Science of the Total Environment 487 (2014) 771–777

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

The interaction of polymer-coated magnetic nanoparticles with seawater Enikö Kadar a,⁎, Íris L. Batalha b, Andrew Fisher c, Ana Cecília A. Roque b a b c

Plymouth Marine Laboratory, Prospect Place, The Hoe, Plymouth PL1 3DH, UK REQUIMTE, Departamento de Química, Faculdade de Ciências e Technologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal School of Geography, Earth and Environmental Sciences, University of Plymouth, Drake Circus, Plymouth PL4 8AA, UK

H I G H L I G H T S • • • • •

Hydrodynamic diameter (HD) of MNPs increase over time regardless of carrying solution. The relative changes in HD are carrying solution- and coating-dependent. Polydispersity indexes of the freshly suspended MNPs are below 0.5 for all coated MNPs. Initially high colloidal stability of MNP suspensions decreases regardless of carrying solution. Soluble iron leaching from MNPs is 3–9% of the total iron-content of the initially added iron.

a r t i c l e

i n f o

Article history: Received 7 September 2013 Received in revised form 16 October 2013 Accepted 12 November 2013 Available online 5 December 2013 Keywords: Iron oxide Magnetic nanoparticles Aggregation Dissolution Bioavailability

a b s t r a c t Laboratory studies were conducted to evaluate the interaction between bare and polymer-coated magnetic nanoparticles (MNPs) with various environmentally relevant carrying solutions including natural oceanic seawater with and without addition of algal exopolymeric substances (EPS). The MNPs were coated with three different stabilising agents, namely gum Arabic (GA-MNP), dextran (D-MNP) and carboxymethyl-dextran (CMD-MNP). The colloidal stability of the suspensions was evaluated over 48 h and we demonstrated that: (i) hydrodynamic diameters increased over time regardless of carrying solution for all MNPs except the GA-coated ones; however, the relative changes were carrying solution- and coat-dependent; (ii) polydispersity indexes of the freshly suspended MNPs are below 0.5 for all coated MNPs, unlike the much higher values obtained for the uncoated MNPs; (iii) freshly prepared MNP suspensions (both coated and uncoated) in Milli-Q (MQ) water show high colloidal stability as indicated by zeta-potential values below − 30 mV, which however decrease in absolute value within 48 h for all MNPs regardless of carrying solution; (iv) EPS seems to “stabilise” the GA-coated and the CMD-coated MNPs, but not the uncoated or the D-coated MNPs, which form larger aggregates within 48 h; (v) despite this aggregation, iron (Fe)-leaching from MNPs is sustained over 48 h, but remained within the range of 3–9% of the total iron-content of the initially added MNPs regardless of suspension media and capping agent. The environmental implications of our findings and biotechnological applicability of MNPs are discussed. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Magnetic nanoparticles (MNPs) are materials composed by magnetic elements, such as iron, cobalt and nickel, and present characteristic properties such as superparamagnetism, high coercivity, high magnetic susceptibility and low Curie temperature (Wu et al., 2008). Iron oxide MNPs may exist in different forms, such as magnetite, maghemite, wüstite, hematite, and goethite or their combinations, depending on the Fe(II)/Fe(III) ratio, which influences size, composition, morphology, and magnetic properties of the particles (Laurent et al., 2008). In order to keep their superparamagnetism particles must have a core diameter below 20 nm, but they may aggregate in larger particles of different ⁎ Corresponding author. Tel.: +44 1752 633450 (direct); fax: +44 1752 633101. E-mail address: [email protected] (E. Kadar). URL: http://www.pml.ac.uk (E. Kadar). 0048-9697/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.scitotenv.2013.11.082

sizes (Ho et al., 2011). To improve applicability MNPs are coated or complexed with materials that increase their selectivity or affinity to desired target molecules (Li et al., 2009). With the latest developments in nanotechnology (Dahl et al., 2007), MNPs can be applied for magnetophoretic separation of a wide range of materials (Li et al., 2009) as they provide several advantages over other available separation methods including recyclability for multiple usage, high-throughput process, low operational cost, high efficiency, flexible implementation and scalability (Kang and Park, 2005). These features make MNPs very attractive in numerous fields including biotechnology/biomedicine (Gupta and Gupta, 2005; Batalha et al., 2010; Haun et al., 2011; Dias et al., 2011), environmental remediation (Li et al., 2011; Xu et al., 2011; Lim et al., 2012; Toh et al., 2012; Huang et al., 2012), data storage (Hyeon, 2003) and other emerging fields. However, unavoidable problems may occur associated with particles in this nanosize range such as their intrinsic instability/ tendency to form aggregates and their interaction with the (often high

772

E. Kadar et al. / Science of the Total Environment 487 (2014) 771–777

ionic strength) media. The inherent magnetic forces contribute to the attractive forces among NPs which may lead to aggregation (Rosicka and Sembera, 2011). Moreover, uncoated metallic NPs are highly chemically reactive and easily oxidised, generally resulting in loss of magnetism and colloidal stability. Metal leaching may also occur due to physicochemical interactions between surface bound ions and the surrounding media (Kadar et al., 2013). Stabilisation of MNPs may be achieved by coating the particles, either by chemical bonding or physical adsorption, with different capping agents. A variety of coating agents have been used for the purpose of providing stabilised, monodisperse and biocompatible MNP suspensions. Such dispersants include both natural polymers such as gum Arabic (GA) (Roque et al., 2009; Batalha et al., 2010) dextran (Laurent et al., 2008) or carboxymethylated dextran (Liu et al., 2011) and synthetic materials including Naacrylic copolymer applied to the commercial product Nanofer 25S® (Kadar et al., 2011, 2012, 2013; Rosicka and Sembera, 2011) and amphiphilic triblock copolymers such as Pluronic F127 and Pluronic F68 (Lim et al., 2009). GA consists of three fractions, where the major one is a highly branched polysaccharide, about 10% (wt) of the total is a high molecular weight arabinogalactan-protein complex and around 1% (wt) of the total is protein (Dror et al., 2006). Despite such extensive research with Fe-rich nanoscale materials in general, their fundamental aggregation properties and stability under environmentally relevant conditions have not yet been fully established and there is virtually no such information available on the transformations that take place when MNPs come in contact with natural seawater. We have addressed this by studying the colloidal stability of both coated MNPs (gum Arabic, dextran and carboxymethyl dextrancoated) and uncoated MNPs incubated in natural oceanic seawater, in the presence or absence of algal exopolymeric substance (EPS) extract. The focus is on how surface coating influences MNP size, surface charge and chemical reactivity under a range of environmental conditions. EPS are ubiquitously released by microorganisms in aquatic environments and have a key role in the formation of marine biofilms and in colloid and trace element scavenging. Although originally the term EPS was used to describe extracellular polysaccharides, it is now acknowledged that these matrixes are more complex, including lipopolysaccharides, glycolipids, lipids, proteins or peptides and nucleic acids (Decho et al., 2010). These anionic colloid polymers can assemble into microgels possibly by means of hydrophobic and ionic interactions (Chen et al., 2011). Changes in the EPS assembly kinetics can be induced by engineered NPs (Chen et al., 2011) suggesting a previously unaccounted indirect ecological impact of synthetic NPs which may influence the environmental applicability of MNPs, especially when magnetic manipulations are intended on EPS-producing cells. Furthermore, putative EPS-mediated NP uptake mechanism was recently reported in some marine phytoplankton (Kadar et al., 2012) with potential impact on biomass production. Although we have started to recognise that marine algae can utilise nanoforms of the metal (Kadar et al., 2012), the role of EPS in uptake and its influence of on the environmental fate of MNPs is not yet fully understood. Thus, here we have used an EPS extract from a common phytoplankton species – Nannochloropsis salina – and added it to natural oceanic sea water in order to investigate the physicochemical transformations of MNPs in contact with seawater so that we can better understand their environmental fate and behaviour. Specifically, we have studied the Fe dissolution, particle aggregation/size distribution and zeta potential of MNPs in contact with seawater under a range of environmentally realistic conditions (i.e. typical in oceanic waters, in the presence of EPS particles and allowing ageing for 48 h). To the best of our knowledge, this is the first research attempting to systematically investigate the changes in physico-chemical properties of MNPs upon dispersion in seawater with and without EPS, which shed new light on the environmental fate and behaviour of these promising nanomaterials for environmental applications.

2. Experimental 2.1. Synthesis of MNPs Iron oxide magnetic nanoparticles were synthesised using the coprecipitation method. Briefly, a solution of 2.5 M ammonium hydroxide (Roth) was degassed by purging nitrogen for 20 min in a closed stirred reactor. After that period, 25 mL of an aqueous solution of 5.4 g of FeCl3·6H20 and 2 g of FeCl2·4H20 (Sigma-Aldrich) was added drop wise to the ammonium hydroxide solution under strong agitation of approximately 1000 rpm. Immediately after, 25 mL of 0.08 g/mL polymer solution was added dropwise to the reactor. The polymers used were, either gum arabic from acacia tree, dextran from Leuconostoc mesenteroides or carboxymethyl-dextran (Fluka). The reaction proceeded for 2 h under nitrogen flow. The resulting (nano)particles were extensively washed with deionised water by magnetic decantation. 2.2. EPS extraction from phytoplankton culture and quantification Exopolymer particles were extracted by cross-flow ultrafiltration according to the method previously reported by Zhang and Santschi (2009), from a 20 L stationary culture of N. salina (CCAP 849/3) grown in a 450 L photo bioreactor to a density of 16 × 106 cells mL−1. Culture conditions were: 20 °C; 16/8 light/dark regime; and f/2 medium dissolved in Instant Ocean®. The phytoplankton culture was centrifuged at 4000 rpm for 30 min and the supernatant was used to collect free dissolved EPS. Briefly, 20 L supernatant fraction was filtered (b0.45 μm) and the ultra-filtered on a 1 kDa cartridge (GE Healthcare UK) to ~300 mL retentate. The cartridge was rinsed with 200 mL MQ water, and then soaked for 6 h following washing with 200 mL MQ twice. The first retentate, the rinse and the two washing solutions were combined resulting in 1 L EPS extract which was used in subsequent experimentations as described below. The extract isolated by this method from both phytoplankton cultures and natural waters was reported to mainly contain glucuronic acid, fucose and galactose (Zhang and Santschi, 2009). The transparent exopolymeric substance (TEP) fraction of the EPS, operationally defined as Alcian Blue staining particles greater than 0.2, was quantified (Cunliffe et al., 2009) as the xanthan gum equivalents and was 4.66 mg L− 1 ± 0.81, which converted to carbon gives 2.94 mg C L−1 ± 0.51. These are typical concentrations in very productive waters, such as estuaries and thus a 1:20 (v/v) addition of this stock EPS to oceanic, 30 kDa-filtered seawater is an environmentally realistic dose. 2.3. Nanoparticle–seawater interactions Triplicate suspensions of the four types of MNPs were obtained by adding 1:10 v/v each working stock (e.g. ~500 mg L−1, which was the 100 × dilution of the original synthesised nanomaterial containing ~5 × 104 mg L−1 Fe) to the 3 environmentally relevant carrying solutions (total number of 45 suspensions) to a final volume of 50 mL: MQ water, natural oceanic seawater and seawater with EPS extract; these suspensions were continuously shaken for 48 h and kept at 20 °C. Subsamples of 10 mL were taken from each suspension both immediately and 48 h after addition of MNPs and were processed for nanoparticle characterisation, i.e. Fe dissolution (10 mL for centrifugal ultrafiltration using 3 kDa pore size Vivaspin®) and colloidal stability using dynamic light scattering (DLS) (2 mL for particle size distribution and 2 mL for zeta potential). 2.4. Fe dissolution from (nano)particles and total Fe content Triplicate samples (10 mL) were taken at different times (immediately after suspension of nanoparticles – referred to as “fresh”, and after 48 h – referred to as “aged”) and centrifuged (Beckman) for

E. Kadar et al. / Science of the Total Environment 487 (2014) 771–777

30 min (5000 g and 20 °C) using vivaspin 3 kDa centrifuge tubes (Amicon Ultracel, Millipore, USA). Aliquots of 4.5 mL of the filtrates were collected and acidified with 0.5 mL of concentrated HCl until analysis on Inductively Coupled Plasma Mass Spectrometry (ICP-MS, X Series 2, Thermo Scientific, Hemel Hempstead, UK) to determine the iron concentration. Seawater samples were diluted five-fold using 2% nitric acid. Calibration standards were matrix matched with the samples, i.e. they were prepared in five-fold diluted seawater that had not been exposed to Fe. Freshwater samples were analysed without further sample pre-treatment. Standards for this analysis were prepared in dilute (2%) nitric acid. For all analyses, the dwell time was 10 ms, sweeps were 50, ICP forward power was 1400 W, coolant gas flow rate was 14 L min− 1, auxiliary gas flow rate was 0.7 L min− 1 and nebuliser flow rate was 0.82 L min− 1. A V-groove nebuliser and a Sturman-Masters spray chamber were used for sample introduction. The polyatomic interference from 40Ar16O+ on the 56Fe+ isotope was effectively overcome by the addition to the collision cell of 7% hydrogen in helium at a flow rate of 3.5 mL min−1. Total Fe in each stock MNP solution was also determined on acid digests without the filtration step using a Varian SpectrAA 50 flame atomic absorption spectrometry instrument (air–acetylene flame was used and the lamp current was 5 mA; slit width was 0.2 nm; wavelength used was 248.3 nm and digests were diluted × 10). Samples and standards were prepared in HDPE tubes. All equipment had been carefully acid washed prior to use. All reagents were from Sigma Aldrich and at least p.a. grade. All solutions were prepared in MQ water (Millipore). 2.5. (Nano)particle colloidal stability Dynamic light scattering (DLS) methods were used to investigate the changes of zeta potential, size distribution and polydispersity of MNPs using Zetasizer Nano ZS (Nano series, Malvern Instruments, UK). MNPs were dispersed in seawater with or without addition of EPS extract and in MQ water. Samples were taken immediately and 48 h following dispersion and were analysed within 1 h of sampling. 2.6. (Nano)particle micromorphology Particle micro-morphology and core diameter of both bare and GAcoated MNPs were determined by transmission electron microscopy (TEM). TEM samples were prepared by evaporation of the dilute

773

suspensions onto formvar-coated grids. Electron micrographs were taken using a Hitachi 8100 TEM with a Rontec standard EDS detector and digital image acquisition. Owing to the TEM sample processing limitations (i.e. aggressive vacuum effect on the aggregates), there was no change detectable by TEM between the coated and uncoated MNPs (Fig. 4) or before and after exposure samples (data not shown). 2.7. Statistical analyses Statistical analyses were performed using Minitab v16. All results are presented as means ± standard error from mean (SEM). Significant differences between groups were determined using one way ANOVA followed by Tukey's comparisons. All statistical analyses used the default 5% rejection level. 3. Results and discussion 3.1. Colloidal stability of the MNPs in seawater and in the presence of microalgal EPS All MNPs (both coated and uncoated) freshly suspended in MQ water present average hydrodynamic diameters (HDs) below 700 nm in the following ascendant order: CMD-MNP (144 nm), GA-MNP (306 nm) (bare), MNP (597 nm) and D-MNP (677 nm) (Fig. 1). The "real" size however is smaller since, when dispersed particles move through a liquid medium, a thin electric dipole layer of the solvent adheres to their surface. This layer influences the movement of the particle in the medium, which will ultimately determine the HD. Thus the HD gives us information of the inorganic core along with any coating material and the solvent layer attached to the particle as it moves under the influence of Brownian motion. Nevertheless HD is an important parameter for understanding and optimising the nanoparticles' performance in biological assays as well as understanding the in vitro migration of the particles. GA is a complex polymer composed of 44% galactose, 27% arabinose, 16% glucuronic acid, 13% rhamnose, and 2–3% peptide moieties. It encompasses three fractions: a highly branched polysaccharide backbone of 1,3-linked β-galactopyranose units and 1,6-linked galactopyranose side-chains terminating in glucuronic acid terminating in glucuronic acid or 4-O-methylglucuronic acid residues; 10% (wt) of arabinogalactan-protein complex rich in hydroxyprolines, prolines and serines; and 1% (wt) fraction with the highest protein content rich in aspartic acid, serine, leucine and glycine. Dextran is a branched glucan

Hydrodynamic diameter (nm)

FRESH

AGED

Average particle diameter (nm)

12000

10000

8000

6000

4000

2000

GA-MNP

D-MNP

CMD-MNP

SW+EPS

SW

MQ

SW+EPS

SW

MQ

SW+EPS

SW

MQ

SW+EPS

SW

MQ

0

MNP

Nanoparticle type and carrying solution Fig. 1. Hydrodynamic diameter (based on DLS) of magnetic nanoparticles (MNPs) stabilised with different capping agents including gum Arabic (GA-MNP), dextran (D-MNP), CM-dextran (CMD-MNP) and uncoated (MNP) and suspended in media with distinct composition, e.g. ultrapure water (MQ), natural seawater (SW) with and without exopolymeric substances (EPS); columns represent average hydrodynamic diameter (nm) ± SEM, N = 5.

774

E. Kadar et al. / Science of the Total Environment 487 (2014) 771–777

with EPS, it is possible that rapid aggregation lead to gravitational removal of the large particles, which may have accounted for the apparent reduction in HDs measured by DLS. Zeta potential values did reach stability value of −30 mV (Malvern Instruments technical note on Zeta Potential theory) for all MNPs freshly suspended in MQ, but decreased in time and depending on carrier solution. This result was expected due to the increased ionic strength of seawater media and may be explained by Debye–Hückel theory (Debye and Hückel, 1923). However, further in depth rheological studies are required to better understand the colloidal transformations taking place when MNPs interact with seawater and EPS (Fig. 3). In summary, the collective data from the DLS methods applied indicate that GA is providing the best stability in both high and low ionic strength media, remaining stable over 48 h. Transmission electron micrographs indicate that the coating layer does not alter the core diameter of individual particles, which is approximately 10 ± 1 nm (Fig. 4). Microalgal EPS addition increases HDs of MNPs, which is in agreement with previous reports (Chen et al., 2011). An apparent “stabilising” effect of EPS on the bare MNPs as indicated by the average HDs of the aggregates in suspension that decreases over 48 h may be due to the bias of the method towards smaller particles when aggregation is fast. This, however, was not the case for the D- and CMD-MNPs which increased in HD over 48 h in the presence of EPS in seawater. 3.2. Particle dissolution in seawater — the effect of EPS and stabilising agent Iron leaching from magnetic nanoparticles, expressed as a percentage weight of the initial starting MNP iron weight (Fig. 5), were determined, i.e. 0.563 mg L− 1 in the GA-MNP, 0.512 mg L− 1 in the bare MNP, 0.612 mg L−1 in the D-MNP and 0.586 mg L−1 in the CMDMNP. The amount of Fe leached from MNPs was constant regardless of capping agent when dispersions were prepared in MQ water, and represent ~3% of the total Fe in the original MNPs. This result was consistent over the 48 h, with the exception of the GA-stabilised MNPs that released the twice as much Fe after 48 h in suspension in MQ water (Fig. 5). A similar ~ 3% of Fe release was also observed when MNPs (both coated and uncoated) were suspended in seawater, but the dissolution increased significantly (to ~ 10%) for both the bare- and the dextran-coated MNPs within 48 h. Curiously, the presence of EPS in seawater did not significantly influence Fe dissolution over 48 h, which remained b 6% for all MNPs. There was some EPS-mediated increased dissolution observed for the uncoated and the CMD-coated MNPs but

Polydispersity index

having mainly α-1,6 glycosidic linkages with α-1,3 linkages at branching point and CM-dextran is a carboxymethylated derivative of dextran (Dias et al., 2011).Both CM-dextran and GA possess carboxylic groups on their surfaces, unlike dextran which is a neutral polysaccharide. The negatively charged surface of the particles at neutral pH will create repulsive forces between them, decreasing the effect of particle agglomeration. The HDs remained unchanged after 48 h in the coated MNP when suspensions were prepared in MQ water, except for D-MNP that doubled in size, whilst the bare MNPs showed a ten-fold increase (Fig. 1). This may be explained by the fact that although dextran is a neutral coating agent, it still stabilises the particles by decreasing the effect of the attractive magnetic forces between the core of the particles but to a lesser degree than the negatively charged carboxylic groups of GA and CM-dextran. Dextran stabilises the particles via steric phenomena. When the MNPs were suspended in natural seawater they tended to aggregate into particles with HDs 2–3 times bigger than those in MQ and continued to grow over time except for GA-coated MNPs which maintained the HD. However, when the uncoated MNPs are suspended in natural seawater the “ageing-associated” aggregation is prevented, unlike in MQ, i.e. the average HD of aggregates did not increase after 48 h, which is probably explained by the suppression of the electrical double layer and thus smaller HD in a higher conductivity media rather than a real size difference when the particles are uncoated. The effect of EPS was somewhat more ambiguous in that when it is freshly mixed with NP suspensions it tended to diminish or remain similar HDs of the coated particles in suspensions (e.g. 710 ± 35 vs 780 ± 16 nm in the case of CMD-MNP) unlike the bare ones. In time however, both D-MNP and CMD-MNPs have clustered exhibiting increased HDs by 48 h, contrasting the bare MNPs, which presented a slight decrease in their average HD compared with those in seawater without EPS, after 48 h in suspension. GA-MNPs remained unaltered in time indicating little electrostatic interaction with EPS and good colloidal stability regardless of ionic strength of carrying solution. Polydispersity Index (PDI) corroborate some of the above results providing supportive evidence for the following: stability of the GA-MNP suspensions is maintained in all media and also in time; PDIs of the other two coated MNPs (dextran- and CM-dextran coated) when freshly suspended indicate fairly even cluster formation, which become more variable in size after 48 h incubation (Fig. 2); as the uncoated MNPs showed the highest PDIs both in MQ and in seawater

Nanoparticle type and carrying solution Fig. 2. Polydispersity index (PDI) measured by DLS of magnetic nanoparticle suspensions (MNPs) stabilised with different capping agents including gum Arabic (GA-MNP), dextran (D-MNP), CM-dextran (CMD-MNP) and uncoated (MNP) and suspended in different carrying solutions, e.g. ultrapure water (MQ), natural seawater (SW) with and without exopolymeric substances (EPS), immediately upon dispersion and 48 h after. Vertical bars represent average PDI ± SEM, N = 5.

775

Zeta potential (mV)

E. Kadar et al. / Science of the Total Environment 487 (2014) 771–777

Nanoparticle type and carrying solution Fig. 3. Zeta potential (mV) measured by DLS of magnetic nanoparticles stabilised with different capping agents including gum Arabic (GA-MNP), dextran (D-MNP), CM-dextran (CMD-MNP) and uncoated (MNP) measured by DLS analysis in ultrapure water (MQ), in natural seawater (SW), with and without exopolymeric substances (EPS); vertical bars represent average zeta potentials ± SEM, N = 5.

the differences were not statistically significant at 95% confidence level (Fig. 5). Our dissolution ranges are somewhat higher than the 0.16% Fe release reported in the literature (Hoskins et al., 2012) for MNPs stabilised with common polycationic capping agents used in drug delivery or other clinically approved iron oxide nanoparticles (Arbab et al., 2005). MNPs appeared to degrade slightly more in seawater than in MQ (Fig. 5) after 48 h, which effect was statistically significant for both the bare and the dextran coated MNPs. Dissolution rates in general are controlled by the diffusion coefficient of the solute molecule, particle surface area, and diffusive sublayer thickness, when solution volume and other environmental conditions such as pH and temperature are fixed. High dissolution rates could therefore be expected for the MNPs with low HD as a result of their extremely high specific surface area, which was the case of our MNPs in suspensions in MQ water. Even when they are aggregated into larger particles in seawater, as indicated by the doubled HDs of aged MNPs as compared with analogues in MQ water (Fig. 5), dissolution of free Fe from aggregates is sustained. It has been previously shown with other metal oxide NPs (Miao et al., 2010) that, despite rapid aggregation into micrometre-size particles, the dissolution rate of free ions from aggregates remains at a similar level with that of primary NPs until they become more compact with

time and thus their specific surface area is reduced concurrently with the dissolution rate. Natural chelating molecules present in seawater, i.e. dissolved organic material (DOM), particularly the coloured fraction (CDOM), may also have contributed to iron dissolution especially in the case of bare MNPs. The origin and photochemistry of DOM and the subsequent generation of reactive species play an important, but not well quantified role (Fan, 2008) in determining the stability and dissolution of natural as well as engineered NPs in the aquatic environment. The mechanisms involved include environmentally important processes like the photolysis, reduction and thermal reduction of Fe(III) complexes in all of which DOM can act as electron donor (Meunier et al., 2005). The photo reactive formation of H2O2 via the superoxide intermediate capable of reducing Fe(III), thus aiding dissolution, was also suggested by other laboratory experiments focussed on the redox cycling of Fe in the presence of both artificial and natural polysaccharide extracts from diatoms (Steigennberger et al., 2010). The ecological significance and scale EPS–nanoparticle interactions in aquatic environments is derived from the pervasive presence, i.e. about ~40–60% of the planktonic photosynthetic product is released as EPS, and its key role in biofilm formation and colloid and trace element scavenging (Fogg, 1983) and in the global carbon cycling.

Fig. 4. Transmission electron micrographs of A) bare and B) GA-coated MNPs. Note the lack of visible change between coated vs uncoated MNPs.

776

E. Kadar et al. / Science of the Total Environment 487 (2014) 771–777

References b b b ab ab a a

aaa

aaa

a

ab a

a a

ab

a

a

a a

a

Fig. 5. Dissolution of Fe from magnetic nanoparticles, representing the % of the b3 kDa filtered fraction of the total Fe content in MNPs stabilised with different capping agents including gum Arabic (GA-MNP), dextran (D-MNP), CM-dextran (CMD-MNP) and uncoated (MNP) in ultrapure water (MQ), in natural seawater (SW), with and without exopolymeric substances (EPS); vertical bars are average Fe concentrations ± SEM, N = 3. Similar letters on error bars mean no significant difference detected at the P ≤ 0.05 level according to the one-way ANOVA and Tukey's comparison.

Distinct dissolution of MNPs with different coatings in SW is related with the type of interaction that exists between the coating agent and the particles. Dextran for instance, will be simply adsorbed to the surface of the particles mainly through electrostatic interactions. On the other hand, both in CM-dextran and GA there is probably coordination between carboxylic groups of the polysaccharides and the iron centres of the particles (Han et al., 2010). It is likely that as a result of the polyanionic quality of the microalgal EPS conveyed by its functional groups that remain to be identified, the EPS–NP co-aggregation in seawater resulted in more compact aggregates than those in the absence of EPS, resulting in an insignificant degradation of MNPs over time as shown in Fig. 5. To conclude, coating with organic polymers such as dextran, Gum Arabic and CM-dextran does improve colloidal stability of MNPs especially in MQ water and also holds barrier to extensive dissolution over 48 h, which remained within the range of 3–9% for all four types of MNPs regardless of suspension media. Based on this evidence, these magnetic nanomaterials with improved stability can be applied sustainably for magnetic harvesting of microalgae both in cultures and from the environment (Lim et al., 2012; Toh et al., 2012; Xu et al., 2011). Given the rapid time scales reported from magnetic removal experiments (i.e. 90% cell removal within 5 min) the dissolution reported here is negligible point toward safe biological applications.

Acknowledgements The Natural Environment Research Council (NERC) funded facility FENAC at the University of Birmingham is acknowledged for their help with characterisation of MNPs (FENAC/2012/02 grant to EK). The NERC-funded Western Channel Observatory is acknowledged for providing physicochemistry data on the L4 seawater samples. Dr Mike Allen is acknowledged for providing the microalga culture and infrastructure for the EPS extraction. Authors also thank the financial support from Fundação para a Ciência e a Tecnologia through grant no. PEst-C/ EQB/LA0006/2013 and contract nos. PTDC/EBB-BIO/102163/2008, PTDC/EBB-BIO/098961/2008, PTDC/EBB-BIO/118317/2010 and SFRH/ BD/64427/2009 for I.L.B.

Arbab AS, Wilson LB, Ashari P, Jordan EK, Lewis BK, Frank JA. A model of lysosomal metabolism of dextran coated super paramagnetic iron oxide (SIPO) nanoparticles: implications for cellular magnetic resonance imaging. NMR Biomed 2005:383–9. Batalha IL, Hussain A, Roque ACA. Gum Arabic coated magnetic nanoparticles with affinity ligands specific for antibodies, oct. J Mol Recognit 2010;23:462–71. Chen CS, Anaya JM, Zhang S, Spurgin J, Chuang CY, Xu C, et al. Effects of engineered nanoparticles on the assembly of exopolymeric substances from phytoplankton. PLoS One 2011;6:1–7. Cunliffe M, Salter M, Mann PJ, Whiteley AS, Upstill-Goddard RC, Murrell JC. Dissolved organic carbon and bacterial populations in the gelatinous surface microlayer of a Norwegian fjord mesocosm. FEMS Microbiol Lett 2009;299:248–54. Dahl JA, Maddux LS, Hutchinson JE. Toward greener nanosynthesis. Chem Rev 2007;107: 2228–68. Debye P, Hückel E. Zur Theorie der Elektrolyte. I. Gefrierpunktserniedrigung und verwandte Erscheinungen (The theory of electrolytes. I. Lowering of freezing point and related phenomena). Phys Z 1923;24:185–206. Decho AW, Norman RS, Visscher PT. Quorum sensing in natural environments: emerging views from microbial mats. Trends Microbiol 2010;18:73–80. Dias AMGC, Hussain A, Marcos AS, Roque ACA. A biotechnological perspective on the application of iron oxide magnetic colloids modified with polysaccharides. Biotechnol Adv 2011;29:142–55. Dror Y, Cohen Y, Yerushalmi-Rozen R. Structure of gum Arabic in aqueous solution. J Polym Sci B Polym Phys 2006;44:3265–71. Fan SM. Photochemical and biochemical controls on reactive oxygen and iron speciation in the pelagic surface ocean. Mar Chem 2008;109:152–64. Fogg GE. The ecological significance of extracellular products of phytoplankton photosynthesis. Bot Mar 1983;26:3–14. Gupta AK, Gupta M. Synthesis and surface engineering of iron oxide nanoparticles for biomedical applications. Biomaterials 2005;26:3995–4021. Han GH, Ouyang Y, Long X-Y, Zhou Y, Li M, Liu Y-N, et al. (Carboxymethyl–dextran)modified magnetic nanoparticles conjugated to octreotide for MRI applications. Eur J Inorg Chem 2010;34:5455–546. Haun JB, Yoon T-Y, Lee H, Weissleder R. Molecular detection of biomarker cells using magnetic nanoparticles and diagnostic magnetic resonance. Biomed Nanotechnol Meth Prot 2011;726:33–49. Ho D, Sun X, Sun S. Monodisperse magnetic nanoparticles for theranostic applications. Acc Chem Res 2011;44(10):875–82. Hoskins C, Cuschieri A, Wang L. The cytotoxicity of polycationic iron oxide nanoparticles: common endpoint assays and alternative approaches for improved understanding of cellular response mechanism. J Nanobiotechnol 2012;10:15–26. Huang H, Cho HH, Jacangelo JG, Schwab KJ. Mechanisms of membrane fouling control by integrated magnetic ion exchange and coagulation. Environ Sci Technol 2012;46: 10711–7. Hyeon T. Chemical synthesis of magnetic nanoparticles. Chem Commun 2003:927–34. Kadar E, Tarran GA, Jha AN, Al-Subiai SN. Stabilisation of engineered zero-valent nanoiron with Na-acrylic copolymer enhances spermiotoxicity. Environ Sci Technol 2011;45: 3245–325. Kadar E, Rooks Paul, Lakey Cara, White Daniel A. The effect of engineered iron nanoparticles on growth and metabolic status of marine microalgae cultures. Sci Total Environ 2012;439:8–17. Kadar E, Al-Subiai SN, Dyson O, Handy RD. Are reproduction impairments of free spawning marine invertebrates exposed to zero-valent nano iron associated with dissolution of nanoparticles? Nanotoxicology 2013;7(2):135–43. Kang JH, Park J-K. Technical paper on microfluidic devices — cell separation technology. APBN 2005;9:1135–46. Laurent S, Forge D, Port M, Roch A, Robic C, Elst LV, et al. Magnetic iron oxide nanoparticles: synthesis, stabilization, vectorization, physicochemical characterizations, and biological applications. Chem Rev 2008;108:2064–110. Li YG, Gao H-S, Li W-L, Xing J-M, Liu H-Z. In situ magnetic separation and immobilization of dibenzothiophene-desulfurizing bacteria. Bioresour Technol 2009;100:5092–6. Li J, Zhou Y, Li M, Xia N, Huang Q, Do H, et al. Carboxymethylated dextran-coated magnetic iron oxide nanoparticles for regenerable bioseparation. J Nanosci Nanotechnol 2011;11(11):10187–92. Lim JK, Majetich SA, Tilton RD. Stabilization of superparamagnetic iron oxide core-gold shell nanoparticles in high ionic strength media. Langmuir 2009;25(23):13384–93. Lim JK, Chieh CJ, Jalak SA, Toh LA. Rapid magnetophoretic separation of microalgae. Small 2012;8:1683–92. Liu G, Hong RY, Guo L, Li YG, Li HZ. Preparation, characterization and MRI application of carboxymethyl dextran coated magnetic nanoparticles. Appl Surface Sci 2011;257: 6711–7. Meunier L, Laubscher H, Hug JS, Sulzberger B. Effects of size and origin of natural dissolved organic matter compounds on the redox cycling of iron in sunlit surface waters. Aquat Sci 2005;67:292–307. Miao A-J, Zhang X-Y, Luo Z, Chen C-S, Chin W-C, Santchi PH, et al. Zinc oxide-engineered nanoparticles: dissolution and toxicity to marine phytoplankton. Environ Toxicol Chem 2010;29:2814–22. Roque ACA, Bicho A, Batalha IL, Cardoso AS, Hussain A. Biocompatible and bioactive gum Arabic coated iron oxide magnetic nanoparticles. J Biotechnol 2009;144(4):313–20. Rosicka D, Sembera J. Influence of structure of iron nanoparticles in aggregation their magnetic properties. Nanoscale Res Lett 2011;6:527–36. Steigennberger S, Statham PJ, Volker C, Passow U. The role of polysaccharides and diatom exudates in the redox cycling of Fe and the photoproduction of hydrogen peroxide in coastal waters. Biogeosciences 2010;7:109–19.

E. Kadar et al. / Science of the Total Environment 487 (2014) 771–777 Toh PY, Yeap SP, Kong LP, Ng BW, Chan DJC, Ahmad AL, et al. Magnetic removal of microalgae from fishpond water: feasibility of high gradient and low gradient magnetic separation. Chem Eng J 2012;211–212:22–30. Wu W, He Q, Jiang C. Magnetic iron oxide nanoparticles: synthesis and surface functionalization strategies. Nanoscale Res Lett 2008;3:397–415.

777

Xu L, Wang F, Zheng S, Liu CZ. A simple and rapid harvesting method for microalgae by in situ magnetic separation. Bioresour Technol 2011;201:10047–51. Zhang S, Santschi PH. Application of cross-flow ultrafiltration for isolating exopolymeric substances from a marine diatom (Amphora sp.). Limnol Oceanogr Methods 2009;7:419–29.