Nanoparticle Silver Released into
Water from Commercially Available
T R O Y M . B E N N * A N D P A U L W E S T E R H O F F
Civil and Environmental Engineering, Arizona State
University, Box 5306, Tempe, Arizona 85287-5306
Received December 31, 2007. Revised manuscript received
March 18, 2008. Accepted March 24, 2008.
Manufacturers of clothing articles employ nanosilver (n-Ag)
release from commercial products are unknown. The quantity
and form of the nanomaterials released from consumer
products should be determined to assess the environmental
risks of nanotechnology. This paper investigates silver released
from commercial clothing (socks) into water, and its fate in
wastewater treatment plants (WWTPs). Six types of socks
contained up to a maximum of 1360 µg-Ag/g-sock and leached
as much as 650 µg of silver in 500 mL of distilled water.
Microscopy conducted on sock material and wash water
revealed the presence of silver particles from 10 to 500 nm in
diameter. Physical separation and ion selective electrode
(ISE) analyses suggest that both colloidal and ionic silver leach
from the socks. Variable leaching rates among sock types
suggests that the sock manufacturing process may control the
release of silver. The adsorption of the leached silver to
WWTP biomass was used to develop a model which predicts
that a typical wastewater treatment facility could treat a
high concentration of influent silver. However, the high silver
The burgeoning nanotechnology industry is quickly produc-
products. As of 2007, the Project on Emerging Nanotech-
nologies at the Woodrow Wilson International Center for
Scholars had compiled a list of more than 500 consumer
products that claim to include some form of engineered
nanoparticles. Socks, paints, bandages, and food containers
incorporate nanosilver (n-Ag) to exploit its antimicrobial
properties. In clothing such as socks, n-Ag may restrict the
growth of odor-causing bacteria (2-10).
Despite the growing commercialization of n-Ag, little is
known about the environmental effects of widespread use
of products containing silver nanoparticles (11). Ionic silver
is highly toxic to aquatic organisms (12-14), and the United
States Environmental Protection Agency (USEPA) has set
water quality criteria values for silver in salt and fresh water
a secondary drinking water standard for silver of 100 ppb.
Toxicity and exposure data for nanoparticle silver, however,
is currently lacking (15-17). Studies have demonstrated the
toxicity of nanoparticle silver to bacteria (3, 5, 6, 8, 10),
suggesting that the antimicrobial effects of silver may be
to characterize (as colloidal or ionic) and quantify the silver
released from commercial products.
The ubiquitous use of commercial products containing
n-Ag could potentially compromise the health of many
ecosystems. For example, household washing of clothing
more than 70% of the U.S. population is served by public
enter a municipal wastewater treatment plant (WWTP). The
n-Ag present in sewage may partition onto wastewater
biomass and be removed at a WWTP, only to re-enter the
treatment biosolids. If n-Ag proves to be difficult to remove
in a wastewater treatment system, n-Ag remaining in the
treated effluent stream may enter surface water environ-
ments, potentially disrupting numerous biological ecosystems.
This paper investigates n-Ag release from commercial
clothing (specifically, socks) into water, as well as the form
quantified before determining the concentration and form
washings of the socks with distilled water. Batch adsorption
isotherm studies were conducted with wastewater biomass
and two sources of silver: (1) silver released from the socks
into the wash water (nanoparticle or ionic), and (2) reagent
model for wastewater treatment to evaluate the amount of
silver that could be present in the treated effluent or WWTP
Materials and Methods
Acid Digestions of Sock Fabric. Six brands of commercially
available socks were purchased (Table 1) based on the
of silver. A modified digestion method was used to quantify
An air-dry mass (100- 500 mg-dry) of each sock was
submerged in a solution of 5 mL of ultrapure reagent grade
nitric acid (6901-05, JT Baker, Phillipsburg, NJ) and 5 mL of
deionized water. After a watch glass was placed over the
digestion beaker, the solution was heated to approximately
100 °C and allowed to react. Nitric acid was added in 2 mL
hydrogen peroxide (HX0635-2, EMD Chemicals Inc., Gibb-
stown, NJ) was added to complete the digestion process.
Again the digestion beaker was heated to 100 °C, and
hydrogen peroxide was added in 1 mL aliquots until
effervescence was minimal, indicating completion of the
a glass fiber filter (Qualitative #2, Whatman) and diluted to
Washing of Socks in Water. Socks were placed in 1 L
Inc.). The bottles were agitated for either 24- or 1-h contact
times on an orbital shaker table at approximately 50 rpm.
The 24-h contact time was chosen to allow sufficient
opportunity for the socks to leach silver. The 1-h contact
cycle, though the quantities of leached silver from both
* Corresponding author phone: 480-965-3589; fax: 480-965-0557;
Environ. Sci. Technol. 2008, 42, 4133–4139
10.1021/es7032718 CCC: $40.75
Published on Web 04/09/2008
2008 American Chemical SocietyVOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY94133
contact time, the socks were removed, excess water was
wrung out into the glass bottle, and the socks were placed
in new bottles for the next washing. Each brand of sock was
washed at least 3 consecutive times for either 1 or 24 hours.
3, EMD, Gibbstown, NJ) or 10% HNO3, triple rinsed with
distilled water, and air-dried prior to use.
One brand of sock (1b) was washed with City of Tempe
tap water (conductance ∼1000 µmhos/cm) using the above
procedure and a single 24-h contact time. This was done for
comparison with socks washed in ultrapure water.
domestic wastewaters was not addressed in this research.
washing their clothes, the goal of this study was to obtain
data on the interaction of n-Ag from socks with distilled and
of detergents on the quantity and form of silver released
from socks into domestic wastewater streams.
Separation of Nanoparticle and Ionic Silver Species.
Three approaches were employed to separate the nanopar-
ticle form of silver from ionic silver and to quantify both
forms. First, washwater samples were centrifuged at 15 000
rpm (F ) 24 900g) for 20 min, but this procedure did not
in the washwater samples was size-separated using mem-
brane filters (Pall) of 0.4, 0.1, and 0.02 µm pore diameter in
either a 25 mm syringe filter or a 45 mm vacuum pump
clogging the smaller-pored filters. The 0.02 µm filter is the
smallest pore size commercially available for a syringe filter
between ionic and nanoparticle silver. Third, a silver ion
specific electrode (ISE) (Accumet Silver/Sulfide, Fisher) was
used in combination with a pH/mV meter (Φ 250 series,
Beckman) to measure free Ag+ions of the unfiltered 1-h
washes of socks 1b and 3.
Scanning and Transmission Electron Microscopy Analy-
ses. Scanning electron microscopy (FEI XL30 EFSEM with
EDX capabilities) and transmission electron microscopy
(JEOL JEM-2010F TEM/STEM with EDX capabilities) were
used to confirm the presence of silver nanoparticles in the
sock material and in the washwater samples, respectively.
The sock material was ashed at 550 °C in a programmable
muffle furnace (Fisher Scientific), then prepared on an SEM
stub. Two methods of stub preparation were used: (1) the
ashed sock material was lightly dusted onto the carbon tape
suspended in distilled water and subsequently evaporated
in droplets onto the carbon tape of the SEM stub. Energy-
dispersive X-ray analysis (EDX) was used to confirm the
elemental presence of silver in the electron micrographs.
to concentrate the nanoparticles, thus increasing the prob-
ability of identifying them on the TEM stub. Drops of the
concentrated sock washwater were then evaporated on the
of silver in the micrographs.
Adsorption Experiments with Wastewater Biomass.
of silver (nanoparticle and ionic, combined) from the
washwater by wastewater treatment system biomass (acti-
of silver-containing water solutions: the sock washwaters
and a reagent ionic silver solution. The latter was prepared
using a plasma standard solution of Ag+(1000 ppm Ag+, 5%
HNO3, Cat. no. PAGN-100, Manchester, NH) and distilled
and prepared in two ways. First, biomass was sampled from
a bench-scale reactor. This reactor was conditioned with
return activated sludge (RAS) from a local WWTP and
operated without wasting of biosolids (i.e., long sludge age).
Second, a fresh sample of biomass was collected from the
RAS line of a local wastewater treatment facility. A 5 mM
it into the batch isotherm experiments.
For each batch sample in the isotherm test, 40 mL of the
silver solution of interest was spiked with a dose of the
biomass stock solution, and the sample was diluted up to 50
mL with ultrapure water. The initial silver concentrations in
the batch experiments ranged from 60 to 500 ppb. Doses
between 0.5 and 6 mL (0.17-2.23 mg of dry biomass) of the
biomass stock solution were used in the samples.
All experiments were conducted within a pH range of
5.8-7.4. The pH values of the experiments using sock
solution experiment was adjusted from pH 3.5 to pH 6.1
were allowed to mix on a shaker table at 45 rpm for 1 h of
contact time, making these quasi-equilibrium experiments,
allowing ample time for adsorption to take place while
limiting the time for the biomass to produce additional
surface area for adsorption. Then they were filtered through
a 0.4 or 0.45 µm membrane filter (Pall) to remove any
suspended biosolids. The filtrate was then analyzed for total
silver by ICP-OES.
Quantification of Silver in Sock Material and SEM Con-
firmation of Nanoparticle Size. Acid digestion of the sock
material indicated that 5 of the 6 types of socks contained
detectable levels of silver ranging from 2 to 1360 µg-Ag/g-
sock (Table 1). SEM confirmed the presence of silver
nanoparticles in socks 1a, 3, and 4 (not shown). Figures 1
TABLE 1. Sock ID/Characterization and Silver Content
acid digestion analysis
mass of silver
per mass of sock
(µg Ag/g sock)
released after 4 24-h
Fox River (Xstatic)
Arctic Shield (E47)
lounge sock (green)
lounge sock (blue)
4134 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 11, 2008
socks 3 and 1a, respectively, after ashing. The carbon and
oxygen peaks of the EDX analyses can be attributed to the
surrounding residual sock material and/or the carbon tape
used for SEM stub preparation. Particles of elemental silver
with diameters of 100-500 nm exist within the three types
of socks. The silver particles in sock 1a do not appear nearly
spherical like those of sock 3, but are irregularly shaped and
FIGURE 1. SEM image of ashed sock 3 material showing spherical silver particles on the order of 100 nm in diameter. Inset: Represen-
tative EDX analysis of the spherical silver particles marked with white arrows.
FIGURE 2. SEM image of ashed sock 1a material showing agglomerated silver nanoparticles. Inset: Representative EDX analysis of
points within the mass confirms a majority of silver particles. The carbon peak is attributed to residual sock fabric and/or the
carbon tape used to mount the sample.
VOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4135
silver is 962 °C, the ashing of the sock material at 550 °C may
have sintered smaller n-Ag particles into larger diameter
spheres. However, this preparatory step was necessary to
remove the bulk of the sock fabric to obtain a clear image
of the n-Ag.
Release of Silver into Washwater. Three of the six sock
sample types (1a and 1b, 3, 6) detectably leached silver into
the ultrapure wash water (1- or 24-h contact time). Figure
sequential 24-h washes of the four sock samples. Silver was
whereas socks 1a and 6 had leached almost all of their silver
by the fourth wash. The silver content in the 500 mL
The 24-h wash simulations leached comparable (i.e., same
After three washes, two socks of 1b leached a total of 1245
A comparison of socks based on the amount of silver
leached relative to the silver content of the sock (Table 1)
processes of the socks control the amount of silver that is
released into the washwater. For example, socks 3 and 4
µg, respectively), yet released only small percentages (<1%)
of their total silver into the wash water, while socks 1a, 1b,
and 6 released nearly 100% of their silver content in four
into the washwater at different rates. Socks 1 and 6 released
These data would suggest that the various sock types have
as an antimicrobial agent. For example, because most of the
silver, if not all, contained in socks 1a and 1b is leached in
the first four washes, one might assume that these socks
would not perform as well as socks 3 and 4 at preventing
odor-causing bacteria growth.
A fresh sock 1b sample was washed once with City of
Tempe tap water (19) (conductance ∼1000 µmhos/cm) for
24 h to investigate the effect of water quality on the release
of silver from the sock. The potential of water to corrode
metals is related to many water quality parameters, but in
general, as buffering capacity and alkalinity increase, water
corrosivity decreases (20). After one wash, sock 1b had
µg of silver released into the ultrapure water (Figure 3). This
result may indicate that tap water is less aggressive than
ultrapure water at stripping silver from the sock fabric.
Therefore, these experiments in ultrapure water may be an
overestimate of the amount of silver that could be leached
solution could not be characterized as nanoparticle or ionic
because of interferences with the ISE and the probable
formation of silver salts during SEM/TEM sample preparation.
and EDX analysis of the colloids in the wash water of sock
1a (Figure 4) indicated the presence of silver material with
diameters from one to a few hundred nm. These particles
particles in the SEM image of the sock 1a material. Thus, at
least some of the n-Ag is released into the washwater as
nanoparticles; not just as dissolved ionic silver.
Table 2 summarizes the colloidal and ionic characteriza-
tion of the silver leached into the washwaters. Very little
as ionic by the ISE depending on the number of washings,
which suggests that 75-100 µg/L of n-Ag may be present in
of the silver initially released from socks is in the dissolved
The first wash of sock 3 released nearly all silver as
colloidal, as confirmed by the agreement of the filtration
and ISE data. Subsequent washes of sock 3 contained
water, the ionic silver in solution increased over time as
measured by ISE. This suggests that n-Ag is oxidized into a
dissolved ionic form when subjected to prolonged exposure
Partitioning of Ionic and Nanoparticle Silver to Waste-
water Biomass. Quasi-equilibrium batch adsorption iso-
solutions or reagent ionic Ag+solution. Silver partitioned
onto the biomass and was fit with the Freundlich isotherm
equation (q ) KC1/n), where C is the equilibrium silver
5). The values for the Freundlich adsorption capacity
parameter, K, ranged from 3.4 to 17 (µg-Ag/g-biomass)(L/
FIGURE 3. Cumulative mass of silver released from three sock
types (four socks total) into four consecutive 24-h washings in
TABLE 2. Nanoparticle and Ion Separation for Silver in 1-hr Washes via Filter Analysis and Ion Selective Electrode
percent of total silver
µm filtersock ID
total silver in
wash water (µg)
1b, first 1-hr wash
1b, second 1-hr wash
1b, third 1-hr wash
3, first 1-hr wash
3, second 1-hr wash
3, third 1-hr wash
4136 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 11, 2008
µg-Ag)1/n. The slopes of all of the isotherms are very similar,
yielding an average Freundlich adsorption intensity param-
eter, 1/n, of 0.66 (unitless).
Although the presence of colloidal silver was confirmed
in washwater samples, the isotherm experimental data
suggest that the silver leached from the socks adsorbs to
biomass in a manner similar to that of ionic silver. The
measurement (Table 2). Sock 3 leached low percentages of
Ag+in washes 1 and 2, but the isotherm experiment for this
sock was conducted with washwater that had been stored
for 4 weeks, possibly giving the colloidal silver time to
solubilize into ionic silver. Removal of Ag+from wastewater
can be attributed mainly to precipitation with chloride and
adsorption to biomass, which can be hindered by complex-
ation with dissolved organic matter (21).
Two wastewater biomass preparations were used for the
isotherm experiments. Figure 5 presents data for isotherms
conducted with biomass from a full-scale WWTP and a
conducted on wash water from socks 1b and 1a used full-
scale WWTP and laboratory-scale biomass, respectively, the
difference between the closed and open data points is most
difference in isotherm results could also be attributed to
interactions between dye from the socks and the leached
silver because socks 1b and 1a differ in color.
Although adsorption characteristics may change in real
sewage, etc.), this adsorption data can be used to estimate
the performance of a WWTP. The Freundlich isotherm
in wastewater biomass or in treated effluent. A steady-state
mass balance of a WWTP with nonlinear sorption of an
adsorbate can be expressed as
C ) C0-(XτKC1⁄n
where C is the effluent concentration of silver, C0 is the
adsorption parameters, and X, τ, and θ are operational
parameters of a WWTP (mixed liquor suspended solids,
hydraulic and solids retention time, respectively). The
common values used for these model parameters are
provided as Supporting Information. Figure 6 presents
FIGURE 4. TEM image of colloidal material from sock 1a washwater. Inset: EDX confirmation that the dark particles within the
circle are predominantly silver.
FIGURE 5. Batch adsorption isotherms for the wash solutions of
three socks (1b, 3, 1a) and an ionic silver solution (Ionic Ag).
Initial silver concentrations varied from 61 to 490 ppb, and pH
values ranged from 5.8 to 7.4.
VOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4137
sludge reactor using common WWTP design conditions
Using a common municipal WWTP influent silver con-
centration of 5 µg/L, the model results in an effluent silver
concentration of 0.01 µg/L, and the wasted biosolids silver
concentration is 2.8 mg-Ag/kg-biosolids. The effluent from
the wastewater treatment facility would exceed the USEPA
silver concentrations of about 2900 and 4250 ppb, respec-
tively. These influent concentrations are 3 orders of mag-
nitude higher than those commonly observed for municipal
WWTPs (21). The treated effluent would not exceed the 100
ppb secondary drinking water standard for silver until the
influent concentration reached approximately 45 400 ppb
(not shown graphically).
Based upon the existing water quality criteria for silver,
the model suggests that wastewater treatment plants are
capable of removing a much higher silver load from a
two WWTP operation parameters: the MLSS concentration
was varied from 2000 to 4000 mg/L and the ratio of θ:τ was
varied from 5 to 20. Using the 1.9 ppb salt water CMC as the
maximum allowable effluent concentration, a MLSS con-
centration of 2000 mg/L with an θ:τ of 20 would treat an
influent silver concentration of 1460 µg/L. Similarly, a MLSS
concentration of 4000 mg/L with an θ:τ of 5 would treat an
influent silver concentration of 11 600 µg/L (a table of the
model results can be found in the Supporting Information).
of some consideration. The model suggests at an influent
silver concentration of 180 µg/L, the silver concentration in
Leaching Procedure (TCLP) by the USEPA. An increase in
consumer use of n-Ag could therefore restrict municipal
as fertilizer for agricultural lands.
New analytical techniques that distinguish between nano-
materials (metal, metal-oxide, quantum dots, etc.) and
dissolved ionic species at relevant concentrations in envi-
ronmental matrices are important for the advancement of
nanotechnology. Methods such as capillary electrophoresis
(23) and size-exclusion chromatography (24) have the
potential to separate nanomaterials from ionic species, but
detection methods suited for environmental matrices and
to separate and quantify nanomaterials, exposure data can
be obtained during ecotoxicity studies to facilitate environ-
risks are currently preventing scientists and the public from
fully supporting the advancement of nanotechnology (25),
but new analytical techniques can answer these questions,
to all stakeholders.
This research was supported by the National Science
State University (Award Abstract 0504248) and the United
States Environmental Protection Agency (Grants RD831713,
Environment Research Foundation also provided support.
We gratefully acknowledge the use of facilities within the
Environmental Laboratory at Arizona State University. The
cooperation of the staff at the Mesa Wastewater Treatment
Facility is also much appreciated.
Supporting Information Available
WWTP model parameters and a table summarizing the
multiple simulations of the model. This material is available
free of charge via the Internet at http://pubs.acs.org.
(1) Woodrow Wilson International Center for Scholars. Nanotech-
nology Consumer Product Inventory; http://www.nanotech-
project.org/44; October, 2007.
Nanotechnol. 2007, 3 (2), 203–208.
S. H.; Park, Y. K.; Park, Y. H.; Hwang, C. Y.; Kim, Y. K.; Lee, Y. S.;
Jeong, D. H.; Cho, M. H. Antimicrobial effects of silver
nanoparticles. Nanomed.-Nanotechnol. Biol. Med. 2007, 3 (1),
P. K. H.; Chiu, J. F.; Che, C. M. Silver nanoparticles: partial
12 (4), 527–534.
(5) Morones, J. R.; Elechiguerra, J. L.; Camacho, A.; Holt, K.; Kouri,
J. B.; Ramirez, J. T.; Yacaman, M. J. The bactericidal effect of
silver nanoparticles. Nanotechnology 2005, 16 (10), 2346–2353.
(6) Panacek, A.; Kvitek, L.; Prucek, R.; Kolar, M.; Vecerova, R.;
nanoparticles: Synthesis, characterization, and their antibacte-
rial activity. J. Phys. Chem. B 2006, 110 (33), 16248–16253.
(7) Shahverdi, A. R.; Fakhimi, A.; Shahverdi, H. R.; Minaian, S.
Synthesis and effect of silver nanoparticles on the antibacterial
activity of different antibiotics against Staphylococcus aureus
and Escherichia coli. Nanomed.-Nanotechnol. Biol. Med. 2007,
3 (2), 168–171.
agent: a case study on E-coli as a model for Gram-negative
bacteria. J. Colloid Interface Sci. 2004, 275 (1), 177–182.
(9) Tian, J.; Wong, K. K. Y.; Ho, C. M.; Lok, C. N.; Yu, W. Y.; Che,
C. M.; Chiu, J. F.; Tam, P. K. H. Topical delivery of silver
nanoparticles promotes wound healing. Chemmedchem 2007,
2 (1), 129–136.
(10) Yoon, K. Y.; Byeon, J. H.; Park, J. H.; Hwang, J. Susceptibility
constants of Escherichia coli and Bacillus subtilis to silver and
copper nanoparticles. Sci. Total Environ. 2007, 373 (2-3), 572–
(11) Morris, J.; Willis, J. U.S. Environmental Protection Agency
Nanotechnology White Paper; U.S. Environmental Protection
Agency: Washington, DC, February, 2007.
FIGURE 6. Model results illustrating the removal of influent
silver for a typical WWTP (model parameters: Freundlich K and
1/n ) 9.0, 0.7; τ ) 0.5 d; θ ) 5 d; X ) 2000 mg/L). The silver
concentration in the treated effluent would exceed the USEPA
salt and fresh water Criteria Maximum Concentrations (CMCs)
at influent concentrations of 2900 and 4250 ppb, respectively.
The concentration of silver in the waste activated sludge flow
is represented by the dashed line and the secondary y-axis.
4138 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 11, 2008
(12) McGeer, J. C.; Playle, R. C.; Wood, C. M.; Galvez, F. A Download full-text
toxicity of waterborne silver to rainbow trout in freshwaters.
Environ. Sci. Technol. 2000, 34 (19), 4199–4207.
(13) Ratte, H. T. Bioaccumulation and toxicity of silver compounds:
A review. Environ. Toxicol. Chem. 1999, 18 (1), 89–108.
1981, 31 (124), 83–91.
Nanoscale Materials; National Nanotechnology Coordination
Office (NNCO): Arlington, VA, September, 2006.
(16) Davies, J. C. EPA and Nanotechnology: Oversight for the 21st
Century; Woodrow Wilson International Center for Scholars:
Washington, DC, 2007.
and the Environment: Report of a National Nanotechnology
Committee on Technology, and Subcommittee on Nanoscale
Science, Engineering, and Technology: Arlington, VA, May 8-
(18) U.S. Census Bureau. Statistical Abstract of the United States,
2000; U.S.Census Bureau: Washington, DC, 2000; p 234.
(19) City of Tempe Arizona Typical Water Quality Values; http://
(20) McNeill, L. S.; Edwards, M. Iron pipe corrosion in distribution
systems. J. Am. Water Works Assoc. 2001, 93 (7), 88.
(21) Wang, J.; Huang, C. P.; Pirestani, D. Interactions of silver with
wastewater constituents. Water Res. 2003, 37 (18), 4444–4452.
(22) Rittmann, B. E.; McCarty, P. L. Environmental Biotechnology:
Principles and Applications, 4th ed.; McGraw-Hill Higher
Education: New York, 2001.
the size/shape separation and optical properties of silver
nanoparticles by capillary electrophoresis. J. Chromatogr., A
2005, 1062 (1), 139–145.
(24) Al-Somali, A. M.; Krueger, K. M.; Falkner, J. C.; Colvin, V. L.
Recycling size exclusion chromatography for the analysis and
separation of nanocrystalline gold. Anal. Chem. 2004, 76 (19),
E.; Guston, D. H. Scientists worry about some risks more than
the public. Nat. Nanotechnol. 2007, 2, 732–734.
VOL. 42, NO. 11, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4139