This study was conducted with B. subtilis AP91, isolated in Brazil from leaves of long grain rice. This bacterium has the potential to bioprecipitate CaCO₃ according previous researches [23]. Initially the bacteria was reproduced from two culture medium: Mueller Hinton (MH) (Sigma-Aldrich, USA) and Trypton Soy Broth (TS) (Sigma-Aldrich, USA). These two culture medium were tested to verify in which one the bacteria have the more effectively reproduction ability. From the analysis of this tests, it was select the best culture medium, which provided the highest bacterial growth, enabling to obtain the experiment.

The culture media (MH and TS) were formulated to 500 mL and autoclaved for 15 minutes at 121°C and 1 atm. After sterilization, the bacteria were inoculated in the culture medium and samples were placed in shaker TE 400 (Tecnal) for 66 hours at an average temperature of 29.1ºC to induce bacterial growth.

After the stir period, the solution with the bacteria was washed in order to remove the culture medium and to obtain a more concentrated solution of bacteria. For this, the culture media was centrifuged in Falcon tubes for 20 minutes at 3600 rpm spin, the supernatant was removed and replaced with 0.85% saline solution [1177]. The experiment was repeated twice (n = 2).

Then, the solution containing bacteria was stored for two days at the temperature of 8ºC to induce spore formation. It is important to mention that spores formation are fundamental for the survival of the bacteria in the cementitious materials.

Quantification of the bacteria grown in MH and TS were performed with the spectrophotometer UV340GX (Gehaka), with reading at 600 nm. For this test, three samples were used: the control with 0.85% of saline solution, the bacteria grown in MH dissolved in 0.85% saline solution and the bacteria grown in ST dissolved in 0.85% saline solution. To calculate the concentration of bacteria in the medium through the absorbance, Equation 4 was applied [17].

Where:

X = absorbance (A);

After complete this stage, the culture medium that resulted in the highest cell growth was identified and then the procedure was repeated with this medium to obtain the required concentrations in the experiment, which are 0.3x10⁸ and 1.2x10⁸ spores/mL according to previous experiment with the bacteria B. pasteurii and Pseudomonas aeruginosa [17].

For the production of concrete mix, the following materials were used: Portland cement, high early strength (CPV ARI, Brazilian denomination); fine aggregate: medium quartz sand; coarse aggregate: gravel 19 mm of basalt, water and B. subtilis spores dissolved in 0.85% of saline solution. The aggregates were characterized according to Brazilian standard [24].

The cement specifications and the requirements of NBR 5733 (Brazilian standard) [25] are presented in Table 1.

Table 1

Source: Brazilian standard [25].

Mixture proportions are presented in Table 2. The mix proportions used were 1:1:2 and 1:2:3 (cement: sand: gravel), in mass, with water to cement ratio of 0.33 and 0.45, respectively. For each composition used, two treatments were made with the addition of B. subtilis spores to the mixing water in the concentration of 0.3x10⁸ and 1.2x10⁸ spores/mL. Besides, in the mixing water it was added 0.85% saline solution to maintain the medium isotonic for the bacteria.

Table 2

Source: The authors.

The concrete mixture was performed in a vertical shaft mixer and bacteria concentrations were added to the mixing water.

For each treatment, six cylindrical specimens (100x200 mm) were made, being three for visual analysis of crack filling (n=3) over the curing time and the other three for the compressive strength test at 28 days (n=3). It should be noted that, for each mix compositions, reference specimens were made without bacterial spores, with 0.85% sodium chloride to ensure the same conditions as the specimens with spores.

Specimens were molded in 100x200 mm cylindrical samples according to NBR 5738/2015 (Brazilian standard) [26]. The specimens were demolded after 24 hours and stored in a dripping wet chamber at 20ºC ± 2ºC and relative humidity greater than 95% until the test ages, which was 7 days for cracking, 8, 15, 22, 29, 36 and 43 days for the monitoring visual of crack filling, and 28 days for the compressive strength test.

In order to evaluate the cracks filling by spores addition compared to reference, a 2 centimeters slice was extracted from each concrete (n=3). These slices were sent to the hydraulic press on the seventh day of healing to apply a strength to simulate a crack, as shown in Fig. 1.

Figure 1 Source: The authors.

The visual monitoring of cracks during the curing period, at 8, 15, 22, 29, 36 and 43 days, was conducted by images taken from an optical microscope with an attached camera (Olympus), with magnification of 100x. To ensure that the analysis was always held in the same position of the crack, it was demarcated. It is worth mentioning that from this method of cracking simulation it was not possible control it’s thickness.

The compressive strength test was performed with non cracked specimens of 100x200mm at the age of 28 days (n = 3) according to a NBR 5739/2007 (Brazilian denomination) [27].

A statistical analysis was performed only for the evaluation of compressive strength, from a completely randomized design experiment (CRD), in a 2x3 factorial scheme, consisting of two concrete mixtures (1:1:2 and 1:2:3) and three bacteria content (0, 0.3x10⁸ and 1.2x10⁸). In this statistical scheme, the two concrete mixtures were evaluated with all bacterial contents, with three replications.

For the statistical analysis, the software Sisvar was used to perform an analysis of variance [28].

From the two culture medium used in the experiment, the spores quantification of each medium was performed in accordance with Eq. 1., Fig. 2 shows the results of absorbance and concentration of spores/mL of culture media MH and TS.

Figure 2 Source: The authors.

Through the analysis of Fig. 3, it was observed a higher absorbance values and concentration of spores in the medium MH, of 1.28 and 1.21x10⁸ spores/mL, respectively, while under the same conditions, the medium TS presented, respectively, 0.29 and 1.61x10⁷ spores/mL. From this result, only the culture medium MH was selected for this experiment.

The reproduction procedure was repeated only with the medium MH, which presented the highest concentration of spores/mL, until it reached the concentrations of 1.2x10⁸ and 0.3x10⁸ spores/mL, which was quantified by spectrophotometer.

Sand was performed by Particle size analysis [24]. The results of these test were presented in Fig. 3.

The fine aggregate used was a medium sand. The particle size analysis is shown in Fig. 3. It was found that the grading curve is inside the limits established by NBR 7211/2009 (Brazilian standard) [29]. Analyzing the grading curve it should be noted that it has a continuous particle size, with desirable characteristic for use in cementitious materials due to the better packaging of the grains in the structure of the material. For the grading curve, it were calculated the maximum size, specific mass and unit mass. The values found were 2.36 mm, 2.62 g/cm³ and 1.57 g/cm³, respectively. Besides that, the gravel has maximum size of 19 mm.

The visual analysis of crack filling was performed with one crack per specimen until 43 days of curing, being that three specimens of each concrete were analyzed (n=3). The images obtained from an optical microscope with attached camera are shown in Figs. 3 and 4.

Figure 3 Source: The authors.

Figure 4 Source: The authors.

It can be seen that in all the compositions in which bacteria were added, there was a crystal precipitation from the 15th day on, possibly calcium carbonate (CaCO₃) according to literature [20,22,30-35]. Besides that, the same strain was used in another study [23] that performed an analysis in vitro of the bioprecipitation of calcium carbonate by pH change during biofilm formation. The authors found that the B. subtilis AP91 were able to precipitate crystals of CaCO₃ with the same morphology as other researches [11,36-40].

Figure 5 Source: The authors.

Other investigators conducted the studies in the same way, first studying whether the bacterium has the ability to precipitate CaCO₃ and in the sequence analyzing how that addition would behave in the cement matrix [9,41].

The filling of cracks up to 0.19 mm were visible in concretes with addition of spores of B. subtilis. However, in the samples without spores addition, precipitated material was not visible and the crack remained with the same feature throughout the curing period analyzed.

Although the cracks have different thickness sizes, it was visually found that in the concretes with addition of B. subtilis spores there was filling of the cracks by precipitation of crystals, even if the crack were of greater thickness compared to the reference concrete, without spores. Besides, in the reference concretes this filling was not visualized in this analysis. It is noteworthy that this analysis was just visual. To a better detailing a microstructural analysis would be necessary.

Fig. 6 shows the crystals formed by calcium carbonate precipitation promoted by the bacterium B. alkalinitrilicus added with a nutrient source and curing totally submerged under water for the healing regime [3]. It was observed that there was closure of the cracks from 0.13 to 0.40 mm thickness after 100 days of curing. Similar results were found in this research. Besides the above referenced authors, similar results using the bacterium B. sphaericus have been presented [18].

Figure 6 Source: [3].

Thus, it was observed that the crystals precipitation promoted by the addition of B. subtilis spores under the conditions analyzed was visually satisfactory in crack filling. However, for more specific studies a microstructural analysis would be required.

The compressive strength test was performed on the 28th day of curing and the results are shown in Fig. 7. There was a difference in resistance between the compositions due to the proportions of materials used and the water/cement ratio (w/c), with higher resistance for the mix proportion 1:1:2 compared to w/c of 0.33 and without addition of bacteria, and the lowest resistance for the mix proportion 1:2:3 with w/c of 0.45.

Figure 7 Source: The authors.

Through the analysis of the results, it was noted that there was a decrease in the mean resistance of samples with added spores when compared to the reference. This decrease in the mix proportion 1:1:2 was of 20.06 and 16.58% for the treatments 0.3x10⁸ and 1.2x10⁸ spores/mL, respectively. For the mix proportion 1:2:3, the difference was greater, corresponding to 63.63 and 23.28% for the treatments 0.3x10⁸ and 1.2x10⁸ spores/mL.

There was an increase in the resistance of the samples with 1.2x10⁸ spores/mL compared to the ones with 3.0 x10⁷ spores/mL in both mix proportions, demonstrating that the amount of bacteria added to concrete influences the compression strength. However, it is necessary to test a larger number of bacteria concentrations for a better understanding. According to the literature [31], calcium hydroxide (Ca(OH)₂) is the most important source of calcium for concrete, but it does not alter mechanical resistance. On the other hand, calcium carbonate (CaCO₃) contributes to increase this resistance, which makes the bioprecipitation a positive contribution [31]. It should be noted that the crystals precipitated by bacteria in cementitious materials, calcium carbonate stand out [12, 20, 22-23, 30-35]. But, to confirm that the crystal precipitated by B. subtilis was calcium carbonate more studies need to be conducted, since this research is a preliminary test.