**1. Introduction**

Antimony (Sb) is an environmentally relevant potentially toxic element (PTE), usually combined, in alloys, with metals such as lead and zinc [1]. Sb, generally detected as a trace element in the Earth's crust (0.2–0.3 g per metric ton) and water (less than 1 μg <sup>L</sup>−1) [2], reached worrying levels of contamination in different world areas in the last decades [3]. This has been mainly due to anthropogenic activities, such as mining and the processing of Sb-containing ores [4]. High Sb levels in soils can also be due to vehicular traffic and recreational shooting [5–7]. Moreover, Sb is extensively used as a flame retardant in plastics, and as catalyst in the production of polyesters fibers [8]. As a consequence of its ubiquity and toxicity (Sb is recognized as carcinogenic and clastogenic agent), Sb is included in the list of high-priority pollutants by the U.S. Environmental Protection Agency and the European Union [9,10]. In most natural systems, Sb mainly occurs as reduced (trivalent Sb(III)) or oxidized (pentavalent Sb(V)) inorganic species [11]. Inorganic compounds of Sb are more toxic than its organic species, and Sb(III) compounds are predicted to be 10-fold more toxic than Sb(V) [1]. However, as emphasized by Filella et al. [12], this cannot be generalized, since toxicity depends on many factors (e.g., the target organism, the route of exposure, and the presence of other pollutants). Sb oxidation state, its reactivity, and its potential bioavailability in soil are largely dependent on the pH, redox conditions, and concentrations of co-occurring reducing and oxidizing agents [13]. Sb(V) is the prevalent form in aerobic conditions, and mainly occurs as octahedral antimonate ion, Sb(OH)6<sup>−</sup>. This latter form is considered the most stable form, compared to the thermodynamically unstable Sb(III), which mainly occurs in anoxic conditions, as Sb(OH)3 [1,12,14]. On the other hand, Sb(V) mobility is greater than Sb(III) in the 5.0–8.5 pH range, due to its negative

**Citation:** Diquattro, S.; Garau, G.; Garau, M.; Lauro, G.P.; Pinna, M.V.; Castaldi, P. Effect of Municipal Solid Waste Compost on Antimony Mobility, Phytotoxicity and Bioavailability in Polluted Soils. *Soil Syst.* **2021**, *5*, 60. https://doi.org/ 10.3390/soilsystems5040060

Academic Editor: Jorge Paz-Ferreiro

Received: 9 August 2021 Accepted: 24 September 2021 Published: 1 October 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

charge (pKa HSb(OH)6 = 2.55; pKa Sb(OH)3 = 11.8) [1,15]. Sb in soil is mostly associated with Fe, Mn and Al (hydr)oxides and organic matter [16,17], and their occurrence and abundance in soil seems to control Sb mobility and bioavailability, e.g., [18–20].

As a PTE, Sb can affect soil microbial communities and their functionality, as well as plant growth [21,22]. Diquattro et al. [21] showed a significant Sb impact on the composition and general catabolic activity of soil microbial communities, while Yu et al. [23] recorded reduced carbon mineralization and nitrification in the presence of Sb. Moreover, in situ Sb immobilization significantly stimulated the growth of *Helichrysum italicum* [22], indirectly showing a certain impact of this PTE on plant growth, which, however, was also confirmed in other studies, e.g., [24–26].

In the search for eco-friendly materials to restore Sb-contaminated soils, and/or to limit Sb ecotoxicological impact, many organic and inorganic amendments have been studied. For instance, Garau et al. [22] observed that the addition of a municipal solid waste compost (MSWC), together with an iron(Fe)-rich water treatment residue (WTR), favored the chemical and biological recovery of a subalkaline soil that was contaminated with Sb (~110 mg kg−<sup>1</sup> soil). Wang et al. [27] showed that ferrous Fe and nitrate promoted the formation of Fe plaques in rice, decreasing Sb bioavailability. Teng et al. [28] showed that Fe-modified rice husk hydrochar was effective at immobilizing Sb in soil, while the same was shown by Almas et al. [29], using Fe-rich slag in combination with FeSO4.

Despite the existence of substantial literature on effective amendments that are able to immobilize Sb in soil, very limited and often conflicting information can be found on the impact of compost on Sb mobility, and its toxicity and availability for plants. Nakamaru and Peinado [30] observed an increase in Sb availability in contaminated soils that were amended with compost. Verbeeck et al. [31] observed that the complexation capacity of dissolved organic matter could increase Sb mobility under different redox conditions. Lewinska et al. [32] showed that a long incubation time with compost (i.e., 140 d) did not always lead to Sb immobilization in different shooting range soils. Finally, Clemente [33] showed that adding greenwaste compost mulch to PTE-contaminated soil increased Sb mobilization. By contrast, Abou Jaoude et al. [34] showed that compost addition to Sbcontaminated sub-alkaline soils reduced water-soluble and exchangeable Sb, and increased its residual (non-extractable) fractions. Likewise, Garau et al. [35] showed that compost reduced the concentration of labile Sb in soil. Taken together, the results from these studies highlight the need for further investigations, aimed at clarifying the impact of compost on Sb-contaminated soils, and its potential use in the remediation of polluted environments. In particular, in view of a remediation intervention, it would be useful to establish the effects of compost on soil microbial communities, enzyme activities, plant growth, and Sb uptake in Sb-contaminated soils, other than establishing its impact on Sb mobility. Accordingly, the aim of the present study was to evaluate the effect of an MSWC, added at two different rates, on Sb mobility, phytotoxicity, and bioavailability in two polluted soils, as well as its impact on soil microbial activity, catabolic diversity, and enzyme activity.

#### **2. Materials and Methods**

#### *2.1. Soil and MSWC Origin, Characteristics and Mesocosms Set Up*

Different soil samples were randomly collected (0–30 cm depth) from two soils, named SA (soil A) and SB (soil B), located in North-Western Sardinia (Italy) (soil SA: 40◦5615.7 N 8◦5330.4 E; soil SB: 40◦4332.77 N 8◦2448.6 E). These soils were selected because they had never been cultivated (they were not treated with agrochemicals), were not close to Sb-contaminated areas, did not contain Sb (i.e., Sb < 1 mg kg−1), and exhibited different physical–chemical properties, which were previously reported in detail (Table S1; [21]). In the laboratory, soil samples were air dried, sieved to <2 mm and pooled together according to their origin (SA and SB). SA was a loamy coarse sand while SB was a sandy clay loam soil (USDA classification).

SA and SB soils were then spiked with Sb(V) (i.e., KSb(OH)6; CAS 12208-13-8; Merck) to obtain soils with medium–low (100 mg kg−1; S-100) or high (1000 mg kg−1; S-1000) Sb(V) pollution level as previously reported [21]. Spiked soils were kept at 40% of their water-holding capacity (WHC) using deionized water and equilibrated for nine months at 25 ◦C. During this period, soils were periodically mixed once a week and maintained at constant humidity. Afterwards, triplicate mesocosms from each soil type (SA and SB) and contamination level (S-100 and S-1000) were treated as follows: T0—polluted untreated soil; T1—polluted soil amended with 1% MSWC; T2—polluted soil amended with 2% MSWC. A total of thirty-six mesocosms (each consisting of 5 kg soil; 2 soil types × 2 contamination levels × 3 amendment treatments × 3 reps) were prepared using SA (n = 18) and SB (n = 18) soils. The MSWC rates were selected based on the specific Sb-immobilizing capabilities of compost highlighted in previous studies [17,22]. The MSWC was provided by Verde Vita Srl (Sassari, Italy) and was sieved to <2 mm before addition to mesocosm soils. The main characteristics of the MSWC were previously reported [17] and resumed in Table S2. Briefly, the MSWC had sub-alkaline pH (i.e., 7.93) and 27.3% organic matter content (OM); it had high cation exchange capacity (CEC, 92.3 cmol(+) kg−1), dissolved organic carbon (DOC, 0.82 mg kg−1) and humic acids content (14.2%). After amendment with MSWC, soils were carefully mixed, brought to 40% of their WHC and equilibrated for three months at 25 ◦C. During this period, soils were periodically mixed once a week and maintained at constant humidity.

#### *2.2. Soil Characterization and Sb Mobility in Treated and Untreated Soils*

After the equilibration period, selected physico-chemical properties were determined in treated and untreated polluted soils (Table 1). The pH, electric conductivity (EC), total organic carbon (TOC) and nitrogen (TN) were determined for treated and untreated soils following the national standard guidelines [36]. The DOC content was estimated as previously described by Brandstetter et al. [37]. The available phosphorus was quantified following the Olsen method (P Olsen) and the CEC was determined using the BaCl2 and triethanolamine method [36].

**Table 1.** Chemical characteristics of Sb-polluted SA and SB soils amended or not (control) with MSWC (dry matter basis). SA/SB-100 and SA/SB-1000 denote soil type and pollution level, i.e., 100 and 1000 mg Sb kg−<sup>1</sup> soil. For each soil type and Sb pollution level, different letters (e.g., a, b, c) denote statistical differences (Tukey–Kramer, *p* < 0.05) between treatments.


Total Sb concentration was quantified in all soils after digestion with aqua regia reverse solution (HNO3/HCl 3:1 ratio) and microwave mineralization (Milestone MLS1200), using graphite furnace atomic absorption spectroscopy (GFAAS; PerkinElmer AAnalyst 400- HGA 900, Software AA-WinLab32). A standard reference material (NIST-SRM 2711A) was included for quality assurance and quality control. The Sb mobility in polluted (amended and not) SA and SB soils, i.e., its chemical reactivity with the soil matrix, was evaluated through the sequential extraction procedure proposed by Wenzel et al. [38] with minor modifications. Briefly, triplicate soil samples (1 g each) from each mesocosm were first treated with water (25 mL) to estimate water-soluble Sb (step 0, this was the only additional

step with respect to the original procedure), then they were treated with 25 mL of a 0.05 M (NH4)2SO4 solution to quantify the readily exchangeable Sb fraction (step 1) and 25 mL of a 0.05 M (NH4)H2PO4 solution (step 2) to estimate surface-bound Sb, while Sb associated to amorphous and crystalline Al- and Fe-(hydr)oxides was quantified after extraction with 25 mL of a 0.2 M NH4- oxalate solution at pH 3.25 (step 3) and with 25 mL of a 0.2 M NH4-oxalate + 0.1 M ascorbic acid solution at pH 3.25 (step 4), respectively. After each step, soil samples were centrifuged at 3500 rpm for 10 min, filtered using Whatman filter No. 42 to separate the liquid and solid phases, and Sb concentration in the supernatant was quantified using GFAAS as previously described. A standard reference material (NIST-SRM 2711A) was included for quality assurance and quality control.

#### *2.3. Biolog Community-Level Physiological Profiles and Soil Enzyme Activities in Treated and Untreated Sb-Polluted Soils*

The activity and catabolic diversity of microbial communities inhabiting the different Sb-polluted soils (amended or not) was investigated using the Biolog community-level physiological profile (CLPP) approach using Biolog EcoPlates ™ (Biolog Inc., Hayward, CA, USA). In particular, triplicate soil samples (10 g) from each mesocosm were added with (90 mL) sodium pyrophosphate solution (2 g <sup>L</sup>−1) and serial ten-fold dilutions were obtained using 0.89% NaCl solution. The obtained soil suspensions were centrifuged for 5 min at 500 rpm and the clear supernatant was used to inoculate the wells of the Biolog EcoPlate ™ (120 μL per well). The Biolog EcoPlates ™ are ready to use 96-well microtiter plates containing a different carbon source of soil/environmental relevance in each well [39]. A total of 31 different carbon sources and a control well with no carbon (all replicated three times) were present in each Biolog EcoPlate ™ [40]. Inoculated plates were incubated at 28 ◦C for 6 days (144 h) and purple color formation in each well, due to the reduction of a tetrazolium dye and indicative of C-source catabolism, was recorded daily by measuring the absorbance at 590 nm (OD590), using an automatic Biolog MicroStation ™ reader. All OD590 data were first blanked against the absorbance at time 0 and further subtracted from the respective control well (with no carbon source). Finally, they were processed to obtain a measurement of the potential catabolic activity of the microbial community, i.e., the average well color development (AWCD). The latter was calculated as follows: AWCD = [ Σ (Ri − C)]/31 where C represents the absorbance value of control well, Ri is the absorbance of each response well, and 31 is the number of carbon substrates in the plate [41]. The richness value, or the number of substrates catabolized by each microbial community, was also determined as the number of wells with OD590 >0.15 [42]. Also, the Shannon–Weaver diversity index (H') was used to estimate the catabolic diversity of microbial communities and calculated as follows:

H = − ∑(Pi × ln Pi), where Pi is the ratio between the absorbance value in the blank subtracted ith well (1 to 31) and the total absorbance values of all wells [43].

All Biolog-derived parameters (AWCD, richness and H') refer to the 120 h incubation time as this time point provided the best discrimination between treatments.

Enzymatic activities, such as dehydrogenase (DHG), β-glucosidase (GLU) and urease (URE), were determined colorimetrically in triplicate soil samples collected from each mesocosm as described by Alef and Nannipieri [44]. Briefly, the DHG activity was spectrophotometrically quantified (OD480 nm) after incubation of soil samples (10 g) with a triphenyltetrazolium chloride solution and expressed as triphenyl formazan (TPF) formed per g soil (dry weight basis). GLU activity was determined spectrophotometrically (OD400 nm) after incubation of soil samples (1 g) with p-nitrophenyl glucoside and expressed as p-nitrophenol released per g soil (dry weight basis). Finally, URE activity was determined spectrophotometrically (OD690 nm) after incubation of soil samples (5 g) with urea and expressed as ammonia released per g soil (dry weight basis).

#### *2.4. Sb Phytotoxicity and Bioavailability in Treated and Untreated Sb-Polluted Soils*

The influence of MSWC on Sb phytotoxicity was determined using triticale plants (× *Triticosecale* Wittm. cv. Universal) grown in treated and untreated Sb-polluted soils. This plant species was selected since in previous studies its growth was significantly affected by the presence of PTE in the growing medium [45–48]. A total of thirty-six pots each containing 1.5 kg of soil deriving from the different mesocosms were set up, i.e., 3 replicated pots × 3 amendment treatments (T0, T1, T2) × 2 Sb contamination levels (100 and 1000 mg kg−1) × 1 plant species × 2 soil types (SA and SB). Seven triticale plants were planted in each pot after their germination in the dark at 25 ◦C. Planted pots were arranged in a completely randomized design and plants were grown over 8 weeks in a naturally lit greenhouse under controlled conditions (20–25 ◦C temperature, 60–70% relative humidity). At harvest, shoots were separated from roots, carefully washed with deionized water and dried at 55 ◦C for 72 h. Plant growth, i.e., root and shoot dry weight values, was used to estimate soil Sb phytotoxicity.

Sb bioavailability, i.e., the Sb uptaken by triticale plants, was determined after mineralization of roots and shoots with 2 mL suprapure H2O2 and 9 mL of HNO3 and ultrapure H2O (ratio 1:1), using a Microwave Milestone MLS 1200. The total Sb concentration in the mineralization solutions was determined using GFAAS as previously reported. Peach leaves were used as standard reference material (NIST-SRM 1515). The Sb translocation factor (TFSb) was calculated as the ratio between Sb concentration in shoots and that present in roots.
