3.1 Batch test
The extraction rate of total Cr and Cr (Ⅵ) by saponin increased significantly with the increased in concentration (Fig. 3a and Fig. 3b), which is also observed in previous studies (Cao et al. 2013; Li et al. 2009). However, when the saponin concentration is too high, the increase in the size of micelles will reduce the ability to adsorb heavy metals, and collisions between micelles can prevent complex formation (Mulligan et al. 1999). The increase in the concentration of other surfactants did not significantly improve the extraction rate of total Cr and Cr (Ⅵ) (Fig. 3c). In addition to SDS, the four surfactants were shown to have higher extraction rates of total nitrogen at low concentrations, implying that increasing surfactant concentration may reduce nitrogen adsorption sites on the surface of soil particles (Wang et al. 2017). For ammonia nitrogen, several surfactants showed good removal effects (Fig. 3d).
Unlike surfactants, the extraction rate of total Cr by EDTA and EDDS decreased with the increase of concentration (Fig. 3f), which was consistent with the results of previous studies. As shown in Fig. 3g, the extraction rate of Cr (Ⅵ) by EDDS increased significantly with the increase in concentration. The extraction rate of Cr (Ⅵ) by EDDS with a concentration of 20 g/L was as high as 95%. This trend has also been confirmed in recent studies (Dong et al. 2017). The extraction rate of Cr (Ⅵ) by EDTA was the opposite to that of EDDS, which was different from a previous research result (Guidotti et al. 2015). The reason might be that the test in this paper is conducted under neutral conditions (pH = 7). The extraction rate of total Cr by CA was only slightly higher than that by deionized water, and the flushing effect of Cr (Ⅵ) was not as good as the other two chelating agents. The extraction rate of total nitrogen by the chelating agents was mostly less than 30%, and they were not sensitive to the concentration (Fig. 3h). The extraction rates of ammonia nitrogen by the chelating agents were between 41% and 62%, and the flushing effect was similar (Fig. 3i).
The agents with high COD values will lead to high COD in the effluents, which is positively correlated with the concentration. SDS, SDBS, and EDTA were the least biodegradable, resulting in secondary pollution of the soil (Fig. 3j). Tween 80, RL, and saponin with low COD and biodegradable were more recommended as agents used in soil flushing.
The best extraction rates by the above 8 agents and deionized water are shown in Table 3. Considering the biodegradability, Tween 80, RL, and saponin can be screened out as preferred agents. It can be found from the results of Fig. (a-d) that the surfactant with a concentration of 20 CMC does not significantly improve compared with that of 10 CMC. Therefore, the solution concentration was designed to be 10 CMC in the soil column test.
Table 3
The best extraction rate by each flushing agent
Agents | Extraction rate (%) |
total Cr | Cr (Ⅵ) | total nitrogen | ammonia nitrogen |
Water | 14.0 | 7.7 | 6.5 | 14.5 |
SDS | 61.0 | 63.1 | 72.7 | 28.2 |
SDBS | 68.0 | 63.0 | 72.5 | 24.8 |
Tween 80 | 61.0 | 56.3 | 82.5 | 35.8 |
RL | 61.0 | 67.8 | 72.9 | 30.7 |
Saponin | 90.0 | 90.0 | 82.2 | 39.5 |
EDTA | 64.0 | 68.9 | 54.7 | 40.0 |
EDDS | 50.0 | 95.0 | 61.9 | 33.2 |
RA | 20.0 | 21.6 | 60.9 | 12.7 |
3.2 Preferred agents used soil column tests
The breakthrough curves of the contaminants are illustrated in Fig. 4. The results of the soil column show that the contaminants in effluents first rise rapidly to reach their peak and then gradually decrease, which is consistent with the previous research results (Lo et al. 2011; Song 2014). Previous studies have confirmed the good removal effect of saponin on heavy metals and hydrophobic organic compounds (HOCs) (Hong et al. 2002; Li et al. 2019). In this soil column test, the removal for all contaminants in landfill contaminated soil is the highest of the three selected agents. The extraction rate of total Cr and Cr(Ⅵ) is more than 70%, and the extraction rate of total nitrogen and ammonia nitrogen is more than 55%. This indicates that saponin can wash the contaminated soil efficiently, and there was no antagonistic effect between heavy metals and organic contaminants. Saponin can simultaneously remove different contaminants because it forms complexes at different sites of saponin molecules (Song et al. 2008).
In this soil column test, the extraction rate of total Cr by the three agents is basically similar to that in the batch test. The extraction rate of total nitrogen by saponin is more than 50%, which is higher than that in batch test. The extraction rate of ammonia nitrogen by the three agents is lower than that measured in the batch test. The COD in effluents flushed by saponin increases almost linearly, which is different from the COD measured in the batch test. The reason for this discrepancy is related to the adsorption of contaminants in the soil and the transport of flushing solutions in the soil column.
3.3 Injection modes used in soil column tests
Although the injection modes are different, the peak of contamination removal always appears before 3 PV (Fig. 5a). The step-gradient injection has the highest peak of contamination removal (Fig. 5b), which is due to the high concentration of saponin injected at the beginning. This indicates that the concentration in the initial injection has a non-negligible effect on the removal efficiency. Continuous injection requires more flushing time and metal complexes may be reabsorbed in the soil, which reduces the flushing efficiency (Finzgar and Lestan 2008). A single-pulse injection for the extraction of total Cr with a broad main peak can extract more total Cr in a short time. (Fig. 5c). Fluctuations in contaminant removal by multi-pulse injection diminish with the reduction of contaminants (Fig. 5d). The operation process of multi-pulse injection is relatively complex, but the rinsing step with water can flush out the complexes in a timely manner.
As shown in Fig. 6a, the cumulative extraction rate in the soil column is relatively constant under the continuous injection mode and the step-gradient injection mode. The continuous injection is the best mode for the removal of ammonia nitrogen, with an extraction rate of 64.72%. In addition, the removal effect of this method on total nitrogen is also good, with a removal rate of 61.74%. The single-pulse injection was more effective than the multi-pulse injection for the removal of ammonia nitrogen and total Cr, which was different from the previous results (Pociecha and Lestan 2009). This may be related to the complexing capacities of saponin and chelating agents for total Cr. The cumulative COD in effluent was highest under the multi-pulse injection mode (Fig. 6b). The flushing step at each pulse could help flush out more organic contaminants. Although shorter operation time and less flushing solution are advantages of single-pulse, this mode does not provide a long enough time to allow more organic contaminants to dissolve (Lo et al. 2011).
Based on the above results, if the main contaminants in the landfill leachate-contaminated soil are organic matter, then the muti-pulse injection mode is a good choice. If total nitrogen or ammonia nitrogen is the priority target, the step-gradient injection can demonstrate better removal efficiency. For Cr-dominated contaminated soils, continuous injection is recommended.