Toxicity of Nano-Zero Valent Iron to Freshwater and Marine Organisms
Citation: Keller AA, Garner K, Miller RJ, Lenihan HS (
Toxicity of Nano-Zero Valent Iron to Freshwater and Marine Organisms
Arturo A. Keller 0
Kendra Garner 0
Robert J. Miller 0
Hunter S. Lenihan 0
Wei-Chun Chin, University of California Merced, United States of America
0 University of California Center for Environmental Implications of Nanotechnology, Bren School of Environmental Science and Management, University of California Santa Barbara , Santa Barbara, California , United States of America
We tested whether three commercial forms (uncoated, organic coating, and iron oxide coating) of nano zero-valent iron (nZVI) are toxic to freshwater and marine organisms, specifically three species of marine phytoplankton, one species of freshwater phytoplankton, and a freshwater zooplankton species (Daphnia magna), because these organisms may be exposed downstream of where nZVI is applied to remediate polluted soil. The aggregation and reactivity of the three types of nZVI varied considerably, which was reflected in their toxicity. Since levels of Fe2+ and Fe3+ increase as the nZVI react, we also evaluated their toxicity independently. All four phytoplankton species displayed decreasing population growth rates, and Daphnia magna showed increasing mortality, in response to increasing levels of nZVI, and to a lesser degree with increasing Fe2+ and Fe3+. All forms of nZVI aggregated in soil and water, especially in the presence of a high concentration of calcium ions in groundwater, thus reducing their transports through the environment. However, uncoated nZVI aggregated extremely rapidly, thus vastly reducing the probability of environmental transport and potential for toxicity. This information can be used to design a risk management strategy to arrest the transport of injected nZVI beyond the intended remediation area, by injecting inert calcium salts as a barrier to transport.
Funding: This work was supported by the National Science Foundation and the U.S. Environmental Protection Agency under Cooperative Agreement #
NSFEF0830117, and by National Science Foundation Grant EF-0742521. Any opinions, findings, and conclusions or recommendations expressed in this material are
those of the authors and do not necessarily reflect the views of the National Science Foundation or the U.S. Environmental Protection Agency. The work was also
supported by a grant from AECOM Environment. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the
Competing Interests: The authors have read the journals policy and have the following conflicts: The authors received a $86,000 grant from AECOM to study
the toxicity, mobility and persistence of nZVI, to inform their risk assessment. Arturo Keller is member of the Advisory Board of AECOM and receives a small
($3,000/yr) honorarium from AECOM. None of the authors have been employed by AECOM, and there are no patents, or conflicts with products in development
or marketed products. In particular, the authors have no commercial or other conflict of interest with the manufacturer of the nano-scale zero valent iron particles
used in this study. This does not alter the authors adherence to all the PLoS ONE policies on sharing data and materials.
Zero valent iron (ZVI) is an excellent electron donor that is used
to transform via reduction or indirect oxidation many common
contaminants in soil and groundwater . The development of
stable nano-scale ZVI (nZVI) products has generated significant
interest in environmental remediation applications, with at least 80
pilot and field scale studies completed or underway . Stable
refers to the incorporation of a coating to the nZVI that reduces
the rate of aggregation  and may also slow down the rate of
release of Fe2+ from the core ZVI [4,915]. Maintaining a stable
small particle diameter is important to achieve sufficient mobility
to reach the target contaminants. Reducing the rate of oxidation
maximizes the electrons that are donated for the intended
reactions. A number of commercial ZVI products are now
available that contain stabilized nanoparticles.
While nZVI holds considerable promise for many remediation
applications, the environmental risks are still poorly understood. In
particular, the bioavailability and ecotoxicity of nZVI in different
environmental media has not been studied in detail, and what we
understand about ZVI toxicity is based mainly on studies of
nonnano ZVI, or Fe0. There is an implicit assumption that nZVI is
relatively non-toxic because Fe0 simply oxidizes to Fe2+ and then
to Fe3+, both of which are common chemical species in the
environment that most organisms are well adapted to deal with.
However, ZVI applications can increase the concentration of Fe2+
and/or Fe3+ substantially at a local level in the short term. ZVI
oxidation can also lead to the production of reactive oxygen
species (ROS), such as hydroxyl radicals (OH?) from superoxide
(O2?2) and hydrogen peroxide (H2O2) in living cells . Fe ions
enter the cytoplasm of cells and induce oxidative stress, which,
among other impacts, can damage cell membranes leading to
leakage of intracellular contents and cell death .
To date, only a few studies have evaluated the toxicity of nZVI,
and most have focused on microbes (See Table 1 for summary).
Uncoated nZVI, 35 nm (range 1080 nm) in diameter, were toxic
to Escherichia coli (ATCC strain 8739), displaying greater toxicity in
hypoxic than aerobic conditions in soils and water . Lee et al.
also determined that ROS generation was responsible for E. coli
death when exposed to Fe2+ at 5.6 mg L21 Fe, but it required
$56 mg L21 Fe3+ to kill E. coli. Another study found mechanisms
other than ROS by which nZVI can kill soil-based E. coli,
including mitochondrial membrane damage, but also revealed that
toxicity declines with the length of exposure because of a strong
tendency for nZVI to form large aggregates (320630 nm),
regardless of soil pH . Uncoated nZVI with 728% Fe0
content is toxic to E. coli at a concentration of about 5 mg L21, but
toxicity was not observed below 100 mg L21 for humic acid
coated-nZVI; below 140 mg L21 for polyaspartate-coated nZVI;
and below 516 mg L21 for poly(styrene\sulfonate)-coated nZVI,
thereby indicating that electrostatic repulsion provided by
negatively charged coatings inhibits toxicity . Other forms of
toxicity, or a lack thereof, have been identified for fungi , viri
, human cells  and rodent cells . In aquatic
ecosystems, polyaspartate-coated nZVI was found toxic to the
amphidromous (seawater-freshwater inhabiting) medaka fish at
concentrations beginning at ,5 mg L21 .
Here we extend our understanding of the potential ecological
risks of nZVI and its chemical byproducts to aquatic biota,
specifically those inhabiting freshwater streams and coastal
marine/estuarine waters, ecosystems that are connected to nZVI
remediation sites via the seepage of groundwater . We focus
on the potential toxicity to primary producers, specifically
phytoplankton (4080 mm in size), and a primary consumer
Daphnia magna, a freshwater zooplankton herbivore (12 mm). We
chose these species because planktonic species are ecologically
important as basal species in aquatic food webs, and are at
substantial ecological risk from nanomaterial (NM) exposure due
to potential exposure associated with terrestrial runoff or
groundwater seepage into freshwater stream and ponds, as well
as coastal bays, lagoons, and estuaries . In addition, metal ions
dissolved from some NMs may be readily bioavailable and harm
phytoplankton cells, leading to declines in population growth rates
and abundance . Dissolution rates, and therefore the effective
toxicity of NMs, often decreases with increasing ionic strength of
the surrounding aquatic media (freshwater or seawater) because it
leads to nanomaterial aggregation. The presence of ions also
increases the rate of NM sedimentation, decreasing exposure to
pelagic organism while increasing the probability of exposure to
benthic organisms .
In light of these complex biological and chemical features,
predicting under what conditions nZVI poses risks to planktonic
organisms is challenging. To meet this important environmental
challenge, we tested the following hypotheses: (1) the aggregation
rate and aggregate size of nZVI varies with the type of coating, in
the rank order of uncoated nZVI .polymer coated-nZVI . iron
oxide coated-nZVI; (2) the aggregation rate and aggregate size for
all forms of nZVI is greater in seawater than in freshwater; (3)
following the aggregation behavior, the toxicity of nZVI is greater
in freshwater than in seawater; and (4) based on oxidative capacity,
toxicity of polymer coated-nZVI . iron oxide coated- nZVI .
Fe+2. Fe+3. Uncoated nZVI was not used for the toxicity studies
due to excessive aggregation which would have thoroughly
confounded our results.
Particle Size Analysis
SEM imaging of Nanofer 25S revealed aggregates of primary
nano zero valent iron (nZVI) of approximately 80120 nm
diameter (Figure 1). Since the material was received as a slurry,
the sample was dried before SEM imaging; thus the aggregation in
the SEM images may not accurately reflect the size of the original
material. Nanofer STAR was composed of aggregates of nZVI of
,100200 nm in diameter (Figure 1). The Nanofer 25, an
E. coli (ATCC strain 8739)
E. coli (ATCC strain 8739)
E. coli (ATCC strain 8739)
E. coli (ATCC strain 8739)
E. coli (ATCC strain 33876)
E. coli (ATCC strain 33876)
E. coli (ATCC strain 33876)
E. coli (ATCC strain 33876)
Uncoated 35 nm nZVI
Uncoated 50 nm nZVI
poly(styrene sulfonate) coated nZVI
polyaspartate coated nZVI
humic acid coated nZVI
Polyaspartate coated 30 nm nZVI
Polyaspartate coated 30 nm nZVI
90 mg L21 (inactivation)
9 mg L21 (inactivation)
5.6 mg L21 Fe (inactivation)
56 mg L21 (inactivation)
under aerated conditions
under deaerated conditions
Studies conducted at pH 55.5, toxicity
observed after 1 hour contact
gram positive under aerobic conditions
gram negative under aerobic conditions
Survival ranged from 90100% at all
inactivation within 5 minutes
pH 77.6. gill samples showed increasing
deposition of black particles, swelling of
the epithelium cells and missing scales
no significant effect observed on
development at varied pH levels
no significant effect observed
uncoated material, appeared more agglomerated (see Figure S1 in
Supporting Information), making it more difficult to determine the
primary particle size.
We evaluated particle aggregation in synthetic water and
natural (freshwater, groundwater, seawater) conditions. Nanofer
25S particles were stable at a hydrodynamic diameter (peak
intensity) of around 250650 nm across a range in pH from 411
(Figure 2A). Nanofer STAR particles were also relatively stable
between 700 and 1800 nm across the same pH range (Figure 2B).
The Nanofer 25 particles were rather large (several mm) at any pH
from 4 to 10.5 (Figure S2). These experiments were conducted at
low ionic strength (,1 mM). At higher ionic strength (IS), the
Nanofer 25 particles started out very large (.3 mm), but remained
fairly stable in size (Figure S3). There was not much difference
between a monovalent ionic solution (NaCl) and a divalent ionic
solution (CaCl2). Nanofer 25S aggregates, in contrast, remained
stable even at high IS when NaCl was used, but began aggregating
rapidly when CaCl2 was used to increase the IS to 10 or 100 mM
(Figure 3A). At higher IS, the size of the Nanofer STAR particles
remained relatively constant over the 30-minute experiments,
though they were somewhat larger at higher ionic strengths
(Figure 3B), regardless of the nature of the cations present. Thus,
rapid aggregation in hard groundwater would be expected for all
particles, particularly for Nanofer 25S, since its initial particle size
is much smaller than for the other materials.
In the majority of natural water samples studied here
aggregation of the particles was enhanced compared to the
synthetic waters. Nanofer 25 particles aggregated to .3 mm very
rapidly in the three natural water samples, and generally exhibited
further aggregation over time (Figure S4). The Nanofer 25S
particles were stable in freshwater (pH 7.5) at an aggregate size of
around 280650 nm, but aggregated in groundwater and seawater
(Figure 4A). The Nanofer STAR particles were somewhat stable in
freshwater and seawater, but formed large aggregates in other
water samples (Figure 4B).
Generally, the zeta potential of the Nanofer 25 particles varied
and was closer to neutral than the zeta potential of the Nanofer
25S particles, which was near 240 mV (Table S1). The zeta
potential of the Nanofer STAR was around neutral, similar to the
Nanofer 25, but with greater variation (Table S1). Neutral
particles tend to aggregate faster unless a stabilizing coating is
added to the nZVI. Aggregation of the particles was high when the
charge was small, below around 615 mV. Thus, the zeta potential
of particles in a given media can be used to predict whether the
particles will be stable or not.
Toxicity to Phytoplankton
Population growth of the marine phytoplankton species Isochrysis
galbana was significantly depressed at concentrations of Nanofer
25S $3 mg L21 (Figure 5A) compared with controls. Growth was
reduced to near zero above 6 mg L21 (Figure 5A). In contrast,
growth of I. galbana was not significantly affected by Nanofer
STAR at any concentration (Figure 5B). Ionic iron species did not
reduce growth of I. galbana at concentrations below 50 mg L21 for
Fe2+ and below 75 mg L21 for Fe3+ (Figures 5C, D). Since it
exhibited the highest toxic potential of the two particles, we tested
Figure 3. Nanofer 25S and Nanofer STAR particle size as a function of ionic strength at 100 mg L21 and pH 7, over time.
the effects of Nanofer 25S on two additional species of marine
phytoplankton, Dunaliella tertiolecta and Thalassiosira pseudonana.
Population growth of both species was depressed at low
concentrations of this nanomaterial: 1.3 mg L21 for D. tertiolecta
and 0.4 mg L21 for T. pseudonana (Figure 6). Population growth of
the freshwater phytoplankton species Pseudokirchneriella subcapitata
was not significantly affected by Nanofer 25S at concentrations
,8 mg L21 (Figure 7A). However, unlike the case for I. galbana,
Nanofer STAR significantly impacted P. subcapitata at
concentrations $12 mg L21 (Figure 7B). P. subcapitata was also more
sensitive to Fe2+ and Fe3+, which significantly reduced its growth
rate at concentrations of 10 mg L21 for Fe2+ and 25 mg L21 for
Fe3+ (Figures 7C, D).
Toxicity to Zooplankton
Exposure tests (96-hour) showed that Daphnia magna survival was
dramatically impacted by both Nanofer 25S and Nanofer STAR
at total Fe concentrations $0.5 mg L21 (Figures 8A, B). To
determine whether the observed toxicity for the nZVI was
attributable only to nanoparticle-associated Fe(0), we evaluated
the toxicity to Daphnia magna of Fe2+ and Fe3+ amended growth
media (Figure 8C, D). Higher concentrations of Fe2+ and Fe3+
were reached before significant mortality effects: 4 mg L21 for
Fe2+ and 15 mg L21 for Fe3+, although there were indications of
decreased survival at the lowest concentrations also (Figures 8C,
D). Daily survival data as the experiments progressed indictated
that at concentrations above ,1 mg L21 Nanofer particles caused
significant die-offs within the first 2448 hours (Figure 9A,B), and
a similar response was observed for Fe2+ (Figure 9C). D. magna
responded more slowly to Fe3+ at all but the highest concentrations
.15 mg L21 (Figure 9D).
Our results show that commercial formulations of nZVI can be
toxic to aquatic organisms that may be exposed to the material
downstream of remediation sites, either in freshwater streams,
ponds, or in the coastal marine environment (Table 2). These
values represent the no observed effect concentration (NOEC),
and an assessment factor would have to be applied to estimate the
Predicted No Effect Concentration (PNEC) based on OECDs
guidelines. For reference, we summarize the previous results with
nZVI and related materials. We found that the toxicity is strongly
dependent on the form of nZVI, which is an important
consideration when using these nanomaterials in remediation
applications. Coatings can profoundly affect the toxicity of nZVI,
and in general we found support for our first two hypotheses that
(1) the aggregation rate of uncoated nZVI .polymer coated-nZVI
. iron oxide coated-nZVI, and (2) that the aggregation rate for all
forms of nZVI is greater in seawater than in freshwater. Uncoated
particles, in this case Nanofer 25, aggregated so rapidly as to make
them unsuitable for remediation applications. Nanofer 25S, which
was coated with polyethylene glycol sorbitan monostearate,
showed minimal aggregation in pH 7.5 freshwater, although
aggregation of this particle was still generally high in hard
groundwater (Figure 4). Nanofer STAR was capped with a 2 nm
Fe-O shell, and although it was stable, the initial aggregate size
was large in all tested media (Figure 4). The latter two particles
aggregated in seawater, as expected, but aggregation of the
surfactant-coated Nanofer 25S was faster than for Nanofer STAR
As predicted, the smaller aggregate size of Nanofer 25S
increased its toxicity over Nanofer STAR ; the passivated
iron oxide surface of the Nanofer STAR also reduces its reactivity
Figure 5. Growth Rate for I. galbana exposed to (a) Nanofer 25S, (b) Nanofer STAR, (c) Fe2+, and (d) Fe3+.
and toxicity. The NOEC of Nanofer 25S on population growth of
the freshwater phytoplankton P. subcapitata was 8.2 mg L21 versus
12.4 mg L21 for Nanofer STAR (Figure 6). Both particles were
highly toxic to D. magna at low concentrations, although Nanofer
25S induced mortality more rapidly (Figures 8 and 9). We
expected that the high level of aggregation of the nZVI in salt
water would result in lower toxicity of the particles, but that was
not always the case. Indeed, Nanofer 25S significantly depressed
growth of the marine microalgae I. galbana at concentrations as low
as 3 mg L21; Nanofer STAR, however, showed no effect on
I. galbana even at very high concentrations close to 100 mg L21,
suggesting that aggregation in seawater did affect its toxicity
(Figure 5). Exposure tests with two additional species of marine
phytoplankton also showed toxicity of Nanofer 25S at relatively
low concentrations. Since the pH remained fairly constant
throughout the exposures (Table S2, freshwater range from 7.5
Figure 7. Growth Rate for P. subcapitata exposed to (a) Nanofer 25S, (b) Nanofer STAR, (c) Fe2+, and (d) Fe3+.
8.1 and seawater range from 8.18.3), it is unlikely to be a factor in
the toxicity. However, light transmission did decrease by around
10% for the concentration of iron around 10 mg L21 and more
than 95% when the concentration of iron is greater than 100 mg
L21, which can be an important factor affecting growth of
phytoplankton at these higher iron concentrations (Figure S5).
However, when the concentration of iron is less than 5 mg L21,
this mechanism is likely to be minor or negligible in overall
We predicted that the hierarchy of oxidative capacity among
the forms of Fe tested, organic coated-nZVI (Nanofer 25S) . iron
oxide coated nZVI (Nanofer STAR) . Fe+2. Fe+3, would be
mirrored in the toxicity results. In general this was the case.
Nanofer 25S consistently exhibited toxicity at lower concentrations
than either Nanofer STAR or ionic Fe in both freshwater and
Figure 8. D. magna survival after 4 days in the presence of (a) Nanofer 25S, (b) Nanofer STAR, (c) Fe2+, and (d) Fe3+.
marine organisms. Fe+2 consistently showed toxicity at lower
concentrations than Fe+3. Nanofer STAR, however, did not
always show higher toxicity than Fe+2; indeed, this particle showed
no evident toxic effects on growth rate of the marine
phytoplankton species I. galbana and its toxicity to the freshwater microalgae
P. subcapitata was lower than that for Fe+2 (Figure 6). Nevertheless,
Nanofer STAR was highly toxic to the freshwater
suspensionfeeder D. magna, causing mortality at a much lower dose than Fe+2
or Fe+3 (Figure 8). Two potential explanations for decreased
toxicity of metal oxide nanoparticles with increasing charge have
been put forth: 1) the corresponding increase in energy needed for
release of dissolved ions, and 2) the decrease in ionization potential
with increasing charge . The first scenario is unlikely in this
case, since the nanoparticles in general exhibited toxicity at lower
concentrations than the salts despite their particulate nature.
Ionization potential, and the increased power of Fe to catalyze
production of hydroxyl radicals with lower charge, is more likely
the cause of the relationship seen here. Intracellular iron, indeed,
Aggregation at pH 7, 30 min
Statistically Significant Toxic Effect*
*Results are for observed statistically significant toxic effect.
2 nm Iron Oxide Shell
has been shown to be a potent cause of hydrogen
peroxideinduced DNA damage .
Our focus in this study were freshwater and marine organisms
that may be exposed to groundwater with residual yet elevated
concentrations of nZVI, Fe2+ and Fe3+. Other work on aquatic
organisms is thus far fairly limited. nZVI had a significant impact
on medaka (Oryzias latipes) fish and their embryos .
Commercial nZVI (primary size 30 nm) coated with 4 wt% of sodium
polyaspartate were used at concentrations ranging from of 0.5 and
50 mg L21 of nZVI, at pH 7 to 7.6, with a hardness of 200 mg
L21 as CaCO3. The embryos exhibited changes in enzymatic
activity in response to the ROS at 5 and 50 mg L21; even after
only 0.5 day exposure, with increasing changes in enzymatic
activity as the exposure time increased to 8 days . In the
adults, gill samples showed increasing deposition of black particles,
swelling of the epithelium cells and missing scales at concentrations
of 5 and 50 mg L21, after 14 days of exposure. Swelling and black
particle accumulation was also observed in the intestines at these
higher concentrations. No effect was observed in liver or brain
cells. Under natural 0.2 mm-filtered seawater conditions, no
significant effect was observed for 50 nm Fe2O3 nanoparticles or
an FeCl3 solution on the development of a mussel, Mytilus
galloprovincialis, at varied pH levels  and concentrations up to
10.0 mg L21 for the ferric oxide nanoparticles or 0.80 mg L21 of
Although there was clearly a toxic effect from dissolved Fe2+ and
Fe3+, the nZVI exhibited additional toxicity due perhaps to the
nanoparticles, their aggregates, or the H2 released during the
transformation of the nZVI. In most cases, the response at 1 mg
L21for ferrous and ferric iron was not statistically different from
the control, and even the effect at 5 mg L21 and in some cases
even 10 mg L21 were not as deleterious as observed from Nanofer
25S. It is likely that the nZVI attaches to the cell surfaces and
transfers electrons to different biochemicals at the surface, leading
to undesired reactions. The concentration-response curves based
on the growth rate of the phytoplankton population indicated that
the marine phytoplankton species I. galbana was more tolerant of
either Fe2+ or Fe3+ than the freshwater algae P. subcapitata. The
growth rate of I. galbana was statistically the same as the control
(lowest Fe, present in seawater) up to around 15 mg L21for either
Fe2+ or Fe3+ (Figure 5). However, I. galbana tolerated Fe3+ better
than Fe2+. In the case of P. subcapitata, effects were noticeable at
.1 mg L21for both Fe2+or Fe3+, and there was almost no
difference between the iron species (Figure 6).
Given that pilot and full remediation tests have used
concentrations of approximately 4.5 to 300 g L21 of nZVI slurry , the
concentrations that one may expect in the aquatic environment
influenced by the discharge from a remediation site could range
from mg L21 to mg L21. Nanofer 25S exhibited toxicity at 0.5
1.0 mg L21 in freshwater media to the freshwater phytoplankton
P. subcapitata and the water flea D. magna, and was also toxic for
three species of marine phytoplankton at 0.33.0 mg L21, similar
to the case for freshwater. The toxicity likely stems in part from the
oxidation products released from the ZVI particles, namely Fe2+
and Fe3+ ions. Additional studies may show that at the surface of
the interaction between the ZVI and the organisms, oxidation
reactions from the oxidation of Fe(0) to Fe2+ also result in localized
damage which can ultimately affect growth and even survival. In
many cases nZVI will be injected into the subsurface at a
significant distance from freshwater or coastal receptors, resulting
in considerable dilution of the concentrations of Fe2+ and Fe3+
ions, or precipitation of iron compounds. However, it would be
important to monitor the concentration of these ions
downgradient from an nZVI injection site, to determine whether there is
sufficient dilution or precipitation. Uncoated nZVI aggregate too
rapidly to transport significantly, but even nZVI with either an
organic surfactant coating or an iron oxide protective layer tend to
aggregate with time, particularly in the presence of a high
concentration of calcium ions in hard groundwater. This
information can be used to design a risk management strategy to
arrest the transport of injected nZVI beyond the intended
remediation area, by injecting Ca salts as a barrier to transport.
Three commercial nZVI were evaluated, Nanofer 25, Nanofer
25S, and Nanofer STAR (all from NANO IRON s.r.o., Rajhrad,
Czech Republic). The materials were received by air shipment,
with Nanofer 25 and Nanofer 25S as aqueous suspensions, and
Nanofer STAR as a powder. According to the manufacturer, the
iron content of all three Nanofers is 7090% nZVI and 1030%
iron oxides when produced. Nanofer 25S is coated with
polyethylene glycol sorbitan monostearate, a surfactant. Nanofer
STAR particles are coated with 2 nm iron oxide shell to reduce
their oxidation, allowing in-situ preparation of the suspensions.
Particle Size and Aggregation Studies
The size of the nZVI was determined using Scanning Electron
Microscopy (SEM) and Dynamic Light Scattering (DLS). Particles
were imaged using aZL40 Sirion FEG Digital Scanning
Microscope w/EDS (FEI, USA). Aggregation studies using DLS
(Zetasizer, Malvern Instruments, Ltd., UK) were conducted over
120 min periods in different waters, including a surface water,
groundwater and seawater. Groundwater was considered to
understand the potential mobility of the nZVI after injection.
Freshwater and seawater were considered in the toxicity studies.
Since the initial pH of the freshwater and groundwater were low,
aggregation was studied at an adjusted pH of 7.5 using 0.1 M
NaOH. A detailed characterization of these waters is provided in
the File S1. The charge on the ZVI particles was also measured
using the Zetasizer.
Three species of marine phytoplankton were used: Thalassiosira
pseudonana (centric diatom, Bacillariophyceae: Centrales); Dunaliella
tertiolecta (Chlorophyceae: Chlamydomonadales), and Isochrysis
galbana (Prymnesiophyceae: Isochrysidales). Axenic cultures were
obtained from the Provasoli-Guillard National Center for Culture
of Marine Phytoplankton (Bigelow Laboratory for Ocean
Sciences, West Boothbay, Harbor, Maine, USA), and were maintained in
standard media (f/2, 23, 24) made with 0.22 mm filtered natural
seawater, which was autoclaved prior to inoculation. The
background total Fe in the seawater media averaged 0.04 mg
L2160.04 mg L21. To provide inoculant for exposure
experiments, the phytoplankton were incubated under cool white
fluorescent lights (14:10 light:dark) at 20uC with aeration for 57
days until growth reached log phase. Cell densities were measured
by hemacytometer (Reichert, Buffalo, NY). Experiments were
conducted at 20uC, 34 parts per thousand salinity (%), under the
same fluorescent lights. All equipment was acid-washed, rinsed
with nanopure water, and autoclaved before use. For media, f/2
was used, [23,24] with only major nutrients added and no trace
metals, to avoid adding EDTA that would complex free metal
ions. Cells to inoculate the experiments were first filtered
(0.22 mm) and rinsed three times with filtered autoclaved seawater
to remove EDTA, and resuspended in EDTA-free growth media.
Experiments were run in 500 ml Erlenmeyer flasks, media volume
200 ml, and were mixed at ,150 rpm on a rotary shaker (New
Brunswick Scientific Co., NJ, USA). The nZVI concentrations
tested ranged from 0.2 to 100 mg L21 total Fe, with five replicates
per treatment. Five replicates per nZVI treatment were conducted.
Flasks were inoculated with 12 ? 105 cells ml21. Cell densities
were monitored every 24 hours for 96 hours by fluorometer
(Trilogy, Turner Designs, Sunnyvale, CA).
One species of freshwater phytoplankton, Pseudokirchneriella
subcapitata (Chlorophycea: Sphaeropleales) was tested. Starter
cultures were obtained from Carolina Biological Supply
(Burlington, NC, USA), and were maintained in standard freshwater
media  made with ultrapure filtered water (0.2 mg, Nanopure),
which was autoclaved prior to inoculation. All other conditions
were the same as for the marine phytoplankton. The background
concentration of total Fe in the media was 0.01 mg L21.
Nanofer 25 aggregated so rapidly (see Results) that it would not
be useful in remediation, thus our toxicity studies focused on
Nanofer 25S and Nanofer STAR. All four phytoplankton species
(three marine and one freshwater) were exposed to Nanofer 25S,
but only one marine (I. galbana) and the freshwater phytoplankton
(P. subcapitata) were exposed to Nanofer STAR because we had a
limited supply of the nanomaterials. We expected Nanofer to
dissolve and produce dissolved iron, which is naturally present in
freshwater and seawater. To test whether toxicity was due to the
nanomaterial or the dissolved iron that accumulates in the media
with the dissolution of the Nanofer, we compared toxicity of
Nanofer 25S and Nanofer STAR and dissolved iron, which we
mimicked using iron chloride salts (FeCl2 and FeCl3) at
concentrations of Fe2+ and Fe3+. with one marine (I. galbana) and
one freshwater (P. subcapitata) phytoplankton species.
Toxicity for phytoplankton was measured as a reduction in
population growth rates, which were estimated for each replicate
flask as the slope of log-transformed cell count data, obtained
through least-squares regression . One-way ANOVA was used
to test for an overall effect of NP toxicity on growth rates.
Homogeneity of variances was tested with Levenes test, and when
heterogeneous, data were transformed. When ANOVA revealed
significant differences among treatments, post-hoc tests were
conducted with Dunnetts method , which tests for pair-wise
differences between each treatment and the control.
The toxicity of the nZI to Daphnia magna, a freshwater
zooplankton grazer, was tested by measuring the survival of
young (neonate) individuals as per EPA Method 2021 .
Cultures of adult D. magna were obtained from Sachs Systems
Aquaculture (St. Augustine, Fl). Pregnant females were separated
until neonates were present, which were then collected and
transferred to test Petri dishes. Duplicate studies were done for
treatments 15 mg L21 total Fe and below. Exposure was
conducted by pipetting sufficient Nanofer 25S suspension to
achieve the desired total Fe concentration from 0.2 to 100 mg L21
total Fe. Survival of neonates was monitored daily for 96 hours.
To determine the concentration of total Fe in the test media,
4 ml of trace-metal-free nitric acid was added to a 1 ml sample of
the media used for each phytoplankton trial. This sample was then
digested in a HACH DRB200 digester (Hach, USA) at 80uC for
60 minutes and cooled for 30 minutes, diluted to 50 ml in a
volumetric flask using nanopure water, and analyzed via ICP-AES
(iCAP 6300, Thermo Scientific, Waltham, MA). NIST-traceable
standard solutions for total Fe (Fluka Analytical, Switzerland) were
used to generate calibration curves ranging from 0.01 to 100 mg
L21 for comparison.
Nanofer 25 particles imaged with SEM. Scale
Nanofer 25 particle size at a function of pH,
Nanofer 25 particle size in different waters,
Figure S5 Transmission of light in freshwater and
seawater for Nanofer 25S, STAR, and dissolved Fe2+
and Fe3+ at different nominal Fe concentrations.
Conceived and designed the experiments: AK RM. Performed the
experiments: KG. Analyzed the data: KG AK RM HL. Contributed
reagents/materials/analysis tools: AK. Wrote the paper: AK KG RM HL.
1. Zhang W-X ( 2003 ) Nanoscale Iron Particles for Environmental Remediation: An Overview . Journal of Nanoparticle Research 5 : 323 - 332 .
2. Karn B , Kuiken T , Otto M ( 2009 ) Nanotechnology and in Situ Remediation: A Review of the Benefits and Potential Risks . Environmental Health Perspectives 117 : 1823 - 1831 .
3. Comba S , Sethi R ( 2009 ) Stabilization of highly concentrated suspensions of iron nanoparticles using shear-thinning gels of xanthan gum . Water Research 43 : 3717 - 3726 .
4. He F , Zhao D ( 2007 ) Manipulating the Size and Dispersibility of Zerovalent Iron Nanoparticles by Use of Carboxymethyl Cellulose Stabilizers . Environmental Science & Technology 41 : 6216 - 6221 .
5. He F , Zhang M , Qian T , Zhao D ( 2009 ) Transport of carboxymethyl cellulose stabilized iron nanoparticles in porous media: Column experiments and modeling . Journal of Colloid and Interface Science 334 : 96 - 102 .
6. He F , Zhao D , Paul C ( 2010 ) Field assessment of carboxymethyl cellulose stabilized iron nanoparticles for in situ destruction of chlorinated solvents in source zones . Water Research 44 : 2360 - 2370 .
7. Li Z , Greden K , Alvarez PJJ , Gregory KB , Lowry GV ( 2010 ) Adsorbed Polymer and NOM Limits Adhesion and Toxicity of Nano Scale Zerovalent Iron to E . coli. Environmental Science & Technology 44 : 3462 - 3467 .
8. Phenrat T , Liu Y , Tilton RD , Lowry GV ( 2009 ) Adsorbed Polyelectrolyte Coatings Decrease Fe0 Nanoparticle Reactivity with TCE in Water: Conceptual Model and Mechanisms . Environmental Science & Technology 43 : 1507 - 1514 .
9. Saleh N , Kim H-J , Phenrat T , Matyjaszewski K , Tilton RD , et al. ( 2008 ) Ionic strength and composition affect the mobility of surface-modified Fe0 nanoparticles in water-saturated sand columns . Environmental Science & Technology 42 : 3349 - 3355 .
10. Bennett P , He F , Zhao D , Aiken B , Feldman L ( 2010 ) In situ testing of metallic iron nanoparticle mobility and reactivity in a shallow granular aquifer . Journal of Contaminant Hydrology 116 : 35 - 46 .
11. Lin Y-H , Tseng H-H , Wey M-Y , Lin M-D ( 2010 ) Characteristics of two types of stabilized nano zero-valent iron and transport in porous media . Science of The Total Environment 408 : 2260 - 2267 .
12. Phenrat T , Long TC , Lowry GV , Veronesi B ( 2008 ) Partial Oxidation (''Aging'') and Surface Modification Decrease the Toxicity of Nanosized Zerovalent Iron . Environmental Science & Technology 43 : 195 - 200 .
13. Shin M-C , Choi H-D , Kim D-H , Baek K ( 2008 ) Effect of surfactant on reductive dechlorination of trichloroethylene by zero-valent iron . Desalination 223 : 299 - 307 .
14. Wang W , Zhou M , Jin Z , Li T ( 2010 ) Reactivity characteristics of poly(methyl methacrylate) coated nanoscale iron particles for trichloroethylene remediation . Journal of Hazardous Materials 173 : 724 - 730 .
15. Xiu Z- m, Gregory KB , Lowry GV , Alvarez PJJ ( 2010 ) Effect of Bare and Coated Nanoscale Zerovalent Iron on tceA and vcrA Gene Expression in Dehalococcoides spp . Environmental Science & Technology 44 : 7647 - 7651 .
16. Li H , Zhou Q , Wu Y , Fu J , Wang T , et al. ( 2009 ) Effects of waterborne nanoiron on medaka (Oryzias latipes): Antioxidant enzymatic activity, lipid peroxidation and histopathology . Ecotoxicology and Environmental Safety 72 : 684 - 692 .
17. Lee C , Kim JY , Lee WI , Nelson KL , Yoon J , et al. ( 2008 ) Bactericidal Effect of Zero-Valent Iron Nanoparticles on Escherichia coli . Environmental Science & Technology 42 : 4927 - 4933 .
18. Diao M , Yao M ( 2009 ) Use of zero-valent iron nanoparticles in inactivating microbes . Water Research 43 : 5243 - 5251 .
19. Klaine SJ , Alvarez PJJ , Batley GE , Fernandes TF , Handy RD , et al. ( 2008 ) Nanomaterials in the environment: Behavior , fate, bioavailability, and effects. Environmental Toxicology and Chemistry 27 : 1825 - 1851 .
20. Keenan CR , Goth-Goldstein R , Lucas D , Sedlak DL ( 2009 ) Oxidative Stress Induced by Zero-Valent Iron Nanoparticles and Fe(II) in Human Bronchial Epithelial Cells . Environmental Science & Technology 43 : 4555 - 4560 .
21. Miller RJ , Lenihan HS , Muller EB , Tseng N , Hanna SK , et al. ( 2010 ) Impacts of metal oxide nanoparticles on marine phytoplankton . Environmental Science & Technology 44 : 7329 - 7334 .
22. Keller AA , Wang H , Zhou D , Lenihan HS , Cherr G , et al. ( 2010 ) Stability and Aggregation of Metal Oxide Nanoparticles in Natural Aqueous Matrices . Environmental Science & Technology 44 : 1962 - 1967 .
23. Phenrat T , Long TC , Lowry GV , Veronesi B ( 2008 ) Partial Oxidation (Aging) and Surface Modification Decrease the Toxicity of Nanosized Zerovalent Iron . Environmental Science & Technology 43 : 195 - 200 .
24. Puzyn T , Rasuley B , Gajewicz A , Hu X , Dasari TP , Michalkova A ( 2011 ) Using Nano-QSAR to predict the cytotoxicity of meta oxide nanoparticles . Nature Nanotechnology 6 : 175 - 178 .
25. Barbouti A , Doulias PT , Zhu BZ , Frei B , Galaris D ( 2001 ) Intracellular iron, but not copper, plays a critical role in hydrogen peroxide-induced DNA damage . Free Radical Biology and Medicine 31 : 490 - 498 .
26. Kadar E , Simmance F , Martin O , Voulvoulis N , Widdicombe S , et al. ( 2010 ) The influence of engineered Fe2O3 nanoparticles and soluble (FeCl3) iron on the developmental toxicity caused by CO2-induced seawater acidification . Environmental Pollution 158 : 3490 - 3497 .
27. Gavaskar A , Tatar L , Condit W ( 2005 ) Cost and performance report nanoscale zero-valent iron technologies for source remediation . Naval Facilities Engineering Command.
28. Methods for Measuring the Acute Toxicity of Effluents and Receiving Waters to Freshwater and Marine Organisms . Fifth Edition, October 2002 . EPA Volume 821-R-02-012.