Reaction mechanism
During the solidification of CFA by using Sporosarcina pasteurii,bacterial metabolism produces urease with cell dry weight up to 1%, which can decompose urea into carbonic acid (CO32−) and ammonia (NH4+) (formula 1). Because of the negative charge on the cell wall, Ca2+ in the CFA sample is adsorbed around the cell. With increasing amount of adsorption, calcium carbonate crystal with gelation is formed around the cell as the crystal nucleus (formula 2), which gradually increases and accumulates in the process of sample curing. Subsequently, a microbial-induced calcium carbonate block with certain strength is formed in CFA.
CO(NH2)2+2H2O→CO32−+2NH4+ (1)
CO32−+ Cell-Ca2+→ Cell-CaCO3↓ (2)
In particular, it can be seen from Table 1 that the particle diameter of CFA is generally small. The particles with diameter between 0.075 − 0.005 mm or smaller than 0.005 mm accounts for 81.2% and 10.2% of the total mass, respectively. The particle size is in the same order of magnitude with the monomer size of Sporosarcina pasteurii (length 2–3 µm, diameter 0.5–1.5 µm). The particle size of microbial induced calcium carbonate is close to that of CFA particles. Hence, it is more effective for connection and filling of fly ash particles and enhance the curing strength of CFA.
Minerals and SEM analysis
Mineral crystals exhibit specific X-ray diffraction patterns. The intensity of the characteristic peaks in the patterns is correlated positively with the mineral content in the specimens, and the mineral content can be calculated from the intensity of the characteristic diffraction peaks.
To compare the mineral compositions of UT- and MICP-treated CFA specimens of MC and NE, according to SY/T 5163 − 2010 (2010), the X-ray diffraction spectra of non-clay minerals were obtained on a Rigaku TTR III instrument. A comparison of the UT, MC and NE specimens is shown in Fig. 9, and the ordinary non-clay minerals analysis results are shown in Table 3.
After MICP treatment, for rich nutrients of calcium ion and nitrogen sources, the metabolism of Sporosarcina pasteurii can precipitate calcium carbonate crystals of calcite in the CFA. Figure 9 compares the UT specimens, and shows that the calcite content increased for the MC and NE curing conditions. As shown in Table 3 the content of calcite in the UT CFA was 7.0%. After MICP treatment, the calcite content reached 18.9% and 15.3%, respectively, which is an increase of 170% and 119%, respectively.
SEM/EDS images were used to explore the presence and pattern of CaCO3 precipitation. In Fig. 10 (a), it can be seen that, a great number of pores are formed in the CFA samples, and the calcium carbonate crystals induced by microorganisms are distributed among the CFA particles, with varying patterns at different positions, generally in the form of bundles (Fig. 10 (b)) and short columns (Fig. 10 (c)). Since the particle size of the microbial induced calcium carbonate is close to that of the CFA particles, the calcium carbonate crystals can connect and fill the surrounding fly ash particles more effectively, and improve the curing strength. The EDS analysis confirms that the mineral observed is calcium carbonate (Fig. 10 (d)).
Water-content change
Table 4 lists the water-content change statistics of CFA specimens in the different test groups. After 7 days of curing, under the MC curing condition, the water content of the CFA specimens with only water added decreased from 50–33.75% and 33.72%, which is a decrease of 32.50% and 32.56%, respectively. In contrast, the highest water content decrease of the MICP-treated specimens from 50–36.06%, occurred in the test group with a 0.25 mol/L nutrient concentration, which is a decrease of 27.88%, with the lowest reduction in the 0.1 mol/L group, from 50–47.97%, which is a decrease of 4.06%.
Under the NE curing condition, the water content decrease in each group of specimens was similar to that in the MC, but the decrease was larger. The water content of the specimens with only water added decreased from 50–4.90% and 4.36%, respectively, which is a decrease of 90.20% and 91.28%. The highest water content decrease of the MICP-treated specimens was from 50–24.63% in the 1.25 mol/L test group, which represents a decrease of 50.74%. The lowest reduction occurred in the 0.25 mol/L group, from 50–36.70%, which is a decrease of 26.60%. The decrease in water content indicates that the use of MICP in CFA could strengthen the interparticle connection, and reduce the internal water loss. MICP-treated CFA achieved a better water retention,could worked well in the dust suppression of CFA storage yards.
Unconfined compression strength
Unconfined compression tests were performed on all 7-day cured CFA specimens in accordance with ASTM D2166-00. Tests were performed on a CMT 4304 universal testing machine (MTS Systems Corporation). The axial loading rate was 1.0 mm/min. Loading continued until the load values decreased with an increasing strain, or until a 15% strain was reached during the test.
The different colors represent different nutrient concentrations in the stress–strain curves that were obtained from the tests. The three parallel CFA specimens in the same test groups were expressed by solid line a and dotted line b and c, and the corresponding concentrations of 0.10, 0.25, 0.50, 0.75, 1.00, 1.25 and 1.50 mol/L were marked. The control specimens with no numerical value are also shown in which no bacterial solution, but only water, was added.
Figure 11 shows the test result for curing in MC. The good cementing ability of calcium carbonate produced by MICP allowed loose CFA particles to be cemented or filled during crystallization, which can improve the compressive strength simultaneously.
The comparison shows obvious peak points in the stress–strain curves of the MICP-treated specimens. A higher peak strength yields a higher corresponding strain, and the overall tendency is similar. During the initial loading stage after the test commences, the stress increases rapidly with an increase in strain and the stress–strain curve increases approximately linearly. After reaching a maximum, the specimen reaches a limit of bearing capacity with a further increase of strain, and the stress decreases rapidly with certain brittle failure characteristics (Fig. 12). For the experiment where only water is added to the control specimens, in contrast, when the axial stress reaches its peak value, the stress decreases slowly with the increase in strain, and the brittle failure characteristics are not obvious.
The unconfined peak compressive strength of the MICP-treated specimens with different nutrient concentrations is different under the MC curing condition. Table 5 shows that the maximum peak stress of 97.63 kPa occurs in the test group with a 0.50 mol/L nutrient concentration and the average peak stress of this test group was 80.70 kPa. The minimum peak stress was 23.84 kPa in the 0.10 mol/L test group with an average stress of only 26.14 kPa. The average peak stresses for the 0.10, 0.25, 0.50, 0.75, 1.00, 1.25 and 1.50 mol/L nutrient concentration test groups were 26.14, 53.45, 80.70, 68.08, 64.74, 50.53 and 45.92 kPa, respectively. The results showed that the peak stress of CFA specimens increased initially and then decreased with the increase of nutrient concentration from 0.10 mol/L. According to the maximum compressive strength, an optimum nutrient concentration existed for the MC curing condition. At this nutrient concentration, the MICP reacted more fully and the peak strength was higher.
Compared with the MICP-treated specimens, the unconfined compressive strength of the control specimens with only water added was lowest under the same curing conditions. The maximum was only 18.20 kPa, the minimum was 14.91 kPa, and the average was 16.56 kPa, which is only 63.35%, 30.98%, 20.52%, 24.32%, 25.58%, 32.77% and 36.06% of the average peak stress of the 0.10, 0.25, 0.50, 1.00, 1.25 and 1.50 mol/L of the MICP-treated specimens, respectively.
Under the NE curing condition, the stress–strain curves from the unconfined compressive tests are shown in Fig. 13. Although the stress–strain curves have obvious peak points, the overall tendency changed partially. Compared with Fig. 11, during the initial loading stage, with the increase in strain, 1.00a and 1.25a show a rapid growth of stress and an approximately linear growth. Before reaching the peak value of 1.00b and 1.50a, the growth of stress is initially fast and then becomes slow; after failure, the stress decreases rapidly with an increase in strain. However, with the increase in strain, specimens 1.00b and 1.50a appeared and another lower peak appeared after the peak value. Unlike Fig. 11, after reaching their peak value, the control specimen stress decreased rapidly with an increase in strain, and specimen b shows a lower peak value after the maximum.
Compared with the specimens cured in MC, the peak of the axial stress appears near 2% of the relatively low strain, and all specimens showed more obvious brittle failure characteristics regardless of whether bacterial solution was added or not (Fig. 14).
Under the NE curing condition, the unconfined compressive strength with different nutrient concentrations was significantly different. Table 6 shows that a maximum peak stress of 78.46 kPa occurred in the 1.00 mol/L nutrient concentration test group, with an average stress of 69.82 kPa. The minimum peak stress was 17.07 kPa in the 0.10 mol/L test group, with an average stress of 28.50 kPa. The average peak stresses of the 0.10, 0.25, 0.50, 0.75, 1.00, 1.25 and 1.50 mol/L test groups were 28.50, 35.92, 37.27, 46.06, 69.82, 51.75 and 50.04 kPa, respectively. The peak stress increased first and then decreased with an increase in nutrient concentration from 0.10 mol/L. According to the maximum compressive strength, an optimum nutrient concentration of 1.00 mol/L existed in the MICP-treated CFA under the NE curing condition. The MICP reacted more fully and the peak strength was higher for this nutrient concentration.
Compared with the MICP-treated specimens under the same curing condition, the strength of the control specimens with only water added was lowest, with a maximum of only 20.92 kPa, a minimum of 16.49 kPa, and an average is 18.70 kPa, which is only 65.61%, 52.06%, 50.17%, 40.60% 26.78%, 36.14%and 37.37% of the average peak stress of 0.10, 0.25, 0.50, 0.75, 1.00, 1.25 and 1.50 mol/L of the MICP-treated specimens.