The sample as lou soil was collected at uphill about 2 km far from the north campus of Northwest A&F University, Yangling, Shaanxi Province, China. The lou soil was air-dried and passed through a 10 mesh nylon sieve. The Cation Exchange Capacity (CEC) of soil is 178.49 mmol/kg.

AR-grade BS-12 (Dodecyl dimethyl betaine) surfactant in this experiment was provided by Xingguang Auxiliary Factory, Tianjin City, China. The molecular structure of amphoteric modifier was shown in Figure 1.

The weight of soil was modified by dodecyl dimethyl betaine (BS-12) equivalent 50% and 100% as analytical reagent with 30% (W/V) through calculating with Cation exchange Capacity (CEC) of lou soil. Firstly, the BS-12 was dissolved with deionized water between soil’s weight and deionized water [1:10] at 40˚C, and added soil under being stirred for 3 h. Then the sample was centrifuged and washed 3 times with deionized water, dried at 60˚C and passed through 10 mesh sieves.

The amount of modification of BS-12 was calculated according to the following formula:

W = m × C E C × M × 10 − 6 × R / b (1)

where: W is the quality of the modifier, (g); m is the mass of the soil sample, (g); CEC is the cation displacement of the modified soil sample, (mmol/kg); M is the molar mass of the modifier, g/mol; b is the modifier product content (BS-12 take 30%).

The transport experiment was conducted saturated-column and column was 5 cm in length and 5 cm diameter with a thickness of 0.5 cm. The two covers of the column were one hole to let solution flow in and out for testing concentration. To place 1000 mesh sieve filters in the top and bottom of the column was to protect falling down of soil. In the experiment, the head pressure of moving liquid was set 1.5 cm and was controlled the distance between the level surface of the big bottle at the top and level surface of the column. The soil in the column was prepared as unmodified soil (CK), 50BS-12 modified soil (50BS-12), and 100BS-12 modified soil (100BS-12). The basis parameters of unmodified and modified soils were listed in Table 1.

This experiment was used Convection Dispersion Equation (CDE) for studying the transport of Cl⁻ and phenol:

R ∂ c ∂ t = D ∂ 2 c ∂ z 2 − v ∂ c ∂ z (2)

where C is pollutant concentration ppm (mg/l), t is time transportation (h), z is vertical depth (cm), D is dispersion coefficient (cm²/h), v is the average of pore water velocity (cm/h), v = q/θ_s, where q is water flux in unit area (cm/h), and θ_s is soil porosity.

R is a retardation factor to indicate linear equilibrium adsorption transport as:

R = 1 no interaction between soil and solutes;

R < 1 ion exclusion;

R > 1 chemical reaction-adsorption;

Using the STANMOD program by CXTFIT (version 2.1) model was determined the parameter in experiment v, D and R are fixed for transporting Cl⁻ and phenol.

Using CURVEEXPERT software was simulated penetration of curve phenol such as saturation amount, washout amount, the residual amount (saturation amount-washout amount), residual retention (residual amount/soil’s weight in the column) and holding rate (residual amount/saturation amount).

The Breakthrough Curves (BTCs) of Cl⁻ was shown and it was plotted between concentration (C/Co) and pore volume (V/Vo) in Figure 2. The parameter of D and R of Cl⁻ was fitted by using the equilibrium model (EM) and R² > 0.99*** (P < 0.01) in Table 2. The order of v average pore water velocity of Cl-migration in soil was CK > 100BS-12 > 50BS-12 and the order of D value was CK > 100BS-12 > 50BS-12. Linear correlation equation between v and D of BS-12 modified soil was D = 0.4432v + 0.578, R² = 0.8239, so it indicated that v and D value have related each other to the physical structure the soil column after modification [18] [23] . Therefore, the flow velocity was decreased gradually and the D value becomes smaller, respectively. The retardation factor R-value of Cl⁻

in Table 2 showed in order of 100BS-12 > 50BS-12 > CK and all R-value were smaller than 1. So that, it indicated that the transport of Cl⁻ was non-reactive ion in the soils as the same result [15] [18] [23] .

The BTCs of transport and washout of phenol in different modified soil were shown in Figure 3, and the result of parameter D and R from fitting the BTCs by the non-equilibrium model (NEM) of CXTFIT model was shown in Table 3. The number pore volume (NPV) of phenol migration was 17.89 (CK), 29.65 (50BS12) and 55.69 (100BS-12), respectively. The pore volume increased while the modification ratios increased and it indicated the effect of phenol on modified soil was enhanced. The NPV of blocking phenol was in order of CK < 50BS12 < 100BS12 at V/Vo as 1.78, 14.83 and 17.98 at C/Co (0 - 0.009) in Figure 3. Table 3 showed the R-value of CK, 50BS-12 and 100BS-12 (50 ppm, pH = 7, and ionic strength = 0.1 mol/l) was bigger than 1 and R² was closer to 1, so it was proved that the phenol in soil was blocked. The R-value of modified soil was larger while modification ratios increased. The linear correlation equation R and T_0.5 of modified soil were R = 0.0566T_0.5 +1.1396, R² = 0.9942, so it indicated that the duration of transport was longer while R became larger. The R-value of CK

Note: *** indicates that the correlation coefficient is significant at P < 0.01.

Note: *** indicates that the correlation coefficient is significant at P < 0.01.

was smaller than (50BS-12) 3.05 and (100BS-12) 5.27 times. Because the BS12 modified soil with long chain carbon was changed the surfactant from hydrophilic to hydrophobic and the adsorption mechanism was enhanced with increasing modified proportion [24] . By adding more details, researchers proved that high organic matter content, microstructure and surfactant in soil greatly impacted on pollutants migration [25] [26] [27] . Therefore, modified soil greatly enhanced blocking ability while modification ratios were increased.

The BTCs of various phenol’s concentrations of CK and 100BS-12 was shown in Figure 4, and its results of transport parameters were also shown in Table 3. The NPV at various phenol’s concentration in CK was 17.89 (50 ppm), 11.24 (200 ppm) and 11.94 (500 ppm); 100BS-12 was 55.69 (50 ppm), 44.15 (200 ppm) and 47.26 (500 ppm).

The R-value of CK didn’t significantly change in each concentration. This was because the unmodified soil was a weakness to block and easily polluted. The R-value of 100BS-12 was bigger than R of CK as 6, 7 times, while high concentration was 500 ppm and the R-value was in order of R_50ppm > R_200ppm > R_500ppm. It indicated R-value became smaller while the high concentration was larger.

Therefore, the retardation ability of modified soil strongly blocked with high concentration.

The BTCs of various pH (4, 7 and 10) on phenol in CK and 100BS-12 was shown in Figure 5. As seen the results in Table 3, the v value of CK decreased and R-value increased while the pH was high. It showed that pH had little effect on CK. The parameter of R-value for pH on phenol in 100BS-12 didn’t change remarkably but only v value increased with increasing pH value, respectively. Thus, the various pH on the transport of phenol through CK and 100BS-12 only

changed the number of charges in soils [15] [28] . At the same of pH value, we observed that the R-value, T_0.5 of CK and 100BS-12 were different and it indicated that the 100BS-12 strongly blocked phenol. However, the blocking phenol was strongly retarded in 100BS-12 at pH 7.

The BTCs of the various ionic strength of phenol through CK and 100BS-12 was shown in Figure 6. Table 3 showed that the R-value of CK and 100BS-12 both was R_0.1 > R_0.01 > R_0.001 and various ionic strength of phenol migration obviously

affected. The parameter of v in CK and 100BS-12 both was faster when the ionic strength decreased. This is because the mechanism increasing of ionic strength could increase aggregation adsorbent and decreased the electrostatic force of pollutants migration [29] [30] [31] [32] [33] . Therefore, the decreasing of ionic strength could penetrate faster and the adsorption of phenol was reduced.

When phenol almost penetrated in soils, the solution was changed to NaCl to wash out phenol. Figure 2 showed the pore volume in CK, 50BS-12 and 100BS-12 started at 19.86, 31.5 and 57.49. Using the CXTFIT model as a non-equilibrium model (NEM) for fitting the washout curve and the parameters of phenol washout was shown in Table 4. The R_w value of washout was bigger than R-value of transport, and it indicated the phenol keeping in soil was difficult to wash out. The R_w value in soils was in order of CK < 50BS-12 < 100BS-12, and the R_w of 100BS-12 was bigger than CK as 17.27 times. The linear analysis equation between residual retention in soil and saturation amount in soil was y = 6.5541x + 0.2064, R² = 0.771* (P < 0.01), it indicated that the residual retention in soil and saturation amount in soil was related each other. So that, the high modification ratio in soil could hold phenol and amount of pollutant not only was increasingly adsorbed but also the mass of soil was increased with the amount of pollutant after washing out [18] [28] . The residual retention in 50BS-12 and 100BS-12 was bigger than CK 6.98 and 7.21 times. It demonstrated that the surface was changed to amphoteric surfactant by covering BS-12, and BS-12 surfactant obviously affected phenol migration in the column experiment. Unit holding rate in each soil was CK (41.37%), 50BS-12 (83.13%) and 100 BS-12 (52.06%).

Last, the lou soil with high modification ratio had a better adsorption capacity for phenol migration. As seen the result with high blocking, high residual retention, and duration, it was confirmed that modified surface soil significantly reduced and blocked the pollutant migrant and also provided high organic matter to soil.