DEQ-CFW_00000125

Perfluorinated Compounds in the
Cape Fear Drainage Basin in North
Carolina
SHOJI NAKAYAMA, MARK J. STRYNAR,
LAURENCE HELFANT, PETER EGEGHY,
XIBIAO YE, AND
ANDREW B. LINDSTROM*
National Exposure Research Laboratory, U.S, Environmental
Protection Agency, Research Triangle Park,
North Carolina 27711
Concern over perfluorinated organic compounds (PFCs),
e.g., perfluorooctane sulfonate (PFOS) and perfluorooctanoic
acid (PFOA), is due to a number of recent studies which
show that the PFCs are persistent, bioaccumulative, and
toxic in animals. Despite sustained interest in this topic,
little information is available concerning the environmental
distributions of the compounds. In this study, a new
method was developed for the analysis of 10 target PFCs
and its performance was examined in a systematic
evaluation of surface water in the Cape Fear River Basin
in North Carolina. One hundred samples from 80 different
locations were collected during the spring of 2006.
Detectable levels of the target PFCs were found in all
samples, and were comparable to values reported previously,
with maximum PFOS at 132 ng/L, PFOA at 287 ng/L,
perfluorononanoic acid (C9) at 194 ng/L, and perfluorohep-
tanoic acid (C7) at 329 ng/L. In general, the lowest
concentrations of the PFCs were found in the smallest
tributaries while the highest levels were found in middle
reaches of the Drainage Basin. Variability of PFC
concentrations suggests a series of source inputs
throughout the Basin. Seventeen sample sites (22%) had
PFOS concentrations greater than 43 ng/L, a conservative
safe water concentration estimated to be protective of
avian life. In addition, a total of 26 sites (32%) had PFOA
concentrations above 40 ng/L.
Introduction
Increasing worldwide attention is being focused on a group
of persistent organic compounds known as the perfluorinated
compounds (PFCs). Perfluor000tanoic acid (PFOA) and
perfluorooctane sulfonate (PFOS) are probably the two best
known PFCs, but there are a great number of other structurally
related compounds which share the unique physical and
chemical characteristics of this class of materials. Concern
over these compounds is in part due to a number of recent
studies that have indicated serious health effects associated
with PFOS and PFOA in various animal models (1, 2).
Consequently, this has led to voluntary cessation of the
production of PFOS in the United States and reductions in
factory emissions of PFOA and its residuals in finished
products (3). There are still many companies worldwide
which produce and/or use a wide range of different PFCs in
*Corresponding author phone: [phone]; fax: [phone];
e-mail: li ndstrom. andrew@epa. gov.
a great variety of products (4). While most residents tested
in the industrialized countries have detectable levels of many
PFCs in their blood (5), the routes of exposure and the
associated risks are largely unknown.
A series of studies in Japan has suggested a relationship
between PFOS and PFOA levels in water supplies and in the
blood of residents living in some of the most heavily
industrialized areas of that country (6, 7). Likewise, in the
United States, PFOA in human blood was found to be
correlated to the consumption of contaminated well water
and homegrown fruits and vegetables in one particularly
contaminated area (8). Other studies have documented that
the PFCs are ubiquitous in aquatic food webs and that they
tend to be concentrated in the fish that may be eaten by
humans (9, 10).
Although mounting evidence indicates the importance
of aquatic systems in the global transport of many of the
PFCs (11,12), there are still few data that have been published
describing PFC distributions in the aqueous environment.
In the small number of studies which have been published,
many aspects of the collection and analysis procedures are
poorly described. Few contain adequate detail on the
performance characteristics (i.e., precision and accuracy) of
the methods employed, making it difficult to interpret the
data. In a recent worldwide interlaboratory study (13), only
31% of the participating laboratories demonstrated sub-
stantial agreement in the analysis of an aqueous PFOS sample.
The need for more rigorous standardized testing procedures
for the PFCs will increase as our understanding of the issue
increases (14, 15).
This study was undertaken to establish a new rnethud fur
the collection and analysis of 10 PFCs in surface water and
to provide the details on the performance characteristics
that are needed for interpretation of the resulting data. This
method was applied in a pilot -scale evaluation of the Cape
Fear River Basin in North Carolina to demonstrate its utility
and to provide preliminary information about the PFCs in
this watershed,
Materials and Methods
Standards and Reagents. Potassium salts of perfluorobutane
sulfonate (PFBS, 98% purity), perfluorohexane sulfonate
(PFHS, 93%), and perfluorooctane sulfonate (PFOS, 93%) were
provided by 3M Company (St. Paul, MN). Perfluorohexanoic
acid (C6, 97%), perfluoroheptanoic acid (C7, 99%), perfluo-
rooctanoic acid (C8 or PFOA, 96%), perfluorononanoic acid
(C9, 97%), and perfluorodecanoic acid (C10, 98%) were
purchased from Sigma -Aldrich (St. Louis, MO). Perfluoroun-
decanoic acid (Cl 1, 96%), and perfluorododecanoic acid (C12,
96%) were purchased from Oakwood Products (West Co-
lumbia, SQ. 1802-Ammonium perfluorooctane sulfonate
(180-PFOS) was purchased from Research Triangle Institute
(Research Triangle Park, NC). 1,2-13C2-labeled PFOA (13C-
PFOA) was purchased from Perkin-Elmer Life and Analytical
Sciences (Boston, MA). Deionized (DI) water, HPLC-grade
methanol, and ammonium acetate were determined to be
free of PFCs prior to use.
Water Collection. One hundred water samples were
collected from 80 different sites (more than 10% dupli-
cates) in the Cape Fear River Basin in central North Carolina
on 6 different dates during the spring of 2006 (Figure 1).
Sample sites were subjectively selected to reflect water
quality throughout the Basin. The Cape Fear is the lar-
gest drainage basin within North Carolina with an area of
23 700 km2. The Haw and Deep Rivers originate in the
north central part of the state and their confluence
10.1021/es070792y Not subject to U.S. Copyright. Publ. xxxx Am. Chem. Soc. VOL, xx, NO. xx, xxxx / ENVIRON. SCI. & TECHNOL. ■ A
Published on Web 07/04/2007 PAGE EST: 5.8
DEQ-CFW 00000125

High
East coast of
United States
'SNOW Y o i k
f > Washington. DC
boro
Durham p
Haw fI r `
North Carolina
--
7
Deep River
Little River
v
f.l PFt IS, rig/L.
0 PFOS, ng/L
Little River 9.55..
3.36 tR 0
141,
2.19 40.3 J
26A
132
.84
Pope, Air {,vi,56.3
Fort Bragg Military Force Base .: C
Reservatlon Cape Fear Rlver
CAPE FEAR BASIN
i
Ca�eFear River
1 `>
4.10
U40,7 ,
0 1 �\1 W11mIr�at�
FIGURE 1. Cape Fear River Drainage Basin, North Carolina. The solid circles and triangles represent sampling locations on the main stream
and tributaries, respectively. Eleven numbered locations along the watershed show the sites with the highest total PFC concentrations
measured in this survey (See Table 3).
forms the Cape Fear River. The Little River joins the Cape
Fear River just north of Fayetteville. Local water author-
ities estimate that as many as 1.7 million residents obtain
drinking water from surface water resources within this
basin. While the watershed is principally rural and agricul-
tural in nature, possible sources of PFCs include use of
fire -fighting foams, metal -plating facilities, textile and
paper production, and other industries found within this
basin.
Samples were collected in pre -cleaned (methanol rinsed)
1 L high -density polyethylene (HDPE) bottles (Nalge Nunc
International, Rochester, NY) using either a Kemmerer
stainless steel sampler, an open water grab sampler (Wildlife
Supply Company, Buffalo, NY), or a homemade dip sampler.
All samples were collected approximately 15-30 cm below
the surface of the water. Samples were returned to the
laboratory and stored at room temperature for no longer
than 3 days prior to analysis.
Field Quality Assurance Sample Preparation. On each
sampling date, a 1 L bottle was filled with deionized (DI)
water and carried into the field as a field blank. Independent
quality control (QC) samples were spiked with known levels
B ■ ENVIRON. SCL & TECHNOL. / VOL. xx, NO. xx, xxxx
of the PFCs (typically at two different levels) and transported
to the field each day. These field blanks and QC samples
were returned to the laboratory and analyzed at the same
time as the field samples.
Solid -Phase Extraction (SPE). Oasis HLB Plus (225 mg)
cartridges (Waters Corporation, Milford, MA) were condi-
tioned with 10 ml, of methanol and DI water at a flow rate
of approximately 10 mL/min. Water samples were divided
into two 500 mL aliquots and spiked with 100 µL of a 100
pg/µL solution (10 ng) of the internal standards (18C-PFOA
and 180-PFOS), and then loaded onto the pre -conditioned
cartridges at a flow rate of 10 mL/min with a positive pressure
pump (Sep -Pak Concentrator, Waters Corporation). The
cartridge was then washed with 10 mL of DI water and dried
completely by purging with nitrogen gas. The target analytes
were eluted from the cartridge with 2 mL of methanol at a
flow rate of 1 mL/min. The eluate was reduced in volume to
500 µL with a TurboVap II nitrogen evaporator at 60 °C
(Caliper Life Sciences, Hopkinton, MA). Finally, a 200 µL
aliquot of this reduced eluatc was mixed with 50 µL of 2 mM
ammonium acetate to match the initial HPLC mobile phase
conditions.
DEQ-CFW 00000126

TABLE 1. Method Performance Characteristics
recoverya
precisionb
accuracy°
10 ng/L, N = 5
100 ng/L, N = 5
intra, ng/L, N = 6
inter, ng/L, N = 4
5 ng/L, N
= 4
50 ng/L, N
= 4
compound
% recovery
RSD°
% recovery RSD
mean
RSD
mean
RSD
% accuracy
RSD
% accuracy
RSD
C12
55.3
0.100
59.8 0.098
3.56
0.096
46.4
0,058
80.3
0.160
101
0.045
C11
66.9
0.086
78.4 0,100
27.6
0.051
49.6
0,044
90.3
0.069
103
0.038
C10
84.6
0,088
90.1 0.048
76.5
0.069
51.6
0.080
102
0,064
102
0.050
C9
92.4
0.035
97.4 0,018
147
0,041
52.7
0.043
103
0,120
105
0.039
C8
96.5
0.019
101 0.007
197
0.026
50.5
0.019
96.2
0,003
99.5
0.013
C7
91.1
0.024
104 0,027
60.3
0.047
48.4
0.034
96.2
0,043
98.4
0.012
C6
90.1
0.026
95.6 0.035
12.4
0.054
49.2
0.111
103
0.064
101
0.064
PFOS
95.1
0.033
96.5 0,018
30.9
0.031
48.9
0.048
92.0
0.048
94.6
0.011
PFHS
92.8
0,024
102 0.038
7.15
0,052
48.2
0.044
94.1
0,029
97.2
0.036
PFBS
83.2
0.028
93.0 0.054
2.50
0.041
48.2
0.108
95.3
0.051
96.3
0.092
a Matrix
matched recovery.
b Intra-day
variation of samples
from a
location and inter -day variation of spiked samples.
® Percent accuracy using
deionized water spiked with
target compounds (measured value/spiked amount). ° Relative standard
deviation.
Instrumental Analysis. Samples were analyzed using an
Agilent 1100 high-performance liquid chromatograph (HPLC,
Agilent Technology, Palo Alto, CA) coupled with an API 3000
triple quadrupole mass spectrometer (Applied Biosystems,
Foster City, CA).
The HPLC consisted of a membrane degasser, binary high-
pressure gradient pumps, an auto -sampler, and column
heaters. A Wakopak Fluofix-II 120E column (5 um), 3.0 x 100
mm (Wako Pure Chemical Industries, Osaka, Japan) con-
taining a fluorinated stationary phase was used for the
analysis at 40 °C. Samples (10 µL) were injected using a
gradient mobile phase consisting of mixture of 2 mM
ammonium acetate and methanol at a flow rate of 200 yL/
min. The gradient program was optimized for the separation
of all analytes and matrix interferences (Table S 1, Supporting
Information).
The API 3000 was operated in the electro-spray ionization
(ESI) mode using multiple reaction monitoring (MRM). The
operational parameters are described in Table S1 (Supporting
Information). Ionization and collision cell parameters were
optimized for each individual analyte. The MRM transitions
for each analyte are indicated in Table S2 (Supporting
Information).
Quantitation. Six -point calibration curves were produced
for each analytical batch by spiking blank DI water with
varying amounts of the target PFCs and fixed levels of the
two internal standards ('BO-PFOS and 13C-PFOA) such that
the quantifiable range for this study was from 1 to 500 ng/L.
Curves were prepared by plotting the area ratio of analytes
to internal standards versus the concentration of the PFC
standards. Quantitation was performed with the Analyst I AJ
software (Applied Biosystems) using a quadratic "1/x"
weighted regression fit with a coefficient of correlation greater
than 0.99.
In a separate series of experiments, the instrumental
quantitation limit (IQL) and lower limit of quantization
(LLOQ) of the method were calculated by using a series of
solvent standards and fortified DI water samples (0.01-250
pg/µL and 0.01-250 ng/L, respectively). The IQL was
determined to be 0.5 pg on column and the method LLOQ
was determined to be 0.2 ng/I, for all PFCs. At these levels
the signal-to-noise ratio was at least 10:1 with precision of
±15% and accuracy of 100% 1 20%.
Recoveries and QC Values. Recoveries were calculated
based on a matrix matched method (details in Supporting
Information). The intra-day precision was calculated based
on analysis of the 6 replicated samples collected from a single
location on a single day. The inter -day precision was
determined by comparing DI water spiked with 50 ng of PFC;
mixture on 4 different days. The relative standard deviations
(RSD), or coefficient of variation, were calculated from these
measurements. Accuracy was calculated by analyzing low
and high (5 and 50 ng/L) QC spikes into DI water on 4 different
days. The QC samples were treated with the same procedure
as the other samples and calibrated by the standard curves
described above.
Statistical Analyses. All statistical analyses were per-
formed using SAS/STAT software (SAS Institute, Cary, NC)
with the level of significance set at 0.05. If duplicate samples
were collected at any given location, mean values were used
in all subsequent analyses.
Results and Discussion
Method Validation. None of the field blanks were determined
to have measurable PFC contamination. The recovery,
precision, and accuracy for the 7 perfluorinated carboxylates
and 3 sulfonates targeted in this method are listed in
Table 1. Recoveries for most compounds ranged from 80%
to 104%. The C11 and C12 acids had lower recoveries
but excellent overall precision with RSDs no higher than
10%. The greatest variance was found for the inter -day C6
acid at a concentration of 50 ng/L (11% RSD). The accuracy
of spiked samples was 80-105% with less than 16
RSD. None of the duplicate samples taken from 8 different
locations on 5 different days had significant variation either
within -day or between -day analyses (p values > 0.1, paired
t-test).
Several aspects of this method provided enhanced ac-
curacy and precision. The use of a positive -pressure dual
piston pump, Sep -Pak Concentrator, allowed a relatively large
volume of water to be run through the SPE cartridge without
filtration at a steady flow rate and pressure in an automated
manner, thereby contributing to reproducible SPE loading
and overall sample consistency. Most surface water samples
were found to contain complex organic materials which
coeluted from the SPE cartridge and were present as
interferences in the final eluate. To minimize this interference,
a Fluofix-II column with a bonded fluorinated stationary
phase was used to separate these organic interferences from
the target PFCs (Figure S1, Supporting Information). The
combination of the HLB cartridge and the Fluofix analytical
column provided excellent accuracy and precision for the
measurement of PFCs in surface water.
Low levels of background contamination in the analytical
instrumentation will also contribute to low LOQs and
improved accuracy and precision. Some researchers have
suggested that all HPLC fittings and parts containing poly-
tetrafluoroethylene (PTFE) should be replaced with stainless
steel and/or polyetheretherketone (PEEK) materials to avoid
potential PFC contamination (16, 17). While this may be
necessary for ultralow-level determinations, any potential
VOL. xx, NO. xx, xxxx / ENVIRON. SCL & TECHNOL. ■ C
DEQ-CFW 00000127

TABLE 2. Summary of Measurements
mean
median
GMa
max.
min.
% above
compound
ng/L
ng/L
ng/L
ng/L
ng/L
LOW
% NO,
C12
2.17
1.95
1.93
4.46
< LOQ
19.0
53.2
C11
10.4
5.67
6.25
52.1
< LOQ
43.0
17.7
C10
22.1
13.2
8.35
120
< LOQ
62.0
15.2
C9
33.6
5.70
9.73
194
< LOQ
74.7
10.1
C8
43.4
12.6
16.2
287
< LOQ
82.3
7.6
C7
38.7
14.8
14.0
329
< LOQ
55.7
32.9
C6
7.38
5.14
5.41
23.0
< LOQ
44.3
45.6
PFOS
31.2
28.9
20.0
132
< LOQ
97.5
0
PFHS
7.29
5.66
5.73
35.1
< LOQ
73.4
1.3
PFBS
2.58
2.46
2.34
9.41
< LOQ
39.2
38.0
Geometric mean. b Limit of quantitation (1 ng/L, samples below
this level were excluded from the calculation of mean and GM). c Not
detected, less than 0.05 ng/L.
contamination was eliminated in this study by flushing the
entire HPLC system (degasser, pumps, tubing, and valves)
with 100% methanol for more than 3 days as well as by
avoiding injection of more than 1 ng of any specific PFC on
column at any time.
PFC Concentrations in Surface Water. The PFOS and
PFOA were found to be above the LOQ (1 ng/L for all
compounds) in 97.5%n and 82.3%n of the samples, respectively
(Table 2). Of the other compounds, the C7-10 acids and
PFHS were the most prevalent, being found above the LOQ
in more than 50%n of samples. The median concentrations
were all below30 ng/i, for each cnmpnund, with PFOS levels
at 28.9 ng/i,, PFOA at 12.6 ng/L, the C7 acid at 14.8 ng/L, and
the C10 acid at 13.2 ng/L. However, the peak levels of each
compound (see Table 3 and discussion below) were rela-
tively high when compared with previnusly ptthlished data,
For example, maximum PFOS was measured at 132 ng/L,
PFOA at 287 ng/L, the C10 acid at 120 ng/L, the C9 acid at
194 ng/L, and the C7 acid at 329 ng/L. As all data were found
to be log -normally distributed, Spearman's correlation
analysis was conducted indicating that the carboxylates were
strongly correlated with each other and that PFOS was
correlated with the C8 and C9 acids, PFHS, and PFBS (Table
S3, Supporting Information). A significant correlation was
also found between PFBS and the C6-C9 acids. These cor-
relations suggest common sources among these groupings.
Different PFC profiles were observed at each sampling
location along the entire length of the watershed (Figure S2,
Supporting Information). Figure 2 shows plots of the PFOA
and PFOS concentrations from the mouth of the Cape Fear
River up into the headwater tributaries more than 400 river
km inland. These plots reveal that lower concentrations of
the PFCs were found in the smallest upland tributaries and
the broad lowland costal sections of the river. The highest
250
E 200
0
8 150
8
0 100
a
m
U 50
0 100 200 300 400
River kilometer
s
80
3
ID
80 0
f
40 Q
0
rn
20 3
N
(8))F1
ia Flow
120Little
♦ Concentration
0
I,
80 m
100T
c
Haw River
�
8oAli
5I
3
50 w
eo�II
f
8
�7
Wildlife
plsclvorbeepRiver40
i
40
CapeII
t tJ
20
I i
20 3
100 200 300 400
River kilometer
FIGURE 2. Concentrations of PFOA (A) and PFOS (R) Its river km.
The Haw and Deep River join at river kilometer277 to form the Cape
Fear River. River kilometer 0 is the mouth of the Cape Fear River
near Wilmington, NC.
concentrations were found in the middle reaches of the Cape
Fear River and its two major tributaries (the Haw and Deep
Rivers). Figure 2 also shows the long-term median flow rate
of the main streams. Together these data indicate that the
highest concentrations and the greatest degree of variation
tend to occur in the low -volume, middle and upper reaches
of these rivers. Source inputs in these areas apparently have
a greater influence here than in the downstream costal areas
with substantially greater water volume. It is of great interest
to determine if this decline in PFC concentration is due to
dilution, biological uptake, or sequestration in sediments or
other abiotic pools.
Table 3 lists the measured concentrations of each PFC at
the eleven sampling locations (Figure 1) with the highest
aggregate (sum of all target compounds) PFC levels. The
maximum concentrations of the C8, 9, 10, 11, and 12 acids
TABLE 3. Measured Concentrations at the Eleven Sites with the Highest Total Concentrations of PFCs in the Cape Fear River
Basin"
(See Figure 1 for locations)
C12
C11
C10
C9
Ca
C7
C6
PFOS
PEAS
PFBS
no.
river
(ng/L)
(ng/L)
(ng/L)
(ng/L)
(ng/L)
(ng/L)
(ng/L)
(ng/L)
(ng/L)
(ng/L)
total(ng/L)
1
Haw River
4.46
52.1
120
194
287
118
21.7
127
8.43
9.41
942
2
Haw River
3.20
28.7
112
157
200
66.8
14.5
33.4
7.87
2.61
626
3
Haw River
3.29
27.6
109
157
191
59.2
13.7
36.4
9.49
3.04
609
4
Haw River
1.98
20.0
88.2
151
201
58.2
13.2
31.5
7.49
2.88
574
5
tributary to Cape Fear
2.26
15.0
19.6
71.2
58.6
329
23.0
30.0
3.36
ND
531
6
Haw River
1.18
8.87
31.0
72.1
152
58.3
13.5
31.2
7.70
ND
376
7
Cape Fear River
< LOQ
3.34
13.2
34.8
70.3
24.0
7.84
66.7
5.59
ND
227
8
Cape Fear River
1.14
6.39
17.2
35.7
71.5
26.9
9.35
50.4
4.82
ND
223
9
Cape Fear River
1.23
6.75
17.1
38.0
72.7
23.7
7.05
40.7
4.10
ND
211
10
Cape Fear River
< LOQ
7.55
19.3
31.2
46.8
13.9
4.62
56.3
6.84
2.12
189
11
Little River
< LOQ
< LOQ
2.17
2.24
12.6
3.38
3.23
132
26.4
3.20
185
Italicized values show maximal concentrations of each compound.
D ■ ENVIRON. SCL & TECHNOL. / VOL. xx, NO. xx, xxxx
DEQ-CFW 00000128

as well as PFBS were found at sampling point 1 in Figure 1.
Peak levels of PFOS and PFHS were found in the Little River
at sampling point 11. The highest levels of the C6 and C7
acids were found in a small tributary of the Cape Fear River
at sampling point 5. These data indicate the presence of
many different PFC sources within the Basin. Further
evaluation of these areas could be undertaken to identify the
various sources.
Comparison to Other Findings. In general, these results
are similar to PFOS and PFOA levels measured in 9 major
freshwater lakes and rivers throughout New York State U8).
In that study, median PFOS levels were all below 7 ng/I,
except Lake Onondaga (a listed Superfund site) where it was
found to be 756 ng/L. Median PFOA levels ranged from 14
to 49 ng/L with a high value of 173 ng/L. In the Cape Fear
River Basin, median PFOS was 28.9 ng/L with a maximum
of 132 ng/L, and median PFOA was 12.6 ng/Lwith a maximum
of 287 ng/L, One difference noted between these two studies
is that the New York State effort measured only PFOS, PFOA,
and PFHS with a 4 target compound method. In the Cape
Fear Basin, all 10 of the target compounds were routinely
quantified, with an average of 6 compounds being above
LOQ at each location.
Another study examined the impact of a fluorochemical
production facility on the Tennessee River in Alabama (19).
In that study, PFOS and PFOA levels remained below 55 and
25 ng/L, respectively, before the discharge site of the
fluorochemical plant. After a 10 km mixing distance down-
stream of the discharge, the PFOS and PFOA concentrations
remained fairly constant, averaging 114 ng/L and 394 ng/L,
respectively, for the remaining 55 km of the river that was
studied. The authors pointed out that this pattern was
consistent with a single source that influenced the main body
of the river for a considerable distance after the input. In
contrast, the current study revealed evidence of many
unidentified sources of PFCs in the Cape Fear Basin leading
to much greater overall variability in water concentrations
(Figures 2 and S2).
Comparing these results with a nationwide survey in Japan
(20, 21), the PFOS and PFOA levels from the present study
were at least 3.5-6 times higher than all of the Japanese
regions surveyed except the heavily industrialized area
around Osaka, where the peak levels of PFOS were found to
be 526 ng/ L and PFOA was as high as 67 000 ng/L. The authors
determined that the PFOA source was a water reclamation
facility which receives waste from a number of industrial
facilities operating in the area. The elevated PFOS concen-
trations were found in a tributary draining the Osaka
International Airport with the concentrations as high as 526
ng/L (roughly 500 times higher than typical background
concentrations in that study). The authors noted that use of
fire -fighting foams at airports has been known to cause PFC
contamination of ground and surface waters (22, 23) and
they speculate that this may be the source of contamination
here as well. In light of these findings, it is interesting to note
that the highest PFOS concentration measured in the current
study (132 ng/L) was from the Little River which runs along
the northern boundary of Fort Bragg and Pope Air Force
Base (Figure 1). The highest PFHS concentration (26.4 ng/L)
was also recorded at this location. In Figure 1 both com-
pounds increase to their maximum concentration as the Little
River flows along the northern boundary of this military
reservation and it makes its confluence with the Cape Fear
River. According to the NC Department of Environment (24),
the Base is permitted to pump 30 300 kL of wastewater per
day into the Little River in this area. This finding is consistent
with past or current use of PFOS-containing materials in this
area.
Another recent study measured PFCs in the Rhine River
and some of its tributaries in Germany (25), In general, the
median levels of PFOA and PFOS on the Cape Fear were
higher than most of the sampling locations on the main body
of the Rhine River. One exceptionally contaminated tributary
to the Rhine was identified in an area that had received surface
application of organic wastes containing PFC material.
Further testing of finished drinking water supplies coming
from this highly contaminated area showed little evidence
of effective removal of the PFCs by conventional activated
carbon filters. Like the Japanese work discussed above, this
study underscores the worldwide nature of this issue, and it
also shows how the systematic application of an effective
collection and analysis method can be used to trace and
identify PFC sources within a watershed.
Exposure Aspects. A U.S. EPA Great Lakes Initiative
methodology (26) was used to estimate an avian wildlife value
for PFOS of approximately43 ng/L (17). PFOS concentrations
below this level are estimated to be protective of trophic
level IV bird species which consume aquatic organisms at
equilibrium with PFOS in the water. Because of uncertainties
in the estimate, the authors (17) consider this value to be
"probably overly conservative, possibly by 50-100 fold." It
is interesting to note that 17 (22%) of the sampling sites in
this study had PFOS concentrations above this 43 ng/L
threshold (Figure 213). The New York State study (18) and a
recent Korean study (27) also found limited areas where this
threshold was exceeded.
While this study only measured surface water, a health -
based guidance level recommended by the State of New Jersey
for PFOA in drinking water provides a reference point for
interpretation of some of the data from the current study.
The State of New Jersey Department of Environmental
Protection has recommended that PFOA levels in drinking
water not exceed 40 ng/L in order to be protective of both
non -cancer effects and cancer at the one in one million risk
level (15), In the current study, 26 sites (32%) had PFOA levels
above 40 ng/L. While no drinking water measurements were
made in this study, these findings indicate the potential for
exposures above this threshold if PFOA is not effectively
removed by drinking water treatment plants using the Cape
Fear River and its tributaries as source water. The removal
of all the PFCs by water treatment processes should be
evaluated.
In conclusion, this method for 10 target PFCs in surface
water specifically identifies the key performance character-
istics (accuracy, precision, and sensitivity) that are needed
to design and conduct sampling surveys which will adequately
document surface water quality. This pilot study of the Cape
Fear Drainage Basin found ample evidence of potential
sources of PFCs, with PFOS and PFOA being the most
prevalent compounds identified. The CT C9, and C10 acids
and PFHS were also commonly detected, suggesting other
sources of these materials as well, In general, the indication
of a wide variety of PFC sources indicates that much further
work will be required to evaluate this river system and the
potential impacts on drinking water sources, wildlife species,
and potential human exposures. This study contributes to
the growing body of data that suggests that PFC contamina-
tion in the waterways of the industrialized world is pervasive
and as yet poorly characterized.
Acknowledgments
We thank Waters Corporation for their contribution through
Cooperative Research and Development Agreement (CRADA
#392-06) and Wako Pure Chemical Industries, Ltd, for their
contribution through the CRADA (#399-06), We also thank
the following individuals for contributing to this work: Dr.
Laurence I.ibelo, USEPA; Professor Akio Koizumi, Kyoto
University, Japan; Dr, Norimitsu Saito, the Research Institute
for Environmental Sciences and Public Health of Iwate
Prefecture, Japan; and Dr. JerryVarns, USEPA. This research
VOL. xx, NO. xx, xxxx / ENVIRON. SCI. & TECHNOL. ■ E
DEQ-CFW 00000129

was supported in part by an appointment to the Research
Participation Program at the National Exposure Research
Laboratory administered by the Oak Ridge Institute for
Science and Education through an interagency agreement
between the U.S. Department of Energy and the U.S.
Environmental Protection Agency. The United States Envi-
ronmental Protection Agency through its Office of Research
and Development funded and managed the research de-
scribed here. It has been subjected to Agency administrative
review and approved for publication. Mention of trade names
or commercial products does not constitute endorsement
or recommendation for use.
Supporting Information Available
Tables showing LC/MS/MS conditions, mass transitions of
each analyte, and Spearman's correlation coefficients be-
tween analytes. Additional Figures showing representative
chromatograms and summaries of all measurements made
at each location. This material is available free of charge via
the Internet at http://pubs.acs.org.
Literature Cited
(1) USEPA. Revised Draft, Hazard Assessment of Perfluorooctanoic
Acid and its Salts; U.S. Environmental Protection Agency:
Washington, DC, November 4, 2002.
(2) OECD. Hazard Assessment of Perfluorooctane Sulfonate (PFOS)
and its Salts; ENV/JM/ RD(2002)17/ FINAL; Organisation for
Economic Co-operation and Development: Paris, France, 2002.
(3) USEPA. 2010115 PFOA Stewardship Program; http://www.ep-
a.gov/oppt/pfoa/pubs/pfoastewardship.litm (December 20,
2006)
(4) Prevedouros, K.; Cousins, I. T.; Buck, R. C.; Korzeniowski, S. H.
Sources, fate and transport of perfluorocarboxylates. Environ.
Sci. Technol, 2006, 40, 32-44.
(5) Kannan, K.; Corsolini, S.; Falandysz, J.; Fillmann, G.; Kumar, K.
S.; Loganathan, B. G.; Mohd, M. A.; Olivero, J.; Van Wouwe, N.;
Yang, 1. U.; Aldoust, K. M. Perfluorooctanesulfonate and related
fluorochemicals in human blood from several countries. En-
viron. Sci. Technol. 2004, 38, 4489-4495.
(6) I-larada, K.; Saito, N.; Sasaki, K.; Inoue, K.; Koizumi, A. Per-
fluorooctane sulfonate contamination of drinking water in the
Tama River, Japan: Estimated effects on resident serum levels.
Bull. Environ, Contain. Toxicol. 2003, 71, 31-36.
(7) Harada, K.; Saito, N.; Inoue, K.; Yoshinaga, T.; Watanabe, T.;
Sasaki, S.; Kamiyama, S.; Koizumi, A. The influence of time, sex
and geographic factors on levels of petfluorooctane sulfonate
and perfluorooctanoate in human serum over the last 25 years.
J. Occup. Health 2004, 46, 141-147.
(8) Emmett, E. A.; Shofer, F. S.; Zheng, FL; Freeman, D.; Desai, C.;
Shaw, L. M. Community exposure to perfluorooctanoate:
Relationships between serum concentrations and exposure
sources. J. Occup. Environ, Med. 2006, 48, 759-770.
(9) Giesy, J. P.; Kannan, K. Global distribution of perfluorooctane
sulfonate in wildlife. Environ. Sci. Technol. 2001, 35, 1339-
1342.
(10) Martin, J. W.; Whittle, D. M.; Muir, D. C.; Mabury, S. A.
Perfluoroalkyl contaminants in a food web from Lake Ontario.
Environ. Sci. Technol. 2004, 38, 5379-5385.
(11) Simcik, M. F.; Dorweiler, K, J. Ratio of peril uorochemical
concentrations as a tracer of atmospheric deposition to surface
waters. Environ. Sci. Technol. 2005, 39, 8678-8683.
(12) Scott, B. F.; Spencer, C.; Mabury, S. A.; Muir, D. C. G. Poly and
perfluorinated carboxylates in North American precipitation.
Environ. Sci. Technol. 2006, 40, 7167-7174.
(13) van Leeuwen, S. P.; Karrman, A.; van Bavel, B.; de Boer, J.;
Lindstr6m, G. Struggle for quality in determination of perflu-
orinated contaminants in environmental and human samples.
Environ. Sci. Technol. 2006, 40, 7854-7860.
(14) USEPA. EPA —DuPont 2006 order on consent; http://www.ep-
a.gov/region03/enforcement/dupont_order.pdf.
(15) Post, G. Guidance for PFOA in Drinking Water at Pennsgrove
Water Supply Company (h ttp:llwww. state, nj. usldeplwatersupplyl
pfoa.htm); State of New Jersey, Department of Environmental
Protection: Trenton, NJ, February 13, 2007.
(16) Yamashita, N.; Kannan, K.; Taniyasu, S.; Horii, Y.; Petrick, G.;
Gamo, T. A global survey of perfluorinated acids in oceans.
Mar. Pollut. Bull, 2005, 51, 658-668.
(17) So, M. K.; Taniyasu, S.; Yamashita, N.; Giesy, J. P.; Zheng, J.;
Fang, Z.; Im, S. H.; Lam, P. K. Perfluorinated compounds in
coastal waters of Hong Kong, South China, and Korea. Environ.
Sci. Technol. 2004, 38, 4056-4063.
(18) Sinclair, E.; Mayack, D. T.; Roblee, K.; Yamashita, N.; Kannan,
K. Occurrence of perfluoroalkyl surfactants in water, fish, and
birds from New York State. Arch. Environ. Contain. Toxicol.
2006, 50, 398-410,
(19) Ilansen, K. I.; Johnson, 11. O.; Eldridge, J. S.; Butenhoff, J. L.;
Dick, L. A. Quantitative characterization of trace levels of PFOS
and PFOA in the Tennessee River. Environ. Sci. Technol. 2002,
36, 1681-1685.
(20) Saito, N.; Sasaki, K.; Nakatome, K.; Harada, K.; Yoshinaga, T.;
Koizumi,A. Perfluorooctane sulfonate concentrations in surface
water in Japan. Arch. Environ. Contain. Toxicol. 2003, 45, 149-
158.
(21) Saito, N.; Harada, K.; Inoue, K.; Sasaki, K.; Yoshinaga, T.; Koizumi,
A. Perfluorooctanoate and perfluorooctane sulfonate concen-
trations in surface water in Japan. J. Occup. Health 2004, 46,
49-59.
(22) Moody, C. A.; Field, J. Perfluorinated surfactants and the
environmental implications of their use in fire -fighting foams.
Environ, .Sci., Technol, 2000, 34, 3864-3870.
(23) Moody, C. A.; Hebert, G. N.; Strauss, S. H.; Field, J. A. Occurrence
and persistence of perfluorooctanesulfonate and other perflu-
orinated surfactants in groundwater at a fire -training area at
Wurtsmith Air Force Base, Michigan, USA. J. Environ. Monit.
2003, 5, 341-345.
(24) NCDENR. 2000 Cape Fear River Basin wide Water QualityPlan:
Appendix 1, NPDF.S Dischargers and Individual Stormwater
Permits in the Cape Fear River Basin; The State of North
Carolina: Raleigh, NC, 2000.
(25) Skutlarek, D.; Exner, M.; Farber, H. Perfluorinated surfactants
in surface and drinking waters. Environ. Sci, Pollut. Res. 2006,
13, 299-307.
(26) USEPA. Final water quality guidance for the Great Lakes system.
Fed. Regist. 1995, 60, 15366-15425.
(27) Rostkowski, P.; Yamashita, N.; So, 1. M.; Taniyasu, S.; Lam, P.
K.; Falandysz, J,; Lee, K. T.; Kim, S. K.; Khim, J. S.; Im, S. I-L;
Newsted, J. L.; Jones, P. D.; Kannan, K.; Giesy, J. P. Perfluorinated
compounds in streams of the Shihwa industrial zone and Lake
Shihwa, South Korea. Environ. Toxicol. Chem. 2006, 25, 2374-
2380.
Received for review April 3, 2007. Revised manuscript re-
ceived May 24, 2007. Accepted May 29, 2007,
ES070792Y
F ■ ENVIRON. SCI. & TECHNOL. / VOL. xx, NO. xx, xxxx PAGE EST: 5.8
DEQ-CFW 00000130