Effect of Exposure Environment and Calcium Source on the Biologically Induced Self-Healing Phenomenon in a Cement-Based Material

Abstract

Microbially induced calcium carbonate precipitation (MICP) presents a sustainable, environmentally friendly solution for repairing cracks in cement-based materials, such as mortar and concrete. This self-healing approach mechanism enables the matrix to autonomously close its own cracks over time. In this study, specimens (50 mm in diameter and 25 mm in height) were exposed to submersion and a wet–dry cycle environment. The solution considered a nutrient-rich suspension with calcium lactate, urea, calcium nitrate, and Bacillus subtilis or Sporosarcina pasteurii in a biomineralization approach. The self-healing efficiency was assessed through optical microscopy combined with image processing, focusing on the analysis of the superficial crack closure area. S. and B. subtilis exhibited notable capabilities in effectively healing cracks, respectively, 8 mm² and 5 mm² at 35 days. Healing was particularly effective in samples placed in a submerged environment, especially with a 69 mM concentration of calcium lactate in bacterial suspensions containing B. subtilis, where 87.5% of a 4 mm² crack was closed within 21 days. In contrast, free calcium ions in the solution, resulting from anhydrous cement hydration, proved ineffective for S. pasteurii biomineralization in urea-rich environments. However, the addition of an external calcium source (calcium nitrate) significantly enhanced crack closure, emphasizing the critical role of calcium availability in optimizing MICP for bio-agents in cement-based materials. These findings highlight the potential of MICP to advance sustainable self-healing concrete technologies.

1. Introduction

In the global construction industry, cement holds a predominant position, driving substantial energy demand, especially in the heating of clinker furnaces. This process, reliant on fossil fuels, contributes significantly to carbon dioxide emissions; it means cement manufacturing is responsible for more than 3 Gt/year of CO₂ emissions [1,2]. Reducing CO₂ emissions in clinker production can occur at different stages, presenting opportunities for advancements and alternatives [3]. Strategies to diminish CO₂ emissions also extend to the manufacturing of various cement types and their utilization. Supplementary cementitious materials play a pivotal role by decreasing clinker content while enhancing cement performance [4]. Moreover, alternative approaches like self-healing concrete align with carbon dioxide capture concepts and offer solutions for enhancing building durability [5,6].

The self-healing phenomenon relies on the material’s inherent ability to seal cracks on its own, whether through autogenous or autonomous processes [7]. Autogenous self-healing involves mechanisms such as anhydrous cement hydration and carbonation of free calcium hydroxide, while autonomous self-healing utilizes substances like vascular networks, hydrogels, capsules, and bacteria to promote crack closure [8,9,10,11,12]. The use of bacteria, such as Sporosarcina pasteurii and Bacillus subtilis, in self-healing approaches has been extensively discussed [13,14,15,16]. S. pasteurii and B. subtilis, are Gram-positive spore-forming bacteria, noteworthy for thriving in highly alkaline environments [17,18].

Studies on bacterial contributions to crack healing emphasize the importance of selecting suitable bacterial strains and optimizing conditions such as pH, temperature, and nutrient availability to enhance healing efficiency [19]. The system saturation index, which is affected by microorganisms and organic matter content, plays a crucial role in microbiologically induced calcium carbonate precipitation (MICP) [20]. The type of nutrient provided significantly influences the metabolic pathways involved in calcium carbonate production, with calcium lactate and urea commonly used for their specific benefits in promoting calcite precipitation [14,15,16,21].

The exposure environment significantly influences bio-agent-modified concrete self-healing, affecting calcium carbonate crystal formation [5,22]. The authors in [23] highlight water’s impact on crack self-healing, observing the complete crack closure when submerged in deionized water for one year. Wetting and drying cycles simulate conditions for bacterial activation and dormancy, aiding calcium carbonate precipitation. Conversely, constant submersion enhances bacterial activity but may dilute the healing agent [24]. Understanding these dynamics is crucial for optimizing self-healing efficiency.

Anhydrous cement hydration not only strengthens concrete but may also provide a continuous source of calcium at early ages [25]. Bacteria-mediated calcium carbonate precipitation relies on free calcium ions, which can be endogenous (released from the cement matrix during anhydrous cement hydration) or exogenous (supplied externally, e.g., with calcium lactate), acting as precursors for calcite crystal formation [26]. Nevertheless, the impact of endogenous calcium ions on bio-self-healing remains unexplored [27]. Consequently, this research aims to fill this gap by investigating how endogenous calcium ions from cement hydration influence the effectiveness of bacteria-mediated self-healing processes in concrete.

The central question driving this study was the following: Is the calcium source from cement hydration alone sufficient to promote crack closure in cement-based materials? The study investigates the interactions between exposure environment, calcium ions, and nutrient type on the self-healing properties of a bio-agent-modified cement-based matrix using two bacteria: Sporosarcina pasteurii and Bacillus subtilis. Analyzing wet–dry cycle and submersion conditions, and nutrient sources, including calcium lactate versus urea with calcium nitrate, aims to elucidate optimal conditions for MICP. Additionally, it examines the role of calcium ions from anhydrous cement hydration in bio-mineralization, contributing to advancements in bio-agent-modified concrete technology and sustainable civil engineering. Thus, the main objectives are to (a) analyze the influence of exposure environments, and (b) investigate the effect of different calcium sources on the precipitation of calcium carbonate related to self-healing in cement-based materials.

2. Materials and Methods

2.1. Microorganisms

Microorganisms Sporosarcina pasteurii (ATCC 11859) [28] and Bacillus subtilis (ATCC 6633) [29] were provided by Tropical Culture Collection—André Tosello Foundation, following ATCC (American Type Culture Collection) guidelines.

2.1.1. Culture Medium and Solutions

Bacteria were cultivated in two different culture media, B. subtilis in Luria Bertani (LB) medium and S. pasteurii in Tryptic Soy Broth (TSB) supplemented with 83 mM urea (TSB+urea), both incubated at 130 rpm for 24 h at 30 °C. LB Medium containing 1% tryptone, 0.5% yeast extract, and 0.5% NaCl was used. Tryptic Soy Broth medium with urea 0.5% (TSB+urea) composed of casein peptone 1.7%, dipotassium hydrogen phosphate 0.25%, glucose 0.25%, sodium chloride 0.5%, and soy peptone 0.3%. Subsequently, cell concentrates were obtained by centrifugation at 7000 rpm for 7 min at 20 °C, washed, and centrifuged twice with 0.85% NaCl solution. Calcium lactate and urea concentrations were based on previous studies [30,31], which are influential works in the bio-self-healing research area.

2.1.2. Cell Concentrations

Cellular concentrations were measured by spectrophotometry in a Thermo Fischer BioMate 35 UV–visible spectrophotometer equipment, manufactured in Waltham, MA, USA. Solutions containing bacterial cells and nutrients were prepared and adjusted to OD600 of 0.2. Cell count was confirmed by colony-forming units (CFU) in the respective culture mediums, achieving around 3 × 10⁷ cells/mL [32].

2.2. Mortars

2.2.1. Materials and Mix Proportion Design

The cement composition excluded supplementary cement materials since their presence can impact carbonate precipitation by bacteria in the self-healing process, in both physical characteristics and chemical interactions [33], such as the free calcium ions in the sample environment as discussed by [34]. There are 1.06% MgO and 4.69% SO₃ in the chemical composition and 3.09% loss of ignition in the cement. The particle size distribution has an average diameter of 11.35 μm, with D10 and D90 values of 0.31 μm and 32.17 μm, respectively. The cement’s mechanical strengths at 3, 7, and 28 days are 29.32 MPa, 35.15 MPa, and 42.08 MPa following the [35] requirements.

Quartz sand is a fine aggregate with a fineness modulus of 1.89, maximum size of 2.36 mm, specific gravity in the dry state equal to 2.54 g/cm³, and water absorption of 1.30%, following, respectively, [36,37,38]. The tap water used in the mixtures was supplied by the municipal public company and had an average pH of 6.

Table 1. Mortar composition.

Cement	Sand	w/c	fc3 ± SD (MPa) *	fc28 ± SD (MPa) *	Consistency (cm)
1.00	1.37	0.40	33.56 ± 1.68	43.76 ± 1.78	35.0

* fc represents the concrete’s strength at 3 days (fc3) and at 28 days (fc28).

shows the mortar composition in mass kg/m³.

The mortar consistency index was tested according to the [39] requirements. Mortar composition was previously used in works by the research group [40,41,42]. The mix proportion includes the mortar fraction of a concrete mix, along with an adjustment for water due to the coarse aggregate’s absorption.

2.2.2. Test Specimens, Characterizations, and Crack Creation

The mixture preparation followed [43], which is equivalent to [44]. Cylindrical specimens were produced with a diameter of 50 mm and a height of 100 mm for compressive strength tests. For analyzing the self-healing phenomenon, each specimen was cut into four slices, each measuring 25 mm. The top and bottom slices were discarded, leaving the two middle slices for use in the experimental program. Figure 1 shows the procedure to create cracks artificially, map the crack mosaic, the exposure environment, and three stages of the experimental program. Cracks were created by diametrical compression [45].

Samples underwent a diametral splitting procedure to induce a crack in their central portion, as shown in Figure 1A. After the cracking process, the samples were dried, and the crack was divided into eight segments using nylon threads placed at the widest crack opening. This step mapped the crack length and prepared the sample for optical microscopy analysis.

2.3. Exposure Environment

This research compared different exposure environments, including submersion and wet–dry cycles (Figure 1B,C). Samples were consistently submerged in two conditions: Bacteria Submerged (BS) and Control Submerged (CS). In BS, samples were immersed in a liquid mixture of 100 mL nutrient-rich bacterial suspensions, while CS involved submergence in 100 mL distilled water.

For wet–dry cycles, samples underwent two days of submersion followed by five days in an open-air laboratory environment. In its first cycle, the Bacteria Cycle Unique (BCU) environment received a 100 mL nutrient-rich bacterial suspension and 100 mL distilled water in subsequent cycles. The Bacteria Cycle Multiple (BCM) environment received a renewed 100 mL nutrient-rich bacterial suspension each wet period. Control Cycle (CC) samples underwent wet–dry cycles with 100 mL of distilled water renewed at each wet cycle. Seven cycles (35 days) were conducted for each analysis stage.

For all exposure environments containing microorganisms, B. subtilis was the microorganism, and calcium lactate was the nutrient and calcium source, respectively, 3 × 10⁷ cells/mL and 23 mM.

The samples were placed in individual containers, meaning each sample had its own container. All containers were kept at room temperature, approximately 23 ± 2 °C.

2.4. Nutrient Type and Concentration

This experiment was designed to analyze the influence of various nutrient sources and concentrations in a submerged exposure environment (Figure 1D). Samples are separated into two submersion groups: control (CS), in which the samples were submerged in 100mL of nutrient solution without bacteria (nutrients and saline solution); and with bacteria (BS), in which the samples were submerged in 100 mL of nutrient-rich bacterial suspension (bacterial cells, nutrients, and saline solution). B. subtilis and S. pasteurii were used at a concentration of 3 × 10⁷ cells/mL. Calcium lactate (L) was used only with B. subtilis at concentrations 23, 46, and 69 mM (L23-69). Urea without and with calcium nitrate (U and UN, respectively) at equal molarity were used only with S. pasteurii at concentrations 167, 333, and 666 mM (U167-666 and UN167-666). CS samples without nutrients (saline only) are indicated with 0. Numbers in acronyms indicate the concentration of nutrients. For example, one sample with B. subtilis and calcium lactate 23 mM is indicated as BS-L23. Solutions were prepared using powered ingredients by mass and water.

2.5. Cement- and Nutrient-Derived Calcium Sources

The experiment assessed the impact of calcium ions derived from cement hydration and nutrient sources (Figure 1E). S. pasteurii was selected due to its ability to grow and produce carbonate with only urea, allowing the exploration of cement-derived calcium ions in calcium carbonate precipitation. Calcium nitrate was used as an exogenous calcium source. The same samples initially exposed to only a source of urea (Figure 1D) were dried and, after six months, exposed to a submerged environment rich in urea and calcium nitrate (UN samples).

2.6. Self-Healing Phenomenon Evaluation and Image Processing Analysis

Crack closure was assessed through optical microscopy for cracked area measurements. Zeiss microscope model Stemi 508, a Greenough Stereo Microscope with 8:1 zoom, manufactured in Germany, was utilized for microscopy analyses. Unlike conventional methods used in the literature, which often analyze hand-picked points or short segments, our approach involved assessing the full extension of the crack. This comprehensive microscopy analysis captures nuances that other methods may overlook, leading to a much more robust approach. Sample preparation and image processing with ImageJ software, version 1.8.0, were followed by the automated procedure described by [46], as shown in Figure 2.

2.7. Statistical Analysis

Statistical analyses of cracked areas were conducted with PASW Statistics 18 and GraphPad Prism 8.0. PASW Generalized Estimating Equations (GEE) with Bonferroni correction were employed to assess the significance of the time and treatment effects on crack closure patterns. Graphs, along with the 95% Confidence Band analysis, were generated using GraphPad Prism software, version 10.3.0.

2.8. Thermal Analysis (TGA/DTG)

Thermogravimetry is a technique utilized to identify compounds in cementitious materials during the observation of the self-healing phenomenon [47]. Calcium carbonate exhibits a characteristic behavior, with decomposition occurring at temperatures above 600 °C [48]. Three samples were analyzed: a control sample of CaCO₃, obtained from Sigma Aldrich with a purity of 99% or higher, and samples consisting of crystals formed at the bottom of the reservoirs used in the Nutrient type and concentration tests, and the cement- and nutrient-derived calcium sources tests for each microorganism (B. subtilis and S. pasteurii). The deposited material underwent washing with distilled water and drying in an oven at 35 °C until a constant mass was achieved. Approximately 10 mg of powdered samples were placed in open alumina crucibles and exposed to airflow. The temperature gradually increased at a rate of 20 °C/min, ranging from room temperature to 1000 °C under airflow. Thermogravimetric analysis was conducted using a TGA 2 analyzer equipment from Mettler Toledo (Columbus, OH, USA).

3. Results

3.1. Observations About the Exposure Environment

A combination of B. subtilis with calcium lactate as a nutrient and calcium source was employed for calcium carbonate precipitation. This approach shows the cracks at 0 (initial crack age) and 28 days of mortars in the CS, CC, BS, BCU, and BCM exposure environments. Regarding the profile of the crack, as shown in Figure 3, they are all very similar: a larger opening at one end, comprised of a single crack segment, straight, and uniformly distributed across the sample. The BCM sample has a slightly wider crack opening compared to the others Figure 3. As discussed later, the BS environment was the one that most favored the superficial crack closure, where not only moisture but also the activity of the bio-agent and nutrient played a pivotal role.

Figure 4 illustrates the crack areas from the exposure environment evaluation experiment, presented in two ways: (A) the raw area and (B) the variation in the cracked area relative to the initial crack area. Initial crack areas across all samples are notably similar (3.2–4.9 mm²), as shown in Figure 4A. In Figure 4B, the dashed line delineates reductions and increases in the cracked area, with negative values indicating crack healing. Comparing healing profiles between all samples at days 0 and 28, BS exhibited the most significant increase in crack closure. In control environments, CS and CC, crack closure yields are comparable but did not achieve complete crack closure. Samples in wet/dry cycles with B. subtilis suspensions (BCU and BCM) exhibited minimal change in their crack areas and did not exhibit higher healing yields than the control (CC). The exposure environment significantly influences the crack closure rate, regardless of the presence of a healing agent [41]. The findings align here with those of [41], indicating that submerged conditions are ideal for MICP in self-healing approaches. Submerged conditions often yield favorable results in crack closure analysis [49] and durability-related assessments, such as chloride penetration [50]. Wet and dry cycles may be evaluated case by case, depending on the system under analysis, due to factors such as the activation of hydration reactions [42] or variations in wetting periods [40]

Based on the results of the exposure environment experiment, the submerged exposure environment (BS) was chosen to assess the nutrient type and concentration. This test aimed to evaluate the nutrient content and the efficiency of the pyruvate (calcium lactate) and ureolytic (urea) metabolic pathways for calcium carbonate production by the microorganisms B. subtilis and S. pasteurii, respectively.

3.2. Findings About the Nutrient Type and Concentration

Figure 5 displays the cracks at their initial stage (0 days) and after 35 days of submersion in each exposure environment. The crack profiles within treatment groups (calcium lactate and urea) are highly consistent: similar width distribution, with a larger opening at one end, consisting of a single crack segment, straight and uniformly distributed across the sample. Although some samples exhibit bifurcations and parallel crack segments, such as “Urea 167 mM sample N”, these variations did not impact the analysis.

All cracks in the calcium lactate solution exhibited closure to some extent (Figure 5). Figure 6A displays data in three different groups (indicated by colors black, blue, and red) in an ascending crack closure order: CS-0, CS-L23, CS-L46, CS-L69, BS L-23, BS-L46, and BS-L69. Initial crack openings with smaller dimensions showed more significant or total closure at the final analysis age, a behavior also observed by [50]. According to [51], crack closure in the presence of nutrients alone may be sufficient for openings between 0.2 and 0.3 mm; however, combining microorganisms and nutrients enhances efficiency, closing openings between 0.3 and 0.4 mm. These findings align with those of [15], validating that an aqueous solution containing calcium lactate enhances the closure of cracks measuring up to 0.3 mm in thickness within mortars with B. subtilis. Moreover, the observed crack closure behavior in Figure 6B could be attributed to the calcium content and the microbiologically induced calcium carbonate.

Analyzing the influence of calcium concentration within each system (CS-0, CS-L23, CS-46, and CS-L69) provided by calcium lactate, crack closure extent shows a proportional relationship with its concentration, with CS-L69 showing the highest surface crack closure (Figure 6B). In environments with lower molarities of calcium lactate (23 and 46 mM), no statistical differences were observed between the presence and absence of bacteria until day 28. Conversely, a high calcium lactate content with bacterial suspension in BS-L69 demonstrated a statistical difference from the initial comparison, observed 7 days after the crack opening age. By day 35, BS-L69 displayed the most effective crack closure performance among the analyzed environments, achieving almost complete crack closure. These findings suggest that calcium content within the environment may influence crack closure and extend to matrix densification, as [52] outlined. In their study, mortars containing various concentrations of calcium lactate and bacterial cells were analyzed, specifically S. pasteurii and Bacillus sphaericus, establishing a significant correlation between mechanical strength and the nutritional ratio.

In Figure 6C,D, concerning the samples in urea environments, there are three distinct groups (indicated by colors orange, green, and black) arranged in ascending order based on crack closure: BS-U333, BS-U666, BS-U167, CS-U167, CS-U666, CS-U333, and CS-0. Unlike the calcium lactate environments, the CS-0 environment exhibited the highest performance. At 35 days, the CS-U167, CS-U333, and CS-U666, without bacterial suspension, demonstrated a closure like that of the CS-0 environment. In conditions featuring S. pasteurii suspension, there was either no change or an increase. Considering these insights, it is possible to attribute the varying behaviors to a primary factor: the influence of byproducts resulting from bacterial carbonic acid production.

Ureolytic bacteria utilize urea as an energy source by enzymatically breaking it down through ureases, producing ammonia and CO₂ as byproducts. Subsequently, the CO₂ is converted into carbonic acid, which reacts with free calcium ions in the solution, forming calcium carbonate. However, ammonia harms the cementitious matrix, acting as a “corrosive agent” [23,53]. The supersaturation of carbonic acid, caused by a deficiency of calcium ions, lowers the system’s pH and inhibits the formation of cement hydration products. Thus, crack closure can be hindered in environments with higher bacterial conversion of urea into ammonia and a deficiency of calcium ions. This phenomenon was observed in conditions featuring urea and bacteria (BS-U167, BS-U333, BS-U666). Among these, at 35 days, the CS-U333 condition exhibited the most positive change in the cracked area, indicating a more ineffective healing process (Figure 6D). This behavior can be attributed to the optimal urea concentration of 333 mM, as [30] noted, which promotes substantial S. pasteurii growth and activity. Consequently, a high cell concentration can efficiently degrade a significant amount of urea, leading to increased ammonia and carbonic acid levels in the environment. A closure profile like CS-0 was observed in conditions without bacteria but with urea, indicating that urea alone has minimal to no influence on the healing process.

Regarding environmental calcium content, Ref. [12] identified B. subtilis as capable of precipitating calcium carbonate solely from a calcium source within the hardened matrix. This aligns with our findings, as no healing was observed without a calcium source in the urea conditions with bacteria. Fortunately, for the S. pasteurii samples, an external calcium source, such as calcium nitrate, can assist in the microbiologically induced precipitation of CaCO₃ [20].

3.3. Influence of the Cement- and Nutrient-Derived Calcium Sources

This experiment assessed whether supplementing equimolar amounts of calcium nitrate and urea is sufficient to induce calcium carbonate precipitation in the presence of ureolytic bacteria. Figure 5 displays the cracks at 0 days (day 35) and after 28 days (day 63) of submersion in various exposure environments. The mortars used for this experiment were the same as those evaluated in the “nutrient type and concentration” experiment. This approach aimed to minimize the influence of cement-derived calcium ions on assisting the autogenous self-healing process [25]. By supplementing these samples with urea and calcium nitrate, only the influence of nutrient-derived calcium on the self-healing phenomenon can now be observed.

Figure 6F showcases data in three different groups (indicated by colors black, purple, and grey) in ascending order: CS-0, CS-UN666, CS-UN333, CS-UN167, BS-UN167, BS-UN666, and BS-UN333. As expected, healing was observed only in conditions with bacteria (BS-UN333, BS-UN666, BS-UN167) (Figure 6E,F). These results suggest that even “older” cracks can be closed without the contribution of matrix autogenous healing. The observed crack closure behavior could be attributed to the calcium content and microbiologically induced calcium carbonate formation. Additionally, the source of calcium can influence the activity of calcite-promoting agents, as noted by [11] when analyzing CaCO₃ precipitation induced by S. pasteurii in two nutrient-rich environments: urea with calcium lactate and urea with calcium nitrate. Mineral production was twice as high in the calcium nitrate environment after 7 days, consistent with the findings comparing calcium lactate and urea with calcium nitrate, where the latter exhibited significantly higher performance. This difference can be attributed to two main factors: the concentration of available calcium ions for calcium carbonate formation and the metabolic route involved in carbonate production. The highest concentration of calcium lactate is nearly 10 times smaller than calcium nitrate, significantly influencing the self-healing process. Furthermore, diverse microorganisms may exhibit varying behaviors when exposed to different calcium sources. The authors [25] observed that Bacillus alcalophilus demonstrated increased CaCO₃ precipitation in an environment rich in calcium lactate compared to conditions with calcium nitrate. This difference can be attributed to the lack of nitrate-degrading genes in the genome of B. alcalophilus, rendering it unable to use nitrate as a nutrient source for carbonate production.

Figure 6I shows the differences between treatment yields. The graph depicts the initial and final areas, with the solid red line indicating no change in the area (Δ = 0). Points above this line indicate crack opening, while points below it represent closed cracks. The segmented red line corresponds to the lower limit of the 95% prediction band for control samples, meaning everything below this line differs from the control regarding healing. Therefore, the graph can be interpreted as follows: all points between the dotted line and the solid red line exhibit healing equal to that expected in a sample with only an autogenous healing contribution. Points below the dotted line demonstrate healing beyond what is expected with autogenous healing alone, suggesting the presence of external agents that enhance healing, thereby contributing to autonomous healing.

Despite the more significant closure of cracked areas, conditions with nutrients but without bacteria did not perform differently from the control without nutrients within the 95% prediction band, falling within the expected range for samples undergoing only autogenous self-healing (Figure 6I). The samples that statistically differed from the control performance were those with nutrients and bacteria, with nutrient-derived calcium ions added to the environment, representing autonomous self-healing conditions. Among these autonomous healing performances, all results with calcium lactate were much closer to the statistical difference limit than those with urea and calcium nitrate, which exhibited almost complete closure.

3.4. Thermal Analysis (TGA/DTG) of Calcium Carbonate Behavior

In Figure 7, calcium carbonate (CaCO₃) powder samples originated from four distinct sources: (1) pure calcium carbonate (98% purity), (2) B. subtilis + calcium lactate, (3) S. pasteurii + urea, and (4) S. pasteurii + urea + calcium nitrate. Decarbonation of calcium carbonate occurred at 765 °C for pure CaCO₃, 800 °C for B. subtilis, 775 °C for S. pasteurii + urea, and 793 °C for S. pasteurii + urea + calcium nitrate, consistent with the mineral behaviors reported by [47]. Notably, a peak at 199.33 °C is evident in the S. pasteurii samples, suggesting probable urea decomposition [54], indicating residues from the urea-rich environment. These results underscore that the observed differences in yields resulted from the previously discussed influences, as the resulting crystals exhibited the same composition despite their heterogeneous formation sources.

4. Conclusions

Self-healing in cement-based materials has a wide range of applications, including bridges, high-rise buildings, and structures where maintenance is difficult or very expensive. Based on the comprehensive investigation presented in this study, focusing on the impact of exposure environments, cement- and nutrient-derived calcium ions, nutrient type, and concentrations on the self-healing capabilities of cement-based materials utilizing microbiologically induced calcium carbonate precipitation (MICP), the following conclusions can be drawn:

• Submerged conditions demonstrated the most effective crack healing for MICP compared to wetting and drying cycle exposure environments. A crack area of 4mm² was 87.5% healed at 21 days.
• Crack closure was positively correlated with the concentration of calcium lactate, indicating the pivotal role of calcium content in the self-healing process. Bacterial suspensions enhance crack closure efficiency, particularly in environments with high calcium lactate content.
• The use of only urea as a nutrient source and hydration calcium ions as a calcium source showed less effectiveness in facilitating crack closure due to the adverse effects of ammonia, inhibiting the formation of cement hydration products. This behavior highlights the limited role of cement-derived calcium ions from cement hydration in MICP.
• Employing nutrient-derived calcium sources, such as calcium nitrate, in conjunction with ureolytic bacteria showed promising results in inducing calcium carbonate precipitation and achieving crack closure.
• Calcium nitrate significantly outperformed calcium lactate as a nutrient source, attributed to the higher concentration of available calcium ions and nitrate as an energy source for bacteria, facilitating more efficient calcium carbonate formation and self-healing. A crack area of 8 mm² was fully healed at 35 days.
• TGA analysis shows that while calcium carbonate samples from different sources have varying decarbonation temperatures due to distinct formation environments, their chemical composition remains consistent.
• The study elucidated the complex interplay between nutrient type, bacterial activity, and calcium source, emphasizing the need for a nuanced approach to optimizing the self-healing process in concrete structures.

These conclusions highlight the critical roles of microbial activity, nutrient availability, and environmental conditions in self-healing on cement-based materials. Further research is essential to understand these mechanisms better and optimize the self-healing process for practical applications, using an approach with more cell concentrations and molarities in nutrient-rich environments.