1 Introduction
Cement concrete is a widely used construction material and is integral to civil engineering projects [
1,
2]. Its versatility and affordability make it a popular choice for various infrastructure developments such as bridges, pavements, and tunnels [
3,
4]. Producing per tonne cement has relatively low energy consumption compared to that of other materials like metals, ceramics, and polymers [
5–
7]. With the growing world population and rapid urbanization and industrialization in developing countries like China, India, and Malaysia, the demand for cement is expected to increase by nearly 200% [
8,
9]. Cement concrete offers several advantages, such as high compressive strength, excellent durability, low cost, ease of preparation, compatibility with rebars, and ease of casting, making it an ideal material for diverse infrastructure applications [
10–
12].
However, the widespread application of cement concrete is limited due to other inherent characteristics, such as brittleness, low tensile strength, and poor deformation performance, which contribute to the development of cracks inside the concrete [
13–
16]. These cracks in cement concrete structures can significantly weaken their integrity and bearing capacity, compromising structural safety, resilience, and serviceability, and have drawn much attention from industry and academia, especially as current infrastructure requires larger scale, better resilience, and more complex structures under more aggressive environmental conditions that can exacerbate cracking [
17]. Cracks in cement concrete can occur during both the plastic and hardened states of the material, caused by matrix movement and plastic shrinkage from rapid water loss during the plastic state, as well as chemical reactions, drying shrinkage, thermal stress, and external loads during the hardened state [
18–
22]. Microcracks can propagate, bringing significant potential risks to the functional integrity of concrete structures [
23,
24]. Moreover, harmful solutions and gases can penetrate the concrete through these cracks, possibly leading to the failure of the infrastructure. Additionally, the reinforcement bars can react with the outer environment (e.g., air and water), leading to reinforcement corrosion [
25]. To enhance the durability of concrete, conventional passive maintenance methods are typically conducted to extend the concrete structure’s lifespan. Passive treatment generally refers to using polymers or chemical additives, etc., applied to the damaged area to repair the concrete after damage (e.g., microcracking) has occurred. The detected cracks are injected or sprayed with sealers of chemical mixtures such as polyurethane, acrylics, epoxy resins, and siloxane [
24,
26]. Though passive treatments are used for numerous concrete structures, these treatments have limitations, such as moisture sensitivity, poor bonding with concrete, susceptibility to degradation, and different thermal expansion coefficients, hindering their further application [
27–
29]. Additionally, the direct cost of crack maintenance might be $147/m
3, although initial concrete production costs are in the range of $65–80/m
3 [
30]. More importantly, repairing cracked buildings can be difficult and even impractical, due to the location of the cracks, the size of the cracks, and the requirements for use of infrastructure such as roads and underground structures [
31]. Nowadays, a large amount of investment has been put into the maintenance and rehabilitation of infrastructure in many countries, which is even more expensive than initial manufacturing cost [
32,
33]. It has been reported that the Netherlands spends 1/3 of its annual budget for major construction work on inspection and maintenance, compared to over 45% in the UK [
34]. Besides, passive treatments can only heal the surface of concrete structures, which does not fundamentally mitigate the cracking problem [
35]. Furthermore, the cement and concrete industry has a huge carbon footprint [
34]. Therefore, a more sustainable technique to heal concrete cracking is urgently needed to ensure sustainable use of materials, to save total energy, and to minimize the environmental impact of building structures [
36,
37].
The concept of self-healing concrete was first proposed and studied by the French Academy of Science in 1836 to address the challenges faced by traditional concrete structures [
38]. Since then, extensive research has been conducted on self-healing concrete. This study aims to provide a comprehensive review of recent developments and advances in self-healing concrete for resilient infrastructure.
2 Assessment of self-healing efficiency and regained properties of concrete
In general, the self-healing ability of concrete can be evaluated by filling the damaged region with a self-healing agent, which can lead to an improved microstructure with narrower and fewer cracks, restored mechanical strengths, and regained durability [
39,
40]. Various tools are employed to assess the repair of concrete in this situation, including cameras, microscopes (such as optical and scanning electronic microscopes (SEMs)), X-ray computed tomography (XCT), and mechanical and durability tests [
41–
44]. These evaluation methods are summarized in Fig.1.
2.1 Crack observation
The gradual evolution and expansion of microcracks in cement-based materials under external loads over time can lead to structural failure, typically resulting from loss of strength and durability [
45,
46]. In laboratory settings, crack damage in concrete is typically induced through mechanical loading or exposure to corrosion solutions, and cracks tend to close upon the addition of healing agents or hydrates [
40,
47]. Therefore, cracking characterization is a direct method to assess the self-healing ability of concrete.
The characteristics of cracks, including their number, width, length, and depth, are important indicators for evaluating the self-healing capability of cement composites [
39,
40,
45]. These properties can be easily measured using various imaging techniques, including cameras, microscopes, and SEM. For instance, Xu et al. [
48] used a high-power microscope to assess the crack characteristics of concrete before and after healing, and they found that the crack tended to close after being filled with shape memory (Fig.2(a)). However, their results are limited to a qualitative assessment of cracks and cannot quantify the self-healing capacity. Further, a camera or optical microscope can be used in conjunction with image analysis software to quantitatively characterize the cracks’ number, width, and length (Fig.2(b)), as performed by Su et al. [
49]. Additionally, SEM equipped with an energy dispersive spectrometer (EDS) can further confirm the healing agent/product of the repaired region in concrete [
50,
51], in addition to acquiring the above quantitative information on microcracks. However, the techniques mentioned above are limited in their ability to detect and characterize the depth of cracks in concrete. In this regard, XCT is a perfect candidate tool for observing the internal features of concrete samples. XCT is a nondestructive method that is easy to operate and has been extensively used in recent years to quantify crack healing efficiency. XCT can precisely capture the behavior and monitor the status of admixing healing agents such as hollow fibers and capsules, making it an ideal tool for evaluating the self-healing capacity of cement composites [
44,
52].
Overall, traditional techniques such as cameras and optical microscopes can only provide qualitative analysis of cracks near the surface of a specimen and cannot detect damage inside the specimen. In contrast, XCT is a nondestructive and easy-to-operate method, making it an excellent candidate for quantifying crack healing efficiency. XCT is particularly useful for monitoring the behavior and status of admixing healing agents such as hollow fibers and capsules, and can precisely capture the depth and extent of cracks inside concrete specimens.
2.2 Re-acquisition of mechanical strengths
The mechanical strength of cementitious materials is a crucial factor in determining their practical applications [
53–
55]. Therefore, the recovery of mechanical properties of concrete after healing is a key indicator of its self-healing ability [
39,
40]. As shown in Fig.2(c), microbes can further restore ~4% of the flexural strength than the fiber-enhanced mortar [
56], mainly ascribed to microbial mineralization improves the interfacial transition zone between the fiber and the matrix. Siad et al. [
57] used 0.25 wt.% and 0.50 wt.% carbon nanotubes (CNTs) to modify engineered cementitious composites (ECC), the experimental results indicated that the flexural strength of 0.25 wt.% CNTs-modified composite and 0.50 wt.% CNTs-modified composite can increase by 7.3% and 27.1%, respectively, as compared to ECC without CNTs. Interestingly, Azarsa et al. [
58] reported that the use of crystalline admixtures can result in a significant increase in the 28-d compressive strength of concrete samples (dimensions of Φ100 mm × 200 mm), with a 10% improvement compared to plain concrete. However, the complex treatment and harsh experimental conditions required for the preparation of such self-healing concrete have limited its widespread development and application.
2.3 Recovery of durability
Cracks in cement concrete can allow the penetration of water molecules or corrosive media, which can cause steel rebar corrosion and ultimately result in the failure of the parent concrete [
59–
61]. Therefore, the significance of healed cracks in concrete lies in improving the gas and water-tightness, thereby enhancing durability [
62]. The repair efficiency of concrete is often defined by its transport properties. For instance, Azarsa et al. [
58] developed concrete samples containing crystalline admixtures, and they found a 50% lower water penetration depth than control samples. Similarly, Aspiotis et al. [
63] found that the combination of dicarboxylic acids and expansive additive performed best at long-term age (in terms of capillary water absorption, chloride ingress, and wet–dry cycles tests), which promises considerable durability. In addition, the use of fly ash (FA) and super absorbent polymer (SAP) as admixtures in concrete has been shown to significantly decrease the rapid chloride permeability of concrete, especially with SAP [
64–
66]. This is because SAP can directly block cracks and provide sufficient water for unhydrated cement, thus filling the cracks with hydrates. However, it should be noted that the effectiveness of SAP is primarily influenced by the alkalinity of the cement solution, and the presence of admixing SAP may leave air bubbles inside the concrete after drying [
67–
69].
In summary, a combination of various techniques is recommended to fully evaluate the self-healing ability of concrete. Optical microscopy or SEM can be used for quantitative analysis of crack characteristics, while SEM-EDS can distinguish the products nearby the crack region. XCT is crucial for observing and analyzing the distribution and action of healing materials inside the cement matrix. Mechanical strengths and durability tests provide a comprehensive assessment of the healing effect. However, the accuracy of this method needs further confirmation through extensive experimentation. Overall, these techniques provide useful evidence for the practical application of self-healing concrete, which can significantly improve the gas/water-tightness and durability of concrete structures.
3 Strategies for achieving self-healing cement concrete
3.1 Autogenous healing
Autogenous healing is a natural healing process that occurs in cement-based materials, similar to asphalt, which is another self-healing material commonly used in civil engineering [
70–
74]. Microcracks, typically smaller than 60 µm, can be spontaneously healed by the continuous hydration of residual clinker components such as C
2S and C
3S or the carbonation of dissolved calcium hydroxide (CH) [
75,
76]. Additionally, the swelling and cracking of calcium silicate hydrate (C-S-H) gels and the plugging of cracks by particles may also contribute to autogenous healing, although these mechanisms have received less attention.
The autogenous healing process of concrete is a complex phenomenon that involves three main stages, as reported in previous studies [
41,
77–
79]. The first stage involves the continuous hydration of unhydrated cementitious materials present in the system. In the second stage, the CH present in the concrete dissolves and carbonates due to the exposure to atmospheric CO
2. Finally, in the third stage, portlandite is deposited within the microcracks. These three stages work in concert to fill and heal the microcracks in the concrete. However, during hydration, the hydration products tend to accumulate around the surface of the hydrated cement particles, which can inhibit the entry of water into the anhydrous particles and slow down the rate of hydration. When cracks form, the residual anhydrous cement particles can rehydrate upon exposure to water. In the later stage of autogenous healing, carbonation and precipitation of the hydrates are considered the primary mechanisms for concrete healing [
77,
78]. In contrast, the matrix itself releases very few carbonate ions, and the incoming water will be the only source of carbonate (
) and bicarbonate (HCO
3−). As such, a concentration gradient of
ions is formed from a high value near the crack, which fades to a low value in the matrix as a result of the low diffusion of
ions in water. As a result, more calcium carbonate crystals are formed in the cracks (Fig.3) [
80].
3.2 Autonomous healing
Numerous attempts have been made to create autonomous healing concrete [
81–
83]. To enable cracks in concrete to heal on their own, a specific environment must be created. There are primarily eight approaches to achieving autonomous healing based on the healing mechanism of concrete [
39,
81]: 1) supplementary cementitious materials (SCMs); 2) mineral admixtures; 3) nanomaterials; 4) fibers; 5) polymeric adhesive agents; 6) curing agents; 7) bacteria; and 8) electrodeposition technology. Among these methods, using SCMs, mineral admixtures, nanomaterials, and bacteria are intrinsic to the host-host chemistry that occurs to bridge the microcracks [
79,
80]. In addition, polymer adhesives are typically used to provide an external seal, but in some cases, they can also provide intrinsic sealing properties [
84,
85]. For instance, certain epoxy resins used in polymer-modified concrete may react with alkaline substances or hydroxide ions to promote self-healing [
65].
In summary, self-healing cement concrete includes both autogenous and autonomous healing methods, autonomous healing is a controllable and efficient approach that can be applied in practical applications under the premise of reasonable design, and is capable of healing larger cracks compared to autogenous healing.
4 Development of self-healing concretes and their performance
4.1 Self-healing concrete based on supplementary cementitious materials
In recent years, SCMs, such as FA, silica fume (SF), and ground granulated blast-furnace slag (GGBS), have been widely used to replace some proportion of cement (by weight) to develop low-carbon cement composites and high-performance concrete [
75,
86,
87]. These potentially active SCMs are also used to develop self-healing concrete [
77,
79,
82,
88,
89]. For instance, Maddalena et al. [
90] studied the self-healing property of concrete incorporating 10% SF, 30% FA, and 50% GGBS, and they found that the admixing SCMs can promote the generation of additional C-S-H gel, which can reduce the sample’s water absorption by 68% (SF-modified mixture), 52% (FA-modified mixture), and 61% (GGBS-modified mixture). In addition, according to previous studies [
40,
91], Class-C FA is more appropriate for developing self-healing concrete relative to Class-F FA, because cement concrete prepared with Class-C FA will perform better under freeze–thaw cycles. Notably, it has been reported that GGBS can perform better in terms of self-healing of concrete than using FA at the same substitution rate, mainly due to the higher CaO content and higher pH of GGBS, which contributes to the precipitation of calcite [
79,
92]. Accordingly, the combined usage of FA and GGBS can develop cement concrete with outstanding self-healing ability compared to using FA or GGBS alone; this is mainly attributed to the better particle-filling effect obtained by using binary mixing SCMs [
93–
95].
As such, it can be speculated that binary/ternary hybrid SCMs have great potential to develop self-healing concrete with excellent properties. However, it should be noted that this approach is unable to target heal crack regions in concrete and therefore it exhibits low efficiency; in addition, the low reactivity of SCMs in early-age hydration, which is often dependent on a strongly alkaline environment, also limits the application of SCMs in self-healing concrete.
4.2 Self-healing concrete based on mineral admixtures
In addition to using SCMs in concrete for developing self-healing concrete, adding additives, such as expansive chemicals, crystals, and silica-based materials, also helps to improve the self-healing ability of concrete [
42,
80,
82]. Among these admixtures, expansive chemical additives are considered an ideal candidate to develop self-healing concrete. This is mainly due to the fact that adding chemical additives can fill the voids and cracks of concrete, namely, rapidly increase the volume of the system, which helps to seal cracks, while the formation of crystallization products requires a certain amount of time to produce healing products [
80]. However, it is important to note that the production of healing products is greatly influenced by the environment, including factors such as pH, temperature, and humidity, which can significantly affect the efficiency of concrete repair.
Using a combination of silica-based materials, swelling or crystalline components can be effective in self-healing cement concrete. Previous studies have also shown that a low water-to-binder ratio (W/B) or a high amount of cementitious material can be beneficial for preparing self-healing concrete, as it provides sufficient unhydrated active materials that can undergo further hydration. This can lead to the production of more healing products and improve the self-healing ability of the concrete. Curing conditions can also significantly affect the self-healing behavior of concrete; for instance, continuous water curing performs better in enhancing the chloride ion permeability of concrete than continuous air curing. Higher relative humidity during curing is also beneficial to improve the concrete’s self-healing ability. In addition, concrete with smaller microcracks can be healed better because fewer healing agents or less healing product is needed to easily fill the crack. In general, concrete with cracks exceeding a certain width is difficult to heal by solely adding admixtures.
4.3 Self-healing concrete based on nanomaterials
During the past decade, advances in nanotechnology/nanomaterials have provided further opportunities to increase the self-healing property of cement concrete [
46,
47]. Typically, admixing nanomaterials (e.g., self-healing epoxy nanocomposites, carbon-based nanomaterials, nano-TiO
2, and nano-SiO
2) can significantly enhance the performance, such as mechanical properties, chemical resistance, and transport properties of concrete, with mechanisms as follows. 1) Nucleation effect: nano-additives can promote cement hydration thanks to additional nucleation sites for cement hydration. For instance, Han et al. [
47] reported that the admixing 0D nanoparticles (such as 1% of carbon black), 1D nanofibers (such as 0.1% CNTs), and 2D nanoplates (such as 0.1% graphene) all have a nano-core effect and help strength formation and durability improvement of concrete, as shown in Fig.4(a). The hydrates can also generate around nano-additives, which can accelerate cement hydration, and this positive effect can be improved with dosage increase on the premise of ensuring good dispersion. 2) The admixing nano-additives (such as CNTs) can bridge pores and cracks inside the concrete [
96–
98], as presented in Fig.4(b) [
99]. For instance, Nadiv et al. [
99] found that using a low concentration of CNTs (0.063 vol% and 0.15 vol% for loading straight tungsten di-sulfide CNT and raw CNT, respectively) can improve the mechanical properties of composites by 25%–38%. 3) Cracks near the graphene/cement interface or graphene oxide/cement interface can be deflected due to the high resistance along the crack propagation pathways, creating a crack inhibition and bridging effect [
100–
103]. The admixing nano-additives may have a variety of enhancing effects. As an example, admixing graphene oxide in composites typically has ability of crack inhibition/bridging effect and filler effect [
104,
105]. 4) The admixing active nano-additives, such as nano-SiO
2, has pozzolanic reactivity, which can react with CH to generate additional C-S-H gel, thereby enhancing the compactness of composites [
106,
107], as illustrated in Fig.4(c). Compared with nano TiO
2, nano-SiO
2 is more effective for enhancing the self-healing properties of concrete. For instance, Hou et al. [
106,
107] showed that nano-SiO
2 can enhance the cement hydration rate at a very early age. Still, the rate tends to slow down due to the compact gel structure. In particular, water curing is recommended when admixing nano-SiO
2 to develop self-healing concrete. In addition, high concentrations of Ca ion and high temperatures, are beneficial for enhancing the self-healing ability of concrete. In summary, the above four mechanisms have been proposed to explain how nano-fillers can improve the self-healing properties of concrete. However, the research on concrete containing nano-additives is still limited and needs to be continuously investigated in depth with more advanced equipment.
4.4 Self-healing concrete based on fibers
According to previous studies [
68,
69,
108–
110], adding polymer fibers into cement composite, e.g., polyethylene (PE), polypropylene (PP), and polyvinyl alcohol (PVA), is also an effective way to develop the self-healing concrete. Admixed fibers can limit the expansion of cracks [
111], therefore, improving the self-healing ability of concrete since fewer healing products are needed to fill the cracks [
112–
114]. In addition, fibers play a significant role in bridging cracks through crystallization products, which contributes to the self-healing ability of concrete [
41]. Yang et al. [
13] reported that ECC containing 4.5% PVA fibers exhibited self-healing behavior in cracks with a width of smaller than 150 μm, suggesting that the type and admixture of fibers, as well as the crack width, may be important influencing factors affecting concrete self-healing ability. For instance, Choi et al. [
109] indicated that polar PVA fibers were shown to be better than PE and PP fibers as they restored the water-tightness of the composites and promoted the precipitation of self-healing substances. As noted by Kan and Shi [
110], the number of fibers also significantly affects the self-healing ability because it can affect the number of adherent products. Noted that the fibers’ uniform dispersion is a prerequisite to enhance the self-healing ability of concrete, which is also a hot spot for future research. In addition, the fibers’ length and surface properties greatly influence the self-healing ability of concrete.
4.5 Self-healing concrete based on polymeric adhesive agents
4.5.1 Self-healing concrete based on hollow fibers
In recent years, adding hollow fiber-containing healing agents have shown excellent healing ability in concrete [
108,
115,
116]. As presented in Fig.5(a), Sun et al. [
79] proposed a novel polyvinylidene fluoride hollow fiber structure incorporating oily rejuvenators, which was spun using a single wet-spinning method; they found that admixing fibers can be uniformly dispersed inside the matrix (Fig.5(c)), and the mechanical strength recovery ratio was over 50% when the width of cracks was under 0.3 mm (Fig.5(d)). Similarly, Al-Gemeel et al. [
108] developed a new kind of ECC comprising hollow glass microspheres and PVA fiber. Specifically, the hollow glass microspheres have controllable dimension hollow that prevents air from being encapsulated by the thin glass enclosure of a sphere, and they found that the ECC exhibited greater flexural and compressive strengths than control mixtures.
To summarize, hollow fibers have unique advantages in self-healing cement concrete development due to their ability to carry rejuvenator agents inside, which has been shown to be more effective than solid polymer fibers in enhancing the self-healing ability of concrete. However, the use of hollow fibers can also present challenges, such as poor dispersion due to their shape and brittle structure. During the blending process, fibers can break prematurely, causing the restorative agent to be released before it is needed. Therefore, the use of hollow fibers in self-healing cement concrete should be carefully optimized to achieve the desired results.
4.5.2 Self-healing concrete based on capsules
Adding capsules (containing healing agents) to concrete is another way to develop self-healing concrete [
114,
118–
120]. According to the capsules’ dimensions, they can be classified into micro-capsules and macro-capsules. As shown in Fig.6(a), the preparation of capsulation needs three steps during a typical process [
114]. The healing process of capsules in concrete is as follows: cracks form where damage occurs in the matrix; then cracks rupture the microcapsules, and the healing agent is released into the cracks by the capillary action; finally, the healing agent is in contact with the catalyst to trigger polymerization and seal the crack plane by bonding [
121].
The most commonly used macro capsule containers are spherical macro-capsules, porous carriers, and hollow tubes (glass tubes, ceramic tubes, polymeric tubes, etc.) [
116,
119]. The macro-capsules’ diameter is usually a few millimeters. Qureshi et al. [
84] proposed concentric glass macro-capsules incorporating expansive minerals, which could close large cracks (~400 μm) in cement composites. The crack sealing and the strength recovery of the water-impregnated cement mortar recovered by ~95% and 25% at 28 d, respectively. It is important to note that the granular gelling size may affect the concrete’s performance, especially in fresh mixes, depending on the capsules’ material, amount, and size. For instance, Oh et al. [
123] reported that when a 5% self-healing solid capsule was admixed into a cement composite, 600–2400 μm of capsules did not affect the fluidity of the mixture; however, ~5% fluidity loss was observed when the capsules particles over 300 μm. Recently, Shao et al. [
122] first developed a “smart” capsule and prepared a high flowability mixture for 3D printing. This approach further widens the range of mixture design of cement-based materials used for printing (Fig.6(b)).
In summary, to improve the self-healing efficiency of concrete, the preparation of appropriate capsules is crucial, with considerations given to their shape, size, and content. Ideally, capsules should possess the proper dimensions, healing agents, survivability during blending, good interfacial adhesion, compatibility with the concrete matrix, mechanical and chemical stability, and susceptibility to cracking. However, it is difficult to meet all of these requirements simultaneously, and thus, capsule technology for self-healing concrete needs further optimization. A probabilistic model should be proposed based on the desired healing level to determine the theoretical proportion of capsules required for the matrix. The probability of success can be increased by selecting appropriate capsule size, dosage, shape, and other factors.
4.6 Self-healing concrete based on curing agents
Using internal curing agents is an effective method for developing self-healing cement concrete. Porous lightweight aggregate (LWA) and SAP are the commonly used curing agents [
66–
69,
124]. These agents function as an internal reservoir, absorbing and storing water when sufficient water is available. When moisture gradients occur, the curing agents gradually release the contained water into the unhydrated cement to support continued hydration [
66–
68]. This process significantly reduces the plastic and autogenous shrinkage caused by low
W/
B of concrete. Additionally, the swelling of SAP during water ingress can prevent liquid intrusion, seal cracks, and improve overall water-tightness [
66]. The water storage capacity of LWA is typically 5 wt.%–25 wt.%, while SAP has ultra-high water adsorption properties and can absorb over 1000 times its weight [
69,
115].
According to Powers’ model, when the
W/
B is at or above 0.36, the binder is fully hydrated. However, when the W/B is below 0.36, the hydration extent of a sealed curing cement paste is limited by the lack of water [
125]. For instance, a cement paste with a
W/
B of 0.30 can only reach a hydration degree of 0.73 due to the absence of water (Fig.7(a)). However, addition of extra internal curing water at 3.20% and 7.36% of the total water volume increases the hydration degree to 0.77 and 0.83, respectively (Fig.7(b) and Fig.7(c)). The maximum theoretical hydration degree increases with the volume of internal curing water, but it is virtually unchanged at 0.83 when the internal curing water reaches the limit value (7.36%) due to the absence of pore space (Fig.7(d)). Therefore, the content level of internal curing water in curing agents, such as LWA and SAP, is critical for the success of self-healing concrete.
As shown in Fig.8, the admixing LWA can absorb some water from the fresh cement paste through capillary suction during the first hours (see also the ring-like domain of recipients) [
124]. This water was released from LWA in the later stages of hydration, contributing to developing late-age strength for concrete [
115]. The adsorbed water also facilitates the mixture’s workability. Indeed, Henkensiefken et al. [
125] found that the flowability of the mixture containing 50% expanded glass aggregate can increase by 6.3% compared to the plain mix. However, its compressive and flexural strength dropped by 54.3% and 47.6%, respectively, mainly caused by the low strength of the LWA themselves. Typically, a fine LWA with a larger specific surface area is more conducive to internal curing than a coarse one since it offers a more uniform distribution of additional curing water.
According to previous studies [
66–
69], incorporating SAP can enhance workability, freeze/thaw resistance, and later hydration degree of cement and decrease autogenous shrinkage properties of concrete. SAP has been widely used in developing ultra-high-performance concrete (UHPC), but it is difficult for external moisture to migrate to the interior of the matrix because SAP has only extremely small size defects. The introduction of SAP can offer sufficient water for cement particles hydration, thus significantly reducing shrinkage. For example, Yu et al. [
64] reported that the 7-d of autogenous shrinkage and 28-d total shrinkage of UHPC with 0.15% SAP were 51.3% and 30.1% lower, respectively, than those of plain samples. Compared with control samples, the compressive strength of concrete containing curing agents increased, while that of other concrete decreased. This can be explained by the fact that a suitable amount of internal curing water can increase the binder hydration degree, while excessive curing water results in some spherical capillary pores. To date, there are still some challenges for SAP-enabled self-healing concrete to overcome, relating, for example, to the formation of SAP clusters, the need for additional water to compensate for the swelling of SAP, and the negative effect of macropore formation.
4.7 Self-healing concrete based on microbiology
CaCO
3 precipitation is a common natural phenomenon found in soils, marine water, and cementitious composites, and is associated with various bacterial species [
42,
45,
49,
119]. Bacteria can generate CaCO
3 through several metabolic pathways, including the oxidation of organic compounds, hydrolysis of urea, and nitrate reduction [
126]. The metabolically active bacteria can consume oxygen, contributing to the corrosion resistance of steel in reinforced concrete (RC) structures. Simultaneously, the generated CO
2 can react with CH, ultimately producing a mineral based on calcium carbonate. This property makes it possible to use bacteria to heal concrete cracks by precipitating minerals, mainly CaCO
3. The process can effectively seal the cracks and restore the structural integrity of concrete. Additionally, the resulting CaCO
3 provides protection against further deterioration, making the concrete more durable and resistant to environmental factors. This self-healing mechanism using bacteria is an innovative and promising approach for enhancing the lifespan and sustainability of concrete structures. As suggested in Fig.9 [
127], the self-healing agent, bleaching earth and bacterial suspension (BE + BS), is uniformly distributed in the cement composite, and the bacteria remain inactive and dormant in the pores of BE (BS) when the cement composite is intact (Fig.9(a), II). Once cracks emerge on the cement composite surface, the dormant but living bacteria are reactivated by oxygen and water from the external environment and begin to grow and multiply by utilizing the nutrients in cracks (Fig.9(a), III). Bacteria that produce urease catalyze the hydrolysis of urea to carbonate ions (
) and ammonium (NH
4+). The negatively charged microorganisms tend to adsorb calcium ions (Ca
2+), and
can react with free calcium ions (Ca
2+) in the cracks to form CaCO
3 crystals (Fig.9(a), IV). Finally, these CaCO
3 sedimentations are used to fill cracks in concrete (Fig.9(a), V).
Self-healing of concrete using microbial technology requires the use of bacteria, nutrients, and nitrogen or carbon sources [
44,
126,
129]. The selection of suitable bacteria and calcium/nitrogen sources is crucial for survival in the harsh environment of dry, and alkaline concrete. Several ureolytic bacteria, such as
Bacillus sphaericus,
Bacillus subtilis, and
Bacillus megaterium, are commonly used to develop self-healing concrete. Although their mechanisms are similar, different species of bacteria significantly impact the self-healing properties of concrete, mainly because of the significant differences in their activity and the degree of reaction with cementitious materials. For instance, Soda et al. [
85] developed a novel bacteria via encapsulation and dual coating, and the encapsulated bacterial expanded perlite particles were added to cement mortar in the proportion of 10%, 15%, 20%, and 25% (equal weight to replace fine aggregate), they found that adding 20% expanded perlite improved the healing of microcracks ranging in size from 0.154 to 0.626 mm.
Additionally, González et al. [
128] used calcium lactate bacillus and
pseudofirmus bacteria as a nutrient to study their co-effect on the healing ability of cement composites (Fig.9(b)), and results demonstrated that most of the healing occurs in the first 21 d of curing. This means that some bacteria may not be able to survive in the later stages of high alkalinity hydration. Wang et al. [
44] prepared the self-healing concrete using hydrogel-encapsulated bacteria; they found the healing rate of concrete varied from 70% to 100%, far higher than the control samples. They used 3D high-resolution XCT to further explain the effect of bacteria; as indicated in Fig.9(c), the healing products (yellow marked) are mainly distributed in the superficial layer, while the content of healing products sharply decreased in the subsurface [
44]. This means that bacteria may only be used to heal surface cracks in concrete and are ineffective in healing deep (internal) cracks. It is found that the presence of bacteria in the mix design step and curing solution improves the tensile strength of the shotcrete and reduces its water absorption, permeability, and porosity. However, adding these two bacteria into fresh concrete is not suitable as the pH inside the concrete is higher than the optimum pH for them. Each metabolic pathway has both advantages and disadvantages, so we should carefully choose the most appropriate option in light of the actual situation.
4.8 Self-healing concrete based on electrodeposition
One approach to healing cracks in RC structures involves the application of a weak direct current between the rebar and an external electrode (an anode). This process results in the electrodeposition of insoluble inorganic compounds, such as Mg(OH)
2, CaCO
3, and ZnO, around the rebar, forming a barrier that fills the cracks and seals the concrete surface. This barrier also helps prevent further corrosion of the RC structure. Therefore, this method shows promise for improving the durability and lifespan of concrete infrastructure [
129,
130]. As shown in Fig.10, a typical device includes an external electric field and electrolyte solution. Matsubara and Kamimura [
129] found that the electrodeposits that appeared on the sample surface over time were concentrated along the iron bar, mainly due to the pores of the samples forming a region of high current density during the charging process and the occurrence of local electrolysis around the pores. A high current density zone may have formed around the pores and local electrolytic reactions were intensively and selectively triggered in and around the pores, during the loading process. This mechanism exhibits the potential of using electrodeposition technique as a pore-filling and crack-healing tool.
Some parameters such as
W/
B ratio, current density, electrolyte solution, and applied voltage significantly affect the healing efficiency of concrete [
129,
131]. Typically, high levels of W/B, current density, concentration of the electrolyte solution, and solution temperature, can promote crack closure. Increasing W/B also contributes to crack closure since it increases the porosity and reduces the concrete resistivity, resulting in a higher electric current. According to a previous study [
129], ZnSO
4 and MgCl
2 are the most appropriate solutions for precipitating deposition products inside and outside cracks of cementitious composites. Jiang et al. [
132] evaluated the influence of electrodeposition technology on the self-healing efficiency of concrete. They found that the total voids of porous concrete had little effect on the healing effects of electrodeposition at an early age. It is implied that this technique may be less suitable for self-healing porous concrete. Overall, RC’s mechanical and durability performance can be enhanced to some extent by appropriately adopting the technical parameters of electrodeposition technology.
5 Summary and prospects
Concrete structures are prone to cracking and degradation during service, leading to significant maintenance costs. Bacteria-based autonomous healing is a promising solution to this problem, with potential benefits such as increased infrastructure resilience and reduced costs and environmental burden. However, the size limitation of bacteria is a major barrier to its practical application.
Self-healing cement concrete includes both autogenous and autonomous healing methods, with the former limited to healing microcracks less than 60 μm in width and the latter having the potential to heal larger cracks. However, selecting the best healing method for different concrete infrastructures remains challenging due to the advantages and disadvantages of each approach.
Issues such as high additional costs, low healing efficiency, and long-term reliability instability must also be addressed. Most studies on concrete self-healing have been conducted in laboratory settings and focused on static loading, which differs from actual continuously loaded concrete structures in service. Future research should focus on the self-healing behavior of concrete under dynamic loading in real-world service environments.
Advanced techniques, including multiscale and multiple healing, may facilitate the development of self-healing concrete. Finally, a probabilistic model should be developed to determine the theoretical proportion of capsules required to achieve a target healing level and increase the hit probability.