Sorption behavior of copper nanoparticles in soils compared to copper ions

Geoderma 235–236 (2014) 127–132

Contents lists available at ScienceDirect

Geoderma journal homepage: www.elsevier.com/locate/geoderma

Sorption behavior of copper nanoparticles in soils compared to copper ions Dorit Julich a,⁎, Stefan Gäth b a b

Institute of Soil Science and Site Ecology, TU Dresden, Pienner Str. 19, 01737 Tharandt, Germany Institute of Landscape Ecology and Resources Management, Justus-Liebig-University, Heinrich-Buff-Ring 26C, 35392 Giessen, Germany

a r t i c l e

i n f o

Article history: Received 6 September 2013 Received in revised form 23 June 2014 Accepted 6 July 2014 Available online xxxx Keywords: Copper nanoparticles Copper Sorption Freundlich isotherm

a b s t r a c t Metallic nanoparticles have special physical and chemical properties which determine a particular behavior in environmental systems and organisms. While several studies investigated the differences in the toxicity of metallic nanoparticles compared to their ionic forms or salts, there is little knowledge about processes in complex environmental media. For instance, the sorption processes in soils crucially influence accumulation, transport and/or release into other media (water, biota, etc.). Our study assessed the sorptivity of copper oxide nanoparticles (CuO-NPs) in comparison to copper ions (Cu2+) in batch experiments. The results showed significant differences in the solid to liquid distribution at equilibrium and indicated much stronger sorption of CuO-NPs at soil components compared to Cu2+. The sorption isotherms of both variants were fitted to the Freundlich equation showing clear differences of the Freundlich parameters KF and n. The values for Cu2+ sorption were in the range of agricultural soils in Germany (log KF: 2.6–4.1, n: 0.9–1.6). On the contrary the isotherms for the CuO-NP experiments were strongly shifted to the solid phase (log KF: 4.0–9.0, n: 1.3–3.7). Both Cu2+ and CuO-NP sorptions (expressed as log KF) were significantly correlated (P b 0.05) to pH, carbonates, soil organic carbon and amorphous Fe in the soils. However, a larger data set is needed to generate reliable statistical results. Further, more research is required to identify reasons for the detected differences in sorption behavior between nanoparticulate copper and copper ions. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Nanoparticles (NPs) are characterized by at least one average dimension of b100 nm and have special physical and chemical properties based on their size, distribution, morphology and phase (Christian et al., 2008; Nel et al., 2006). Therefore, NPs may differ considerably from their bulk counterparts resulting in different behaviors in environmental systems and organisms (Taylor and Walton, 1993). As a result of increasing industrial production, the release of engineered nanoparticles (ENPs) into the environment is highly probable (Biswas and Wu, 2005; Ma et al., 2010; Nel et al., 2006). Furthermore, many studies demonstrated adverse effects (toxicity) of NPs on plants and other organism (e.g. Karlsson et al., 2008; Midander et al., 2009; Mishra and Kumar, 2009; Nair et al., 2010; Navarro et al., 2008; Nowack and Bucheli, 2007). Such NP-related adverse effects are very complex and strongly depend on the physico-chemical characteristics and interrelation of these properties. Luyts et al. (2013) therefore linked specific properties of NPs separately to their toxicity. Regarding metallic NPs, additional toxicity may be caused by dissolution of metal ions from the particles or by involved redox-processes. ⁎ Corresponding author. E-mail addresses: [email protected] (D. Julich), [email protected] (S. Gäth).

http://dx.doi.org/10.1016/j.geoderma.2014.07.003 0016-7061/© 2014 Elsevier B.V. All rights reserved.

The present study focuses on copper oxide nanoparticles (CuO-NPs) which are widely used in industrial production (e.g. electrics, ceramics, films, polymers, inks, metallics, coatings) and have specific optical, electrical, and catalytic properties (e.g. Lee et al., 2008). Copper is an essential element for organisms but is toxic above a species-dependent tolerance limit. Several recent studies revealed differences in toxicity of copper salts or ions compared to nanoparticulate copper (Amorim and Scott-Fordsmand, 2012; Gomes et al., 2012; Griffitt et al., 2008; Meng et al., 2007). These studies also indicated that the negative effects were not fully caused by released Cu ions from the particles, but primarily induced nanoparticle-specific. Though dissolved metal ions may contribute to toxicity, more stable particles can accumulate and persist inside an organism (Midander et al., 2009). Nair et al. (2010) summarized the effects of metallic NPs (including CuO-NP) on plant growth and suggested that an aggregation/agglomeration of NPs may block pores and channels resulting in higher phytotoxicity of the metal ions which are more mobile within plants. Nevertheless, studies of Lee et al. (2008) and Stampoulis et al. (2009) indicated that comparatively high concentrations of copper nanoparticles are needed to cause visible effects on plant vitality where plant species was an important influencing factor. However, there is a lack of knowledge about mobility and sorption behavior of metallic NPs in soils as crucial accumulation and transfer zone as well as potential source for NPs in ecosystems (Klaine et al., 2008; Ma et al., 2010). The main problems of investigating metallic

128

D. Julich, S. Gäth / Geoderma 235–236 (2014) 127–132

NPs in the complex medium soil are the separation of natural NPs (colloids) from ENPs and the implementation of an appropriate experimental set-up (homogeneous mixing, prevention of aggregation, etc.). Additionally, there is still no information available on dissolution and transformation processes of metallic NPs after addition to a test medium which considerably will influence their fate and effects in the terrestrial environment (Klaine et al., 2008). In general, soils provide a large and reactive sink for substances with high surface reactivity. Strong sorption processes of metallic NPs in soils are therefore conceivable (Klaine et al., 2008). Several studies assessed the behavior of metallic NPs in porous or artificial soil media and demonstrated the dependency of ENPs-mobility from properties of the nanoparticles, test media and test conditions like solution pH and ionic strength (Christian et al., 2008; Fang et al., 2009; Jones and Su, 2012; Lecoanet et al., 2004; Nowack and Bucheli, 2007). However, porous media cannot sufficiently represent complex soil systems where sorption processes on heterogeneous soil constituents occur. Fang et al. (2011) examined the copper transport in soil columns in the presence and absence of TiO2-NPs and derived pedotransfer functions (Freundlich and Langmuir) to describe the sorption processes. Collins et al. (2012) demonstrated in a field study that Cu- and ZnONPs are not completely adsorbed to soil constituents but are mobile in agricultural soils. The sorption of metallic NPs on soil colloids (e.g. clay, organic matter, iron oxide, other minerals) or incorporation into such colloids may be of particular relevance for metal transport through soil profiles (Gilbert et al., 2009; Klaine et al., 2008). In the presented study we examined the sorption behavior of copper oxide nanoparticles in comparison to copper ions in different soils by batch experiments. Batch experiments are a common method to assess the sorption behavior of heavy metals in soil where the effect of contaminant concentration on sorption processes can be included by conduction tests with varying spike concentrations (Krupka et al., 1999). Besides ionic strength and pH value of the batch solution, which affect the position of the sorption isotherm (Utermann et al., 2005) considerably, the size distribution (aggregation/agglomeration) will influence the results of the experiments with nanoparticles (Bian et al., 2011). However, it has to be considered that batch experiments provide the liquid/solid partitioning at equilibrium (KD value) but contain no information on behavior in flow conditions and on metal bonding forms (ion exchange, chemisorptions, bound to complexes and/or precipitates). The main questions in our study were: (1) are there differences in sorption behavior of Cu2+ ions and CuO-NPs? (2) Can sorption isotherms for both Cu2+ and CuO-NP describe sorption processes? (3) Which soil parameters influence Cu2+/CuO-NP sorption?

2. Materials and methods 2.1. Soil samples In this study, six different top soils from agricultural managed sites in Hesse/Germany were analyzed. The sampling strategy and information about local characteristics are given in Zörner (2010). The main parameters for the selected soils are listed in Table 1. The dried (40 °C) soil

samples were sieved for 2 mm and stored at room temperature. The particle size distribution was analyzed by Köhn-pipette procedure (German standard DIN 18123). Soil pH values were determined in 0.01 M CaCl2 suspension with a soil/solution ratio of 1:2.5 wt/vol and the content of carbonates was measured with a Scheibler-apparatus according to the German standard DIN 19684-5. The analysis of soil organic carbon (SOC) was conducted with a C/N-Analyzer (German standard DIN ISO 10694). Cation exchange capacity CEC was calculated according to Krogh et al. (2000) using measured SOC and clay contents (Eq. (1)). CEC ¼ 0:95 þ 2:90  1:72  SOC þ 0:53  clay:

ð1Þ

The background concentrations of Cu in soil were determined by aqua regia digestion for total Cu contents and by extraction with 0.025 M Na2-EDTA (90 min, soil/solution ratio 1:10 wt/vol) to estimate adsorption involved Cu (potentially mobile fraction) (Welp and Brümmer, 1999; Zörner, 2010). The contents of amorphous iron (Feox), manganese (Mnox) and aluminum (Alox) were determined by extraction with oxalate (Schlichting et al., 1995). All extracts were measured by inductively coupled mass spectroscopy (ICP-MS Agilent 7500ce) (Table 1). 2.2. Copper ions and nanoparticles For the sorption experiments with Cu2+ ions, the spike Cu solutions were prepared from copper(II)–nitrate–trihydrate (Cu(NO3)2 ∗ 3H2O) in deionized water. The CuO-NPs were purchased from IoLiTec (Ionic Liquid Technology GmbH, Germany) as dispersion in water (100 g CuO/L water). The particle size provided by IoLiTec was in the range from 40 to 80 nm. A nanoparticle tracking analysis (NTA) for CuO-NPs diluted in water was conducted on a NanoSight LM14 device (NanoSight Ltd., Amesbury, UK) with a 532-nm laser (50 mW). The results revealed a heterogeneous distribution of the particle sizes of the manufactured NPs with a mean size of 173 nm (sd 75 nm). To achieve well mixed dispersion and to minimize aggregation and agglomeration, the nanoparticle solution was shaken and ultrasonicated (stabilization step) before applied in the batch experiments (cf. Lee et al., 2008). 2.3. Batch experiments The batch sorption experiments were performed by mixing 10 g air dried soil (b2 mm) with 25 mL of 0.01 M Ca(NO3)2 solution, spiked with different amounts of Cu (copper(II)–nitrate–trihydrate) and CuO-NPs respectively, all in duplicates. Since the initial levels in the nanoparticle variants were adjusted to copper oxide-levels, the effectively added Cu concentrations of the single levels are lower compared to the Cu2+ variants (Table 2). The soil–solution mixtures were shaken horizontally for 16 h at room temperature. Subsequently, the samples were centrifuged at 1500 rpm for 30 min and filtered through disposable 0.45 μm syringe filters. For stabilizing the samples until Cu analysis they were immediately acidified with 150 μL HNO3. The Cu concentrations were analyzed with ICP-MS (Agilent 7500ce).

Table 1 Selected soil characteristics of the six test soils (agricultural top soils from Hesse/Germany). Sample

pHCaCl2 [−]

CaCO3 [wt.%]

SOC [wt.%]

CECa [cmolc kg−1]

Sand [wt.%]

Silt [wt.%]

Clay [wt.%]

Cutotb [mg kg−1]

CuEDTAc [mg kg−1]

A B C D E F

7.1 6.7 7.4 6.2 5.4 5.8

0.08 0.05 0.64 0.03 0 0

1.09 1.26 1.49 0.99 0.91 1.18

15.8 18.2 26.3 14.6 15.6 24.7

46.2 17.0 3.3 29.7 44.3 3.6

36.0 62.4 62.9 53.9 36.7 62.7

17.8 20.6 33.8 16.4 19.0 33.7

13.12 15.80 17.77 14.13 13.73 17.32

3.49 5.15 5.44 4.04 2.96 4.11

a b c

CEC calculated from Eq. (1) (Krogh et al., 2000). Total Cu content in soil (aqua regia digestion). Na2-EDTA-extractable Cu content in soil.

D. Julich, S. Gäth / Geoderma 235–236 (2014) 127–132 Table 2 Effectively applied Cu concentration in the batch sorption experiments with Cu2+ ions and CuO nanoparticles. Level

Control I II III IV V VI

Effectively applied Cu concentration(referred to batch solution) [mg L−1]

Effectively applied Cu concentration(referred to soil phase) [mg kg−1]

Cu2+ variants

CuO-NP variants

Cu2+ variants

CuO-NP variants

0 0.2 1 2 10 40 200

0 0.16 0.80 1.60 7.99 31.95 159.77

0 0.5 2.5 5 25 100 500

0 0.40 2.00 4.00 19.97 79.89 399.44

129

The influence of soil parameters on sorption behavior of Cu2 + ions and CuO-NPs was tested by linear regression (statistical significance P b0.05). 3. Results and discussion The background copper concentrations of the six test soils (Table 1) were within the typical range for arable soils in Germany (Scheffer et al., 2002: 18 mg kg−1; Thiele and Leinweber, 2001: 12.3 mg kg−1; Zörner, 2010: 10.0–36.3 mg kg− 1). The EDTA-extractable concentrations reached 22–33% of the total Cu contents indicating an elevated potential of mobilization. 3.1. q/c relation of Cu2+ and CuO-NP

2.4. q/c relation The solid phase Cu concentration qi was calculated as the difference between the added Cu concentration c0 and Cu equilibrium concentration in solution ci (Eq. (2)), added by the original Cu pool in the soils cb (background). Here, the EDTA-extractable metal pool, which in contrast to the total pool is involved in short-term reactions controlling sorption and solubility, proved to be suitable for the correction of the q/c relation (Reiher, 2008; Welp and Brümmer, 1999). qi ¼ c0 −ci þ cb :

ð2Þ

The ratio of the quantity of the adsorbate on solid phase and the quantity of the adsorbate in solution at equilibrium represents the distribution coefficient KD for the individual concentration levels of the batch experiments (Krupka et al., 1999): K D ¼ qi =ci :

ð3Þ

The q/c relation of copper in soil can be described adequately by sorption isotherms. It has been demonstrated that Cu sorption follows either the Langmuir (Eq. (4)) or the Freundlich (Eq. (5)) isotherms (Arias et al., 2006; Bradl, 2004; Elzinga et al., 1999; Usman, 2008): qi ¼

bKci 1 þ Kci

n

qi ¼ K F ci

ð4Þ

ð5Þ

where qi is the adsorbed quantity of an adsorbate i; ci is the equilibrium solution concentration of the adsorbate i; b is the upper limit for qi (maximum adsorption of i); K is the steepness of the isotherm (measure of affinity of the adsorbate for the surface); KF is the Freundlich affinity coefficient, which is represented as y-intercept of the log qi against log ci plot; n is the (non-)linearity parameter (n = 1 linear C-type isotherm) and is derived from the slope of the log–log-plot. 2.5. Data verification and statisticals The analytical and calculated data were verified in three steps: (1) plausibility (exclusion of values q ≤0 or c b detection limit), (2) analytical uncertainty (coefficients of variation of the replicates b25%), and (3) “Freundlich conformity”. According to Utermann et al. (2005), the test on “Freundlich conformity” examines the linear slope of the Freundlich isotherm in the log–log plot. The gradient between the points of the lowest solution concentration should be in the range 0.3–1.2. Outside this range, it was tested which point could be eliminated to receive a higher regression coefficient of the isotherm.

The results of the batch experiments indicated considerable differences in the sorption behavior of Cu2+ ions and CuO-NPs in the tested soils. The q/c relations of Cu2+ ions (Fig. 1a) are more shifted to the solution concentration whereas the Cu concentrations in equilibrium solution of the nanoparticle experiments were far lower and shifted to the solid phase (Fig. 1b). This indicates that CuO-NPs are sorbed quantitatively stronger to soil constituents compared to Cu2+. Nevertheless, the formation of CuO-aggregates or their integration in complexes cannot be excluded. In both cases NPs could be separated from equilibrium solution by the filtering step of the batch experiments resulting in lower residual solution concentrations. Though the experimental approach involved ultrasonication and shaking of the nanoparticle solutions, recent studies indicate aggregation of Cu-NPs during storage and/or transportation, and failure of disaggregation during sample preparation (Jones and Su, 2012; Midander et al., 2009). Furthermore, particles may reagglomerate after the preparation step (Dhawan and Sharma, 2010). Altogether, a wide variety in particle size distribution from small particles in nm-range to greater fraction in μm-range is expected. The nanoparticle tracking analysis (NTA) as pre-test of our batch experiments confirmed this assumption. However, a second test after shaking/ sonication was not conducted in this study. It can be assumed that the preparation procedure is adequate to prevent agglomeration (weak van der Waals Forces) rather than aggregation (strong chemical bonds) (cf. Jiang et al., 2009). The agglomeration/aggregation processes and resulting size distribution further depends on the given medium (solution characteristics) and particle concentration which influence their surface properties. Thus, the surface reactivity increases with decreasing particle size (increasing ratio of surface atoms to total atoms) leading to larger chemical and biological activities (Midander et al., 2009). The KD values defined as ratio of the quantity of the adsorbate on solid phase and the quantity of adsorbate in equilibrium solution were determined individually for all tested concentration levels. The KD for Cu2 + ion values increased at the lower concentration levels but decreased at the highest Cu levels (Fig. 2a). Usman (2008) found decreasing KD values with increasing spiked Cu concentration in batch experiments on Egyptian soil samples with high pH values (7.4–8.3) and high carbonate contents. It is suggested that the high KD values in experiments with low metal concentrations may be associated with sorption sites of high selectivity associated with strong bonding energy. In contrast, unspecific sorption dominates at higher metal concentration because of the increasing occupation of the specific bonding sites leading to lower KD values (Sastre et al., 2006; Usman, 2008). Following this concept, our experiments show that specific Cu2+ sorption occurs at spike concentrations up to 40 mg Cu L−1. For spike concentration between 40 and 200 mg Cu L−1, a change of sorption processes can be observed. Here, most specific bonding sites may be occupied whereas more unspecific bonding occurs resulting in lower KD values. For soil E, the decrease starts at a lower spiked concentration level indicating that less specific bonding sites are available (Fig. 2a). Here, the lower pH value (pH = 5.4) of the soil may have an influence. In the CuO-NP experiments, we found no decrease of KD values in the tested

130

D. Julich, S. Gäth / Geoderma 235–236 (2014) 127–132

Fig. 1. The q/c relation for Cu from six soils in Hesse/Germany (c — equilibrium solution concentration, q — adsorbed concentrations including the EDTA extractable native fraction) resulting from batch experiments with a) Cu ions and b) CuO nanoparticles.

concentration range. On the contrary, the KD values strongly increase with Cu concentration increase in all soils and to a much higher level compared to the Cu2+ experiments (Fig. 2b, note different scales in subpanels a and b). This shows the high sorption capacity for this type of adsorbate in the tested soils. It is possible that Cu nanoparticles also use other bonding sites due to their high surface reactivity leading to predominantly specific and less unspecific bonding in the observed concentration range. In both variants, Cu2+ and CuO-NP, soil C exhibited the highest KD values which is caused by the differing soil characteristics (e.g. higher pH value and content of carbonates) compared to the other soils (cf. Table 1). The influence of soil parameters on the sorption processes is discussed below in Section 3.3. 3.2. Sorption isotherms of Cu2+ and CuO-NP According to the literature the sorption of copper in soils follows the Langmuir as well as the Freundlich isotherm (Arias et al., 2006; Bradl, 2004; Elzinga et al., 1999). However, the fitting of the data obtained from our experiments to the Langmuir isotherms yielded in no acceptable results (results not shown). The adsorbed Cu concentrations strongly increased over the whole range of solution concentrations at equilibrium in both batch variants and in all soils (Fig. 1). Therefore, the Freundlich model yielded a better description of the sorption isotherms in all cases. These results are supported by observations of Fang et al. (2011) who achieved better fitting of the Freundlich equation compared to Langmuir for Cu sorption under unsaturated concentration conditions. The calculated values for the Freundlich coefficient KF (uncorrected) range from 1.5E + 02 to 2.4E + 03 (log KF: 2.2–3.4) for the Cu2 +

experiments and from 1.7E + 04 to 1.2E + 10 (log KF: 4.2–10.1) for the CuO-NP experiments (Table 3, Fig. 3). The arithmetic mean of KF for Cu ions (7.0E + 02) was several orders of magnitudes lower compared to the mean KF for CuO-NPs (1.7E + 08) and the variance between the tested soils was clearly higher in the latter ones (Fig. 3). The mean Freundlich exponents n were 0.93 for Cu2 + and 2.59 for CuO-NP representing the clearly steeper isotherms for the nanoparticle data. The KF and n values for Cu2 + are in range of values published by Reiher (2008), Thiele and Leinweber (2001), and Utermann et al. (2005) for arable soils in Germany. However, the determined Freundlich parameters indicate significant different sorption characteristics of Cu ions compared to CuO-NPs. The coefficients of determination (R2) reveal deviations from the Freundlich isotherm in both test variants (Table 3). This may be caused by uncertainties in the measurement of Cu concentrations in solution, especially when they are close to the detection limit, which also impacts the calculated sorbed Cu contents. A further difficulty in finding an adequate fitting to this type of isotherm is the requirement of constant test conditions during the batch experiments (particularly of pH value and ionic strength in the batch solution). Utermann et al. (2005) stated that the q/c relation is notably influenced from solution pH for values pH b 5. However, the pH values of the test soils were N5 and the parameter was not measured during the batch experiments in our study. The data verification tests excluded all non-plausible and highly uncertain analytical values. The test on “Freundlich conformity” of Cu2+ isotherms determined non-conform data points (slope values outside of the range 0.3–1.2) for three soils, but no substantial improvement of the fitting (higher R2) was achieved by excluding the lowest concentration levels. However, the elimination of the data point of the highest spike concentration in the Cu2+ experiments (cf. Sastre et al., 2006) led to clearly better fitting with a mean R2 of 0.995 (Fig. 3). In the conformity test for the CuO-NP data we found non-conform Freundlich isotherms for all six soils. Here, clearly higher correlation coefficients (R2 = 0.975) were reached when the two lowest concentration levels

Table 3 Freundlich equation parameters KF (Ln mg1 − n kg−1), n and R2 for Cu2+ and CuO nanoparticle sorptions in soil (uncorrected values). Sample

Fig. 2. KD values [L kg−1] resulting from batch experiments with a) Cu ions and b) CuO nanoparticles (ascending initial CO concentration levels from 0 to 200 mg L−1); note different y-axis scales.

A B C D E F

Cu

CuO-NP

KF

n

R2

KF

628 417 2358 297 146 346

1.09 0.91 1.23 0.83 0.66 0.85

0.991 0.978 0.969 0.967 0.955 0.952

1.56E 1.39E 1.19E 1.17E 1.65E 6.70E

+ + + + + +

07 06 10 05 04 04

n

R2

3.13 2.60 4.32 2.05 1.54 1.94

0.897 0.934 0.883 0.973 0.959 0.976

D. Julich, S. Gäth / Geoderma 235–236 (2014) 127–132

131

Fig. 3. Distribution of Freundlich parameters KF, n, and R2 of six top soils, Freundlich equation fitted to sorption data of Cu2+ ions and CuO nanoparticles (whisker-boxplots: line — median, box 25–75% quartiles, error bar — minimal/maximal values).

were excluded (Fig. 3). Altogether, the correction steps resulted in better fitting to Freundlich isotherms for both experiments and in a notable reduction of the mean residual standard deviation in equation fitting. Utermann et al. (2005) considered an isotherm with a residual standard deviation of b0.1 as well fitted to the Freundlich model which is clearly given here for Cu2 + data (0.04) and nearly reached for CuO-NP data (0.12). 3.3. Influence of soil parameters on Cu2+ and CuO-NP sorption In literature several parameters are discussed to influence copper sorption processes in soils. For instance the soil pH value is often mentioned as dominant parameter controlling the partitioning of metals between solid and liquid phases (Bradl, 2004; Welp and Brümmer, 1999). Thiele and Leinweber (2001) found significant correlations between the Freundlich coefficient KF of Cu and soil pH (R = 0.633), CEC (R = 0.710), organic carbon concentration (R = 0.562), and Mndithionite (R = 0.570). Fang et al. (2011) demonstrated that soil properties considerably influence Cu transport in the presence of TiO2-NPs. An increasing soil pH led to enhanced transport of TiO2-associated Cu, whereas CEC and DOC were negatively correlated. Further, the sorption of copper in soils is influenced by the content of soil organic matter as well as Mn and Fe oxides (Bradl, 2004). Naturally occurring polyelectrolytes such as fulvic and humic acids may enhance nanoparticle transport (Lecoanet et al., 2004). Jones and Su (2012) found improved transport rates of elemental Cu-NPs in porous media in the presence of humic and fulvic acid by sorbing to particle surface and reducing attachment efficiency. In the presented study, correlations between Freundlich coefficients (log KF) and the following soil parameters were tested: pH, carbonates, SOC, CEC, clay content, Feox, Mnox, Alox, and total Cu content in soil. Both test variants (Cu2+ and CuO-NP) revealed significant linear correlations (P b 0.05) for pH with a correlation coefficient of R = 0.890 / 0.912, for

carbonates (R = 0.904 / 0.923), for SOC (R = 0.889 / 0.843), and for Feox (R = 0.829 / 0.848). Log KF clearly increased with rising pH values in the observed pH range (5.4–7.4), but remarkable differences between Cu ions and Cu nanoparticles were observed (Fig. 4). The log KF level was higher and the slope was much steeper in the CuO-NP experiment compared to Cu2+. The same effect can be seen for the other parameters in Fig. 4. For SOC, log KF is slightly increasing at higher soil organic carbon values, but the narrow range of concentration of SOC in the six tested soils (0.9–1.5 wt.%) allows no further conclusions. The observed correlation between log KF and Feox showed a negative slope, but the quite narrow Feox concentration range has to be considered here as well. No significant linear correlations were found for Mnox, Alox, CEC, clay and original Cu content. Although Fig. 4 and the results of the correlation tests indicate a strong dependence of log KF from pH and carbonates in the tested range, these correlations may not be linear. Nevertheless, the increase of KF values with increasing pH in our data is in accordance with the other studies and the obviously much stronger sorption of nanoparticulate CuO compared to Cu ions with clearly higher increase at rising pH values is remarkable. 4. Conclusion The sorption behavior of Cu ions and CuO nanoparticles was considerably different in the tested soils. The results of batch experiments indicate much stronger sorption of CuO-NP to the soil compared to Cu2+. For nanoparticles, the occupation of other bonding sites is suspected to lead to a predominantly specific bonding, but less unspecific in the observed Cu concentration range. The Cu2+ as well as the CuO-NP sorption could be sufficiently described by Freundlich isotherms. The Freundlich parameters KF and n are clearly different for Cu2 + and CuO-NP confirming the high sorptivity of Cu nanoparticles. Cu2+ and CuO-NP sorptions (log KF) were significantly correlated to pH, carbonates, soil

Fig. 4. Relation between Freundlich coefficient log KF and soil parameters.

132

D. Julich, S. Gäth / Geoderma 235–236 (2014) 127–132

organic carbon and amorphous Fe. As the statistical analysis was only based on data sets of six soils, the results have to be confirmed with further experiments. A larger data set is needed to generate reliable statistical results. Furthermore, advanced test methods are required to identify the cause for the observed differences including detailed control of the experiment conditions and particles properties. Acknowledgment Many thanks are due to Sascha Setzer and Heike Weller from the Institute of Landscape Ecology and Resources Management/Justus-Liebig University of Giessen for ICP-MS analysis of batch samples and extracts. We thank Dr. Michael Bunge from the Institute of Applied Microbiology/ Justus-Liebig University of Giessen for conducting the Nanoparticle Tracking Analysis of our copper oxide nanoparticles. References Amorim, M.J.B., Scott-Fordsmand, J.J., 2012. Toxicity of copper nanoparticles and CuCl2 salt to Enchytraeus albidus worms: survival, reproduction and avoidance responses. Environ. Pollut. 164, 164–168. Arias, M., Perez-Novo, C., Lopez, E., Soto, B., 2006. Competitive adsorption and desorption of copper and zinc in acid soils. Geoderma 133, 151–159. Bian, S.-W., Mudunkotuwa, I.A., Rupasinghe, T., Grassian, V.H., 2011. Aggregation and dissolution of 4 nm ZnO nanoparticles in aqueous environments: influence of pH, ionic strength, size, and adsorption of humic acid. Langmuir 27, 6059–6068. Biswas, P., Wu, C.-Y., 2005. Nanoparticles and the environment. J. Air Waste Manag. Assoc. 55, 708–746. Bradl, H.B., 2004. Adsorption of heavy metal ions on soils and soils constituents. J. Colloid Interface Sci. 277, 1–18. Christian, P., von der Kammer, F., Baalousha, M., Hofmann, T., 2008. Nanoparticles: structure, properties, preparation and behaviour in environmental media. Ecotoxicology 17, 326–343. Collins, D., Luxton, T., Kumar, N., Shah, S., Walker, V.K., Shah, V., 2012. Assessing the impact of copper and zinc oxide nanoparticles on soil: a field study. PLoS One 7, e42663. Dhawan, A., Sharma, V., 2010. Toxicity assessment of nanomaterials: methods and challenges. Anal. Bioanal. Chem. 398, 589–605. Elzinga, E.J., Van Grinsven, J.J.M., Swartjes, F.A., 1999. General purpose Freundlich isotherms for cadmium, copper and zinc in soils. Eur. J. Soil Sci. 50, 139–149. Fang, J., Shan, X., Wen, B., Lin, J., Owens, G., 2009. Stability of titania nanoparticles in soil suspensions and transport in saturated homogeneous soil columns. Environ. Pollut. 157, 1101–1109. Fang, J., Shan, X.Q., Wen, B., Owens, G., Zhou, S., 2011. Transport of copper as affected by titania nanoparticles in soil columns. Environ. Pollut. 159, 1248–1256. Gilbert, B., Ono, R.K., Ching, K.A., Kim, C.S., 2009. The effects of nanoparticle aggregation processes on aggregate structure and metal uptake. J. Colloid Interface Sci. 339, 285–295. Gomes, S.I.L., Novais, S.C., Scott-Fordsmand, J.J., De Coen, W., Soares, A.M.V.M., Amorim, M. J.B., 2012. Effect of Cu-nanoparticles versus Cu-salt in Enchytraeus albidus (Oligochaeta): differential gene expression through microarray analysis. Comp. Biochem. Physiol. C Toxicol. Pharmacol. 155, 219–227. Griffitt, R.J., Luo, J., Gao, J., Bonzongo, J.-C., Barber, D.S., 2008. Effects of particle composition and species on toxicity of metallic nanomaterials in aquatic organisms. Environ. Toxicol. Chem. 27, 1972–1978. Jiang, J., Oberdörster, G., Biswas, P., 2009. Characterization of size, surface charge, and agglomeration state of nanoparticle dispersions for toxicological studies. J. Nanopart. Res. 11, 77–89. Jones, E.H., Su, C., 2012. Fate and transport of elemental copper (Cu0) nanoparticles through saturated porous media in the presence of organic materials. Water Res. 46, 2445–2456. Karlsson, H.L., Cronholm, P., Gustafsson, J., Möller, L., 2008. Copper oxide nanoparticles are highly toxic: a comparison between metal oxide nanoparticles and carbon nanotubes. Chem. Res. Toxicol. 21, 1726–1732.

Klaine, S.J., Alvarez, P.J.J., Batley, G.E., Fernandes, T.F., Handy, R.D., Lyon, D.Y., Mahendra, S., McLaughlin, M.J., Lead, J.R., 2008. Nanomaterials in the environment: behavior, fate, bioavailability, and effects. Environ. Toxicol. Chem. 27, 1825–1851. Krogh, L., Breuning-Madsen, H., Greve, M.H., 2000. Cation-exchange capacity pedotransfer functions for Danish soils. Acta Agric. Scand. Sect. B Soil Plant Sci. 50, 1–12. Krupka, K.M., Kaplan, D.I., Whelan, G., Serne, R.J., Mattigod, S.V., 1999. Understanding variation in partition coefficient, Kd, values. Volume 1: the Kd model, methods of measurement, and application of chemical reaction codes (EPA J 402-R-99-004A). Lecoanet, H.F., Bottero, J.Y., Wiesner, 2004. Laboratory assessment of the mobility of nanomaterials in porous media. Environ. Sci. Technol. 38, 5164–5169. Lee, W.-M., An, Y.-J., Yoon, H., Kweon, H.-S., 2008. Toxicity and bioavailability of copper nanoparticles to the terrestrial plants mung bean (Phaseolus radiatus) and wheat (Triticum aestivum): plant agar test for water-insoluble nanoparticles. Environ. Toxicol. Chem. 27, 1915–1921. Luyts, K., Napierska, D., Nemery, B., Hoet, P.H.M., 2013. How physico-chemical characteristics of nanoparticles cause their toxicity: complex and unresolved interrelations. Environ. Sci. Process. Impacts 15, 23. Ma, X., Geiser-Lee, J., Deng, Y., Kolmakov, A., 2010. Interactions between engineered nanoparticles (ENPs) and plants: phytotoxicity, uptake and accumulation. Sci. Total Environ. 408, 3053–3061. Meng, H., Chen, Z., Xing, G., Yuan, H., Chen, C., Zhao, F., Zhang, C., Zhao, Y., 2007. Ultrahigh reactivity provokes nanotoxicity: explanation of oral toxicity of nano-copper particles. Toxicol. Lett. 175, 102–110. Midander, K., Cronholm, P., Karlsson, H.L., Elihn, K., Möller, L., Leygraf, C., Wallinder, I.O., 2009. Surface characteristics, copper release, and toxicity of nano- and micrometersized copper and copper(II) oxide particles: a cross-disciplinary study. Small 5, 389–399. Mishra, V.K., Kumar, A., 2009. Impact of metal nanoparticles on the plant growth promoting rhizobacteria. Digest J. Nanomater. Biostruct. 4, 587–592. Nair, R., Varghese, S.H., Nair, B.G., Maekawa, T., Yoshida, Y., Kumar, D.S., 2010. Nanoparticulate material delivery to plants. Plant Sci. 179, 154–163. Navarro, E., Baun, A., Behra, R., Hartmann, N.B., Filser, J., Miao, A.-J., Quigg, A., Santschi, P.H. , Sigg, L., 2008. Environmental behavior and ecotoxicity of engineered nanoparticles to algae, plants, and fungi. Ecotoxicology 17, 372–386. Nel, A., Xia, T., Madler, L., Li, N., 2006. Toxic potential of materials at the nanolevel. Science 311, 622–627. Nowack, B., Bucheli, T.D., 2007. Occurrence, behavior and effects of nanoparticles in the environment. Environ. Pollut. 150, 5–22. Reiher, W., 2008. Entwicklung eines regionalisierten Modells zur Projektion des landnutzungsabhängigen Schwermetallstatus von Oberböden. Boden und Landschaft, 52. Justus-Liebig-Universität Giessen (Dissertation). Sastre, J., Rauret, G., Vidal, M., 2006. Effect of the cationic composition of sorption solution on the quantification of sorption–desorption parameters of heavy metals in soils. Environ. Pollut. 140, 322–339. Scheffer, F., Schachtschabel, P., Blume, H.-P., Scheffer-Schachtschabel, 2002. Lehrbuch der Bodenkunde, 15. Aufl./. ed, Spektrum-Lehrbuch. Spektrum Akad. Verl., Heidelberg. Schlichting, E., Blume, H.-P., Stahr, K., 1995. Bodenkundliches Praktikum - Eine Einführung in pedologisches Arbeiten für Ökologen, insbesondere Land- und Forstwirte und für Geowissenschaftler, 2., neubearbeitete Auflage. Blackwell Wissenschafts-Verlag, Berlin, Wien. Stampoulis, D., Sinha, S.K., White, J.C., 2009. Assay-dependent phytotoxicity of nanoparticles to plants. Environ. Sci. Technol. 9473–9479. Taylor, R., Walton, D., 1993. The chemistry of fullerenes. Nature 363, 685–693. Thiele, S., Leinweber, P., 2001. Parameterization of Freundlich adsorption isotherms for heavy metals in soils from an area with intensive livestock production. J. Plant Nutr. Soil Sci. 164, 623–629. Usman, A.R.A., 2008. The relative adsorption selectivities of Pb, Cu, Zn, Cd and Ni by soils developed on shale in New Valley, Egypt. Geoderma 144, 334–343. Utermann, J., Meyenburg, G., Altfelder, S., Gäbler, H., Duijnisveld, W., Bahr, A., Streck, T., 2005. Entwicklung eines Verfahrens zur Quantifizierung von Stoffkonzentrationen im Sickerwasser auf der Grundlage chemischer und physikalischer Pedotransferfunktionen. Final Report, BMBF-Forschungsvorhaben 02WP0206. Welp, G., Brümmer, G.W., 1999. Adsorption and solubility of ten metals in soil samples of different composition. J. Plant Nutr. Soil Sci. 162, 155–161. Zörner, D., 2010. Prognose des Schwermetallhaushaltes von Landschaften mit dem Modell ATOMIS. Boden und Landschaft, 53. Justus-Liebig-Universität Giessen (Dissertation).