*2.2. Soil Characterization*

The investigated soil was sampled from a land used in the past for agriculture and located in the Litorale Domizio Flegreo and Agro Aversano (Campania Region, Italy) (Figure 1). In particular, this land includes an area of 1076 km<sup>2</sup> (57 municipalities) and has been affected by the illegal disposal of hazardous wastes [23,24]. The soil was sampled manually from the top 20 cm of the soil over an area of about 1 m2, homogenized, and then stored in hermetic containers. Immediately after collection, the samples were dried at 40 ◦C in an oven (Argolab, TCN115) and then kept at room temperature.

**Figure 1.** The sampling point (40◦9605" N, 14◦1184" E).

The particle size distribution analysis was performed according to American Society for Testing and Materials (ASTM) method D 422-63 [25]. Only the fraction of soil that was smaller than 2 mm, which was assumed to be the most contaminated [26], was used in all of the tests and analytical determinations. The pH of the soil was evaluated according to the method of Violante ed Adamo [27]. Organic matter was determined through the loss on ignition (LOI) index [28]. Total PTEs concentration was measured in the liquid phase from soil acid mineralization procedure [29]. Before the analysis, the solid and liquid phases from the acid mineralization were separated through centrifugation at 4600 rpm for 20 min using an IEC CENTRA GP8R centrifuge (Needham Heights, MA, USA). Then, the resulting liquid phase was analyzed by atomic adsorption spectrometry using a Varian spectrometer, Model 55 B SpectrAA (F-AAS) (Varian Australia Pty Ltd., Victoria, Australia) equipped with a flame (acetylene/air) and a deuterium background correction. The limit of detection (LOD) values for each of the analyzed elements were 5 × 10−<sup>2</sup> mg·L−<sup>1</sup> for Cd and Zn, 0.5 mg·L−<sup>1</sup> for Cr, 0.2 mg·L−<sup>1</sup> for Cu, 0.25 mg·L−<sup>1</sup> for Fe, 0.1 mg·L−<sup>1</sup> for Mn, 0.3 mg·L−<sup>1</sup> for Ni, and 1 mg·L−<sup>1</sup> for Pb. A sequential extraction procedure proposed by the Community Bureau of Reference (BCR) [30,31] was performed on the contaminated soil samples to determine the metal partition among different soil fractions. This procedure was based on an initial extraction in 40 mL of acetic acid (0.11 M) (step 1—exchangeable and weak acid soluble fraction). Afterward, a volume of 40 mL of hydroxylammonium chloride solution (0.5 M) was added to the residual soil from "step 1" and acidified by the addition of a 2 M HNO3 solution (step 2—reducible fraction). Successively, 20 mL of hydrogen peroxide (8.8 M) and 50 mL of ammonium acetate (1 M) were used for oxidizing the soil (step 3—oxidizable fraction). The last step consisted in a soil acid mineralization (step 4—residual fraction).

### *2.3. Soil Flushing Lab Scale Tests*

Soil flushing tests were performed following the indication of Hauser et al. [21]. In particular, the columns were prepared using two polypropylene "falcon" test tubes characterized by a diameter and length of 30 and 80 mm, respectively. Tests were carried out by filling each tube with 40 g of the investigated soil. Soil was packed in the columns until no reduction of volume was observed, according to Pontoni et al. [32]. Subsequently, a total bed volume (bv) of 50 mL was obtained in each tube.

During the first series of tests, two EDDS solution molarities ([EDDS]) were tested (0.36 and 3.6 mM). The flushing solution was leached through the columns using IVEK rotary pumps (IVEK Corporation, North Springfield, Vermont) at three different percolation velocities (1.7, 2.1, and 2.7 mL·h−1), which corresponded to three Empty-Bed Contact Times (EBCTs), i.e., 33, 27, and 21 h. The total treatment lasted 20 days. The solution collected at the bottom of the column was sampled at different times of the test and analyzed to measure the concentration of the following metals: Cu, Zn, Fe and Mn.

In order to evaluate metals mobility in the soil column, other tests were performed for selected experimental conditions ([EDDS] = 3.6 mM and EBCT = 33 h). After the flushing process, the soil column was divided into four sections, and the metal concentration was measured in each of them (Soil Flushing Test 1, SF1). Furthermore, sequential extraction was carried out on the soil contained in each layer to evaluate the changes of PTEs fractionation after the treatment. A comparative test using only pure water as a flushing agen<sup>t</sup> was also performed. A second series of tests (Soil Flushing Test 2, SF2) was performed to evaluate metal transportation along the columns. For this purpose, before the flushing process, the columns were divided into four layers, each of them containing 10 g of the contaminated soil. The flushing solution was sampled at the bottom of each layer and analyzed for PTEs and main cations concentration. Flushing operative conditions were the same as also adopted for the previous test. Finally, a third series of tests (Soil Flushing Test 3, SF3) was conducted to study the release of PTEs in the leachate after the soil flushing treatment (i.e., release due to the rain phenomenon). In this case, only 4 bv of the EDDS solution was injected through the columns at the beginning of the test, and then only pure water was added as a flushing agent. The flushing solutions were collected at the bottom of the columns and analyzed for PTEs evaluation. Also in this case, flushing operative conditions set for SF1 and SF2 were adopted.

### *2.4. Soil Washing Lab Scale Tests*

Comparative soil washing tests were conducted on the same soil, at the lab scale, in 50-mL plastic reactors. Two different values of the liquid to solid ratio (LSR) were chosen, namely 5:1 and 10:1 (*v*/*w*). Soil washing parameters were optimized on the basis of the results of previous studies [24,33] conducted on the same soil. In particular, an extracting solution of [EDDS] = 0.36 mM was adopted to study the process efficiency in EDDS deficiency conditions, whereas an extracting solution of [EDDS] = 3.6 mM was used to achieve the best process performance. Two reaction times (i.e., 48 and 96 h) were selected since a previous work [33] proved these times to be efficient for achieving the PTEs equilibrium conditions. All tests were conducted in triplicate to reduce experimental errors. The metal concentration in the exhausted solution was measured by atomic adsorption spectrometry following the procedure described in Section 2.2. One-way and two-way analysis of variance (ANOVA) were used to analyze the statistical differences among treatments. Comparisons were made with the post-hoc Tukey's Honestly Significant Difference HSD test. Statistical significance was assumed at *p* < 0.05. Statistical analyses were conducted in Microsoft® Excel 2013/XLSTAT ©-Pro (Version 7.2, 2003, Addinsoft, Inc., Brooklyn, NY, USA) and GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA).
