YI Haihe, ZHAN Qiwei, YU Xiaoniu

(1.School of Civil and Architectural Engineering, Shandong University of Technology, Zibo 255000, China; 2. School of Civil Engineering and Architecture, Jiangsu University of Science and Technology, Zhenjiang 212003, China; 3.Department of Civil and Environmental Engineering, Nanyang Technological University, Singapore City 639798, Singapore)

Abstract: The crystal structure and morphology of the mineralization products were studied by scanning electron microscopy (SEM) and X-ray diffraction (XRD), and the thermal properties were studied by thermogravimetric-differential scanning calorimetry (DSC) analysis. The changes of microorganism quantity and enzyme activity in pore solution with time were measured. The experimental results show that microorganism quantity and enzyme activity in pore solution reach the maximum at 50-60 h, mineralization curing begins at this time, the strength of microbial mineralized steel slag reaches the maximum. This study provides a good selection basis for selecting the optimum mineralization system for the production of microbial mineralized steel slag products. Bacterial mineralization can accelerate the rate of carbon sequestration in the mineralization process. The compressive strength of steel slag with 1.5% bacterial can reach up to 55.6 MPa. The microstructure and thermal properties of calcium carbonate precipitate induced by the enzymes of bacillus subtilis differs from the chemical precipitation in pore solution of steel slag. Through the analysis of the mineralized products of steel slag, the reaction rate of free calcium oxide and free magnesium oxide in steel slag after the addition of microorganisms is significantly increased, which improves the stability of steel slag as cementitious material. Meanwhile, the production of calcium carbonate, the main mineralized product, is significantly increased.

Key words: steel slag; microorganism; bio-mineralization; enzyme activity; curing system

1 Introduction

China’s emission of steel slag was nearly 105 million tons in 2017, and the utilization rate of steel slag was only 17%. A large amount of steel slag is accumulated and occupied farmland, causing ecological damage and environmental pollution. How to improve the recycling and utilization of steel slag resources and develop new technologies to efficiently and environmentally use of steel slag was an urgent problem to be solved. Steel slag contains free calcium oxide and free magnesium oxide which will cause cracking when it expands with water. It is very harmful when the steel slag was used as building materials and road materials. The main mineral in steel slag is CS, and its hydration rate is very slow which leads to the strength and the utilization rate of steel slag is low.

Carbonization is considered to be a good method to use steel slag. Steel slag can be hydrated to CH and CSH, these hydration products could react with carbon dioxide. Meanwhile, CS can directly react with carbon dioxide, and carbonization can accelerate the reaction rate of main minerals in steel slag. Carbonization could eliminate the free calcium oxide and magnesium oxide. Many researchers have studied the parameters of carbonation process and the carbonation mechanism of steel slag, but there are still many problems restrict the carbonization reaction of steel slag. The carbonization efficiency and strength of steel slag is low and the carbonization process is complex. Qian CX studied the microbial mineralized steel slag cementitious materials and prepared microbial high efficiency carbon fixed steel slag products under the enzymatic action of microorganisms. The compressive strength of the microbial mineralized steel slag is increased 30% compared with the carbonized steel slag when the carbon dioxide pressure is 0.3 MPa for 3 h.

The process of hydration of carbon dioxide to HCOis a rate-limiting step in the conversion of COto CO, which is quite slow, only 1.3×10S. The carbonic anhydrase produced by microbial metabolism greatly accelerates the reaction, which can be increased by 10times and the reaction rate can reach 1×10S. Bacterial cell wall consists of peptidoglycan and outer membrane. The outer membrane contains a large number of negatively charged groups and it considered to be a site for adsorbing cations and causing crystal nucleation. After adsorption and enrichment of Ca, the continuous addition of anions such as COand HCOleads to nucleation and growth of mineral crystals. In the HO-CO-Casystem, the biochemical reaction of basophilic microorganisms on the cell surface to induce deposition of CaCOis expressed by formula (1) to formula (4).

In this paper, the growth characteristics and product metabolism characteristics of the microbial in the steel slag cementing material was explored. Optimized the mineralization system of the microbial mineralized steel slag cementing material, and the best curing and carbonization system was selected. Determine the optimal curing and carbonization scheme for microbial mineralized steel slag cementitious materials.

2 Experimental

2.1 Microbial preparation

The Bacillus mucilaginous used in this experiment was acquired from China Center of Industrial Culture Collection (CCIC). Bacillus mucilaginous was conducted in sucrose culture media. The purified medium of the strain is shown in Table 1, pH value was adjusted by NaOH is about 8.0. placed the purified medium in the triangle bottle and sterilized at 121 ℃ for 25 min. Remove the sterilized triangle bottle and place it in the oven to dry. The purified bacterial strains of carbonic anhydrase were inoculated in the culture medium, and cultured in oscillatory incubators at 30 ℃ for 24 h, the oscillation frequency of oscillatory incubators was 170 r/min. The OD600 value of culture was 1.4, the number of bacteria is 1.52×10cfu· mLand the enzyme activity value is 0.04U mmol·L. Then the harvested microorganisms were kept in a refrigerator in 4 ℃ as a stock culture the prior that their used.

2.2 Steel slag

In this experiment, steel slag powder was produced by Shanghai Baosteel Steel Slag Comprehensive Development Industrial Co., Ltd, the specific surface area was 425 m/kg. X-ray fluorescence spectrometer was used to test the chemical composition. The statistical results are shown in Table 2, the main components of Baosteel steel slag are CaO, MgO, SiO, AlO, and FeO.

Due to the difference of smelting process、ore type and steel type, the chemical composition of steel slag fluctuates greatly. Through literature research, the fluctuation range of chemical composition of steel slag discharged by ten large steel mills in China: Baosteel, Wuhan Iron, Anshan Iron, Tangshan Iron, Jinan Iron, Panzhihua Iron, Ben Iron, Maanshan Iron, Ti Iron, Shou Iron is statistically analyzed. The statistical results are shown in Table 3. The content of calcium oxide is in the range of 30%-60%, silicon oxide in the range of 8%-28%, alumina in the range of 3%-9%, magnesium oxide in the range of 4%-12%.

Table 1 The purified medium of the strain/g

Sucrose NaHPO4·12H20 MgSO4 CaCO3 KCL (NH4)2SO4 Yeast extract 10.00 2. 50 0.25 1.00 1.00 0.50 0.20

Table 2 The chemical composition of Baosteel steel slag/%

Chemical composition SiO2 Al2O3 CaO MgO FeO MnO Other Total Content 17.22 4.86 44.83 8.29 14.29 3.06 6.49 99.04

TOPAS4.0 was used to analyze the mineral content of steel slag samples in Baosteel, the statistical results are shown in Table 4. The mineral composition of steel slag discharged by Baosteel, Wuhan Iron, Anshan Iron, Tangshan Iron, Jinan Iron, Panzhihua Iron, Ben Iron, Maanshan Iron, Ti Iron, Shou Iron was statistically analyzed. the statistical range value is also listed in Table 1. The mineral content of steel slag from different steel mills also varies greatly, but the mineral types are similar. The mainly mineral includes dicalcium silicate (

β

-CS), dicalcium silicate (

γ

-CS), tricalcium silicate (CS), free calcium oxide (

f

-CaO), free magnesium oxide (

f

-MgO), calcium hydroxide (Ca(OH)) and other inert minerals (RO phase, FeO, and CF). The chemical components and mineral components of the steel slag used in this experiment are within the statistical range, and the method used in this paper is applicable to most types of steel slag.

Table 3 Statistical range of steel slag chemical composition in ten major steel factories/%

Composition SiO2 Fe2O3 Al2O3 CaO MgO FeO MnO Statistical range 15.8-27.7 2.3-5.6 2.8-8.9 38.2-60.5 3.5-9.5 2.1-6.5 0.7-1.6

Table 4 Mineral content of steel slag/%

Mineral Baosteel Average of ten steel mills β-C2S 25.8 9.3-31.2 γ-C2S 3.6 1.5-10.3 C3S 7.9 3.5-20.5 f-CaO 5.6 1.0-9.7 f-MgO 3.9 1.2-9.8 C3A 2.6 1.7-8.9 RO 26.8 20.0-35.9 Fe2O3 5.6 1.5-10.3 Total 81.8 -

2.3 Sample preparation

Mix proportions are given in Table 5. The water/binder (steel slag + slaked lime) ratio was 0.5. The raw materials were mixed with water for 4 min, and then were cast into moulds of 40 mm×40 mm×160 mm. All specimens were demolded after 24 h curing at 70% RH and 20 ℃. After completing the 1d hydration curing, samples were cured in different condition. The standard group samples were cured in 70% RH and 20 ℃ for specified time, the carbonation group samples were cured in 0.3 MPa COcuring for 4 h and then were cured in 70% RH and 20 ℃ specified time.

Table 5 Mix proportions/g

SpecimenSteel slagSlaked limeMicroorganismCO2 pressure/MPa C 525 105 0 0.3 BC 525 105 9.5 0.3

The schematic graphic expression of the mineralization setup is shown in Fig.1. The setup includes the compressed cylinder which with 99% purity COgas, the carbonation chamber, a pressure transducer and a pressure regulator. The pressure transducer monitors gas pressure and the regulator could control the chamber pressure at the set value throughout the mineralization process.

Fig.1 Schematic graphic expression of the mineralization setup

2.4 Microbial enzyme activity

The enzyme activity was measured by a colorimetric method. The bacterial enzyme product can specifically catalyze the 4-nitrophenyl acetate and diethylmalonic acid and form p-nitrophenol, the solution of p-nitrophenol appears yellow. Different degrees of color development can be expressed by the ODvalue of the solution, and can be calculated according to the ODvalue of the solution. The enzyme activity can be indirectly reflected according to the amount of p-nitrophenol produced. This requires measuring the ODvalue of different concentrations of p-nitrophenol solution before the enzyme activity measurement, and plotting the standard curve of the relationship between the concentration of p-nitrophenol solution and the ODvalue. According to the standard method, different concentrations of p-nitrophenol standard solution were prepared, and the respective absorbances were measured at a wavelength of 400 nm. The corresponding ODvalues of p-nitrophenol at different concentrations were obtained, and the linear equation was obtained by linear fitting of the scatter plot to obtain a linear equation

y

= 0.058+0.490

x

, as shown in Fig.2.

Fig.2 Relationship between concentration and absorbance of p-nitrophenol

2.5 Pore solution

Using the extraction device of cement slurry hole solution to collect the pore solution of steel slag cementing material. The sample is placed in the inner cavity of the device. The pressurization rate is 1 MPa/s and the pressure maintained for 10 min at 900 MPa. During the pressing process, the upper end of the sample is first compacted, the test specimen forms a dense gradient from top to bottom, driving the water to flow downward, and the pore solution flows into the collection bottle in the collection device. During the collection process, the air in the pore solution collection chamber was removed from COand ensure the pore solution is not carbonized. After the collection is completed, the collection bottle is taken out and immediately sealed for testing.

The pH of the pore solution at 20 ℃ was measured by pH meter with a measurement accuracy of ±0.005. Two-point calibration with mixed phosphate standard solution and borax standard solution before use, repeated calibration three times to minimize the error. The Caof the pore solution at 20 ℃ was measured by inductively coupled plasma spectrometer.

2.6 Test methods

The compressive strength and flexural strength were measured according to Chinese test methods for Portland cement (GB175-2007). Each property was tested for Six samples after curing and the experimental data was averaged.

The micromorphology measured by using FEI Sirion Scanning electron microscopy (SEM), mineralogical compositions analysis was conducted by using Bruker D8-Discover X-ray powder diffractometer (XRD), the thermal analysis on the specimen was conducted by NETZSCH STA449F3 simultaneous thermal analysis meter (DTA/TG) respectively. Caconcentration in the supernatant was determined according to Stocks-Fischer

et al.

The ODvalue was measured with a spectrophotometer (uv-6000pc) every 2 h to characterize enzyme activity, Bacterial counts were measured by flow cytometry (Easy Cyte-5) every two hours.

3 Results and discussion

3.1 Microbial quantity and enzyme activity

From 0 to10 h, the growth of bacteria is in the adjustment period. Since the bacteria has just been inoculated into the culture medium, its metabolic system needs to adapt to the new environment, and the number of microorganisms in this period does not change much. After 10 h, the number of bacteria increased rapidly and reached a higher level. Within 10 to 40 h, the bacteria grew in the logarithmic phase, which was the fastest growing and vigorous metabolism. During 40 to 50 h, due to factors such as nutrient consumption, imbalance of nutrient content, and accumulation of harmful metabolites, the number of bacteria remains relatively stable, and the growth enters a stable period.

It can be seen from the Fig.3 that the variation of microbial enzyme activity is basically consistent with the change trend of quantity of microbial. Microbial enzyme activity reached a maximum at 55 h from 0 to 10 h. The microorganism is in the period of growth adjustment. During 10 to 50 h, populations of microbial grown rapidly, and the enzyme activity increased rapidly. At 55 h, the microbial quantity is maintained at the stationary stage, and the accumulative of metabolites reaches the maximum. After 55 h, due to the factors such as the accumulation of harmful metabolites in the microbial metabolism, the number of microorganisms decreased, and both the number of metabolites and the activity decreased simultaneously. the carbonic anhydrase could accelerate COin the air reacts rapidly with Hin the solution and form HCOand accelerate the mineralization rate of steel slag. The study on the maximum value of enzyme activity has important guiding significance for the selection of starting time of mineralization.

Fig.3 Microbial quantity and enzyme activity

3.2 Pore solution

Fig.4 shows the variation trend of the concentration of OHand Caions with hydration time in the pore solution of steel slag cementing material from 1d to 7 d under standard curing conditions. It can be seen that the OHconcentration in the pore solution increases gradually at the initial curing stage, reaches its maximum value in 3 d, and then decreases slightly. The change rule of Cain the pore solution is similar, and it gradually increases at the initial curing stage, reaches the maximum at 4 d, and then decreases slightly. At the initial stage of hydration, a large amount of OHand Caions enter the pore solution. With the hydration reaction going on, OHand Caions gradually increase. At 3-4 days, the OHand Caions in the pore solution of steel slag cementing material reach the maximum. In the hydration process of steel slag, CH and CSH crystallization are generated on the surface and in the pores of CS. At this time, the saturation of OHand Caions decreased, which reduces the ion concentration in the pore solution. The crystallization product will hinder the hydration reaction, and the hydration reaction rate will turn to deceleration process, and the concentration of OHand Caions in the pore solution tends to be stable.

Fig.4 Concentration of OH- and Ca2+

3.3 Change of calcium ions concentration

The effect of microorganism on the concentration of calcium ion in the pore solution of steel slag cementing material is shown in Fig.5. The change of calcium ions concentration in the solution can directly reflect the formation rate of carbonate and the rate of formation of calcium carbonate. It can be seen from Fig.5 that the calcium ion concentration of the steel slag solution (without microbie) remains basically unchanged, due to the low hydration rate of carbon dioxide, there was not a lot of carbonate ions formed. When the microorganism was added, the calcium ion concentration in the pore solution was significantly decreased, and the reaction rate was significantly increased in the period of 50-60 h. Due to the adsorption of microbial enzymes and microbial cell walls, the rate of decrease in calcium ion concentration in the pore solution is significantly increased.

Fig.5 Change of calcium ion concentration in different systems

Under the action of carbonic anhydrase, COin the air reacts rapidly with Hin the solution and form HCOwhich would combine with calcium ions and form CaCOprecipitate. At the same time, the adsorption of Caby bacteria is also closely related to the action of bacterial polysaccharide capsules and their secreted mucus, proteins and lipids. These extracellular substances have a certain space structure due to their charge, and can adsorb and capture metal ions by ionic bonds, hydrogen bonds. These coagulate and flocculate form a network-like three-dimensional structure precipitate. The above effects cause a rapid decrease in the concentration of Caions in the solution.

3.4 Mineralogical composition

Under different mineralization times, the XRD pattern of the precipitate generated by microbial mineralization in the pore solution of steel slag is shown in Fig.6. According to the analysis, the precipitates obtained in the pore solution of steel slag under different mineralization times are all calcium carbonate crystals with calcite crystal types, corresponding to the PDF card library of 050586. At the same time, XRD patterns of calcite formed under different mineralization time were compared. The main crystal faces were (012), (104), (110), (113), (202), and (018). The positions of the main characteristic peaks on the crystal face (104) were the same, but the strength of the main characteristic peaks was different. With the extension of the mineralization time, the peak strength of the main characteristic peak of the microbe-induced mineralization of calcite crystal began to increase gradually, which indicated that the preferred growth orientation of the microbe-induced mineralization of calcite crystal along the (104) crystal surface increased and the degree of crystallization gradually increased. The activity of microbial enzymes gradually increases by the mineralization time, the amount of organic matter-Caelemental template increases on the (104) crystal surface, so the main characteristic peak of the calcite crystal induced by microorganisms in the pore solution increased on the (104) crystal surface.

Fig.6 XRD patterns of calcite

3.5 Morphology of calcium carbonate

Fig.7 show the morphology of the calcite formed by microbial mineralization at different starting carbonization curing time. By comparison, it can be clearly seen that although the precipitate at different starting carbonization curing time was calcite, their morphologies are completely different. At 0 and 20 h, the calcite crystals formed by microbial mineralization are in an oblique hexagonal lattice structure, and the crystal particle size distribution is relatively uniform and large, with flaky or rhombic single crystals, twins and their aggregates. At 40 and 60 h, the calcite crystals formed by microbial mineralization were doped with polyhedral and spherical particles. Calcite crystals formed at 80 and100 h, the microbial calcium carbonate present as various irregularly shaped particles with different sizes of spherical or spherical aggregates.

Fig.7 Microstructures of calcium carbonate mineralized by microorganism

The morphology of the crystal depends on the relative growth rate of the crystal in all directions. The growth rate depends on the surface process and is highly sensitive to the influence of foreign elements in the surrounding environment. Because different parts of the lattice are exposed to the outside world, different surfaces have different surface structures. After the nucleation of calcite crystal, the crystal surface can be prevented from growing by the organic matrix secreted by the adsorbed microorganism, while the crystal surface not selectively adsorbed can continue to grow. In this way, the crystal morphology is changed by the growth of kinetics. During 0-20 h, the bacteria is in the adjustment period. The effect of microbial mineralization is weak, and the morphology of calcium carbonate is similar to that of chemical calcium carbonate. During 40-60 h, the bacteria grew the fastest, secreted the active metabolism, and the enzyme activity increased rapidly. At 60 h, the microbial quantity reached at the maximize, the accumulative amount of metabolites reaches the maximum. After 60 h, due to the accumulation of harmful metabolites, the number of microorganisms and the enzymatic activity decreased decreased.

3.6 Thermal properties of calcium carbonate

The decomposition temperature of biogenic calcium carbonate at different starting times of mineralization is shown in the Fig.8. The decomposition temperature of calcium carbonate is quite different at different mineralization start time. The decomposition temperature of calcium carbonate at 60 h was 87 ℃ higher than the decomposition temperature at 20 h. One explanation for this phenomenon is that there is a complex interaction between the organic matrix secreted by microorganisms and calcite, there is a new special effect in calcite containing cementing properties. In addition to overcoming the chemical bonds of calcite itself, calcite also has to overcome the bonds of special effects when it decomposes, which means calcite requires higher energy when it decomposes. However, the quantity of organic matrix secreted by microorganisms is directly related to the quantity of microorganisms. If the starting time of mineralization is too early or too late, the quantity and relative activity of microorganisms will be reduced, and the quantity of organic matrix secreted will be reduced. This binding effect will be relatively reduced, so the decomposition temperature will also be reduced.

Fig.8 Thermo-gravimetric analysis of bacterial-induced calcite

3.7 Strength of microbial slag cementitious material

Fig.9 shows the strength changes of the comparison group of microbial mineralized steel slag and unadded microorganisms at different starting carbonization times. The strength of the microbial mineralization group was significantly higher than that of the group without microbial. When the initial mineralization time from 10 to 60 h, the strength of the microbial mineralization steel slag significantly increased and reached the maximum value of 55.6 MPa at 60 h. After 60 h, the strength gradually decreased with the starting time of mineralization. The strength of the carbonized control group without microorganisms increased gradually with the beginning of carbonation time. The strength of steel slag could reach to 33.4 MPa at 120 h. During mineralization process of the steel slag cementing materials, COdiffuses and forms COwith water, which then reacts with the free Cainside the samples and generates CaCO. After mineralization, these reaction products fill the pores, due to which the pore size is relatively smaller compared to that before mineralization. The hydration rate of carbon dioxide to HCOis low in the pore solution of steel slag, the formation rate of COis slow. With the addition of microbe, the formation rate of COin the pore solution is accelerated.

Fig.9 Compressive strength of steel slag

3.8 Mineralization products of steel slag

The steel slag products of the two groups with and without microorganisms are shown in Fig.10. It can be seen that the diffraction peaks of calcium carbonate can be clearly observed in the carbonized and mineralized steel slag, and the diffraction peaks of Ca(OH), CaO, and MgO can be clearly observed in the carbonized steel slag group without microorganisms. It shows that there are a lot of unreacted Ca(OH), CaO and MgO in the modified system. The diffraction peak of Ca(OH)in the mineralized steel slag has not been observed, indicating that the Ca(OH)in the system has completely reacted. The addition of microorganisms can significantly improve the reaction rate of calcium oxide and magnesium oxide, which can cause the stability of steel slag. The addition of microorganisms can significantly improve the safety of steel slag as a cementitious material. It can be seen from the diffraction peak intensity of calcium carbonate that the addition of microorganisms significantly improves the peak intensity, and the mineralization of microorganisms accelerates the carbonation reaction rate. The formation of these products can improve the compactness of the matrix and the strength of the steel slag cementified material.

Fig.10 XRD patterns of steel slag cementitious material paste

3.9 Industrial production of microbial steel slag products

Microbial mineralized steel slag technology has been successfully applied to industrial production of steel slag pavement bricks. The pavement brick size of the microorganism mineralized steel slag produced is 200 mm×100 mm×60 mm. The production picture of the factory is shown in Fig.11. Microbial mineralized products were sampled and found to have the following properties (10 samples for each test): The average compressive strength is 40.5 MPa, the autoclave linear expansion rate is 0.3‰, the average length of the pit is 30 mm, the average water absorption rate is 5.1%, the average strength loss rate after the freeze-thaw cycle is 2.3%, and the appearance is qualified after the freezethaw cycle test. The performance of pavement brick is better than the requirements of the national standard GB 28635-2012 for the performance of road pavement materials.

Fig.11 Production equipments of microbial mineralized steel slag brick

4 Conclusions

In the process of carbonization, the bacillus could transport a steady stream of carbon dioxide into the carbonization reaction, while the secretion of the enzyme can significantly accelerate the hydration rate of carbon dioxide and form more COinside the steel slag. Cahas a higher probability to combine with COto generate CaCO. When the initial mineralization at 60 h, the microbial number and enzyme activity reached the maximum, the strength of the microbial mineralization steel slag reached to the maximum value of 55.6 MPa. Meanwhile, the decomposition temperature and crystallization degree of calcium carbonate are higher, the microstructure appears as regular globules. Through the analysis of the mineralized products of steel slag, the reaction rate of free calcium oxide and free magnesium oxide in steel slag after the addition of microorganisms is significantly increased, which improves the stability of steel slag as cementitious material. Meanwhile, the production of calcium carbonate, the main mineralized product, is significantly increased. The addition of microorganism realizes the safe and efficient utilization of steel slag cementitious material. The study provides a good selection basis for selecting the optimum mineralization system for the production of microbial mineralized steel slag products.