columns decreases by over 20% [17]. Reportedly, the strength decrease of FRP structures is directly related to the degradation of FRP, which is primarily driven by water molecules and aggressive ions (i.e., chloride and hydroxide ions) in corrosive environment [2,16,17]. Therefore, the development of novel matrix materials with excellent hydrophobicity, anti-corrosive properties, and robust mechanical strength seemed an effective strategy to address the challenges presented by insufficient hydrophobicity and alkali resistance.

Epoxy resins are characterized by their outstanding adhesiveness, excellent chemical stability, and low shrinkage, and are therefore widely used as matrices in resin-based composites [18]. However, pure epoxy resins have certain drawbacks such as moisture absorption, brittleness, and inadequate abrasion resistance. Pure epoxy resin is prone to water absorption, which leads to swelling and cracking in harsh environments characterized by high humidity and elevated salinity. Resin hydrolysis is the primary factor contributing to the deterioration of the tensile properties of the composite. Padmaraj et al. [19] found the hygroscopicity to be influenced by the aging conditions. The combined effects of water absorption and aging factors cause the matrix-controlled properties such as compressive strength and interlaminar shear strength to deteriorate. Furthermore, the degradation of the fiber-matrix interface directly diminishes the interlaminar shear strength of the composite. Ferdous et al. [20] investigated the influence of hygroscopicity on the tensile fatigue behavior of polyester and vinyl ester-based composites and discovered that hygroscopicity induced fiber-matrix interfacial debonding, which significantly accelerated the degradation of the fatigue life. However, techniques to modify the resin matrix are complicated. Moreover, the hydrophobicity of composites cannot be fundamentally improved by incorporating nanofillers into the resin matrix, and the hydrophobic modification of composites is poorly understood.

Polydimethylsiloxane (PDMS) is a silicon-based elastomeric polymer with excellent thermal stability, abrasion resistance, chemical stability, and corrosion resistance [21,22], making it a potential choice as a modifying component. However, epoxy resin is polar material, and the low cohesive energy density and lack of polarity of PDMS lead to its poor compatibility with epoxy resin [23]. The compatibility of incompatible polymer components has been enhanced using several commonly employed strategies include grafting [24], copolymerization [25], the incorporation of third-component compatibilizers [26], and the formation of interpenetrating polymer networks [27]. For example, Romo-Uribe et al. [28] synthesized a bifunctional epoxy resin with a rigid structure to modify PDMS. This improved the compatibility of the epoxy resin with PDMS owing to the covalent bonding between the two-phase networks. Dogan et al. [29] found that amphiphilic compatibilizers effectively improve the compatibility between PDMS and epoxy resins, thereby limiting the increase in porosity associated with phase separation. PDMS functionalized with terminal groups results in the formation of phase-separated regions with diameters ranging from approximately 200 nm to 3 μm, which effectively lowers the rate of phase separation to enhance the fracture toughness and reduce the hardness [30,31]. Although the above approaches ensure good compatibility with epoxy-PDMS systems, enhancement of the mechanical and hydrophobic properties of modified polysiloxanes has received little attention.

The objectives of this study were to explore the effect of hydrophobic modification on the long-term behavior of composites in concrete alkaline environment and to study the corresponding deterioration mechanism. The effects of PDMS content and ambient alkalinity on the physical and long-term mechanical properties of the hydrophobic resin were investigated using microscopic methods. The intrinsic mechanism of atomistic evolution and the macroscopic performance of the hydrophobic resin are presented. Moreover, the long-term performance degradation behavior of the hydrophobic composites under different alkaline environments was investigated, and a strength degradation model was established. Molecular dynamics simulations were further employed to reveal the deterioration mechanism of the hydrophobic composites in harsh environments. The current work and findings are expected to offer in-depth atomic insights into the deterioration mechanism of hydrophobic composites in harsh environments and provide promising strategies for the design of durable hydrophobic composites in marine engineering.

2. Experiments and simulation

2.1. Materials and specimens

The raw materials for preparing hydrophobic resin include hydroxyl-terminated PDMS, dibutyltin dilaurate (DBTDL), diglycidyl ether of bisphenol A (DGEBA), γ-aminopropyltriethoxysilane (APS; KH550), polyamide curing agent 650 (PA650), and xylene. The modified epoxy resin is formed via dealcoholization reaction and epoxy-amine cross-linking reactions (Fig. 1). Hydroxyl-terminated PDMS was mixed with KH550 in a ratio of 2:1, and 1 µL of DBTDL was added as a catalyst for the reaction. The mixture was stirred thoroughly at room temperature for 2 h to furnish the siloxane prepolymers with amino groups. Then, 30 g of DGEBA was added, and the mixture was stirred at room temperature for 15 min. The prepolymer and DGEBA were cross-linked by reacting the amino groups of the former with the epoxy groups of the latter to enhance the compatibility. Subsequently, PA650 was added and stirring for another 15 min at room temperature yielded the modified epoxy resin. The total mass of the PDMS, KH550, DGEBA, and PA650 was 60 g. In this study, PDMS was added at different mass fractions to investigate its effect on the physical and mechanical properties of the modified epoxy resin. PE-3 denotes that PDMS constitutes 3 wt% of the modified resin. Based on the proportions and costs, the modified resin cost approximately 10% more than the unmodified resin. The resin specimens were prepared according to the GB/T 2567-2021 standard [32] for performance testing and characterization. Each specimen had a thickness of 4 mm, width of 20 mm, and working section length of 115 mm, resulting in a total length of 220 mm. The resin premix was poured into a mold and smoothed using a scraper. A vacuum pump was used for three rounds of degassing to remove air bubbles. To balance experimental efficiency with complete curing, the specimens were cured for 14 d at 23 °C and 50% relative humidity.

The preparation process of the hydrophobic composites is illustrated in Fig. 2. As the raw materials used are non-corrosive, hydrophobic resins can be applied to FRP with different fiber types. A thoroughly mixed resin was uniformly applied to the basalt fiber cloth on both surfaces. The polytetrafluoroethylene (PTFE) templates were then used to cover the top and bottom sides, followed by manual pressing with an iron bar to completely expel air bubbles. Once this step was completed, a steel plate was placed under pressure on the composite region and left to cure for 1 d before demolding as hydrophobic composite. Subsequently, the specimens are placed in a humidity chamber at a temperature of 23 °C and 50% relative humidity for 14 d. Hydrophobic resins and composites can be easily prepared at room temperature without special equipment, demonstrating their potential for large-scale production. To mitigate the stress concentrations near the clamping region, aluminum or steel sheets were bonded to the ends of the composite specimens after cutting to enhance the friction between the specimen and the chunk. Twelve series of specimens were prepared and categorized according to the immersion environment (three types) and resin formulation (four types). The immersion environment includes a simulated seawater solution at pH = 7.0 (following ASTM D1141-98(2021) standard [33]), a simulated concrete pore solution at pH = 12.8 (following GB/T 34551-2017 standard [34]), and a NaOH solution at pH = 14.0. The following resin formulations were used: PE-0, PE-3, PE-6, and PE-9. Each series comprised four groups corresponding to durability tests conducted over periods of 20, 40, and 60 d. Five specimens were tested repeatedly for each group to ensure the reliability and repeatability of the experimental results.

2.2. Test instrumentation

The surface contact angles of the resins were tested using a 100OSA contact angle tensiometer (LAUDA Scientific GmbH, Germany), with five different measurement points selected to measure the contact angle at that location. The test liquid was pure water with a fixed droplet volume of 2 μL. The degree of swelling of the resin, which reflects the cross-linking density of the polymer, was determined using Eq. (1). The resin was immersed in xylene for 7 d, and the mass was measured before and after immersion.

$Q=\left(1+{\rho }_{2}\times \left(\frac{M}{{M}_{0}{\rho }_{1}}-\frac{1}{{\rho }_{1}}\right)\right)\times 100\%$

where Q is the swelling degree of resin, %; ρ1 and ρ2 are the density of resin and xylene, respectively, g·cm−3; M is the mass of the resin after 7 d of immersion, g; and M0 is the mass of unsoaked resin, g.

The water absorption of the resin was tested according to ASTM D5229/D5229M-20 [35] by immersing the resin in pure water for 7 d. The water absorption rate was determined using Eq. (2).

${W}_{t}=\frac{{A}_{t}-{B}_{0}}{{B}_{0}}\times 100\%$

where Wt is the water absorption of the sample at the immersion time t, %; At is the mass of the specimen at the immersion time t, g; and B0 is the mass of the initial specimen after drying, g.

Tensile test was conducted according to GB/T 2567-2021 [32] at a constant loading rate of 2 mm·min−1. Five parallel specimens were used in each group. To study the effects of various environments on the long-term performance of the resin, three corrosive solutions were selected, including the simulated seawater solution (pH = 7), a simulated concrete pore solution (GB/T 34551-2017 [34]; pH = 12.8), and a NaOH solution (pH = 14.0). The group distribution ratio of the simulated concrete pore solution was 118.5:4.2:0.9, based on the mass ratio of Ca(OH)2, KOH, and NaOH per liter of solution, and the pH value was in the range of 12.7-12.9. The group distribution of the simulated seawater solution is presented in Table 1, and the test periods were 0, 7, and 14 d.

To assess the effect of the hydrophobic-modified matrix on the long-term performance of the hydrophobic composite, the modification protection ratio (Pi,t) was defined to characterize the degree to which the strength retention of the modified composite was enhanced versus that of the unmodified composite at a certain point in time to quantify the effect of hydrophobic modification on the durability performance of the material, as shown in Eq. (3).

${P}_{\text{i,}t}=\frac{{\sigma }_{\text{i,}t}-{\sigma }_{\text{FRP,}t}}{{\sigma }_{\text{FRP,}t}}\times 100\%$

where σi,t(%) and σFRP,t(%) are the strength retentions (%) of the hydrophobic and unmodified composites at 20, 40, and 60 d, respectively.

Fourier-transform infrared (FTIR) spectroscopy (Bruker Equinox 55; Bruker Corporation, Germany) was used to probe the molecular structures of the resins. Resins were assessed using thermogravimetric analysis (TGA; STA6000; PerkinElmer, Inc., USA) at a heating temperature in the range of 35-700 °C and at a heating rate of 10 °C·min−1 under nitrogen atmosphere. Scanning electron microscopy (SEM; Phenom ProX G6; Phenom Scientific Instruments (Shanghai) Co., Ltd., China) was used to examine the surface morphologies of the resins and composites.

2.3. Molecular dynamics simulations

Steered molecular dynamics (SMD) simulations were performed to evaluate the binding properties of the silicate-polymer interfaces in dry and wet environments with and without silicate-to-polymer covalent bonds. As shown in Figs. 3(a) and (b), components of the atomistic model include ① a silicate surface with a cross-section sized 6.1 nm × 5.0 nm in the x-y plane and 2 nm thickness along the z-axis, ② a single strand of silane with amino-, methyl-, silane-, and silanol-terminuses, and ③ a cross-linked cluster of DGEBA and PA650 molecules. The sizes of the initial boxes defined for the pristine and hydrophobic resin are 35.4 Å × 35.4 Å × 35.4 Å and 36.2 Å × 36.2 Å × 36.2 Å, respectively. The cross-linking algorithm was adopted to conduct cross-linking simulations of the resin [31,36]. During the cross-linking process, the reaction radius is progressively extended from 3 to 10 Å in increments of 0.5 Å. Before each subsequent reaction, the cross-linked structure was relaxed under canonical ensemble (NVT) and isothermal-isobaric ensemble (NPT) ensembles to remove structural defects and geometric irregularities. The final cross-linking degrees of the pristine and hydrophobic resins are both 85.0% (Fig. 3(c)). The results showed that the glass transition temperature (Tg), density, and Young’s modulus of the pristine epoxy resin and hydrophobic resin were in good agreement with the experimental data [37]. In the dry case, the components were assembled and sent for equilibration for 2.0 ns, with consistent valence forcefield (CVFF), 1 fs time step, 300 K and 1 atm (1 atm = 0.101325 MPa) NPT ensemble, and a virtual covalent bond connecting the proximate silicon atoms of the silane strand with the silicate surface. The CVFF can describe the interactions between organic and inorganic components [38], [39], [40]. The constant temperature was controlled using a Nose-Hoover thermostat. The cutoff distance was set as 10.0 Å, and the long-range electrostatic interactions are calculated using the particle-particle particle-mesh solver. In the wet case, the coordinates of the SiO2 and resin were first fixed, and the water molecules were relaxed for 1.0 ns in the NPT ensemble. The model was equilibrated for 2.0 ns in the NPT ensemble. Using this approach, we generated four models representing the polymer-epoxy interface in the pristine (removing the virtual bond) and hydrophobic (maintaining the virtual bond) samples in dry and wet ambient environments. The equilibrated state consists of a cluster of cross-linked epoxy molecules wrapping a silane strand lying on a silicate surface (Fig. 3(d)). In the SMD process, a driving force was applied to promote separation between the epoxy cluster and the surface. This driving force was applied only along the z-axis and to the top layer of the epoxy cluster, such that the behavior of the epoxy resin resembles that of a marionette that slowly floats upward with sufficient flexibility [37]. The velocity was set to 2.0 nm·ns−1 and the spring constant of the pulling string was 100 kcal·mol−1·Å−2. Nine pulling trajectories, each with a duration of 2.0 ns, were collected and analyzed. The Jarzynski’s equality was used to calculate the free energy profile during detachment [41]. Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS; Sandia National Laboratories, USA) and Open Visualization Tool (OVITO; Darmstadt University of Technology, Germany) were used for simulation and visualization, respectively [42].

3. Results and discussion

3.1. Contact angle and water absorption

A greater contact angle indicates that the surface of the substrate is more hydrophobic [43]. The relationship between the contact angle and PDMS content is plotted in Fig. 4(a). The contact angle of PE-0 was 89°. After hydrophobic modification, all resins exhibited higher contact angles which gradually increased with increasing PDMS content. Notably, the contact angle of PE-9 was highest (122°), which was attributed to the incorporation of non-polar siloxane chains [44] into the modified epoxy resin system, which lowered the surface energy and tension of the resin. The swelling ratio of the resin after immersion in xylene solution for 7 d is shown in Fig. 4(a). Compared to PE-0, the swelling ratio of PE-3 decreased slightly from 1.17% to 1.14%, because the cross-linking density of the resin increased after modification with a small amount of PDMS [45]. However, the swelling ratios of PE-6 and PE-9 increased with increasing PDMS content. The swelling ratios of PE-6 and PE-9 were 1.25% and 1.36%, respectively, representing increases of 6.8% and 16.2% compared with those of PE-0. This increase can be attributed to excessive amount of PDMS replacing the cross-linking sites within epoxy resin structure, consequently lowering the overall cross-linking density.

The moisture absorption of the resins with different PDMS contents is shown in Fig. 4(b). The water absorption of the resin increased with immersion time, peaking around 28 d before nearing saturation. The water absorption of the modified epoxy resin initially decreased, and then increased with increasing PDMS content. PE-3 absorbed the least water at 49 d by absorbing 65% of that of PE-0. This is because the hydrophobic methyl groups and flexible Si-O main chains in PDMS lower the surface energy of the resins and promote the formation of micro- and nano-rough structures, thereby improving the hydrophobicity of the resins [46]. Moreover, PDMS was cross-linked with the epoxy resin to generate three-dimensional (3D) network structures, and the nanospheres formed by PDMS are embedded in this network and uniformly dispersed within the resins, which increases the cross-linking density of the modified resins [47]. This effectively inhibited the expansion of microcracks, hindered moisture diffusion, and limited the intrusion of aggressive ions to ultimately decrease the water absorption rate of the modified resins significantly. However, the water absorption of PE-9 reached 2.6%, which is 24.8% higher than that of PE-0. This increase was attributed to the incorporation of flexible long-chain PDMS, which replaced some cross-linking points within the molecular structure of the epoxy resin to ultimately lower the overall cross-linking density of the resin [48]. Excessive PDMS causes an uneven distribution of hydrophobic groups within the resin, increasing capillary porosity [49], providing penetration pathways for water and corrosive media, and ultimately promoting water absorption.

3.2. Thermal performance

The effect of PDMS content on the glass transition temperature (Tg) of the resin is shown in Fig. 5(a). The Tg of PE-0 was 61.52 °C. As the PDMS content increased, Tg first increased and then decreased. The Tg of PE-3 reached 134.52 °C, which can be attributed to the effective cross-linking reaction occurring between the amino groups and the epoxy groups present on the PDMS chains of the precursor. Moreover, the uniform distribution of the siloxane phase within the matrix functioned as a protective layer. The rigid alicyclic structure mitigated the adverse effects of flexible PDMS chain segments on the thermal properties [50] to ultimately enhance Tg of the epoxy resin. However, at the maximum PDMS content, PE-9, Tg decreased to 130.02 °C. This was linked to the introduction of excessive Si-O bonds, which enabled the rotation of the molecular chains, decreased the compactness of the molecular stacking, and consequently lowered the glass transition temperature. Mass loss versus temperature (T) curves for modified epoxy resins with different PDMS contents appear in Fig. 5(b). The temperature corresponding to 5% weight loss is referred to as the initial degradation temperature T5%. The peak temperature of the derivative thermogravimetry (DTG) curve, indicating the maximum rate of weight loss, is denoted as Tmax. Cases in which multiple peaks are observed are denoted as Tmax1 and Tmax2. In addition, the mass percentage of the residual resin after the specimen reached its stabilization temperature relative to original specimen mass was defined as the residual carbon rate. The relevant characteristic temperature data are summarized in Table 2.

The primary weight loss of the resin occurs at 300-500 °C, due to the degradation of the epoxy network [51]. The pure epoxy resin had a T5% of 320.89 °C and a Tmax of 413.79 °C, with decomposition essentially complete at 550 °C and a residual carbon rate of 3.05%. The T5% of PE-3 was increased to 325.71 °C, and Tmax1 and Tmax2 were 370.53 and 429.58 °C, respectively, and the residual carbon rate reached 8.03%, indicating that the thermal properties of the epoxy resin had improved. Tmax of the modified epoxy resin increased gradually with the incorporation of PDMS. This improvement is attributed to the higher energy of the Si-O-Si bond in PDMS (460 kJ·mol−1) compared with that of the C-C and C-O bonds in the pure epoxy resin. The PDMS in the modified resin participates in cross-linking reactions and forms siloxane-rich network domains. This delays the scission and degradation of molecular chains at high temperature, thereby increasing Tmax of the modified resin. Tmax1 of the modified resin (350-400 °C) was lower than that of the pure epoxy resin and corresponds to the initial thermal degradation of cross-linked epoxy molecular chains within modified resin. Compared to the pure epoxy resin, the introduction of PDMS altered the chain length of the cross-linked epoxy and reduced the uniformity of the molecular chain distribution. This structural change decreases the degradation temperature of the epoxy-rich domains, which lowers Tmax1. In addition, Tmax2 of the modified resin, at 400-500 °C, corresponds to the breaking of the Si-O-Si chain and subsequent rearrangement of siloxanes into small cyclic oligomers [52]. The degradation of Si-CH3 occurs in the later stages and requires higher temperatures, indicating that PDMS effectively delays the decomposition process. Moreover, the PDMS in the modified resin cross-linked with epoxy to form silicone network structures. During high-temperature pyrolysis, the Si-O-Si segments in the modified epoxy resin facilitate the formation of stable silica-like and ceramic phase structures, which enhance the residual carbon rate [51,52].

3.3. FTIR analysis

In the FTIR spectra of resins before and after modification (Fig. 6), the peak at 1510 cm−1 corresponds to the C-C stretching vibration of the unmodified resin system, and the absorption peak of the epoxy group is observed at 830 cm−1 [53]. The absence of the Si-OH peak of the modified resin at 3295 cm−1 implies that the hydroxyl group of PDMS reacted with the -OC2H5 group of APS. This is consistent with the observation of the grafting reaction of PDMS and APS [54]. Moreover, the intensity of the peak representing the epoxy group at 830 cm−1 weakened, whereas the intensities of the asymmetric stretching vibration peaks for the C-H bonds in methylene (-CH2-) and methyl (-CH3) groups (at 2853 and 2963 cm−1) increased. This confirms that the epoxy rings in modified resin have reacted with the PDMS [55,56]. Furthermore, new peaks at 803 and 1090 cm−1 in modified resin correspond to the stretching vibrations of the Si-CH3 and Si-O-Si bonds, respectively [57]. This further demonstrates that the PDMS chains in the modified resin underwent cross-linking reactions with the epoxy rings to form epoxy-silicone networks [57,58].

3.4. Surface morphology and alkali resistance

The surface morphology varied significantly with the PDMS content. As shown in Fig. 7(a), the fracture surface of PE-0 exhibits a more regular pattern with sharp edges and a scalloped appearance, which is a typical characteristic of brittleness. In contrast, irregular cracks began to appear on the fractured surface of PE-3 (Fig. 7(b)). With a further increase in PDMS content, the fracture surfaces of PE-6 and PE-9 became rough and inhomogeneous round domains produced by phase separation began to appear Figs. 7(c) and (d). This can be attributed to several factors: ① The cross-linking interactions between the epoxy resin and PDMS chains enhance the cohesion among the polymer chains; ② the flexible siloxane chains within the system can deform, absorb, and dissipate impact energy when subjected to external forces [59]; and ③ the circular domains generated during the phase separation of the modified epoxy resin produce a nano-toughening-like effect with crack deflection in the cross-section, indicating increased resistance to crack extension during fracture.

The elemental distributions of PE-3 and PE-9 are shown in Figs. 7(b) and (d). Notably, the light-colored inhomogeneous circular domains observed in the PE-3 section result from the separation of the non-polar siloxane chain segments from the polar epoxy resin network. These circular domains were uniformly distributed throughout the matrix and they increased in size with increasing PDMS content. The circular domains of PE-9 have an approximate diameter of 2 μm, with folds indicative of the fracture process surrounding these domains, which are linked to the energy dissipation mechanisms during fracture. Moreover, the concentration of Si increased with increasing PDMS content, showing a notable accumulation of Si within the inhomogeneous circular domains. This observation suggests a higher concentration of siloxane phases, which remained adhered to the epoxy resin network, even after stretching and fracturing.

The corrosion resistance of the modified epoxy resin in different environments was assessed by immersing the specimens PE-0, PE-3, and PE-6 in seawater solution, concrete pore solution, and NaOH solution. The mechanical strengths and fracture morphologies of the modified and unmodified epoxy resins in the marine environment are shown in Fig. 8(a). The strength retention of PE-3 was 5.7% higher than that of PE-0 after 14 d of immersion. In contrast, the strength retention of PE-6 is slightly lower than that of PE-0. The irregularities and plastic damage of the resin fractures were attributed to the higher PDMS content, which caused the phase separation of the low-surface-energy siloxane phase from the epoxy resin phase. Moreover, the diameter of the phase separation increased with increasing PDMS content, which led to higher porosity and the formation of defects. This enhancement in the porosity facilitates water absorption, thereby facilitating the penetration and diffusion of chemical substances. After 14 d of immersion in concrete pore solution, the strength retention of PE-3 increased by 17.2% compared with that of PE-0, as shown in Fig. 8(b). This is because the addition of an appropriate amount of PDMS effectively enhances the degree of cross-linking and hydrophobicity of the resin, which hinders the diffusion of water molecules and the intrusion of erosive ions, thereby improving the tensile strength of the hydrophobic resin in corrosive environment. After immersion in the concrete pore solution for 14 d, the strength retention of PE-6 was 9.7% higher than that of PE-0. The fracture morphology of PE-6 shows that Ca(OH)2 formed a thin film on the surface of the resin, acting as a barrier that impeded the ingress of water and corrosive ions. Additionally, the low reactivity of the main chain of the siloxane polymer contributes to its excellent water resistance. The co-cross-linking effect between siloxane and the epoxy resin facilitated the formation of integral networks, with siloxane chains migrating to the surface of the resin, thereby enhancing the mechanical strength of the specimens in the concrete pore solution. When immersed in NaOH solution for 14 d, the strength retention of PE-3 increased by 10.6% compared with that of PE-0 (Fig. 8(c)). Conversely, the strength retention of PE-6 was 6.2% lower than that of PE-0. This indicated that excessive PDMS weakened the alkaline resistance of the material. The fracture morphology showed an increase in the irregular fracture pattern of PE-6, which was attributed to the increase in PDMS. Excessive PDMS is prone to phase separation and causes microscopic defects that destroy the network structure of the resin and reduce its mechanical strength and corrosion resistance. The coupling of PDMS with hydroxide ions in the NaOH solution led to increased water absorption and the formation of phase separation defects. This interaction intensifies the water absorption and swelling effects of the resin, thereby worsening the material deterioration. In addition, when exposed to NaOH solution (pH = 14.0), concrete pore solution (pH = 12.8), and seawater (pH = 7.0), the tensile strengths of the specimens were significantly different. The tensile strength retentions of the unmodified specimen (PE-0) were 83.7%, 76.6%, and 68.9% in environments with pH values of 7.0, 12.8, and 14.0, respectively, demonstrating accelerated deterioration with increasing alkalinity. This acceleration is attributed to the higher OH- concentrations, which intensify the hydrolysis of ester or ether bonds within the unmodified epoxy resin, thereby impairing its chemical stability and mechanical properties. In contrast, in environments with pH values of 7.0, 12.8, and 14.0, the strength retention of specimen PE-3 improved significantly to reach 88.5%, 89.8%, and 76.2%, respectively. Under the same conditions, the tensile strength retentions of PE-6 were 82.5%, 84.0%, and 64.6%, respectively. Significantly, the hydrophobic-modified specimens exhibited excellent long-term mechanical properties in the concrete pore solution. The above analysis indicates that the hydrophobic composite experienced the most severe degradation in the NaOH solution, whereas its corrosion resistance in the concrete pore solution was the highest.

3.5. Corrosion resistance mechanism

Fig. 9 shows the corrosion resistance mechanism of hydrophobic modified epoxy resin. In unmodified resins, epoxy molecules crosslink to form two-dimensional (2D) single-chain structures. In harsh environments, the 2D single chain cannot effectively intercept the invasion of water molecules and aggressive ions. As depicted in Fig. 9(a), water molecules and aggressive ions create multiple erosion channels inside the matrix, which result in the rapid deterioration of the unmodified epoxy resin. In contrast, the hydrophobic epoxy resin had excellent corrosion resistance (Fig. 9(b)). The hydroxyl-capped PDMS was functionalized with APS to introduce an amino group, which provided active sites for further cross-linking with the epoxy resin. Subsequently, PDMS and the epoxy resin were co-cross-linked to form a complex 3D network structure. The polarity differences within the modified resins facilitated the formation of a two-phase structure. The circular domains formed by phase separation enhanced the spacing between the siloxane phases, thereby improving the dispersion. The siloxane nanoparticles formed by phase separation enlarge the specific surface area, extend the penetration path of the corrosive media, and ultimately slow the corrosion process. Moreover, the hydrophobic characteristics of the polymer surface are influenced by the surface energy and the micro- and nano-structural features. Exposure of the siloxane chains effectively reduced the surface energy of the resin to further enhance the hydrophobicity of the resin surface. The resulting significant impediment of the penetration and diffusion of chemical substances and water contributes to superior corrosion resistance. Moreover, the resistance of resin system to solvent penetration is also related to the cross-linking density, with a higher cross-linking density corresponding to lower solubility. Although the cross-linking density of the resin system declined slightly after modification, it remained a significant factor in determining the overall corrosion resistance of the system. However, excessive PDMS weakens the corrosion resistance of the hydrophobic composite. The flexible silicate chains in surplus PDMS exhibit poor compatibility with the epoxy matrix, causing uneven hydrophobic group distribution and local stress concentration. Excessive PDMS reduces the number of reaction sites within the matrix and lowers the cross-linking degree of the resin, thereby disrupting the continuous silicate-epoxy skeleton and increasing the capillary porosity. These factors shorten the penetration depth of water molecules and erosive media to ultimately accelerate the degradation of the modified resin.

3.6. Long-term properties of hydrophobic composite

The tensile strength retention of the hydrophobic composites in different environments is shown in Fig. 10 as a function of time. The tensile strengths of uncorroded PE-0, PE-3, and PE-6 were 291.3, 325.9, and 317.5 MPa, respectively. The addition of excessive PDMS caused severe phase separation within the matrix and weakened the bonding at the fiber-matrix interface, thus resulting in the overall performance of PE-9 being significantly lower than that of the unmodified PE-0. In view of this, the high content of PDMS is not advised to be incorporated in the composite. After 60 d of immersion in a marine environment at pH 7.0, the tensile strengths of PE-0, PE-3, and PE-6 decreased by 25%, 21%, and 25%, respectively, as shown in Fig. 10(a). PE-6 experienced the rapid degradation of tensile strength during the initial 40 d of immersion. This phenomenon can be attributed to the hydrophobicity of the resin, which forms a protective barrier against the intrusion of water and ionic species. However, the hydrophobicity was insufficient to counteract water absorption, which was exacerbated by a reduction in the cross-linking density and an increase in the long-term water uptake of the resin. Consequently, these factors detrimentally affect the durability of hydrophobic composites under prolonged corrosive conditions. The long-term strength deterioration rate of the composite decreased after immersion in concrete pore solution (pH = 12.8) for 60 d (Fig. 10(b)). However, the short-term deterioration rates of the tensile strengths of the composites differed slightly. After exposure to the concrete pore solution for 60 d, the tensile strengths of PE-0, PE-3, and PE-6 decreased by 51%, 41%, and 41%, respectively. The tensile strength retention of the hydrophobic specimen was 19.6% higher than that of unmodified specimen. The initial rapid decline in the tensile strength of PE-6 was followed by a gradual reduction until, at 60 d, the strength retention was similar to that of PE-3. This indicates that the hydrophobic composite with low PDMS content is more resistant to alkaline environment. As shown in Fig. 10(c), the deterioration rate of the hydrophobic composite gradually increased with increasing alkalinity. After 60 d of immersion in NaOH solution (pH = 14.0), the tensile strengths of PE-0, PE-3, and PE-6 decreased by 61%, 52%, and 55%, respectively. Notably, compared to PE-0, PE-3 exhibited a lower deterioration rate and superior durability owing to its enhanced hydrophobicity and lower water absorption. The damage pattern of the composite changed with increasing environmental alkalinity, with PE-3 experiencing tearing failure in a weakly alkaline environment and serrated failure in a strongly alkaline environment. The tensile strength of PE-6 deteriorates linearly in a strongly alkaline environment.

In seawater, prolonged exposure to erosion conditions enables moisture to penetrate the resin matrix, which undermines the bonding between the resin and fibers to loosen the fibers (Fig. 10(d)). In the concrete pore solution, the hydrophobic composite demonstrated superior retention of the matrix material on the fiber surface (Fig. 10(e)). This phenomenon can be attributed to the interaction between the modified epoxy resin and the siloxane phase during the fracture process, which resulted in firm attachment of the siloxane phase to the matrix. Moreover, the formation of nanoparticles from the PDMS phase separation inhibited the entry of water, thereby enhancing the durability of the hydrophobic composite. Debonding at the fiber-resin interface intensified and the resin residue on the fiber surface decreased with the alkalinity of the solution. As shown in Fig. 10(f), the deterioration of hydrophobic composite in the NaOH solution was mainly manifested by debonding of the fibers from the matrix and fiber breakage [60]. Hydrophobic modification therefore significantly improves the long-term performance of composites in harsh environments by enhancing bonding at the matrix-fiber interface and hindering water penetration [61]. Given that the cost of hydrophobic composites increases by approximately 10%, the extended service life and reduced maintenance cost demonstrate high industrial feasibility in engineering practice.

3.7. Degradation model

Short-term durability tests are commonly used to forecast the long-term performance of composites. Various models have been devised to predict the degradation behavior of the composite strength in corrosive environments [62], [63], [64], [65]. Among these, the diffusion model in Eq. (4) is based on Fick’s law. This model establishes a correlation between strength and parameters such as the diffusion coefficient (D), ion concentration (C), time (t), and material thickness (r0). However, the diffusion model only considers physical diffusion processes based on moisture absorption and ignores non-diffusion factors such as chemical reactions and interfacial damage. The diffusion model therefore has low prediction accuracy for material properties in a complex service environment.

$Y=100\%·{\left(1-\frac{\sqrt{2DCt}}{{r}_{0}}\right)}^{2}$

where Y is the mechanical strength retention of the composite, %.

The double logarithmic coordinate model assumes that the degradation of the fiber-resin interface governs the performance deterioration of the composites, and considers the composite strength to vary logarithmically over time (Eqs. (5), (6)) [66,67]. However, the double logarithmic coordinate model is derived from the fitting of empirical data, and the prediction accuracy depends heavily on the quality and quantity of short-term aging test data [66], [67], [68]. The residual strength predicted by the double logarithmic coordinate model tended to infinity when the aging time was zero, which contradicted the experimental results. Additionally, the exponential relationship model asserts that the composite strength is exponentially degraded with time, with an initial rapid decline followed by a stabilization phase, as represented by Eq. (7) [69]. Given that the influence of the type of ion in a corrosive environment is not considered, the exponential relationship model cannot predict situations in which the strength retention of the composite may first increase and then decrease during the degradation process [70,71]. Wang et al. [72] integrated a diffusion model with an empirical degradation model (Eq. (8)) to predict the degradation of the mechanical strength at a designated pH level. Against this background, our aim was to establish an enhanced model for predicting the long-term mechanical properties of hydrophobic composites by considering the pH and PDMS content.

$Y=a\mathrm{log}(t)+b$
$\mathrm{log}(Y)=a\mathrm{log}(t)+b$
$ Y=100 \% \cdot \mathrm{e}^{\frac{-t}{\tau}}$
$ Y=\left(100 \% \cdot-Y_{\infty}\right) \cdot 100 \% \cdot \mathrm{e}^{\frac{-t}{\tau}}+Y_{\infty}$

where Y∞ is the strength retention of the composite at stabilization; a, b, and τ are the parameters of the degradation model.

The experimental results indicated that the long-term mechanical strength of the hydrophobic composite depended nonmonotonically on the PDMS content. Excessive PDMS disrupted the stability of 3D spatial networks within the modified resins to diminish the mechanical strength of the hydrophobic composite in a corrosive environment. Moreover, as the immersion time increased, the strength retention of hydrophobic composites exhibited an exponential relationship with the pH values of corrosion environments. As the resistance to alkaline environment of the hydrophobic resin improves, the deterioration of fiber-resin interface emerges as the primary factor contributing to a decline in the mechanical strength of the hydrophobic composite. To improve the accuracy of the degradation model and expand the effective range, 39 data points were collected from the Refs. [31],[69], [70], [71],[73] and integrated into the dataset, which expanded the data to cover longer immersion time (e.g., 90, 180, and 360 d) and additional pH conditions (e.g., pH = 10.1 and 12.9). Based on the experimental data, the mechanical strength degradation model for the composite was established, as shown in Eq. (9). The model quantitatively describes the relationship between time (X1), pH value (X2), and PDMS content (X3) with respect to the strength retention of the composite (Ync). The coefficients (β1-β6) of the proposed model are listed in Table 3. As empirically coefficients, the coefficients of proposed model are determined by multivariate regression analysis based on factors such as environmental conditions, the corrosion time, and material properties, thereby more accurately reflecting the degradation patterns of materials under various service conditions. The coefficient of determination (R2) of the proposed model was 0.931, indicating a strong correlation between the predicted values and experimental results. The proposed model exhibited a maximum prediction error of only 12.7%, and more than 80% of specimens had prediction errors below 10%, as shown in Fig. 11(a).

$ \begin{array}{l} Y_{\mathrm{nc}}=\left(1-\beta_{1} \mathrm{e}^{-\frac{X_{2}-\tau}{\tau} \beta_{3}}\right) \mathrm{e}^{-\frac{X_{1}}{\beta_{2}}} \cdot\left[\beta_{4}\left(\frac{X_{3}}{6}\right)^{2}+\beta_{5} \frac{X_{3}}{6}+\beta_{6}\right]+ \\ \beta_{1} \mathrm{e}^{-\frac{X_{2}-7}{7} \beta_{3}} \end{array}$

Fig. 11(b) presents the relationship between the strength retention, pH, immersion time, and PDMS content calculated by the proposed mechanical model. The strength retention of the hydrophobic composite decreased with increasing pH and immersion time. In addition, the strength retention demonstrated an initial increase, followed by a decline as the PDMS content increased. Notably, compared to the unmodified composite, the hydrophobic composite exhibited enhanced durability under identical alkaline conditions. The trend of hydrophobic composite strength retention with PDMS content in different pH environments at 40 d of erosion was shown in Fig. 11(b). The PDMS content plays a critical role in determining the durability of the hydrophobic composites. Specifically, for PDMS content of 3.2%, the hydrophobic composite exhibited outstanding long-term mechanical properties in diverse pH environments. In this study, 18 data points on the tensile strength retention of composites at various pH conditions were collected from Refs. [18],[72],[74], [75], [76]. The proposed model and other published models were employed to predict the strength retention, which was further compared with experimental data from the literature. The proposed model predicted these data points with errors of less than 10%. In contrast, the maximum prediction error of the double logarithmic coordinate model was 27.3%, whereas that of the exponential relationship model was 26.6%. The proposed model successfully predicted the long-term effect of the pH level on the mechanical properties of the composite. The model provides theoretical support for evaluating the durability of composites at different alkalinity levels and facilitates the assessment of the applicability of hydrophobic composites in harsh environments, as depicted in Fig. 11(c).

3.8. Hydrophobic working mechanism

The above-mentioned experimental results informed that the addition of silane improves the hydrophobicity of the resin in a variety of corrosive scenarios, especially in concrete alkaline environment. In addition, the debonding of the SiO2-resin interface was simulated to explore the enhancement effect of the modified resins on the interfacial corrosion resistance of hydrophobic composites in different environments. The temperature of molecular dynamics simulations in this study was set as 300 K. At this temperature, there is no obvious chemical reaction between SiO2 and the unmodified or hydrophobic resin in the interfacial region [36,77]. The CVFF is used to describe the evolution of the interfacial microstructure and bonding properties in dry and wet environments. The variation in the potential of mean force (PMF) at the unmodified and hydrophobic interfaces with the displacement between the epoxy and the substrate during the debonding process is plotted in Figs. 12(a) and (b). The PMF curves (dark curves) were calculated using Jarzynski’s equality based on the data from nine SMD simulations (light curves). The PMF curve reflects the interfacial free energy of the system during debonding. Higher energy barriers indicate stronger interfacial bonding. Under dry condition, the PMF of the unmodified interface was 112.3 kcal·mol−1, whereas that of the hydrophobic interface was as high as 216.3 kcal·mol−1. The interfacial bond strength of hydrophobic resin to substrate was much higher than that of unmodified resin, up to 92.6%. This is attributed that the introduction of Si-O bonds of PDMS promotes the formation of covalent bonds between the hydrophobic resin and the silicate substrate, leading to strong interfacial interactions. Compared to the hydrophobic interface, although the epoxy resin in unmodified interface first deforms before complete separation, epoxy cluster detaches earlier from the silicate matrix due to the absence of covalent bonding between unmodified resin and silicate substrate (Fig. 12(c)). In other words, the presence of covalent bonds prolongs the period of deformation of the hydrophobic-modified epoxy cluster before full separation, resulting in a more durable hydrophobic epoxy-silicate interface (Fig. 12(d)). This phenomenon was also observed in the wet cases.

The PMF of unmodified interfaces under wet condition decreased by 57.2% compared to that under dry condition (Fig. 13(a)). The erosive factors severely impair the interfacial interaction between unmodified epoxy and substrate, exacerbating the interfacial degradation. Specifically, at 122.1 kcal·mol−1, the PMF of the hydrophobic interfaces under wet condition (Fig. 13(b)) was 8.7% higher than that of the unmodified interfaces under dry condition. This proved that hydrophobic modification effectively improved the interfacial corrosion resistance of the composite. In harsh environments, such as in seawater or alkaline solutions, the epoxy-silicate interface loses its integrity as soon as the ingressive molecules diffuse into the gap layer and block van der Waals or Coulombic interactions. As depicted in Fig. 13(c), the unmodified resin was bonded to the substrate with only a few epoxy molecular chains at 180 ps and reached a critical state of failure at the interface. Conversely, due to the covalent bonding between the hydrophobic resin and substrate, the hydrophobic resin is still filamentally attached to the substrate at 810 ps. Moreover, the introduction of PDMS improved the ability of the resin to undergo elongation in harsh environments, as shown at point C in Fig. 13(d). This explains why the addition of a controlled amount of PDMS (≤ 3 wt%) to improve the hydrophobicity serves to enhance the durability of the composite in harsh environments without affecting the originally dense configuration. Other than enhancing the hydrophobicity, the integrity and bond strength of the inorganic-organic interface are further improved by the presence of silane-silicate bonds in the interfacial region. At point D in Fig. 13(d), a few small molecules (cross-linked dimers) remained on the SiO2 surface because of the cross-linking density of the hydrophobic resin and the strong silane-silicate bonding. Figs. 13(c) and (d) clearly show a shift in the failure mode from direct separation in the unmodified case to stretching-separation at the covalently bonded interface, that is, if epoxy contains components that can form covalent bonds to the substrate, the attack from harsh environments can be alleviated. Moreover, if the harsh environments are cyclic, we might envision a possible recovery in interface binding when ingressive molecules are removed and the covalent bonds are not broken. For example, at point C in Fig. 13(d), the epoxy resin might be pulled backward.

The above phenomena originate from the fact that covalent bonds are generally stronger than the connections formed via van der Waals or Coulombic interactions. This emphasizes the importance of introducing functional groups with the potential to form covalent bonds between dissimilar material phases in composite. The experimental evidence in this study revealed the competing effect of silane addition on the hydrophobicity and the integrity of resin, thus yielding an optimal modification strategy. Moreover, in silico perspectives indicate that promoting the formation of interfacial covalent bonds is another effective strategy for reinforcing resin-fiber and even resin-concrete interfaces. Viewing concrete as a composite material system, silicone agents should find great applications in acting as a mediator between dissimilar organic and inorganic material phases. In addition, the service condition for composites in marine engineering are complicated, and both loading and environmental factors take into effect simultaneously, such as high temperature, high humidity, high salinity, cyclic loading, and sustained loading. Future research is recommended to focus on the long-term mechanical performances of hydrophobic composites under the coupled effects of loading and environmental factors, including temperature, humidity, aggressive salt ions, cyclic loading, and sustained loading. This is expected to provide solid experimental foundation for the practical applications of hydrophobic composites in real marine environment.

4. Conclusions

In this study, a novel hydrophobic modification approach for fiber reinforced composite served in concrete alkaline environment was developed. The physical properties and long-term mechanical properties of novel hydrophobic composite in seawater and alkaline environment were experimentally investigated. The results showed that the hydrophobic performance and thermal stability of the modified resin in the hydrophobic composite in seawater and alkaline environments were significantly improved. The glass transition temperature of the modified resin increased from 61.52 to 136.31 °C, and the contact angle reached 122°, with a water absorption rate of only 1.36%. In corrosive environments, hydrophobic composites with an appropriate amount of PDMS exhibited superior long-term mechanical properties. In the alkaline concrete environment, the tensile strength retention of the hydrophobic composite was 19.6% higher than that of unmodified specimen. The degradation mechanism of hydrophobic composite in corrosive environment was revealed using molecular dynamics simulations and microscopic characterization. Particularly, the nonpolar siloxane chains of PDMS with their low surface energy effectively enhanced the hydrophobicity of the composite and promoted the formation of covalent bonds at the resin-fiber and resin-concrete interfaces to enhance the integrity of dissimilar phases and improve the corrosion resistance of hydrophobic composites. The co-cross-linking effect between PDMS and epoxy resin enhances the hydrophobic matrix-fiber interfacial bonding, effectively inhibiting the intrusion of water molecules and corrosive ions and improving the long-term mechanical strength of hydrophobic composite in corrosive environment. However, excessive PDMS content induces adverse effects. Compared with the unmodified resin, the modified resin with 6% PDMS showed a 6.8% increase in swelling ratio and a 6.2% decrease in tensile strength retention. Excessive PDMS dilutes the reactive groups in the resin, disrupts resin cross-linking, and impedes the formation of a continuous 3D framework. This weakens the integrity of the matrix and generates interfacial defects and micropores that facilitate the penetration of water and erosive ions, ultimately impairing the corrosion resistance and long-term mechanical strength of the composites. Moreover, a model was developed to predict the long-term mechanical properties of the hydrophobic composites by considering pH value and PDMS content. The findings of this study can inspire promising strategies for designing safe and durable composites for marine engineering applications.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]

Lyu C, Yu C, Lu C, Pan L, Li W, Liu J. Long-term performance and microstructural characterization of dam concrete in the Three Gorges project. Engineering 2024; 33:237-62.
[2]

Quaranta E, Davies P. Emerging and innovative materials for hydropower engineering applications: turbines, bearings, sealing, dams and waterways, and ocean power. Engineering 2022; 8:148-58.
[3]

Wang HH, Huo S, Chevali V, Hall W, Offringa A, Song P, et al. Carbon fiber reinforced thermoplastics: GROM materials to manufacturing and applications. Adv Mater 2025; 37(27):2418709.
[4]

Chen Z, Yu T, Yang Z, Wei Z, Li Y, Yang W, et al. Superior mechanical behavior and flame retardancy FRP via a distribution controllable 1D/2D hybrid nanoclay synergistic toughening strategy. Engineering 2024; 40:166-78.
[5]

Xu Z, Wu M, Ye Q, Chen D, Liu K, Bai H. Spinning from nature: engineered preparation and application of high-performance bio-based fibers. Engineering 2022; 14:100-12.
[6]

Feng GY, Zhu DJ, Guo SC, Rahman MZ, Jin ZQ, Shi CJ. A review on mechanical properties and deterioration mechanisms of FRP bars under severe environmental and loading conditions. Cem Concr Compos 2022; 134:104758.
[7]

Zhang W, Li S, Wei D, Shi Y, Lu T, Zhang Z, et al. A fluorine-free and high-robustness photothermal self-healing superhydrophobic coating with long-term anticorrosion and antibacterial performances. J Mater Sci Technol 2025; 210:284-98.
[8]

Zhou A, Chow CL, Lau D. Structural behavior of GFRP reinforced concrete columns under the influence of chloride at casting and service stages. Compos Pt B Eng 2018; 136:1-9.
[9]

Zhou A, Büyüköztürk O, Lau D. Debonding of concrete-epoxy interface under the coupled effect of moisture and sustained load. Cem Concr Compos 2017; 80:287-97.
[10]

Yoo SJ, Lee OS, Parr JLM, Yoon YS. Degradation of flexural bond of CFRP bar in UHPFRC after exposure to elevated temperature. Cem Concr Compos 2024; 152:105627.
[11]

Azzam A, Bassuoni MT, Shalaby A. Performance of nano silica-modified cementitious composites reinforced with basalt fiber pellets under alkaline and salt-frost exposures. Cem Concr Compos 2022; 134:104761.
[12]

Li KX, Liu TJ, Zhang B, Zhou A, Han SW. Axial compressive behavior of low-alkalinity seawater sea-sand concrete column reinforced with hydrophobic longitudinal SFCBs and FRP hoops. Eng Struct 2024; 315:118446.
[13]

Zhang L, Gao C, Zhong L, Zhu L, Chen H, Hou Y, et al. Robust photothermal superhydrophobic coatings with dual-size micro/nano structure enhance anti-/de-icing and chemical resistance properties. Chem Eng J 2022; 446:137461.
[14]

Zhao Y, Xu T, Hu JM. A robust, room-temperature curable and molecular-level superhydrophobic coating with excellent antibacterial and antifouling properties. Chem Eng J 2022; 450:136557.
[15]

Dong Z, Wu G, Lian J. Experimental study on the durability of FRP bars reinforced concrete beams in simulated ocean environment. Sci Eng Compos Mater 2018; 25(6):1123-34.
[16]

Dong Z, Wu G, Zhao X, Zhu H, Lian J. Durability test on the flexural performance of seawater sea-sand concrete beams completely reinforced with FRP bars. Constr Build Mater 2018; 192:671-82.
[17]

Zhou A, Qin R, Chow CL, Lau D. Structural performance of FRP confined seawater concrete columns under chloride environment. Compos Struct 2019; 216:12-9.
[18]

Yi Y, Zhu D, Guo S, Li S, Feng G, Liu Z, et al. Development of a low-alkalinity seawater sea sand concrete for enhanced compatibility with BFRP bar in the marine environment. Cem Concr Compos 2022; 134:104778.
[19]

Padmaraj NH, Vijaya KM, Dayananda P. Experimental investigation on fatigue behaviour of glass/epoxy quasi-isotropic laminate composites under different ageing conditions. Int J Fatigue 2021; 143:105992.
[20]

Ferdous W, Manalo A, Yu P, Salih C, Abousnina R, Heyer T, et al. Tensile fatigue behavior of polyester and vinyl ester based GFRP laminates—a comparative evaluation. Polymers 2021; 13(3):386.
[21]

Zang X, Cao X, Zheng W, Zhu T, Lei Y, Huang J, et al. Meniscus inspired flexible superhydrophobic coating with remarkable erosion resistance for pipeline gas transmission. Chem Eng J 2023; 451:138573.
[22]

Dong Z, Zhu H, Hang Y, Liu G, Jin W. Polydimethylsiloxane (PDMS) composite membrane fabricated on the inner surface of a ceramic hollow fiber: from single-channel to multi-channel. Engineering 2020; 6(1):89-99.
[23]

Shen Y, Li K, Chen H, Wu Z, Wang Z. Superhydrophobic F-SiO2@PDMS composite coatings prepared by a two-step spraying method for the interface erosion mechanism and anti-corrosive applications. Chem Eng J 2021; 413:127455.
[24]

Zhou Z, Tan S, Sun W, Guan X, Xu T, Ji G. A cost-effective photothermal superhydrophobic coating with micro- and nano-graded structures for efficient solar energy harvesting. J Mater Sci Technol 2024; 198:158-65.
[25]

Zhang W, Li S, Wei D, Zheng Z, Han Z, Liu Y. Fluorine-free, robust and self-healing superhydrophobic surfaces with anticorrosion and antibacterial performances. J Mater Sci Technol 2024; 186:231-43.
[26]

Guo Y, Zhao H, Zhang C, Zhao G. Super photothermal/electrothermal response and anti-icing/deicing capability of superhydrophobic multi-walled carbon nanotubes/epoxy coating. Chem Eng J 2024; 497:154383.
[27]

Lu J, Yuan M, Di X, Yuan Q, Ni L, Luo Y, et al. Fast, non-carbonized, ambient-drying PVA/CNF@GO foam: towards super-broadband microwave absorption and structural strength enhancement in aramid honeycomb. Chem Eng J 2024; 489:151385.
[28]

Romo-Uribe A, Arcos-Casarrubias JA, Reyes-Mayer A, Guardian-Tapia R. PDMS nanodomains in DGEBA epoxy induce high flexibility and toughness. Polym Plast Technol Eng 2017; 56(1):96-107.
[29]

Dogan D, Ruthmann S, Seewald O, Bremser W. Tuning of antifouling active PDMS domains tethered to epoxy/amine surface. Prog Org Coat 2022; 170:106977.
[30]

Yang G, Wu Y, Li H, Gao N, Jin M, Hu Z, et al. Effect of shrinkage-reducing polycarboxylate admixture on cracking behavior of ultra-high strength mortar. Cem Concr Compos 2021; 122:104117.
[31]

Li K, Zhou A, Liu T, Zou D, Yu Z. Long-term performance and deterioration mechanism of novel hydrophobic coated fiber reinforced composite in marine environment. Compos A Appl Sci Manuf 2025; 190:108716.
[32]

Standardization Administration of the People’s Republic of China. GB/T 2567-2021: Test methods for properties of resin casting body. Chinese standard. Beijing: Standards Press of China; 2021. Chinese.
[33]

ASTM D1141-98(2021): Standard practice for the preparation of substitute ocean water. ASTM standard. West Conshohocken: ASTM International; 2021.
[34]

Standardization Administration of the People’s Republic of China. GB/T 34551-2017: Test method for alkali resistance of glass fiber reinforced polymer bars in high temperature condition. Chinese standard. Beijing: Standards Press of China; 2017. Chinese.
[35]

ASTM D5229/D5229M-20: Standard test method for moisture absorption properties and equilibrium conditioning of polymer matrix composite materials. ASTM standard. West Conshohocken: ASTM International; 2020.
[36]

Zhou A, Tam LH, Yu ZC, Lau DV. Effect of moisture on the mechanical properties of CFRP-wood composite: an experimental and atomistic investigation. Compos Pt B Eng 2015; 71:63-73.
[37]

Zhao P, Tong Y, Ma N, Han B, Dong X, Qi M. Molecular dynamics simulations and experimental studies of the microstructure and mechanical properties of a silicone oil/functionalized ionic liquid-based magnetorheological fluid. ACS Appl Mater Interfaces 2022; 14(8):10987-97.
[38]

Zhou A, Qiu QW, Chow CL, Lau D. Interfacial performance of aramid, basalt and carbon fiber reinforced polymer bonded concrete exposed to high temperature. Compos A Appl Sci Manuf 2020; 131:105802.
[39]

Liu TJ, Li KX, Zhou A, Yu ZC, Qin RY, Zou DJ. Atomic insight into the functionalization of cellulose nanofiber on durability of epoxy nanocomposites. Nano Res 2023; 16(2):3256-66.
[40]

Zhou A, Yu Z, Wei H, Tam LH, Liu T, Zou D. Understanding the toughening mechanism of silane coupling agents in the interfacial bonding in steel fiber-reinforced cementitious composites. ACS Appl Mater Interfaces 2020; 12 (39):44163-71.
[41]

Park S, Khalili-Araghi F, Tajkhorshid E, Schulten K. Free energy calculation from steered molecular dynamics simulations using Jarzynski’s equality. J Chem Phys 2003; 119(6):3559-66.
[42]

Zhou A, Kang J, Qin R, Hao H, Liu T, Yu Z. Weaving the next-level structure of calcium silicate hydrate at the submicron scale via a remapping algorithm from coarse-grained to all-atom model. Cem Concr Res 2024; 180:107501.
[43]

Zhang B, Duan J, Huang Y, Hou B. Double layered superhydrophobic PDMS-candle soot coating with durable corrosion resistance and thermal-mechanical robustness. J Mater Sci Technol 2021; 71:1-11.
[44]

Chen W, Xu B, Tang Q, Qian S, Bian D, Li H. Preparation and properties of PDMS surface coating for ultra-low friction characteristics. Langmuir 2023; 39 (41):14605-15.
[45]

Bi S, Xu K, Shao G, Yang K, Tian J. Mechanically robust antifouling coating with dual-functional antifouling strategy by infiltrating PDMS into plasma-sprayed porous Al 2O3-Cu coating. J Mater Sci Technol 2023; 159:125-37.
[46]

Kwon JM, Kim WJ, Nam KW, Park SH. Interface nanoarchitectonics of epoxy-siloxane functionalized silica for icephobic and mechanically robust superhydrophobic coating. Appl Surf Sci 2025; 712:164141.
[47]

Liu S, Liu S, Wang Q, Zuo Z, Liang X. Design and synthesis of robust superhydrophobic coating based on epoxy resin and polydimethylsiloxane interpenetrated polymer network. Prog Org Coat 2023; 175:107336.
[48]

Zhang J, Zhang P, Hu Z, Li C, Cui X, Yan B. A fluorinated MXene-doped superhydrophobic coating with mechanochemical robustness, repairable wettability and photothermal conversion for highly efficient anti/de-icing. Chem Eng J 2024; 498:155499.
[49]

Hou N, Guo Q, Zaïri F, Ding N. Mode I crack propagation in polydimethylsiloxane-short carbon fiber composites. Compos Struct 2025; 352:118682.
[50]

Romo-Uribe A, Santiago-Santiago K, Reyes-Mayer A, Aguilar-Franco M. Functional PDMS enhanced strain at fracture and toughness of DGEBA epoxy resin. Eur Polym J 2017; 89:101-18.
[51]

Wan YJ, Tang LC, Gong LX, Yan D, Li YB, Wu LB, et al. Grafting of epoxy chains onto graphene oxide for epoxy composites with improved mechanical and thermal properties. Carbon 2014; 69:467-80.
[52]

Wang Y, Cai Y, Zhang H, Zhou J, Zhou S, Chen Y, et al. Mechanical and thermal degradation behavior of high-performance PDMS elastomer based on epoxy/silicone hybrid network. Polymer 2021; 236:124299.
[53]

Oh CB, Kim BJ, Lee MY. Synergistic interlaminar strengthening of unidirectional carbon fiber-reinforced composites using carbon nanofiber-modified sizing on the surface of PET interleaves. Compos Pt B Eng 2023; 264:110929.
[54]

Jiang X, Zhou C, Su J, Tang S, Li N. Innovative droplet manipulati on techniques on F-L@KH550/HDTMS/PDMS/SiO 2 superhydrophobic surfaces with excellent corrosion resistance using femtosecond laser and chemical modifications. Colloids Surf A Physicochem Eng Asp 2025; 707:135960.
[55]

Liao D, Gu T, Liu J, Chen S, Zhao F, Len S, et al. Degradation behavior and ageing mechanism of E-glass fiber reinforced epoxy resin composite pipes under accelerated thermal ageing conditions. Compos Pt B Eng 2024; 270:111131.
[56]

Sales RCM, Diniz MF, Dutra RCL, Thim GP, Dibbern-Brunelli D. Study of curing process of glass fiber and epoxy resin composite by FT-NIR, photoacoustic spectroscopy and luminescence spectroscopy. J Mater Sci 2011; 46 (6):1814-23.
[57]

Kapridaki C, Maravelaki-Kalaitzaki P. TiO2-SiO2-PDMS nano-composite hydrophobic coating with self-cleaning properties for marble protection. Prog Org Coat 2013; 76(2-3):400-10.
[58]

Sun J, Chen Y, Li J, Su J, Huang S, Zhang Y, et al. Synthesis of transparent and flexible superhydrophobic PDMS-EP/SiO 2 coating with excellent mechanical and chemical durability. Appl Surf Sci 2025; 707:163586.
[59]

Kim J, Kang S, Seong I, Jeon JW, Lee D, Kim JH, et al. Advancing CFRP durability: interfacial and weathering performance of epoxy and acrylic matrices. Compos Part B Eng 2025; 297:112315.
[60]

Harle SM. Durability and long-term performance of fiber reinforced polymer (FRP) composites: a review. Structures. 2024; 60:105881.
[61]

Rosa IC, Firmo JP, Correia JR, Mazzuca P. Influence of elevated temperatures on the bond behaviour of ribbed GFRP bars in concrete. Cem Concr Compos 2021; 122:104119.
[62]

Zhao Q, Zhao XL, Zhang D, Duan L. Effects of exposure in seawater sea-sand concrete pore solution on fatigue performance of carbon FRP bars. Compos Sci Technol 2024; 247:110418.
[63]

Starkova O, Aiello MA, Aniskevich A. Long-term moisture diffusion in vinylester resin and CFRP rebars: a 20-year case study. Compos Sci Technol 2023; 242:110167.
[64]

Li Z, Furmanski J, Lepech MD. Micromechanics modeling and homogenization of glass fiber reinforced polymer composites subject to synergistic deterioration. Compos Sci Technol 2021; 203:108629.
[65]

Liu Y, Wang H, Ding H, Wang H, Bi Y. Effect of hygrothermal aging on compression behavior of CFRP material with different layups. Compos Pt B Eng 2024; 270:111134.
[66]

Litherland KL, Oakley DR, Proctor BA. The use of accelerated ageing procedures to predict the long term strength of GRC composites. Cem Concr Res 1981; 11 (3):455-66.
[67]

Huang J, Aboutaha R. Environmental reduction factors for GFRP bars used as concrete reinforcement: new scientific approach. J Compos Constr 2010; 14 (5):479-86.
[68]

Bank LC, Gentry TR, Thompson BP, Russell JS. A model specification for FRP composites for civil engineering structures. Constr Build Mater 2003; 17(6-7):405-37.
[69]

Chen Y, Davalos JF, Ray I. Durability prediction for GFRP reinforcing bars using short-term data of accelerated aging tests. J Compos Constr 2006; 10(4):279-86.
[70]

Lu Z, Li Y, Xie J. Durability of BFRP bars wrapped in seawater sea sand concrete. Compos Struct 2021; 255:112935.
[71]

Lu C, Ni M, Chu T, He L. Comparative investigation on tensile performance of FRP bars after exposure to water, seawater, and alkaline solutions. J Mater Civ Eng 2020; 32(7):04020170.
[72]

Wang Z, Zhao XL, Xian G, Wu G, Singh Raman RK, Al-Saadi S, et al. Long-term durability of basalt- and glass-fibre reinforced polymer (BFRP/GFRP) bars in seawater and sea sand concrete environment. Constr Build Mater 2017; 139:467-89.
[73]

Guo X, Xiong C, Jin Z, Pang B. A review on mechanical properties of FRP bars subjected to seawater sea sand concrete environmental effects. J Build Eng 2022; 58:105038.
[74]

Chen Y, Davalos JF, Ray I, Kim HY. Accelerated aging tests for evaluations of durability performance of FRP reinforcing bars for concrete structures. Compos Struct 2007; 78(1):101-11.
[75]

Lu Z, Xian G, Li H. Effects of elevated temperatures on the mechanical properties of basalt fibers and BFRP plates. Constr Build Mater 2016; 127:1029-36.
[76]

Micelli F, Nanni A. Durability of FRP rods for concrete structures. Constr Build Mater 2004; 18(7):491-503.
[77]

Ma L, Liu H, Wen X, Szymań ska K, Mijowska E, Hao C, et al. Polyhydric SiO2 coating assistant to graft organophosphorus onto glass fa bric for simultaneously improving flame retardancy and mechanical proper ties of epoxy resin composites. Compos Pt B Eng 2022; 243:110176.
PDF (4878KB)

2934

Accesses

0

Citation

Detail
Sections
Recommended

/

〈 〉