Inactivation of Bacteriophages via
Photosensitization of Fullerol
A P P A L A R A J U B A D I R E D D Y ,†
E R N E S T M . H O T Z E , *, ‡
S H A N K A R C H E L L A M ,† , §
P E D R O A L V A R E Z ,|A N D
M A R K R . W I E S N E R‡
Department of Civil and Environmental Engineering and
Department of Chemical and Biomolecular Engineering,
University of Houston, Houston, Texas 77204-4003,
Department of Civil and Environmental Engineering, Duke
University, Durham, North Carolina 27708-0287, and
Department of Civil and Environmental Engineering, Rice
University, Houston, Texas 77251-1892
The production of two reactive oxygen species through UV
photosensitization of polyhydroxylated fullerene (fullerol) is
shown to enhance viral inactivation rates. The production
of both singlet oxygen and superoxide by fullerol in the
presence of UV light is confirmed via two unique methods:
electron paramagnetic resonance and reduction of nitro
blue tetrazolium. These findings build on previous results
both in the area of fullerene photosensitization and in the
area of fullerene impact on microfauna. Results showed
doubled due to the presence of singlet oxygen and
increased by 125% due to singlet oxygen and superoxide
as compared to UVA illumination alone. When fullerol and
NADH are present in solution, dark inactivation of
viruses occurs at nearly the same rate as that produced
by UVA illumination without nanoparticles. These results
suggest a potential for fullerenes to impact virus populations
in both natural and engineered systems ranging from
surface waters to disinfection technologies for water and
Several fullerenes have been shown to be photosensitizers,
producing reactive oxygen species (ROS) such as singlet
oxygen (1O2) and superoxide (O2•-) (1-5). Data from the
medical literature detail the ability of C60and C70fullerenes
to cleave DNA and inactivate viruses and bacteria and kill
tumor cells (6-10). We have previously suggested that the
the basis for developing processes for oxidation and disin-
fection in water treatment (5). In addition, the implications
of possible ROS generation by fullerene nanoparticles on
require further study (11). In this paper, we describe
(fullerol) and the subsequent inactivation of waterborne
bacterial viruses. This inactivation is proposed to occur by
superoxide and singlet oxygen, respectively (Figure 1) (12).
The MS2 bacteriophage was chosen as a target for
photosensitization for two reasons: the phage is principally
and it is similar in morphology to the hepatitis A virus and
(315-400 nm) irradiation alone when a fullerol suspension
is present and shows that these nanoparticles could poten-
consequences and engineered water treatment. Moreover,
we show that viral inactivation effectively serves as a probe
for the presence of different ROS generated under both
photocatalytic conditions and by the transfer of electrons
from an appropriate donor.
Materials and Methods
Culture and Analysis of MS2 Phage. The double-top agar
as the host. E. coli was first cultured for 18-24 h in a tryptic
soy broth (TSB; Difco, Detroit, MI) and later transferred to
fresh TSB and grown to a mid-log phase for 3-6 h at 37 °C.
Stock MS2 was diluted in phosphate-buffered saline (PBS,
composition 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4‚
of 106PFU/mL. Next, the E. coli suspension (0.9 mL) and
phage dilution (0.1 mL) were mixed in 3 mL of soft overlay
agar and poured onto presolidified trypticase soy agar (TSA,
was then added to the Petri dishes, which resulted in
solution containing the MS2 phages was decanted and
centrifuged at 5000g for 10 min to remove bacterial debris.
by filtering the supernatant through a track-etched ultra-
Pleasanton, CA) and then ultra-centrifuging the filtrate
(Beckman LB-70) at 103 000g for 3 h. The phage pellet was
resuspended in PBS and stored at 4 °C. Fresh phage stock
was prepared in this manner prior to each experiment.
MS2 samples from experiments were cultured using the
method described previously. Following serial dilutions,
plates were prepared in triplicate, and only plaques in the
colony counter (Bel-Art Products, Pequannock, NJ). Control
* Corresponding author phone: (919)660-5030; fax: 919-660-
5219; e-mail: email@example.com.
†Department of Civil and Environmental Engineering, Univer-
sity of Houston.
§Department of Chemical and Biomolecular Engineering, Uni-
versity of Houston.
FIGURE 1. Proposed pathways of fullerol photosensitization via
type I (right side of dotted line) and type II (left side of dotted line)
resulting in superoxide and singlet oxygen, respectively.
Environ. Sci. Technol. 2007, 41, 6627-6632
10.1021/es0708215 CCC: $37.00
Published on Web 08/15/2007
2007 American Chemical Society VOL. 41, NO. 18, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY96627
organism (E. coli) did not vary during the time scale of our
experiments (see Supporting Information Table S1).
Chemicals. Superoxide dismutase (bovine erythrocytes)
(SOD), 5,5-dimethyl-1-pyrroline-1-oxide (DMPO), 2,2,6,6-
tetramethyl-4-piperidinol (4-oxo-TEMPO), adenosine 5′-
(trihydrogen diphosphate) (NADH), 2,2,6,6-tetramethyl-4-
piperidone (TEMP), nitro blue tetrazolium chloride (NBT),
MO). Fullerol (C60(OH)22-24), was purchased from MER
(Tucson, AZ). Deuterium oxide (D2O) was purchased from
Cambridge Isotope Laboratories (Andover, MA). Ultrapure
organic carbon concentration <3 µg/L was autoclaved prior
to use in all experiments. All glassware was washed with
distilled water and autoclaved for 15 min.
adding approximately 0.07 mg/mL powdered fullerol to
ultrapure water. This solution was made up in a volumetric
flask and placed in a sonication bath for 2 h. During this
time, the suspension became gradually more gold as more
fullerol was suspended. Once removed from sonication, the
suspension was filtered through 0.45 µm mixed cellulose
easter (MCE) filters via vacuum to remove unsuspended
particles. The final suspension of the fullerol aggregates had
a number mean diameter of 218 nm and was found to be
stable over a period of at least 3 months as determined by
dynamic light scattering. Characteristics of the suspension
have been previously published (22).
Irradiation and Inactivation Protocols. All experiments
requiring irradiation were performed in the presence of two
15 W fluorescent ultraviolet bulbs (Philips TLD 15W/08).
These bulbs had an output spectrum ranging from 310 to
400 nm and a total irradiance of 24.1 mW/m2with a peak at
365 nm in the UVA region as measured using a Li-Cor 1800
spectroradiometer (See Supporting Information Figure S1).
Two borosilicate glass vials with PTFE-lined caps were
MS2 phage in PBS (pH 7.3 ( 0.1) and other necessary
was exposed to UV irradiation, and the other was double-
wrapped in aluminum foil, serving as a control sample. A
reducing agent (5 mM NADH) was used to promote super-
oxide (O2•-) formation,?-carotene (26µM) was employed as
a singlet oxygen (1O2) scavenger, and SOD (10 U/mL) was
employed to scavenge O2•-, depending on the specific
experimental conditions. Viruses were sampled following 0,
the plaque assay technique. Data points and error bars in
and standard deviation of results from two or three separate
experiments conducted under identical conditions on dif-
ferent dates over the duration of this study.
ROS Measurement Instrumentation. EPR spectra were
recorded at room temperature with a Varian E-6 spectrom-
eter. The conditions for all measurements were as follows:
frequency, 9.27 GHz; power, 5 mW; modulation amplitude,
4 G; and modulation frequency, 100 kHz. UV-vis spectrom-
etry was preformed using a Hitachi U-2000 spectrophotom-
EPR Procedure. The spin-trapping reagent 4-oxo-TEMP
limit and a relatively long adduct lifetime when compared
to other spin traps, making it an ideal trap for suspensions
containing low concentrations of singlet oxygen that must
a mixture of 40 µM fullerol suspension and 80 mM TEMP
was shaken in a 5 mL volumetric flask. This was poured into
removed using a capillary tube, capped with clay, placed in
7.3. In each case, the EPR parameters were held constant as
was the TEMP concentration. Separate samples were irradi-
standard product of TEMP and singlet oxygen (TEMPO) to
determine the singlet oxygen generation rate (6).
Determination of Superoxide Concentrations by NBT
Reduction. NBT reduction was employed to measure the
production of superoxide (24). The reduction of NBT results
in an increase in optical density that can be used to quantify
with and without a quencher for superoxide, for which SOD
allows non-superoxide related NBT reactions to be ac-
was kept free of SOD, and the other contained 10 U/mL
vials in 5 mL portions and capped with Teflon-lined septa.
+ UV + SOD, suspension + DARK, and suspension + DARK
+ SOD. The former two provide a measure of superoxide
of superoxide activity in the dark. Light samples were
irradiated for up to 60 min. A sample was taken via syringe
of several hours while being stored in the refrigerator. All
containing samples was taken to be proportional to the
Results and Discussion
the Supporting Information. In the absence of UV light,
fullerol or controls with any of the added chemicals did not
inactivate the MS2 host (E. coli) (Supporting Information
Table S1), validating the plaque assay technique employed
herein. Furthermore, none of the various chemical agents
was observed in the absence of fullerol. However, MS2 was
inactivated by photosensitized fullerol.
Inactivation with Fullerol in the Presence of UV Light.
Figure 2 depicts MS2 inactivation under three scenarios: in
the presence of UV radiation alone, UV sensitized fullerol,
and fullerol in the dark. In all cases, inactivation could be
photoactivated fullerol. The same concentration of fullerol does
not have an effect in the dark, while the rate of inactivation nearly
doubles in the presence of UV and fullerol rather than UV alone.
are shown in all the graphs.
66289ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 18, 2007
to be 0.038 ( 0.004 min-1. The inactivation rate increased
significantly to 0.071 ( 0.004 min-1when fullerol was added
and the contents were exposed to UVA light. After 60 min of
UV irradiation, the presence of fullerol increased MS2
negligible inactivation of MS2, with a disinfection rate
constant that is statistically indistinguishable from zero.
Enhanced inactivation of MS2 by UV in the presence of
fullerol would therefore appear to be linked, in this case,
with the ability of fullerol to act as a photosensitizer. Next,
viral inactivation experiments and evaluate the conditions
that favor the production of various forms of ROS.
Previous work has shown that fullerol suspensions can
produce ROS under conditions of acidic pH (5). Enhanced
in Figure 2 suggests that ROS was produced by fullerol at a
the ROS production and speciation as either singlet oxygen
or superoxide depending on the conditions of illumination
and the presence of an electron donor.
Fullerol UV Irradiation Produces Singlet Oxygen. The
are superoxide and singlet oxygen. Both species can be
of PBS, 40 µM fullerol, and MS2 virus, the triplet EPR signal
grows after being irradiated for 1, 2, and 3 h. The constant
When either UV light or fullerol is absent, no signal is
generated. These observations are consistent with the fact
that there was no viral inactivation with fullerol in the dark
in viral inactivation by fullerol.
spin trap for the detection of superoxide. No signal was
generated in the presence of DMPO, an indication of
negligible superoxide activity (data not shown).
Viral Sensitivity to Singlet Oxygen. Additional viral
?-carotene and SOD (quenchers for singlet oxygen and
superoxide, respectively) to further determine the ROS
responsible for inactivation. Results from experiments con-
ducted with fullerol in the presence of a singlet oxygen
scavenger, ?-carotene, are depicted in Figure 4. MS2 viruses
were inactivated at a rate of 0.033 ( 0.002 min-1when
from the kinetics of inactivation obtained when the viruses
were exposed to UV alone (0.038 ( 0.004 min-1in Figure 2).
to that observed in Figure 2. Hence, additional MS2 inac-
to result predominantly from singlet oxygen-mediated pho-
tooxidative stress. Negligible inactivation of MS2 was found
when a suspension containing only ?-carotene and NADH
was irradiated (Table S2).
Inactivation experiments conducted in the presence of
SOD (Figure 5) yielded MS2 inactivation rates that were
in the presence or absence (Figure 2) of SOD (0.071 ( 0.002
min-1). Thus, SOD addition did not influence MS2 inactiva-
tion kinetics, suggesting that photoactivated fullerol nano-
irradiated for 1, 2, and 3 h. Triplet signal amplitude (black lines)
can be compared with a 0.5 µM TEMPO standard (dotted lines) to
find a detectible singlet oxygen generation rate.
FIGURE 4. MS2 inactivation with photoactivated fullerol (1 µM) in
the presence of ?-carotene (26 µM). Inactivation is similar to that
of UV alone (Figure 1).
FIGURE 5. MS2 inactivation with photoactivated fullerol in the
presence of SOD (10 U/mL). Inactivation is statistically indistin-
guishable to a similar suspension without SOD (Figure 2). (See
Supporting Information for rate of inactivation of MS2 with SOD
VOL. 41, NO. 18, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY96629
particles either generated inconsequential amounts of su-
peroxide or that superoxide contributed negligibly to MS2
inactivation. The latter explanation can be ruled out as
superoxide is known to inactivate MS2 viruses (25). It is
was negligible in these experiments and that the singlet
primary agent of enhanced viral inactivation. Negligible
only SOD was irradiated (Table S2).
in the production of superoxide require an appropriate
electron donor to be present. NADH was chosen as an
electron donor for two reasons: it is a biologically relevant
donor found in the cell and had been used previously to
of NADH increased the rate of MS2 disinfection by UV
sensitized fullerol to 0.086 ( 0.005 min-1(Figure 6) from
0.071 ( 0.004 min-1(Figure 2). These results suggest that in
the presence of an electron donor (NADH), both singlet
oxygen and superoxide radicals (generated simultaneously
by type II and type I photosensitization) likely contributed
by fullerol and NADH together in the dark. The MS2
inactivation rate under these conditions is coincidentally
similar to UVA alone (Figure 2) and is likely caused by
superoxide production by fullerol in the presence of NADH.
The high electron affinity of fullerenes suggests that ROS
found when a suspension containing only NADH and
added in these experiments, NADH was observed to reduce
the UV dose in this system. At the peak of irradiance (Figure
S1), UV light transmission was reduced by over 50% when
were inactivated as a result. Therefore, the effect of the
NADH-fullerol mixture may be even greater than that
suggested in Figure 6 since the observed increase in
on the MS2 response. Despite this transmission reduction,
on virus inactivation (Table S2).
The production of superoxide via type II photosensitization
(Figure 1) was evaluated in both the UV irradiated and the
non-irradiated conditions utilizing an NBT absorbance
measurement. When suspensions of fullerol were irradiated
as was measured by a change in the optical density at 560
nm (Table 1). Redox conditions favor the transfer of an
of superoxide by NADH was measured, but because the rate
was 2.3 times higher in the presence of fullerol and no
(Table S2), viral inactivation may be attributed mainly to
Consistently, no superoxide was produced by the fullerol
suspension under UV irradiation in the absence of NADH
(Table 1), consistent with a type II reaction (Figure 1)
producing singlet oxygen as the main source of viral
inactivation (Figure 2).
Fullerol Generates Superoxide in the Dark. The same
NBT reaction was used to probe superoxide production in
samples without irradiation. The values of baseline differ-
ences in optical densities in SOD-containing solutions have
activity alone. Fullerol was not observed to produce super-
oxide in the dark when it was the only compound present
in solution. In the dark, NADH alone has a very minimal
superoxide producing power and consequently is only
capable of limited viral inactivation (Table S2). However,
when the suspension contained both NADH and fullerol,
significant superoxide was produced in the dark (Table 1).
We note, however, that estimated redox conditions do not
seem to favor the transfer of an electron from NADH
(approximately -0.32 V vs SHE) to ground state fullerol
(approximately -0.602 V to -0.709 V) (2), suggesting that
other solutes may play a role in mediating this reaction.
Quenchers. We further tested the hypothesis that the
fullerol-NADH combination inactivated viruses by super-
oxide production in the dark by adding SOD to quench any
of the SOD, a negligible virus inactivation in the dark was
produced by the fullerol-NADH mixture (Figure 7), as is
consistent with SOD effectively quenching the superoxide
radicals. A photoactivated mixture of fullerol nanoparticles,
NADH, and SOD gave lower virus inactivation rates (0.057
( 0.004 min-1in Figure 7) as compared to a mixture of
lower rate of inactivation produced by the photoactivated
mixture of fullerol nanoparticles, NADH, and SOD (0.057 (
0.004 min-1in Figure 7) as compared to either the photo-
activated fullerol alone (0.071 ( 0.004 min-1in Figure 2) or
the photoactivated fullerol with SOD (0.071 ( 0.002 min-1
for rate of inactivation of MS2 with NADH alone.)
TABLE 1. Optical Density at 560 nm for NBT Reduced by
∆OD (560 nm)
NADH + fullerol
NADH + fullerol
NADH + fullerol
NADH + fullerol
dark, 4 °C
dark, 4 °C
dark, 4 °C
dark, 4 °C
dark, 4 °C
aConcentrations were 40µM fullerol, 625µM NBT, and 5 mM NADH
in PBS. Vials were measured after being kept under LP UV exposure
or in the dark at 4 °C over the time period indicated. Rate of superoxide
production by fullerol/NADH was at least twice that of NADH alone in
66309ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 18, 2007
oxygen in the photosensitization reaction with the triplet
state of fullerol resulting in lower levels of singlet oxygen
production (Figure 1) or that NADH reduced transmission.
In addition, viral inactivation rates for a UV illuminated
system with fullerol, NADH, a superoxide promoter, and
?-carotene, a singlet oxygen quencher (0.060 ( 0.002 min-1
in Figure 8), were greater than those observed for UV alone
via type I photosensitization of fullerol as the likely inactiva-
tion agents for MS2 as it reduces the likelihood that
superoxide is only generated from the direct transfer of
electrons from NADH to singlet oxygen (26). If this were the
case, UV plots in Figures 4 and 8 would match. Also, as
alone (fullerol and NADH in the dark, 0.041 ( 0.002 min-1
in Figure 6), indicating possible crossover quenching of
superoxide radicals by ?-carotene. As suspected from the
photosensitization theory, this further supports the hypoth-
esis that singlet oxygen is not being produced by fullerol in
a suspension containing only ?-carotene and NADH was
irradiated (Table S2).
These observations of ROS generation and subsequent
both possible impacts on ecosystems and new water treat-
ment technologies. Numerous hurdles must be addressed
before these materials can be used in water treatment
including the development of methods for immobilization
or separation of nanomaterials, evaluation of long-term
efficacy, and cost. Further investigations will be needed to
determine if low levels of UV light incident upon a lake or
river may result in ROS generation that may impact micro-
before an assessment of the impact of these on natural
systems can be performed.
This work was funded in part by the EPA-STAR Grant
91650901-0, a NSF Nano Exploratory Research Grant (BES-
0508207), the Partnership in International Research and
of the NSF, a NSF CAREER award to S.C. (BES 0134301), and
the EPA STAR program. The authors thank Fabian Marian
of the Biochemistry Department at Rice University for his
assistance with EPR spectra acquisition and analysis.
Supporting Information Available
Irradiance of low-pressure UV utilized in this study (Figure
suspensions alone and fullerol with NADH, SOD, or ?-car-
otene (Table S1). Inactivation rates of MS2 phage when
and ?-carotene (Table S2). This material is available free of
charge via the Internet at http://pubs.acs.org.
(1) Arbogast, J. W.; Darmanyan, A. P.; Foote, C. S.; Rubin, Y.;
Diederich, F. N.; Alvarez, M. M.; Anz, S. J.; Whetten, R. L.
(2) Mohan, H.; Palit, D. K.; Mittal, J. P.; Chiang, L. Y.; Asmus, K. D.;
Guldi, D. M. Excited states and electron transfer reactions of
(3) Hamano, T.; Okuda, K.; Mashino, T.; Hirobe, M.; Arakane, K.;
fullerene derivatives: Effect of sequential functionalization of
the fullerene core. Chem. Commun. 1997, 21-22.
(4) Vileno, B.; Marcoux, P. R.; Lekka, M.; Sienkiewicz, A.; Feher, I.;
derivatized C60fullerol. Adv. Funct. Mater. 2006, 16, 120-128.
(5) Pickering, K. D.; Wiesner, M. R. Fullerol-sensitized production
of reactive oxygen species in aqueous solution. Environ. Sci.
Technol. 2005, 39, 1359-1365.
(6) Yamakoshi, Y.; Umezawa, N.; Ryu, A.; Arakane, K.; Miyata, N.;
Goda, Y.; Masumizu, T.; Nagano, T. Active oxygen species
generated from photoexcited fullerene (C60) as potential medi-
cines: O2•-versus1O2. J. Am. Chem. Soc. 2003, 125, 12803-
(7) Kasermann, F.; Kempf, C. Photodynamic inactivation of en-
veloped viruses by buckminsterfullerene 1.Antiviral Res. 1997,
(8) Mashino, T.; Okuda, K.; Hirota, T.; Hirobe, M.; Nagano, T.;
and inhibition mechanism. Biorg. Med. Chem. Lett. 1999, 9,
(9) Tabata, Y.; Murakami, Y.; Ikada, Y. Antitumor effect of poly-
(10) Tsao, N.; Luh, T.; Chou, C.; Wu, J.; Lin, Y.; Lei, K. Inhibition of
Agents Chemother. 2001, 45, 1788-1793.
(11) Wiesner, M. R; Lowry, G. V.; Alvarez, P.; Dionysiou, D.; Biswas,
Sci. Technol. 2006, 40 (14), 4336-4345.
FIGURE 7. Effects of SOD (10 U/mL) addition to a suspension of
fullerol (1 µM) and NADH (5 mM) on MS2 inactivation.
FIGURE 8. Effects of ?-carotene (26 µM) addition to a suspension
of fullerol and NADH (5 mM) on MS2 inactivation.
VOL. 41, NO. 18, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY96631
(12) Foote, C. S. Definition of type-I and type-II photosensitized Download full-text
Oxidation. Photochem. Photobiol. 1991, 54, 659-659.
(13) Ryan, J. N.; Harvey, R. W.; Metge, D.; Elimelech, M.; Navigato,
sand. Environ. Sci. Technol. 2002, 36, 2403-2413.
(14) Jin, Y.; Yates, M. V.; Thompson, S. S.; Jury, W. A. Sorption of
viruses during flow through saturated sand columns. Environ.
Sci. Technol. 1997, 31, 548-555.
adsorption and transport through sandy soils. Appl. Environ.
Microbiol. 1998, 64, 405-410.
(16) Gerba,C.P.;Bitton,G.Microbialpollutants: Theirsurvivaland
transport pattern to groundwater. In Groundwater Pollution
Microbiology; Bitton, G., Gerba, C. P., Eds.; John Wiley and
Sons: New York, 1984; pp 65-88.
Assoc. 2004, 96, 75-87.
(18) Kohn, T.; Nelson, K. L. Sunlight-mediated inactivation of MS2
in natural waters. Environ. Sci. Technol. 2007, 41, 192-197.
(19) Mi, B.; Marinas, B. J.; Curl, J.; Sethi, S.; Crozes, G. Microbial
integrity. Environ. Sci. Technol. 2005, 39, 4270-4279.
(20) Adams,M.H.Bacteriophages;IntersciencePublishers,Inc.: New
(21) Meng, Q. S.; Gerba, C. P. Comparative inactivation of enteric
adenoviruses, poliovirus, and coliphages by ultraviolet irradia-
tion. Water Res. 1996, 30, 2665-2668.
(22) Xia, T.; Kovochich, M.; Brant, J.; Hotze, M.; Sempf, J.; Oberley,
T.; Sioutas, C.; Yeh, J. I.; Wiesner, M. R.; Nel, A. E. Comparison
of the abilities of ambient and manufactured nanoparticles to
Nano Lett. 2006, 6, 1794-1807.
singlet oxygen production. Nature (London, U.K.) 1976, 263,
(24) Bielski, B. H. J.; Shiue, G. G.; Bajuk, S. Reduction of nitro blue
tetrazolium by CO2-and O2-radicals. J. Phys. Chem. 1980, 84,
(25) Cho, M.; Chung, H.; Choi, W.; Yoon, J. Different inactivation
behaviors of MS-2 phage and Escherichia coli in TiO2photo-
catalytic disinfection. Appl. Environ. Microbiol. 2005, 71, 270-
(26) Sparrow, J. R.; Vollmer-Snarr, H. R.; Zhou, J. L.; Jang, Y. P.;
Jockusch, S.; Itagaki, Y.; Nakanishi, K. A2E-epoxide damaged
DNA in retinal pigment epithelial cells: Vitamin E and other
Received for review April 6, 2007. Revised manuscript re-
ceived June 15, 2007. Accepted July 9, 2007.
66329ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 18, 2007