There is a growing development and use of nano-enabled products in agriculture due to their increased activity associated with the small size and to the potential for reduction of the amount of applied active ingredients [1
]. Given its antimicrobial and chemical properties, silver nanoparticles (AgNPs) have been used in a large number of global market products, particularly for disinfection purposes [4
]. Potential applications of AgNPs as fungicides in agriculture have also been investigated [8
]. These were proven effective against pathogenic fungi which commonly affect crops such as corn, barley and rice or against mildew infestation on roses [10
]. Suggested advantages on the use of AgNPs-enabled fungicides compared to AgNO3
or to conventional formulations relate to higher efficacy at lower application doses (15 g AgNPs ha−1
, compared to commonly applied doses of 105 g ha−1
to 6 kg ha−1
, for conventional fungicides), and reduction in human toxicity [8
To develop environmentally safe management practices related to the use of AgNPs in agriculture, however, there is a need to understand geochemical processes regulating the fate of AgNPs in natural media like soil and to assess potential risks on terrestrial systems. Upon entry in the terrestrial ecosystem, a number of transformations occur depending on biogeochemical conditions that control fate, behaviour, and ecotoxicity of NPs in soil [13
] including pH, organic matter, soil texture, clay content and type and amount of metal oxides, or presence of microbes [6
]. Transformations of AgNPs in soil include oxidative dissolution, aggregation to the soil ligands or stabilization in suspension through natural coating by dissolved organic matter (DOM) [6
]. The type of (artificial) surface coating of NPs also affects transformation processes like aggregation and dissolution processes in natural systems [23
] which subsequently might affect their fate in soils. Several studies reported on the fate of AgNPs (and of Ag2
S NPs) in agricultural soils following the application of sludge from wastewater treatment plants (WWTP) and biosolids [16
]. These investigations showed that partitioning of Ag to pore water from AgNPs directly applied to soils is generally limited under natural soil conditions, and dependent on NPs oxidative dissolution; also, DOM stabilizes AgNPs in pore water and suppresses Ag oxidation, with short-chained compounds being preferentially adsorbed over long-chained, aromatic compounds; finally, organic ligands such as those present in root exudates or fertilizers may further facilitate the release of ionic Ag and increase bioavailability of Ag to organisms and plants [16
]. So far, few studies addressed the impact of variable soil conditions and AgNPs properties on their geochemical fate in soil and subsequent potential effects on soil ecological functions and biodiversity. Such assessments are essential in order to promote and regulate the use of NPs in agriculture [29
Traditionally, the chemical fate and ecological risks of metals including those in the form of NPs in soil in the terrestrial environment is assessed using several single extractions of increasing extraction strength. Dilute salt extractions (e.g., 0.01M CaCl2
) or extractions using chelates (e.g., 0.4 M glycine or 0.05 M NH4
] are able to represent the direct (or plant) availability or potential availability of metals (exchangeable and organically bound fraction of metals from soils) quite well [33
]. Here we extend the use of such methods to evaluate the geochemical availability of Ag from both AgNPs and PVP-AgNPs in agricultural soils.
Aside from the geochemical aspects also the impact of NPs on ecosystem functioning is relevant since AgNPs and released silver cations can have a detrimental effect on soil (micro)organisms [15
] and soil enzymes, indirectly affecting microbial-mediated processes including nutrients cycling [38
]. Soil exoenzyme activities are considered relevant indicators of the overall biological functioning of soils. Even more so since activity levels, related to microbial-mediated processes [38
], respond rapidly to changes in soil conditions like those brought forth by the introduction of AgNPs. So far, however, contrasting results were reported on the effect levels of AgNPs once added to soil on exoenzymes activities. Hänsch and Emmerling [36
] tested the effects of AgNPs on the activities of leucine-amino-peptidase, β-cellobiohydrolase, acid phosphatase, β-glucosidase and xylosidase, and observed no effect of AgNPs for an Ag dose up to 0.32 µg g−1
. Shin et al. [40
] evaluated the inhibitory effect of AgNPs on the activities of specific soil exoenzymes related to nutrient cycles including urease, acid phosphatase, arylsulfatase and β-glucosidase, and found that AgNPs were capable of inhibiting exoenzyme activity but within a rather wide range from 1 to 1000 µg g−1
. Shin et al. [40
] attributed the observed adverse effects to that of the AgNPs themselves rather than that of toxicity of dissolved Ag+
ions. Similarly, Peyrot et al. [38
] observed that soil enzyme activities were inhibited by particulate forms of AgNPs. Gaddam et al. [39
] evaluated the effects of indigenous mycogenic AgNPs on soil exoenzymes, in mine-waste contaminated soils and observed an impact on activities of urease, phosphatase, dehydrogenase and β-glucosidase of AgNPs at Ag concentrations of 150 µg g−1
The aim of the present study was twofold: (i) to evaluate the effect of soil properties and of a polymer coating on the geochemical distribution of AgNPs added to soil, and (ii) to assess the inhibitory effects on enzymatic activities of five soil exoenzymes related to nutrient cycles (β-glucosidase, cellulase, acid phosphatase, protease and urease.
In addition, the suitability of various soil extraction tests to characterize environmentally and agriculturally relevant pools of Ag in treated soil was evaluated. Polyvinylpyrrolidone was selected as a representative polymer coating because it has been widely used to improve long-term colloidal stability of AgNPs. A pot experiment was conducted to quantify the soil-pore water partitioning of AgNPs from soils, as well as the potential availability of Ag retained in the soil solid matrix upon addition of both coated and uncoated AgNPs.
2. Materials and Methods
2.1. Sampling and Characterization of Soil Samples
Soil samples from agricultural soils were collected from the 0 to 10 cm layer from non-contaminated grassland fields located in mining, rural and industrial areas across Portugal (located at Southwest, Centre, and Northwest Portugal, respectively) to ensure variability in soil type and soil properties, notably in pH, organic matter, and clay content. Samples were taken using a plastic spade, stored in plastic bags, air-dried at room temperature until constant weight and subsequently sieved at <2 mm (using a Nylon® sieve, Bioblock Tamis Nylon® DIN 4195, by Fisher Scientific SAS, Illkirch Cedex, France). This fraction was used for soil analyses and pot experiments.
The following parameters were determined in soil samples: pH-CaCl2
, organic carbon content, clay content. Soil pH was measured using a glass electrode in a 1:5 (v
) suspension of soils in CaCl2
according to the ISO 10390:1994 procedure. Organic carbon (OrgC) was determined after addition of hydrochloric acid to remove carbonates present (ISO 10694:1995). The clay content was analyzed using a Coulter LS230 laser diffraction particle size. Cation exchange capacity (CEC) was determined in barium chloride solution buffered at pH = 8.1 using triethanolamine as described by Bascomb [41
Pseudo-total concentrations of Ag, Al, Ba, Cd, Cu, Cr, Fe, Mn, Ni, Pb and Zn in soils were determined by aqua regia digestion; 3.0 g of dried soil were extracted with 21 mL HCl and 7 mL HNO3 (ISO 11466:1995)). Levels of chemical elements in the extract were measured by ICP-MS (Thermo X Series) according to ISO 17294-1:2005 and ISO 17294-2:2003.
2.2. Synthesis of AgNPs and Characterization of AgNPs Colloids
AgNPs were synthesized by reduction of silver nitrate (AgNO3
) with an excess of sodium borohydride (NaBH4
), needed to reduce the ionic silver and to stabilize the silver nanoparticles as described by Mulfinger et al. [42
]. A 30 mL of NaBH4
solution (2.0 mM) (NaBH4
, >96%; Sigma Aldrich, Merck KGaA, Darmstadt, Germany) chilled in an ice-bath and under stirred vigorously with a magnetic stir, was slowly mixed (drop by drop) with 10 mL of AgNO3
(1.0 mM) (AgNO3
, 99.9%; Sigma Aldrich, Merck KGaA, Darmstadt, Germany). Stirring was ceased upon completion of the addition of 2 mL of silver nitrate. The resulting suspension which exhibited a light-yellow color was kept in the dark, at room temperature until analysis and used as stock suspension for soil amendment.
PVP-AgNPs were synthesized by the same procedure previously described through a chemical reduction method from the aqueous solution of silver nitrate using PVP as a stabilizing agent in the presence of sodium borohydride as a reducing agent [43
]. The colloidal suspension was stored at room temperature, in the dark until analysis and use in amendment experiments.
The morphology of the AgNPs colloids was determined by transmission electron microscopy (TEM) using a Hitachi H-9000 TEM microscope operating at 300 kV and equipped with energy-dispersive X-ray spectroscopy (EDX) (Hitachi High-Tech Corporation, Tokyo, Japan). Samples for TEM analysis were prepared by putting one drop of AgNPs colloids onto a carbon-coated copper grid and then letting the solvent evaporate. The size of the AgNPs was evaluated by dynamic light scattering (DLS) using a Malvern Zeta-Sizer Nano-ZS (Model Zen3500, Malvern Panalytical Ltd., Malvern, UK), and derived from TEM images (mean NPs diameter was quantified based on measurements of at least 200 individual particles using ImageJ software). UV−Vis absorption spectra for the AgNPs were determined on a Jasco V-560 UV-VIS spectrophotometer (300–700 nm). Quartz cuvettes with 1cm-pathlength were used as sample holder for all UV-Vis analysis. The zeta potential was determine using phase analysis light scattering. Three replicate measurements of zeta potential were performed for each sample.
Bulk concentrations of Ag in AgNPs stock suspensions were determined by elemental analysis by ICP-OES (Jobin Yvon Activa M, Horiba, Palaiseau, France) following microwave-assisted aqua regia digestion using Teflon vessels (microwave oven: MARS model, CEM). A sub-sample of 2 mL of each suspension was digested with 2 mL aqua regia. The limit of quantification for Ag was 4 µg L−1. The bulk average Ag concentration in the two colloidal suspensions was 20 mg Ag L−1 (n = 3).
2.3. Pot Experiments
For each soil, pots were amended with AgNPs, PVP-AgNPs, or used as control (n = 3 for each treatment and the control). The control pots were prepared by adding 140 mL of ultrapure water to 500 g of air dried soil to obtain a water content of 70% of the soil water holding capacity. A volume of 140 mL of either AgNPs or PVP-AgNPs colloidal suspension was added to each AgNP treated pot. Nanoparticle suspensions for each pot were sonicated for 30 s, immediately applied after sonication, and thoroughly mixed with soil using a wooden stick. The dose applied was equivalent to an average concentration of 5.6 mg Ag kg−1
soil assuming a homogeneous distribution over the 500 g of soil in the pot. The actual distribution in soil is measured after completion of the pot experiment. An average concentration of 5.6 mg Ag kg−1
soil is comparable to those used in studies addressing partitioning and toxicity [6
]. During the trial period (30 days) all pots were weighted every 2 to 3 days. Ultrapure water was added from the top of the pots to maintain a constant moisture content.
2.4. Sampling and Analysis of Pore Water
To extract in-situ pore water samples each pot was equipped with RhizonTM solution samplers [31
]. Samplers were installed at 4 cm (“SF sampler”) and 7 cm below the soil surface (“SB sampler”). The pore water samplers had a microfiltration membrane with a nominal pore size of 0.12–0.17 µm with a low sorption capacity and no ion exchange capacity [31
Pore-water samples were collected at t = 24, 48, and 72 h after the addition of AgNPs and PVP-AgNPs to the soil, and once a week after that, until day 30. Pore-water samples were kept at 4 °C and analyzed for Ag by ICP-OES within 1 to 4 days. The limit of quantification for Ag was 4 µg L−1.
2.5. Analysis of Ag Retained in the Soil Solid Matrix
After 30 days, the soil material was removed from all pots, divided into two equal parts (upper–SF-and bottom–SB-layers) and air-dried for 3–5 days (until constant weight). The pseudo-total content of Ag, Al, Fe and Mn in soils were determined by extraction using aqua regia and subsequent analyses of the elements by ICP using the same methods to characterize the soils mg Ag kg−1 soil.
In addition, the following extraction procedures were applied to air-dried soil samples from the SF layer:
Extraction with 0.01 M CaCl2
(1:10 weight to volume ratio). Thirty mL of 0.01 M CaCl2
solution were added to 3 g of soil in a polypropylene screw closure bottles [21
]. The bottles were mechanically shaken with end-over-end rotation for 2 h at room temperature. The extracts were centrifuged during 10 min at 3000 rpm and subsequently filtered using a Millipore filter unit and a filter paper (0.45 µm). A portion of each filtrate was collected and preserved at 4 °C until analysis of Ag, Al, Fe and Mn by ICP-OES.
Extraction with 0.4 M glycine (adjusted to pH 1.5, with 37% HCl stock solution) at a 1:100 weight to volume ratio. Five hundred mg of air-dried soil sample was weighted into a polypropylene screw closure bottle and 50 mL of a 0.4 M solution were added [47
]. The bottles were placed in an orbital incubator shaker at 37 °C for and shaken for 1 h. To ensure that all pH values were within 0.5 units of the starting pH (1.5), the pH was measured in part of the unfiltered extraction fluid. A portion of each filtrate was collected and preserved at 4 °C until analysis of Ag, Al, Fe and Mn by ICP-OES;
Extraction with 0.05 M NH4
-EDTA solution adjusted to pH 7 with 37% HCl stock solution. Thirty mL of 0.05 M NH4
-EDTA solution were added to 3 g of air-dried soil material in polypropylene bottles [48
]. The bottles were placed in an orbital incubator shaker at room temperature and shaken for 1 h. The extracts were filtered using a Millipore filter unit and a filter paper (0.45 µm). A portion of each filtrate was collected and preserved at 4 °C until analysis of Ag, Al, Fe and Mn by ICP-OES.
All chemicals were of analytical grade or better and all solutions were prepared using ultrapure water. All extractions were performed in duplicate. Two extraction blanks were included in each batch of 20 extraction bottles.
2.6. Analysis of Soil Exoenzymes Activities
Soil sub-samples from the SF layer of pots at the end of the 30 days were also collected and stored at 4 °C. Samples were passed through a 2 mm sieve before analysis, and their dry matter content was determined to express the enzymatic activity on an oven-dried soil weight basis (105 °C, 48 h).
Acid phosphatase activity was measured by incubating 1 g of soil with p-nitrophenyl phosphate in modified universal buffer (pH 6.5, 4 mL) at 37 °C, as described by Alef et al. [49
]. After 1 h, 0.5 M CaCl2
(1 mL) was added and the p-nitrophenol (PNP) released was extracted with 0.5 M NaOH (4 mL) and measured spectrophotometrically at 400 nm.
β-glucosidase activity was measured according to Eivazi and Tabatabai [50
] and Alef and Nannipieri [51
], in the same way as the acid phosphatase activity, except that the substrate was p-nitrophenyl-β-D-glucopyranoside and that the PNP released was extracted with 0.1 M tris(hidroxymetil)aminometane-NaOH at pH 12.0. β-glucosidase and acid phosphatase activities were both expressed in µmol PNP g (soil, d.w.)−1
Cellulases activity was determined according to Hope and Burns [52
] and refers to the combined action of endo-1,4-β-D-glucanase, exo-1,4-β-D-glucanase and β-D-glucosidase on Avicel, a purified depolymerised alpha cellulose. The reaction occurred by incubating 1 g of soil for 16 h at 40 °C, in a 0.1 M acetate buffer (pH 5.5, 0.2% NaN3
). The reducing sugars produced were determined spectrophotometrically at 520 nm, after the addition of Cu(II) and molybdo-arsenate reagents. Cellulases activity was expressed in µmol glucose g (soil, d.w.)−1
Urease activity was determined as described by Kandeler and Gerber [53
], measuring the NH3
released after incubating by incubating 5 g of soil with a solution of urea for 2 h at 37 °C, in a borate buffer (pH 10). The ammonium content of the centrifuged extracts was determined spectrophotometrically, at 690 nm, after reaction with sodium dicloroisocyanide 0.1%. Urease activity was expressed in µmol NH4+
-N g (soil, d.w.)−1
Protease activity was measured after the incubation of 1 g of soil with sodium caseynate (2% w
) in Tris-buffer pH 8.1, for 2 h at 50 °C [54
]. Released tyrosine reacted with Folin-phenol reagent to form a blue complex, which was determined spectrophotometricaly at 700 nm. Protease activity was expressed in mmol tyrosine g (soil, d.w.)−1
. All analytical measurements were carried out in triplicate.
2.7. Statistical Analysis
Results for the soil enzymatic activities were analyzed using a Factorial Analysis of Variance (ANOVA) with soil type and AgNP treatment as factors which could explain their variance, and possible interactions of statistical significance. Moreover, results for each soil type were subjected to one-way ANOVA to discriminate if there was an effect of the AgNP treatments on soil enzymatic activities. Whenever significant differences were found (p < 0.05), a post hoc Tukey Honest Significant Difference (HSD) test was used to further elucidate differences among means (p < 0.05). Statistical analysis was carried out with the software Statistica 6.0.
This study revealed that pore water Ag concentrations resulting from the application of non-coated and coated AgNPs to soils was very low (<184 µg L−1 (for AgNPs) and <21 µg L−1 (PVP-AgNPs), corresponding to <0.4% and <0.04% of dosed Ag, respectively). Measured Ag in soil pore water samples was mostly associated with the partition of dissolved Ag ions. Furthermore, dissolved Ag was detected only in topsoil samples from acidic soils and decreased further with an increase in soil organic carbon and clay content revealing the importance of soil properties in the retention of metallic NPs in soil along with their role in controlling NPs dissolution and bioavailability for plants and soil organisms. The type of NPs coating was an additional factor that controlled pore water concentrations and vertical displacement in soil. with the use of PVP-coated AgNPs that possess a larger hydrodynamic diameter resulted in lower pore water concentrations due to the combined effect of straining, reduced solubility and/or increased heteroaggregation with soil organic matter. This combined effect of reduced solubility, or increased retention in the topsoil, resulted in a 40–75% reduction of vertical displacement of PVP-AgNPs when comparing results with those from non-coated AgNPs. This effect was even stronger in soils with a higher clay content, higher pH and organic matter content. The impact of the PVP coating on the release of Ag seems to confirm that polymers can be used for a more controlled release of nanoforms of agrochemical products in soils and to reduce the potential risk of leaching of Ag to deep soil layers.
In view of its low direct availability in well-aerated soils the efficacy of the application of Ag-based NPs via soil for plant protection purposes may be questionable, unless it is intended to be used as a vector for controlled release of ionic Ag over time, e.g., at soil rhizospheres. In the rhizosphere of treated soils, the combined root exudation of protons and of low molecular weight organic ligands (e.g., citrate or malate) or aminoacids by plants under nutrient deficiency may potentially increase the dissolution and bioavailability of ionic Ag from AgNPs. This requires however further investigation. Moreover, targeting AgNPs application to plant rhizospheres must be accompanied by an integrated assessment of ecological risks since it may also cause higher potential localized exposure for sensitive receptors.
As for the soil exoenzymes activities studied, there was no clear effect of the AgNPs and of PVP-AgNPs on the enzymatic activities of β-glucosidase, cellulase, acid phosphatase and protease. In fact, for the Ag concentrations observed in the SF layer of in pots after 30 days of soil amendment (<16 µg g−1 dry soil), the influence of soil properties (particularly the low OrgC content) on the enzymatic activities surpassed that of the application of AgNPs, and no inhibitory effects in these soil enzymatic activities could be observed.