Keywords

1 Introduction

The massive mining of coal resources in China had led to various social and environmental problems such as surface subsidence and gangue accumulation. As one of the important contents of green mining of coal resources, paste backfill technology had developed rapidly in China in recent years [1,2,3,4]. Backfill paste was a Bingham fluid made of gangue, fly ash, cement and admixture mixed with water. Fluidity and strength were important indicators to characterize the performance of backfill paste. The use of water reducing agent could greatly improve the paste concentration, increase the strength of the backfill body [5,6,7]. And achieve step by step improvement of the performance of the backfill paste while ensuring the paste had good fluidity.

The action mechanism of the water reducer was to reduce the adsorption of water molecules by the particles of the cementitious material by adsorbing on the surface of the paste, increase the dispersion, and release a large amount of free water [8,9,10]. So as to reduce the water consumption of the paste and ensure the fluidity of the paste. However, the gangue aggregate in the paste usually had a high mud content [11]. And the gangue powder with montmorillonite and kaolin as the main components will preferentially adsorb the water reducer, greatly affecting the effect of the water reducer [12,13,14]. A small amount of anti-mud agent could effectively increase the action efficiency of water reducing agent [15,16,17], which had been applied in the field of building materials [18, 19], but its effect on backfill paste had not been reported.

In order to study the influence of anti-mud agent on the performance of backfill paste, the competitive adsorption between anti mud-agent and water reducer (FDN) and its mechanism were analyzed, and the influence of anti-mud agent content on the performance of backfill paste was tested. The addition of a small amount of anti-mud agent could greatly reduce the amount of water reducing agent, which provides a new way to reduce the cost of coal gangue gypsum body backfill materials.

2 Analysis of Action Mechanism of Anti-Mud Agent

Because coal gangue powder contains a large amount of montmorillonite, the action mechanism of anti-mud agent was illustrated by taking montmorillonite as an example. The action mechanism of anti-mud agent was shown in Fig. 1. In the cement paste mixed with montmorillonite and water reducing agent, montmorillonite will absorb more water reducing agent. The water reducing agent could not fully combine with the cement particles, which causes the cement particles to absorb a lot of water molecules and agglomerate, affecting the performance of the cement paste [20, 21]. Anti-mud agent was more sensitive to montmorillonite. After adding anti-mud agent, it could replace part of water reducer and montmorillonite for adsorption, releasing a large number of water molecules and water reducer. A large number of water reducer could act on the surface of cement particles, increasing the dispersion of cement particles and releasing water molecules, improving the performance of cement paste [22].

Fig. 1
An illustration of the action mechanism of the anti-mud agent. Anti mud agent is added to a mixture of earth particles with water reducing agent and montmorillonite, water, and cement particles , which results in the splitting of the cement particles.

Mechanism of action of anti-mud agent

Full size image

3 Experiment

3.1 Experimental Materials

Gangue: after secondary crushing of the washed gangue from Changcun Coal Mine of Shanxi Lu’an Environmental Protection and Energy Development Co., Ltd., 4.75 ~ 16 mm gangue was used as coarse aggregate and 0 ~ 4.75 mm gangue was used as fine aggregate. The silt content of fine gangue was up to 5.78%. Fly ash: Grade II ash discharged by Shanxi Lu’an Environmental Protection Energy Development Co., Ltd. Cement: Grade 42.5 ordinary Portland cement produced by Jiaozuo Qianye New Materials Co., Ltd. Water reducer (FDN): naphthalene series superplasticizer produced by Shanxi Yonghong Building Materials Chemical Co., Ltd., with the molecular formula of (C21H14Na2O6S2) n. Anti-mud agent (KN): self-made soluble powder composed of sodium tripolyphosphate and citric acid.

3.2 Experimental Proportion and Method

Competitive Adsorption of FDN and KN on the Surface of Gangue Powder.

Pass the gangue aggregate through a 0.075 mm square sieve, take part of the gangue powder under the sieve, weigh the raw materials according to Table 1, stir and mix them for 30 min, so that FDN and KN could be fully dissolved in water and adsorbed on the soil surface, and then stand for 24 h, after the soil was fully precipitated (Fig. 2.), filter out the supernatant of the precipitation by vacuum suction filtration method, dilute it 10 times with distilled water, and measure the concentration of S element in the diluted solution by inductively coupled plasma emission spectrometer (ICP-OES). The matching numbers in Table 1. were recorded as J1, J2, J3, J4, J5 and J6 respectively. Since KN did not contain S element, while FDN contains S element, the greater the content of S element in the clear night, indicating that the less FDN the soil absorbs, the more significant the effect of KN.

Table 1 Mixing ratios for competitive adsorption experiment between anti-mud agent and water reducer
Full size table
Fig. 2
An illustration of a beaker with precipitate at the bottom and clear night solution at the top.

Solution precipitation

Full size image

Paste Preparation and Performance Test.

The mass concentration of 85% remains unchanged. Weigh the raw materials with the ratio in Table 2, fully mix them evenly, and then test the paste slump and coagulation time according to NB/T 51,070–2017 Test Method for Paste Backfill Materials in Coal Mines. Pour 100 mm × 100 mm × In 100 mm mold, demould after curing for 24 h under standard conditions (20 ℃, 95% humidity), and test the paste strength after curing to the specified age under standard conditions. The matching numbers were recorded as F, F0, F5, F10, F15 and F20 respectively.

Table 2 Experimental ratios.(kg·m−3)
Full size table

4 Experimental Result

4.1 Results of Competitive Adsorption Experiment

The influence of KN on the concentration of element S in the solution was shown in Fig. 3. It could be seen that when FDN and KN were not added, the concentration of S element in the solution was 157 ppm, which was the S element dissolved from the coal gangue powder. After FDN was added, the concentration of S element in the solution reaches 176 ppm, an increase of 12.1%, which was caused by FDN molecules not adsorbed on the surface of coal gangue powder. With the increase of the content of KN, the concentration of S element in the solution shows a linear growth trend. Compared with J1, the concentration of S element in J6 solution increases from 176 to 242 ppm, an increase of 37.5%. This was because KN could preferentially adsorb with montmorillonite [8], releasing FDN. FDN dissolution in the solution increases the concentration of S element in the solution.

Fig. 3
A line graph plots element concentration versus experiment numbers. The line increases between (J 1, 159), (J 2, 180), (J 3, 190), (J 4, 200), (J 5, 205), and (J 6, 240). The y-axis values are approximated.

Effect of KN on the concentration of element S in solution

Full size image

4.2 Paste Fluidity

The influence of KN content on the paste slump was shown in Fig. 4. It could be seen that the paste slump first increases rapidly with the KN content. When the KN content exceeds 0.10 kg·m−3, the growth rate of the slump slows down. The slump of F0 sample was 135 mm, and the slump of F10 fresh paste was increased to 232 mm, with an increase of 71.9%, which meets the pumping requirements [21] and was equivalent to the slump of F sample, indicating that a small amount of KN could greatly improve the fluidity of paste and reduce the amount of FDN. This was because the incorporation of KN releases a large number of FDN molecules adsorbed by coal gangue powder, which improves the water reduction efficiency.

Fig. 4
A scatterplot with error bars has slump in millimeters versus K N dosage. The points for F D N 1 with a connecting line increase between (0.00, 130), (0.05, 180), (0.10, 230), (0.15, 240), and (0.20, 250). The point for F D N 3 is at (0.00, 230). The y-axis values are approximated.

Effect of KN dosage on the slump of paste

Full size image

4.3 Paste Coagulation Time

The influence of KN content on the coagulation time of the paste was shown in Fig. 5. It could be seen that with the increase of KN content, the coagulation time of the paste was gradually extended. Before the content was 0.05 kg·m−3, the coagulation time increases rapidly, and then the increase rate slows down. The coagulation time of sample F0 was 9.0 h, and that of sample F10 increases to 13.1 h, with an increase of 45.6%. The coagulation time of sample F was equivalent to that of sample F10. The extension of coagulation time was due to the increase of free water content in the paste. The incorporation of KN releases FDN molecules adsorbed by coal gangue powder. FDN could disperse agglomerated cementitious materials and release adsorbed free water, which was conducive to improving the controllability of paste pumping. The amount of FDN in F sample without KN should reach 3 kg·m−3, because its FDN molecules were largely adsorbed on the surface of coal gangue powder and could not play a role in reducing water dispersion. In addition, the KN used in this study was a mixture of sodium tripolyphosphate and citric acid, which had a retarding effect [22], and will also extend the coagulation time of the paste.

Fig. 5
A scatterplot with error bars has coagulation time versus K N dosage. The points for F D N 1 with a connecting line increase between (0.00, 9), (0.05, 12), (0.10, 13), (0.15, 14), and (0.20, 15). The point for F D N 3 is at (0.00, 13.5). The y-axis values are approximated.

Effect of KN dosage on the coagulation time of paste

Full size image

4.4 Compressive Strength of Paste

The influence of KN on the compressive strength of paste at 3 d, 7 d and 28 d was shown in Fig. 6. It could be seen that the strength of paste at different ages decreases with the increase of KN content. When the content of KN increases from 0 to 0.20 kg·m−3KN, the paste strength decreases from 1.19 MPa to 0.92 MPa at 3d, by 22.7%, from 1.88 MPa to 1.73 MPa at 7d, by 8.7%, and from 4.05 MPa to 3.65 MPa at 28d, by 9.9%. The strength of sample F and sample F10 at the same age had little change. It shows that KN had a great influence on the strength of the backfill paste at 3 d, and had a little influence on the strength at 7 d and 28 d. The KN used in this study was a mixture of sodium tripolyphosphate and citric acid. With the increase of the dosage, the KN not adsorbed on the surface of coal gangue powder will be adsorbed on the surface of the cementitious material, which had an inhibitory effect on its hydration reaction [23, 24]. Therefore, the strength of the paste decreases significantly in 3 days, the inhibition disappears when the age increases, and the strength decreases in the later period.

Fig. 6
A grouped bar graph with error bars has strength in megapascals versus age in days. The values of bars for F, F 0, F 5, F 10, F 15, and F 20 are above 3.5 megapascals for 28 days, between 1.5 and 2.0 megapascals for 7 days, and between 1.0 and 1.25 megapascals for 3 days.

Effect of KN dosage on compressive strength of paste at different ages

Full size image

The SEM structure observation of F0 sample and F10 sample at the age of 3 d and 28 d was shown in Fig. 7. It could be seen that with the increase of age, the hydration product C-S–H gel of the two groups of samples increases, and the structural pore decreases. Therefore, the strength of the paste increases with the increase of age. Compared with F0 sample, the structural pores of F10 sample at 3 d and 28 d age were larger, and the amount of hydrated product C-S–H gel was less. Therefore, the strength of F10 sample was less than that of F0 sample. The difference between the micro pores of the two groups of samples was no longer obvious at 28 d age, so the strength decline of F10 sample at 28 d age was reduced, which indicates that the KN had little effect on the later strength of the paste.

Fig. 7
4 S E M micrographs a to d are for F 0 and F 10 after 3 and 28 days. The A f T, C S H, and C H particles are marked.

Observation of SEM structure of F0 and F10 specimens at different ages

Full size image

5 Conclusion

  1. (1)

    With the increase of KN content, the adsorption amount of water reducer on the surface of coal gangue powder decreases, the slump of coal gangue gypsum body backfill material increases, and the coagulation time extends.

  2. (2)

    A small amount of KN could greatly improve the fluidity of the paste and reduce the amount of water reducing agent. The slump of coal gangue gypsum body backfill material with 0.10 kg·m−3 KN and 1 kg·m−3 water reducing agent was similar to that without KN and 3 kg·m−3 water reducing agent.

  3. (3)

    With the increase of the dosage of KN, the strength of coal gangue gypsum backfill material decreases in 3 days, but it did not decrease significantly in 7 days and 28 days.