Ti

Na

b

a

ARRA

KFBCTN

1

dnuaa(2m(b

mI

(

0d

Aquatic Toxicology 118– 119 (2012) 1– 8

Contents lists available at SciVerse ScienceDirect

Aquatic Toxicology

jou rn al h om epa ge: www.elsev ier .com/ locate /aquatox

he potential of TiO2 nanoparticles as carriers for cadmium uptaken Lumbriculus variegatus and Daphnia magna

anna B. Hartmanna,∗, Samuel Legrosb, Frank Von der Kammerb, Thilo Hofmannb, Anders Bauna

Department of Environmental Engineering, Technical University of Denmark, Building 113, Kgs. Lyngby, DenmarkDepartment of Environmental Geosciences, University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria

r t i c l e i n f o

rticle history:eceived 1 September 2011eceived in revised form 6 March 2012ccepted 8 March 2012

eywords:acilitated transportioaccumulationadmiumitanium dioxideanomaterials

a b s t r a c t

The use of engineered nanoparticles (e.g. in industrial applications and consumer products) isincreasing. Consequently, these particles will be released into the aquatic environment. Through aggrega-tion/agglomeration and sedimentation, sediments are expected ultimately to be sinks for nanoparticles.Both in the water phase and in the sediments engineered nanoparticles will mix and interact with otherenvironmental pollutants, including metals. In this study the toxicity of cadmium to two freshwaterorganisms, water column crustacean Daphnia magna and sediment oligochaete Lumbriculus variegatus,was investigated both in the absence and presence of titanium dioxide (TiO2) nanoparticles (P25 EvonicDegussa, d: 30 nm). The uptake of cadmium in sub-lethal concentrations was also studied in the absenceand presence of 2 mg/L TiO2 nanoparticles. Formation of larger nanoparticles aggregates/agglomerateswas observed and sizes varied depending on media composition (358 ± 13 nm in US EPA moderately hardsynthetic freshwater and 1218 ± 7 nm in Elendt M7). TiO2 nanoparticles are potential carriers for cad-mium and it was found that 25% and 6% of the total cadmium mass in the test system for L. variegatus andD. magna tests were associated to suspended TiO2 particles, respectively. �XRF (micro X-ray fluorescence)analysis confirmed the uptake of TiO2 in the gut of D. magna. For L. variegatus �XRF analysis indicatedattachment of TiO2 nanoparticles to the organism surface as well as a discrete distribution within the

organisms. Though exact localisation in this organism was more difficult to assess, the uptake seemsto be within the coelomic cavity. Results show that the overall body burden and toxicity of cadmiumto L. variegatus was unchanged by addition of TiO2 nanoparticles, showing that cadmium adsorption toTiO2 nanoparticles did not affect overall bioavailability. Despite facilitated uptake of cadmium by TiO2

nanoparticles in D. magna, resulting in increased total cadmium body burden, no change in toxicity was

observed.

. Introduction

Engineered nanoparticles (ENPs) are increasingly being pro-uced, used and disposed as a result of their applications in aumber of industrial manufacturing processes and consumer prod-cts. As a consequence, nanoparticles will be released into thequatic environment where their fate will depend on factors suchs presence of natural organic matter (NOM), ionic strength and pH

Keller et al., 2010; von der Kammer et al., 2010; Ottofuelling et al.,011; Battin et al., 2009). The complexity of natural freshwatersakes it difficult to predict the environmental fate of nanoparticles

Lin et al., 2010). If particles aggregate/agglomerate or are taken upy aquatic organism, sediments will be an expected sink for ENPs

∗ Corresponding author. Current address: Joint Research Centre – European Com-ission, Institute for Health and Consumer Protection, Via E. Fermi 2749, 21027

spra (VA), Italy. Tel.: +39 0332 78 3947.E-mail addresses: Nanna.Hartmann@ec.europa.eu, nannahartmann@gmail.com

N.B. Hartmann).

166-445X/$ – see front matter © 2012 Elsevier B.V. All rights reserved.oi:10.1016/j.aquatox.2012.03.008

© 2012 Elsevier B.V. All rights reserved.

through sedimentation of aggregates/agglomerates and decayingorganisms (Baun et al., 2008).

Manufactured nanosized titanium dioxide (TiO2) particles areexamples of a nanomaterial, which is already widely used in var-ious applications – many of which are based on their ability toefficiently absorb UV-light and their photocatalytic activity, whichhas been found to increase when particle size is decreased (Gaoand Zhang, 2001). Generally TiO2 is considered an inert mate-rial, but for nanosized particles the increased surface-to-volumeatom ratio enhances activity per mass and may alter interactionswith living organisms and toxicity. The novel properties of TiO2nanoparticles compared to their bulk sized counterparts may alsoincrease the possibilities for interactions with co-existing pollu-tants through sorption (Hartmann and Baun, 2010). For example,sorption of arsenic to TiO2 nanoparticles is greater than to natu-ral soil particles (Zhang et al., 2007) and the sorption of cadmium

to TiO2 nanoparticles is a fast process, reaching equilibrium withinapproximately 30–60 min (Hartmann et al., 2010). Furthermore thesmall size of TiO2 nanoparticles used in consumer products and

2 atic To

padtciia

cuvg(2nhffrmoAwudarb

imtiawItfaw

2

2

Hdu(HS∼4wibea(sa≤mt

N.B. Hartmann et al. / Aqu

rofessional applications may facilitate their uptake into organsnd cells. Aquatic organisms in general – and possibly filter- andeposit-feeding organisms in particular – are therefore importantest organisms in ecotoxicological tests with respect to nanoparti-les (Baun et al., 2008). This also applies to studies elucidating thenteractions between nanoparticles and other co-existing contam-nants with respect to interactions leading to changes in toxicitynd bioconcentration.

The freshwater sediment worm, Lumbriculus variegatus, is aommonly used test organism in ecotoxicological studies. Its nat-ral habitat is mainly shallow waters, where it feeds on decayingegetation and microorganisms. Studies have shown that L. varie-atus ingest particles with sizes of 40–60 �m, while larger particles100–300 �m) would exceed the size of its mouth (Mino et al.,006; Landrum et al., 2002). Due to the added complexity of testinganoparticle toxicity in a sediment matrix tests with L. variegatesave, however, been carried out as water-only tests. The filter-

eeding/grazing freshwater crustacean Daphnia magna is significantor the ecological food web as well as being a key organism foregulatory testing (Baun et al., 2008). By filtration of water D.agna catch particles (including mainly algae) in the size range

f 0.4–40 �m (Geller and Müller, 1981; Gophen and Geller, 1984).s a means to facilitate digestion, daphnids are known to ‘drink’ater (Gillis et al., 2005) whereby smaller particles can be takenp from the surrounding media (Rosenkranz et al., 2009). Not onlyo these two organisms represent different aquatic species, theyre also located in different compartments of the aquatic envi-onment (water column and sediment) and have different feedingehaviours.

The aim of this study was hence to investigate possible changesn toxicity and bioaccumulation of cadmium in L. variegatus and D.

agna in the absence and presence of TiO2 ENPs. The influence ofhe presence of TiO2 nanoparticles on cadmium uptake and toxic-ty to these two organisms was investigated quantitatively throughcute toxicity tests and bioaccumulation studies in combinationith qualitative investigations of intra-organism particle location.

t should be clarified that when we use the term “uptake” we refero increased body burden that may be related to a number of dif-erent exposure routes (diffusive uptake from water, particle–Cdssociation to gill surfaces, ingestion of particulate material fromater phase and grazing/feeding from sedimented material).

. Materials and methods

.1. Chemicals and chemical analysis

Analytical grade cadmium (‘STD Cadmium 1000 ppm’ in 2%NO3 (CAS no. 7440-43-9) was purchased from PerkinElmer. Threeifferent types of TiO2 nanoparticles (CAS no. 13463-67-7) weresed in this study, namely (i) P25 TiO2 NPs from Evonic DegussaNM-105, obtained from the OECD sponsorship programme), (ii)ombikat UV100 NPs and (iii) LW-S TiO2 NPs (both provided byachtleben Chemie GmbH). P25 has a primary particle diameter of30 nm (Jensen et al., 2004) and a specific surface area (SSA) of7 m2/g (BET multipoint) (Hartmann et al., 2010). UV100 is <10 nmith a BET SSA of 288 m2/g whereas and the larger sized LW-S TiO2

s ∼300 nm with a BET SSA of 11.5 m2/g [particle sizes as statedy the producer (Sachtleben, 2005, 2006), SSA from Hartmannt al. (2010)]. Further details on characteristics of these particlesre given in Hartmann et al. (2010) and von der Kammer et al.2010). Cadmium analysis was performed by atomic absorptionpectroscopy (AAS) using a PerkinElmer graphite furnace AAS AAn-

lyst 800. Prior to analysis the pH of the samples was adjusted to2 with concentrated HNO3. The deviations between nominal andeasured values for initial cadmium exposure concentrations in all

ests were <7.5%.

xicology 118– 119 (2012) 1– 8

2.2. Preparation of media and particle suspensions

Moderately hard synthetic freshwater and Elendt M7 mediawere prepared from MilliQ water according to US EPA (2002) andOECD (2004), respectively. TiO2 stock suspensions were preparedby suspending TiO2 particles in test media (M7 or moderately hardsynthetic freshwater) in a concentration of 250 mg/L followed by2 × 10 min bath sonication (Elgasonic, 50W). These suspensionswere kept at 5 ◦C in the dark between tests and sonicated again10 min prior to preparation of test suspensions.

2.3. Characterisation and behaviour of TiO2 nanoparticles in testmedium

Particle sizes and Zeta potentials were measured by dynamiclight scattering (DLS) using a Zetasizer NanoZS (model ZEN3600,Malvern Instruments, UK). All measurements were done directlyin the suspension medium (Elendt M7 or US EPA moderately hardsynthetic freshwater). The concentration of the TiO2 nanoparticlesin the DLS measurements was 2 mg/L, identical to the concentrationused in the toxicity and uptake tests. The z-average hydrodynamicdiameter and the polydispersity index were determined by thecumulant analysis method. Size measurements were done in trip-licate. Each measurement is an average of 11 runs of 10 s. Theelectrophoretic mobility was measured by Laser Doppler anemom-etry that measures the movement of charged particles in an electricfield. The zeta potential is then calculated using the Smoluchowskiapproximation. Zeta potential measurements were done in tripli-cate. Each measurement is an average of 30 runs of 11 s.

The settling behaviour of TiO2 in the two different media wasinvestigated by following the reduction in UV–vis absorbance (CaryBio50 UV-VIS spectrophotometer). Suspensions of 10 mg/L TiO2were prepared by dilution in the respective media (US EPA moder-ately hard synthetic freshwater or M7 medium). The absorbanceof the TiO2 suspensions in a quartz cuvette was measured over360 min at 6 min intervals at � = 338 nm, for which an absorp-tion peak was detected on the UV–vis absorption spectra [as alsoobserved by e.g. Wang et al. (2010)].

2.4. Cultivation of L. variegatus and D. magna

L. variegatus were cultivated in US EPA moderately hard syn-thetic freshwater (US EPA, 2002) in 5-L aquaria with brown papertowel clippings at 20 ± 2 ◦C with 16/8 h light/dark conditions. Theaquarium was cleaned and media was changed once a week. Organ-isms were fed 0.7 g fish food (TetraMin® Tropical Flakes) per week.To ensure that organisms were of the same developmental andphysiological state organisms were synchronised prior to testing.Worms were cut and 100–200 posterior segments were transferredto clean media with brown paper towel cuttings for morphallacticregeneration. After 7 days they were fed with 0.35 g fish food and12–13 days after synchronisation the worms were used in tests.The culture of D. magna originates from Birkedammen (Lyngby,Denmark), and has been cultured in the laboratory since 1978 inaccordance with OECD (2004). Neonates (<24 h) were used in tox-icity tests and for uptake studies animals of 4–5 days were used.

2.5. Acute toxicity tests

Tests to determine acute toxicity of cadmium to L. variegatuswere conducted at 20 ± 2 ◦C in 16/8 h light/dark cycle both withand without the presence of TiO2 nanoparticles. The tests involved

five test concentrations and one control group each including tenreplicates. All exposure solutions were prepared from US EPA mod-erately hard synthetic freshwater (US EPA, 2002). Each replicateconsisted of one synchronised worm in 20 mL test solution in a

atic To

2bwrowq

a(stfiigTe1

vctwTi(c9

2n

wmaei1(ec

e22lto

8a0dtfifa1dtscomd

N.B. Hartmann et al. / Aqu

5 mL glass test vials. The duration of the test was 96 h and the num-er of immobile organisms (dead and/or not moving after agitation)as counted after 24, 48, 72 and 96 h exposure. A preliminary

ange-finding test for cadmium was carried out in concentrationsf 100–1600 �g/L (pH 7.9 ± 0.1). Results showed that the EC50 valueas within the range 200–400 �g/L (data not shown) and subse-

uent tests were carried out in concentrations of 100–506 �g/L.The toxicity of cadmium to D. magna was tested in M7 media

ccording to standardised OECD 202/ISO 6341 immobilisation testsOECD, 2004; ISO, 1998), both with and without addition of TiO2uspensions. The test involved six test concentrations and one con-rol group each with four replicates. Each replicate consisted ofve D. magna neonates (<24 h) in a 100 mL glass beaker contain-

ng 25 mL test solution. The glass beakers were covered with alass lid and tests were incubated for 48 h at 20 ± 2 ◦C in the dark.he number of immobile organisms was counted after 24 and 48 hxposure and the concentrations of cadmium were in the range of0–640 �g/L (pH 7.8 ± 0.15).

The acute toxicity of TiO2 P25 to both D. magna and L.ariegatus was tested separately, under conditions identical toadmium tests, and was found to be non-toxic in a concentra-ion of 2 mg/L after 48 h (data not shown). All acute toxicity testsere repeated twice both with and without the addition of 2 mg/L

iO2 P25. The data was analysed by probit analysis with max-mum likelihood estimation using the Probit Analysis SoftwareVersion 2.3, Swedish Environmental Protection Agency) yieldingoncentration–response curves and EC-values with corresponding5% confidence limits.

.6. Uptake of cadmium with and without the presence of TiO2anoparticles

Uptake studies with L. variegatus were carried out asater-only tests, without the presence of artificial sedi-ent. This was done in order to simplify the test system

nd eliminate possible disturbing factors from the pres-nce of sediment particles that would make it difficult tonterpret the results. Synchronised worms were exposed to00 �g/L cadmium in 20 mL test solution in 25 mL glass test vialspH 7.7 ± 0.15). There were four replicates for each sampling time,ach containing one organism. The study was repeated twice toonfirm the observations.

To examine the uptake in D. magna, 3–4-day-old juveniles werexposed to 100 �g/L cadmium with and without the addition of

mg/L TiO2 nanoparticles (pH 7.8 ± 0.1). Tests were conducted in5 mL test solution in 100 mL glass beakers covered with a glass

ids. There were three replicates for each sampling time, each con-aining five animals. The study was repeated twice to confirm thebservations.

Uptake of cadmium in L. variegatus was measured at times 0, 1, 4,, 24 and 48 h. Clearance was measured at times 1, 4, 8, 24 and 48 hfter transfer to clean media. Uptake in D. magna was examined at, 1, 4, 8 and 24 h to compare the cadmium distribution in the twoifferent test systems and test species. Clearance was measured atimes 1, 4, 8, and 24 h after transfer to clean media. At the speci-ed times organisms were removed and rinsed in 10 �M Na2EDTA

or 5 min in order to remove metals attached to the surface. There-fter organisms were transferred to glass vials and preserved in

mL 3% HNO3 prior to further analysis. The organisms were acidigested by adding 4 mL of 32.5% HNO3 to the samples, subjectinghem to high pressure at 121 ◦C for 30 min in autoclave, after whichamples were diluted to a total volume of 10 mL. Organism metal

oncentrations were calculated for measured average dry weightsf 1 mg/organism for L. variegatus and 0.035 mg/organism for D.agna. The average dry weight of the organisms was measured by

rying 20 L. variegatus or 42 D. magna at 105 ◦C overnight.

xicology 118– 119 (2012) 1– 8 3

During the tests samples were taken to determine cad-mium mass distribution including sorption of cadmium to TiO2nanoparticles. After removal of the organisms, this was done bycentrifugation of the test suspensions (25 mL) for 30 min at 1370 g(Hettich Rotanta 460 R, Tuttlingen, Germany). Water samples(10 mL) were taken from supernatant and pH adjusted to <2 with20 �L HNO3. Additional water was removed resulting in a pel-let consisting of TiO2 nanoparticles (and/or possible debris fromorganisms as well as precipitated media constituents) with approx-imately 1 mL residual media, which was quantified by weighingthe sample. 5 mL H2O and 100 �L HNO3 was then added prior toacid digestion in autoclave and subsequent dilution of the sample.Finally potential adsorption of cadmium to the test vial was deter-mined by rinsing the glass in 5 mL 10 �M Na2EDTA and addition of100 �L HNO3. This was done after 24 h of uptake for both D. magnaand L. variegatus and additionally after 48 h for the L. variegatus test.

To compare the influence of different TiO2 particle types on cad-mium uptake in L. variegatus a study was done with P25 and UV100TiO2 nanoparticles as well as the larger sized LW-S TiO2 particle.The exposure was carried out in three replicates, each beaker con-taining one organism, and three control beakers were included.After 2 days the organisms were rinsed in 10 �M Na2EDTA andprepared for chemical analysis as described above.

To compare the uptake of cadmium in L. variegatus and D.magna with and without addition of TiO2 nanoparticles, a two-way ANOVA with replication was conducted to compare cadmiumuptake across periods and treatments (p = 0.05). The null hypoth-esis, being that the organism cadmium concentration was notstatistically significantly different with and without the presence ofTiO2 nanoparticles, was tested. A one-way ANOVA with replicationwas performed to determine if any of the three TiO2 particle types(P25. UV100 or LW-S) affected cadmium uptake, and Bartlett’s testwas used to test for equal variance.

2.7. Chemical analysis by �XRF

Elemental intra-organism distribution was examined by �XRF(micro X-ray fluorescence) after 24 h of exposure under identicalconditions as in uptake studies both in the absence and presence ofTiO2. The preparation procedure for �XRF analysis included a dehy-dration process of the specimens in ethanol (70% 1 × 10 min, 80%1 × 10 min, 90% 1 × 10 min, 95% 2 × 10 min and 100% 2 × 10 min).This was followed by placing the organisms in HMDS (hexamethyl-disilazane) for 30 min and drying overnight. �XRF measurementswere carried out on a HORIBA XGT-7000 XRF-microscope equippedwith an X-ray guide tube producing a focused and high-intensitybeam with a 10 �m spot size. The X-ray beam is generated using aRhodium X-ray tube at an accelerating voltage of 30 kV with a cur-rent of 1 mA. X-ray emission from the irradiated sample is detectedwith an energy-dispersive X-ray (EDX) spectrometer equippedwith a liquid-nitrogen-cooled high-purity Si detector. Elementsdetected by �XRF analysis ranged from Mg(12) to U(92). The detec-tor resolution was 145 eV at the Mn K� emission line. �XRF analysisdetection limits can reach 100 mg kg−1, depending of the nature ofthe matrix and the atomic weight of the analysed element. The sam-ples were analysed in mapping mode. The characteristics of eachmap obtained are summarised in Table 1. For each pixel in the maps,a complete XRF spectrum is recorded.

3. Results and discussion

3.1. Characteristics of nanoparticles

Characteristics of the TiO2 nanoparticles in the two test mediatested are presented in Table 2. In the US EPA synthetic fresh

4 N.B. Hartmann et al. / Aquatic Toxicology 118– 119 (2012) 1– 8

Table 1Parameter settings for �XRF elemental mapping.

D. magna L. variegatus head L. variegatus body

Size mapping area (mm) 1.024 × 1.024 1.024 × 1.024 1.280 × 0.800Resolution (nb pixel) 128 128 128Pixel width (�m) 8 8 10Survey time per frame (s) 256 256 400Accumulation 20 20 10

Table 2Characterisation of the tested TiO2 P25 nanoparticles in different media.

Moderately hard synthetic freshwatera M7b

z-average hydrodynamic diameter [nm] ± 1 SD of triplicates (CTiO2: 2 mg/L) 358 ± 13 1218 ± 7

Polydispersity index 0.35 0.73Zeta potential [mV] ± 1 SD of triplicates (C : 2 mg/L) −16.3 ± 0.5 −5.3 ± 0.4

wdtzsgazcitrTiif3

3

mv[D5wL1tp7cFh(Htcameefpmi

stants. Faster initial depuration has been observed for organismsexposed through sediments (Dawson et al., 2003). It could there-fore be hypothesised that co-exposure with a particulate phasewould facilitate a faster depuration. This was however not seen

TiO2

a US EPA (2002).b OECD (2004).

ater the z-average of the TiO2 nanoparticles is around 358 nm,emonstrating aggregation/agglomeration of the 30 nm TiO2 par-icles (nominal size). When suspended in Elendt M7 media the-average of the TiO2 particles is much larger (1218 nm), againignifying that the nanoparticles have a clear tendency to aggre-ate/agglomerate in synthetic freshwater media. The difference inggregation/agglomeration size is in agreement with differences ineta potential. In Elendt M7 medium zeta potential was found to belose to zero (−5.3 mV) indicating an unstable suspension, whereasn US EPA moderately hard synthetic freshwater a lower zeta poten-ial was found (−16.3 mV), showing a higher colloidal stabilityesulting in smaller aggregates/agglomerates. The instability of theiO2 suspensions was further confirmed by monitoring changesn UV–vis absorbance (� = 338 nm) over a period of 360 min. Heret was found that reduction in absorbance over time was similaror both media types resulting in 80% of initial absorbance after60 min.

.2. Cadmium acute toxicity to L. variegatus and D. magna

The acute toxicity of ‘cadmium only’ to L. variegatus and D.agna was found to be within the same range (Table 3). EC50

alues for ‘cadmium only’ for L. variegatus varied from 227 �g/L110; 2297]95% to 284 �g/L [234; 377]95% whereas EC50 values for. magna ranged from 299 �g/L [261; 352]95% to 348 �g/L [274;13]95%. Hence no significant difference in cadmium EC50 valuesas found between test repetitions. The toxicity of cadmium to

. variegatus found in this study is higher than the LC50 value of.36 mg/L as determined by Penttinen et al. (2008) under rela-ively comparable conditions (water-only exposure). However, theH was somewhat lower (pH 7) than in the present study (pH.9 ± 0.1) and media composition was different, which influencesadmium speciation and is likely to cause differences in toxicity.or D. magna the observed EC50 values for toxicity are somewhatigher than other values reported in the literature [26.4 �g Cd/LSuedel et al., 1997) to 129.4 �g Cd/L (Stuhlbacher et al., 1993)].owever considerable variations in cadmium bioavailability and

oxicity to D. magna and L. variegatus have also been observed underhanging experimental conditions due to factors such as temper-ture, pH, calcium concentrations and presence of natural organicatter (Clifford and McGeer, 2010; Xie et al., 2008; Guilhermino

t al., 1997; Stuhlbacher et al., 1993). In the study by Guilherminot al. (1997) EC50,48 h values for cadmium toxicity to D. magna were

ound to be higher (approximately one order of magnitude) for testserformed using M7 media compared to tests using more simpleedia without chelators. The M7 EC50 value for cadmium reported

n the study (331.6 �g/L) is similar to EC50 values found in this

study (299 �g/L and 348 �g/L, respectively). Hence the use of dif-ferent media types, possibly combined with other variations in testconditions, may explain these discrepancies. Also the age of theorganisms is likely to influence toxicity. In tests where 2 mg/L TiO2was added to the test suspensions, EC50 values for L. variegatusand D. magna were not significantly different (when 95% confi-dence intervals were compared) from the values obtained in test ofcadmium alone (Table 3).

3.3. Bioaccumulation of cadmium in L. variegatus

3.3.1. Uptake and elimination kinetics of cadmium in L.variegatus

As shown in Fig. 1 neither uptake nor elimination of cadmiumin L. variegatus was affected by the presence of TiO2 nanoparticles.This was further confirmed by statistical data analysis (two-wayANOVA) showing no statistically significant differences (p > 0.05)between samples with absence and presence of TiO2 nanoparticlesfor all measuring times during both uptake and elimination. A lin-ear uptake is seen, which corresponds with what has previouslybeen observed for cadmium uptake in L. variegatus exposed to cad-mium through water (Xie et al., 2008). The slow elimination andhigh retention of cadmium in L. variegatus also corresponds wellto observations by Xie et al. (2008). Due to too few data points itwas not possible to determine uptake and elimination rate con-

Fig. 1. Uptake and clearance of cadmium in L. variegatus in the absence and presenceof 2 mg/L TiO2 P25 nanoparticles. The organisms were exposed to 100 �g/L cadmiumfor 48 h followed by 48 h depuration in clean medium.

N.B. Hartmann et al. / Aquatic Toxicology 118– 119 (2012) 1– 8 5

F xposure) and (B): D. magna (after 24 h of exposure) to 100 �g/L cadmium in the absenceo ystem was 93–111% for test with L. variegatus and 101–105% for test with D. magna.

ipctroohBapa4itvttaaic

imsnca(eacwt

Fig. 3. Uptake of cadmium in L. variegatus in the absence and presence of three dif-

TE

n

ig. 2. Relative distribution of cadmium in (A): L. variegatus (after 24 and 48 h of er presence of 2 mg/L TiO2 P25 nanoparticles. The recovery of cadmium in the test s

n this study when L. variegatus was exposed to cadmium in theresence of TiO2 nanoparticles. During uptake the total cadmiumoncentration in the water phase decreased by ∼10% and ∼22% inhe absence and presence of TiO2 nanoparticles, respectively. Thiseduction is indicating an effect of TiO2 on the co-transportationf cadmium, enhancing the translocation either into the organismsr to the sides or bottom of the test vessels. This would require,owever, an association of cadmium to the TiO2 nanoparticles.y investigating the distribution of cadmium in the test system,

high degree of adsorption of cadmium to the TiO2 nanoparticlehase was in fact observed. This did, however, not lead to increaseddsorption of cadmium to the test beaker (Fig. 2A). After 24 and8 h of test duration approximately 25% of the total cadmium

n the test system was detected in the TiO2 phase after centrifuga-ion of the test suspension. Nonetheless, uptake of cadmium into L.ariegatus was seemingly not affected by this. Organism concen-ration increases from ∼5% to ∼9% of the total cadmium in theest system from 24 to 48 h. It would be logical to assume thatdsorption of cadmium would result in decreased bioavailability,nd thereby toxicity. Conversely, the unchanged uptake and tox-city, despite adsorption, indicates that the adsorbed fraction ofadmium is still bioavailable to the organisms.

It was hypothesised that other types of TiO2 particles mightnfluence uptake of cadmium differently. The influence on cad-

ium uptake in L. variegatus of two additional particle types,maller or larger in particle size, was examined. Results showedo significant difference (one-way ANOVA, p > 0.05) in cadmiumoncentration after 2 days between exposures to cadmium alonend in the presence of 2 mg/L of the three types of TiO2 particlesFig. 3). It should be noted that the standard deviation is small-st for the animals exposed to cadmium alone and highest for the

nimals exposed to the TiO2 particles with smallest primary parti-le size, but that variances between exposures (cadmium alone orith different TiO2 types) were not significantly different (Bartlett’s

est, p > 0.05). In a repeated experiment, including more replicates

able 3ffect concentrations for acute toxicity of cadmium to L. variegatus and D. magna in the ab

L. variegatus

EC10 (96 h) [�g/L] EC50 (96 h)

Cadmium 125 [0; 206]73 [0; 123]

227 [110; 2284 [234; 3

Cadmium + 2 mg/L TiO2 165 [92; 204]94 [1.5; 145]

257 [220; 3235 [189; 2

.q.: not quantifiable.a Outside of the tested range.

ferent types of TiO2 particles in a concentration of 2 mg/L after exposure to 100 �g/Lcadmium for 2 days. The average concentrations for the three replicates are shownwith ±SD.

per treatment, unequal variances were, however, found (Bartlett’stest (�2(3) = 17, p < 0.05; data not shown). This could indicate thatthe presence of TiO2 particles influences the uptake process byincreasing the complexity of factors governing bioavailability eventhough this is not evident from the resulting final average cadmiumconcentrations in the organisms.

3.3.2. Uptake of cadmium in D. magnaIn order to compare the finding for the deposit-feeding

oligochaete L. variegatus to an organism with a different feedingmechanism, tests with the filter-feeding D. magna as test specieswas carried out. One clear difference was revealed from this test,

namely that uptake of cadmium into D. magna is strongly influ-enced by the presence of TiO2 nanoparticles (Figs. 2B and 4). Inthe presence of TiO2 nanoparticles the organism Cd concentrationis in average six times higher compared to organisms exposed to

sence and presence of TiO2 nanoparticles.

D. magna

[�g/L] EC10 (48 h) [�g/L] EC50 (48 h) [�g/L]

297a]77]

196 [112; 237]127 [57; 183]

299 [261; 352]348 [274; 513]

10]97]

213.2 [0; 337]65 [n.q.; 224]

391 [266; 678a]258 [133;n.q.]

6 N.B. Hartmann et al. / Aquatic To

Fig. 4. Uptake and clearance of cadmium in D. magna in the absence and presence of2f

ctWt(ctwNcicsiwAtc(aemlsai(taiinsr

3

�tlldo(t

t

mg/L TiO2 P25 nanoparticles. The organisms were exposed to 100 �g/L cadmiumor 24 h followed by 24 h depuration in clean medium.

admium alone. When exposed to cadmium alone only 0.6% of theotal cadmium in the test system was localised in the organism.

ith the presence of TiO2 nanoparticles, however, this propor-ion increased to 3.3% (Fig. 2B), and hence was significantly higherF(1, 59) = 170.6, p < 0.05) than when the animals were exposed toadmium alone. Again adsorption of cadmium onto TiO2 nanopar-icles was seen. Hence, 6% of the total cadmium in the test systemas associated with the TiO2 nanoparticles in the water phase.on-selective filtration and uptake of particles by D. magna, withadmium adsorbed to their surface, into the gut through filtrations therefore likely to explain the high organismal cadmium con-entrations in the presence of TiO2 nanoparticles. However, as it iseen by comparing the EC50 values and 95% confidence intervalsn Table 3, no significant changes in acute toxicity were associated

ith this increased total organismal concentration (body-burden).n option to investigate further the influence of the uptake and con-

ribution to toxicity from particle bound cadmium versus dissolvedadmium would be to follow the recommendations by Gillis et al.2005) regarding gut purging of D. magna before whole-body metalnalysis. The study concluded that an 8 h purging period in the pres-nce of algae would allow the organisms to purge their gut frometal-contaminated sediments whereafter actual uptake (assimi-

ation) can be determined. The contribution of different exposureources to cadmium assimilation (after an 8 h purging period) haslso been investigated by Barata et al. (2002) demonstrating thatt is possible to separately assess different sources of Cd uptakewater, food, water + food). Results showed that, although Cd isaken up from water and food (algae) in an additive manner, thessimilation efficiency is higher from a dietary exposure. However,t is likely that algae-bound Cd and particle-bound Cd will differn uptake due to the inorganic nature and ingestability of the TiO2anoparticles. The role of different sources of Cd uptake in a testystem with nanoparticles, as well as gut purging, are areas thatequire further attention in future studies.

.3.3. Elemental intra-organism distributionAs total organismal concentrations are quite crude measures,

XRF analysis was undertaken to further elucidate the discrete dis-ribution of TiO2 nanoparticles within the organisms. Due to theow concentrations of cadmium the signal of this element was veryow. Though the presence of cadmium could be confirmed, imagesid not provide much additional information regarding the locationr changes resulting from the presence of TiO2 in the test system

Supporting Information). Figs. 5 and 6 show the distribution ofitanium in L. variegatus and D. magna, respectively.

Supplementary material related to this article can be found, inhe online version, at doi:10.1016/j.aquatox.2012.03.008.

xicology 118– 119 (2012) 1– 8

Fig. 5a and d shows the distribution of sulphur, respectively,in the head and the body of L. variegatus. The distribution of sul-phur is rather homogenous throughout the organism. Fig. 5b and eshows the distribution of titanium. The distribution of titanium ismuch more heterogeneous. For example, in the head end (Fig. 5b)some spot with higher titanium concentration can be identified.This discrete intra-organism localisation points to the fact that TiO2nanoparticles are also taken up by L. variegatus, nevertheless theirlocalisation is less clear. The division of the organism into segmentsis seen from the sulphur map in Fig. 5d. By comparing this with thetitanium distribution in Fig. 5b and e the distribution of titaniumdoes seem to follow a similar pattern to some extent. Howeverdetection of a clear internal distribution is possibly hampered byTiO2 attached to the surface of the organism. It is hypothesisedthat the discrete localisation and indications of segment-relatedpatterns indicate uptake into the coelomic cavity which containsthe gut and other organs.

Fig. 6a shows the distribution of calcium (Ca) in D. magna. Thedistribution of calcium corresponds to the exoskeleton. Fig. 6bshows the distribution of titanium (Ti) in D. magna. Titanium islocalised in totality in the guts of the specimen, showing that TiO2nanoparticles are taken up by that D. magna and accumulated inthe gut.

3.4. General discussion

In the light of the findings of this study it is clear that TiO2nanoparticles do have the potential to act as carriers for cadmiuminto aquatic organisms. During exposure to the two test organismsup to 25% of total cadmium in the test system were associatedwith suspended TiO2 nanoparticles. Especially for D. magna, wherefacilitated transport of cadmium into the gut by TiO2 nanoparti-cles resulted in increased total organismal concentration, this valueaccounts only for Cd adsorbed to TiO2 nanoparticles in the waterphase and does not reflect total adsorption of Cd to TiO2 (in thewater phase and in the gut). This is due to the fact that the degreeof cadmium adsorption to TiO2 nanoparticles inside the gut is notknown from the data in the present study. If the cadmium adsorb-ing to TiO2 in the water phase (6%) is added to the 3.3% localised inthe organisms this adds up to 9.3% which is somewhat lower whatwas observed in the test with L. variegatus. Adsorption isotherms forcadmium adsorption to various TiO2 nanoparticles established byGao et al. (2004) and Hartmann et al. (2010) reveal much higher Kdvalues for cadmium than found in this study. The reason for lowerKd in this study – and differences between the two test systemsdescribed in this study – may be due to the composition and pH ofthe test media used, which deviate from the conditions tested byGao et al., 2004 (pH 6.1, MilliQ water containing electrolytes andbuffer) and Hartmann et al., 2010 (similar pH, algae medium) aswell as the presence of organisms.

The fact that addition of TiO2 and adsorption of cadmium tothe particles did not alter the acute toxicity implies that (I) facili-tated transport of cadmium into the organism does not significantlyalter overall bioavailability and toxicity to either L variegatus orD. magna, (II) since cadmium uptake in L. variegatus is unchangedthe cadmium associated with TiO2 (25% of total mass) is at leastpartly bioavailable and (III) if TiO2 associated cadmium serves asan intra-organism source then reduced bioavailability due to sorp-tion is outweighed by increased bioavailability due to facilitateduptake. The bioavailability of particle bound cadmium is a topicthat is generally not completely understood. One study has foundthat sediment-bound cadmium did not contribute to uptake in L.

variegatus, which depended only on dissolved Cd-species (Piol et al.,2006). However, this is contradictory to other studies finding thatparticle-bound cadmium does, in fact, contribute to toxicity (Boschet al., 2009 and references within). It has been found by Zhang et al.

N.B. Hartmann et al. / Aquatic Toxicology 118– 119 (2012) 1– 8 7

Fig. 5. �XRF elemental maps obtained for L. variegatus: (a) sulphur (S), (b) titanium (Ti) and (c) is an overlay of a and b for the head of the specimen; (d) sulphur, (e) titaniumand (f) is an overlay of (d) and (e) for the body of the specimen.

(a) cal

(cuTssii

bifftteotoacr

Fig. 6. �XRF elemental maps obtained for D. magna:

2007) that TiO2 nanoparticles have a higher sorption capacity foradmium than that of natural soil particles. This resulted in a higherptake of cadmium into carp (Cyprinus carpio) in the presence ofiO2 nanoparticles compared to uptake in the presence of naturalediment particles. Besides measurable cadmium in gut, skin andcales – which could be partly reversible – increased of cadmiumn muscle was also seen, which indicates actual bioconcentrationnto tissue (Zhang et al., 2007).

Facilitated transport of metals by TiO2 nanoparticles has alsoeen observed in C. carpio by Sun et al. (2007, 2009). In the stud-

es the presence of TiO2 nanoparticles (P25 Degussa Evonic) wasound to increase the uptake of arsenate. However, the degree ofacilitated transport will differ between organisms. The reasons forhe observed differences between L. variegatus and D. magna inhe present study are likely to be linked to physiological differ-nces in feeding behaviour and size limits affecting their ingestionf TiO2 nanoparticles. Though no significant (p < 0.05) quantita-ive differences in extend of cadmium uptake in L. variegatus were

bserved in the absence or presence of TiO2 nanoparticles, �XRFnalysis results indicate discrete localisation of Ti which may indi-ate uptake of TiO2 nanoparticles into the organisms. This, however,emains to be investigated further.

cium, (b) titanium and (c) is an overlay of (a) and (b).

4. Conclusion

This study presents new findings on the toxicity of cadmium tosediment organism L. variegatus and crustacean D. magna in thepresence of TiO2 nanoparticles. Sorption of cadmium onto TiO2makes TiO2 nanoparticles potential carriers for cadmium and �XRFanalysis clearly confirmed uptake of TiO2 in the gut of D. magna.For L. variegatus �XRF analysis indicated an attachment of TiO2nanoparticles to the surface of the organism as well as a discretedistribution within the organism. The bioaccumulation and toxi-city of cadmium to L. variegatus was not changed by addition ofTiO2 nanoparticles, a pattern that was also observed for the toxic-ity of cadmium to D. magna despite a facilitated uptake of cadmium,increasing total body burden. The role of nanoparticle bound versussoluble cadmium in relation to contributions to uptake and toxicityneeds further investigations.

Acknowledgements

The authors would like to acknowledge Sinh Nguyen (DTU Envi-ronment) for ICP-OES measurements as well as Susanne Kruse (DTU

8 atic To

Ea

R

B

B

B

B

C

D

G

G

G

G

G

G

H

H

I

J

K

L

N.B. Hartmann et al. / Aqu

nvironment) for assisting with experimental work and chemicalnalysis. Thank you to Torben Dolin for graphical assistance.

eferences

aun, A., Hartmann, N.B., Grieger, K., Kusk, K.O., 2008. Ecotoxicity of engineerednanoparticles to aquatic invertebrates: a brief review and recommendations forfuture toxicity testing. Ecotoxicology 17 (5), 387–395.

arata, C., Markich, S.J., Baird, D.J., Soares, A.M.V.M., 2002. The relative importanceof water and food as cadmium sources to Daphnia magna Straus. Aquatic Toxi-cology 61 (3–4), 143–154.

attin, T.J., Von der Kammer, F., Weilhartner, A., Ottofuelling, S., Hofmann, T., 2009.Nanostructured TiO2: Transport behavior and effects on aquatic microbial com-munities under environmental conditions. Environmental Science & Technology43 (21), 8098–8104.

osch, C., Olivares, A., Faria, M., Navas, J.M., del Olmo, I., Grimalt, J.O., Pina, B., Barata,C., 2009. Identification of water soluble and particle bound compounds caus-ing sublethal toxic effects. A field study on sediments affected by a chlor-alkaliindustry. Aquatic Toxicology 94 (1), 16–27.

lifford, M., McGeer, J.C., 2010. Development of a biotic ligand model to predictthe acute toxicity of cadmium to Daphnia pulex. Aquatic Toxicology 98 (1),1–7.

awson, T.D., Lott, K.G., Leonard, E.N., Mount, D.R., 2003. Time course of metal loss inLumbriculus variegatus following sediment exposure. Environmental Toxicologyand Chemistry 22 (4), 886–889.

ao, L., Zhang, Q., 2001. Effects of amorphous contents and particle size on thephotocatalytic properties of TiO2 nanoparticles. Scripta Materialia 44 (8–9),1195–1198.

ao, Y., Wahi, R., Kan, A.T., Falkner, J.C., Colvin, V.L., Tomson, M.B., 2004. Adsorptionof cadmium on anatase nanoparticles—effect of crystal size and pH. Langmuir:The ACS Journal of Surfaces and Colloids 20 (22), 9585–9593.

eller, W., Müller, H., 1981. The filtration apparatus of Cladocera: filter mesh-sizesand their implications on food selectivity. Oecologia 49 (3), 316–321.

illis, P.L., Chow-Fraser, P., Ranville, J.F., Ross, P.E., Wood, C.M., 2005. Daphnianeed to be gut-cleared too: the effect of exposure to and ingestion of metal-contaminated sediment on the gut-clearance patterns of D. magna. AquaticToxicology 71 (2), 143–154.

ophen, M., Geller, W., 1984. Filter mesh size and food particle uptake by Daphnia.Oecologia 64 (3), 408–412.

uilhermino, L., Diamantino, T., Ribeiro, R., Gonncalves, F., Soares, A.M.V.M., 1997.Suitability of test media containing EDTA for the evaluation of acute metaltoxicity to Daphnia magna Straus. Ecotoxicology and Environmental Safety 38,292–295.

artmann, N.B., von der Kammer, F., Hofmann, T., Baalousha, M., Ottofuelling, S.,Baun, A., 2010. Algal testing of titanium dioxide nanoparticles—testing con-siderations, inhibitory effects and modification of cadmium bioavailability.Toxicology 269 (2–3), 190–197.

artmann, N.B., Baun, A., 2010. The nano cocktail: ecotoxicological effects ofengineered nanoparticles in chemical mixtures. Integrated EnvironmentalAssessment and Management 6 (2), 311–313.

SO (International Organization for Standardization), 1998. ISO6341:1996/Cor.1:1998. Water quality—determination of the inhibition ofthe mobility of Daphnia magna Straus (Cladocera, Crustacea)—acute toxicitytest.

ensen, H., Joensen, K.D., Jørgensen, J., Pedersen, J.S., Søgaard, G., 2004. Character-ization of nanosized partly crystalline photocatalysts. Journal of NanoparticleResearch 6 (5), 519–526.

eller, A.A., Wang, H., Zhou, D., Lenihan, H.S., Cherr, G., Cardinale, B.J., Miller, R., Ji, Z.,2010. Stability and aggregation of metal oxide nanoparticles in natural aqueousmatrices. Environmental Science & Technology 44 (6), 1962–1967.

andrum, P.F., Gedeon, M.L., Burton, G.A., Greenberg, M.S., Rowland, C.D., 2002.Biological responses of Lumbriculus variegatus exposed to fluoranthene-spiked

xicology 118– 119 (2012) 1– 8

sediment. Archives of Environmental Contamination and Toxicology 42 (3),292–302.

Lin, D., Tian, X., Wu, F., Xing, B., 2010. Fate and transport of engineered nano-materials in the environment. Journal of Environmental Quality 39 (6),1896–1908.

Mino, L.A., Folco, S., D’Ángelo, A.M.P.D., Guerrero, N.R.V., 2006. Modeling leadbioavailability and bioaccumulation by Lumbriculus variegatus using artificialparticles. Potential use in chemical remediation processes. Chemosphere 63 (2),261–268.

OECD (Organisation for Economic Co-operation and Development), 2004. Test No.202: Daphnia sp. acute immobilisation test.

Ottofuelling, S., Von Der Kammer, F., Hofmann, T., 2011. Commercial titaniumdioxide nanoparticles in both natural and synthetic water: comprehensive mul-tidimensional testing and prediction of aggregation behavior. EnvironmentalScience & Technology 45 (23), 10045–10052.

Penttinen, O., Kilpi-Koski, J., Jokela, M., Toivainen, K., Väisänen, A., 2008. Impor-tance of dose metrics for lethal and sublethal sediment metal toxicity in theoligochaete worm Lumbriculus variegatus. Journal of Soils and Sediments 8 (1),59–66.

Piol, M.N., López, A.G., Mino, L.A., Dos Santos Afonso, M., Verrengia Guerrero, N.R.,2006. The impact of particle-bound cadmium on bioavailability and bioaccu-mulation: a pragmatic approach. Environmental Science & Technology 40 (20),6341–6347.

Rosenkranz, P., Chaudhry, Q., Stone, V., Fernandes, T.F., 2009. A comparison ofnanoparticle and fine particle uptake by Daphnia magna. Environmental Tox-icology and Chemistry 28 (10), 2142–2149.

Sachtleben, 2005. Product Information Hombikat UV100, Publication no.07.011.03.05.e. Sachtleben Chemie GmbH, Duisburg, Germany.

Sachtleben, 2006. Product Information Hombitan LW-S, Publication no.01.003.07.06.e. Sachtleben Chemie GmbH, Duisburg, Germany, Available from:http://www.sachtleben.de/publications/9103e072.pdf (accessed February 20,2011).

Stuhlbacher, A., Bradley, M., Naylor, C., Calow, P., 1993. Variation in the developmentof cadmium resistance in Daphnia magna straus; effect of temperature, nutrition,age and genotype. Environmental Pollution 80 (2), 153–158.

Suedel, B.C., Rodgers Jr., J.H., Deaver, E., 1997. Experimental factors that may affecttoxicity of cadmium to freshwater organisms. Archives of Environmental Con-tamination and Toxicology 33 (2), 188–193.

Sun, H., Zhang, X., Niu, Q., Chen, Y., Crittenden, J.C., 2007. Enhanced accumulation ofarsenate in carp in the presence of titanium dioxide nanoparticles. Water, Air,and Soil Pollution 178 (1–4), 245–254.

Sun, H., Zhang, X., Zhang, Z., Chen, Y., Crittenden, J.C., 2009. Influence of titaniumdioxide nanoparticles on speciation and bioavailability of arsenite. Environmen-tal Pollution 157 (4), 1165–1170.

Swedish. Environmental Protection Agency, 1990. Probit Version 2.3.US Environmental Protection Agency (US EPA), 2002. Methods for Mea-

suring the Acute Toxicity of Effluents and Receiving Waters to Fresh-water and Marine Organisms, 5th ed. http://www.epa.gov/waterscience/methods/wet/disk2/atx.pdf (accessed February 20, 2011).

von der Kammer, F., Ottofuelling, S., Hofmann, T., 2010. Assessment of the physico-chemical behavior of titanium dioxide nanoparticles in aquatic environmentsusing multi-dimensional parameter testing. Environmental Pollution 158 (12),3472–3481.

Wang, G.-B., Fu, M.-G., Lu, B., Du, G.-P., Li, L., Qin, X.-M., Shi, W.-Z., 2010. Growth ofnanocrystalline TiO2 films by pulsed-laser-induced liquid-deposition methodand preliminary applications for dye-sensitized solar cells. Applied Physics A100, 1169–1176.

Xie, L., Lambert, D., Martin, C., Cain, D.J., Luoma, S.N., Buchwalter, D., 2008. Cadmium

biodynamics in the oligochaete Lumbriculus variegatus and its implications fortrophic transfer. Aquatic Toxicology 86 (2), 265–271.

Zhang, X., Sun, H., Zhang, Z., Niu, Q., Chen, Y., Crittenden, J.C., 2007. Enhancedbioaccumulation of cadmium in carp in the presence of titanium dioxidenanoparticles. Chemosphere 67 (1), 160–166.