ISSN 10619348, Journal of Analytical Chemistry, 2011, Vol. 66, No. 8, pp. 756–762. © Pleiades Publishing, Ltd., 2011.

ARTICLES

Uncertainty in the Quantification of Pentachlorophenol in Wood

Processing Wastewaters by SPMEGCMS1
I. Brása, N. Ratolab, and A. Alvesb
a
Escola Superior de Tecnologia de Viseu, Instituto Politécnico de Viseu, Repeses, Viseu, 3505510 Portugal
b
LEPAE, Departamento de Engenharia Química, Faculdade de Engenharia da Universidade do Porto Rua Dr. Roberto Frias,
Porto, 4200465 Portugal
Received April 2, 2010; in final form, December 30, 2010

Abstract—Pentachlorophenol is a trace contaminant with toxicological effects in environment. When deal

ing with the analysis and quantification of trace amounts in complex matrices, procedures such as extrac
tion/preconcentration are often needed and chromatographic determination are often operated close to the
limit of detection, which contribute to the increase of the uncertainty of the measurement. This matter is of
crucial importance when the results obtained approach the legal limits and is frequently ignored in scientific
work. The aim of this study was the estimation of the global uncertainty associated to the determination of
pentachlorophenol (PCP) in aqueous samples, by gas chromatography with mass spectrometric detection
(GCMS), after solidphase microextraction (SPME). It was concluded that, while in the range of 5 to
40 µg/L the uncertainty did not exceed 20%, in the vicinity of the limit of detection (0.75 µg/L) it raised up
to 64%.

Keywords: analytical uncertainty, pentachlorophenol, SPME, GCMS

DOI: 10.1134/S1061934811080053

1 Pentachlorophenol (PCP) is considered as a dan
gerous substance by the European Union [1] and the
United States Environmental Protection Agency [2],
given its toxicity, persistence and potential for bioac
cumulation [3]. Being a chlorinated hydrocarbon
insecticide and fungicide, it is primarily used as a wood
preservative since 1936, but may also act as a prehar
vest defoliant in cotton, a wide spectrum herbicide,
and as a biocide in industrial water systems [4]. As a
result of over 50 years of use as a pesticide, PCP resi
dues are ubiquitous and harmful to the environment
and humans [5, 6]. The latter are exposed to PCP in
the workplace, in treated homes, indoor and outdoor
air, drinking water and food [7]. It can enter the body
by inhalation, ingestion of contaminated water or food
and skin contact with treated wood [4]. According to
European and Portuguese legislation, the maximum
admissible concentration of PCP in inland surface,
estuary, internal coastal and territorial waters is 2 μg/L
[8].
Several methods have been employed to assess PCP
levels in numerous matrices. In waters, PCP was first esti
mated by Deichmann and Schafer [9], by spectropho
tometry. Other quantification technologies include gas
chromatography (GC) with electron capture detection
(ECD) [10–13], flame ionization detection (FID) [14–
16], Fouriertransform infrared spectroscopy (FTIR) [17]
1 The article is published in the original.
and mass spectrometry (MS) [18–21]; high performance
liquid chromatography (HPLC) with electrochemical de
tection (ED) [22] and MS [23]; capillary electrophoresis
[24, 25]; supercritical fluid chromatography [26] and flow
injection analysis (FIA) [27].
The reference methods for the determination of PCP
in aqueous samples have been employing HPLC or GC
after sample cleanup by liquidliquid extraction (LLE)
[8]. Traditionally, LLEs require extensive training, high
organic solvent and time consumption, often being re
sponsible for increased uncertainties in the quantification
of the analyte [28]. Solidphase extraction (SPE), with
similar drawbacks, was also reported [23, 26].
Solidphase microextraction (SPME) has important
advantages in sample preparations, such as quickness,
sensitivity and versatility [29]. Furthermore, no solvents
or complicated apparatus are required and the problems
related with solvent disposal and operator safety are abol
ished. Several authors employed SPME in PCP determi
nation in aqueous samples, mainly using 85μm poly
acrylate (PA) fibers [11, 12, 14, 18, 20, 22, 30, 31]. Li et al.
[16] studied an alternative calix[4]arene coating. Other
microextraction techniques such as stirbar sorptive ex
traction (SBME) or singledrop microextraction
(SDME) were also recently reported in literature [32, 33].
The quantification of trace compounds in naturally
contaminated samples extracted by SPME is usually
done by matrixmatch calibration, where standards
are extracted in the same conditions as the samples.

756
 UNCERTAINTY IN THE QUANTIFICATION OF PENTACHLOROPHENOL
For such chemicals, the reliability of the final results is
crucial, and in this sense, so is the estimation of the
global uncertainty associated to those results.
Uncertainty is a parameter associated with the
result of a measurement that estimates a range of val
ues that could be reasonably attributed to that result.
This definition is different from the term “error”,
which represents the difference between the result and
the true value [34]. Uncertainly derives from several
sources such as sampling, interferences and matrix
effects, environmental conditions, measuring equip
ment, reference values, approximations and assump
tions incorporated in the method and calculation pro
cedure, and random variation [34]. Often neglected,
this parameter represents the variability and the confi
dence of the analytical method and accounts for most
error sources inherent to a given method, each one in
the form of a “partial” uncertainty. It allows not only
the best interpretation of the results but also the diag
nosis and control of such sources. Hence, it contrib
utes to a decision about the fitness of the methodology
employed for the intended analysis. In fact, for values
close to the limits of detection attained currently by
the sharpest apparatus, the uncertainty of the results
can be unsuitably high [35–39].
When an analytical methodology is implemented,
validation is required. Therefore, a set of parameters
that include characterization (specificity and practica
bility), quantification (linearity and detection limits)
and reliability (precision, accuracy and robustness),
are obtained for that purpose. A different number of
approaches are possible [40], but the global uncer
tainty (U) can be estimated using the method perfor
mance parameters obtained from validation [34].
Depending on the specific methodology, the parame
ters suspected to contribute the most are: the best
available estimate of overall precision; the value
obtained for the average analytical recovery; the data
related with the method linearity (calibration curve),
including the errors (inaccuracy) of the measuring
equipment used in the preparation of the standard
solutions.
This study illustrates how to calculate the global
uncertainty associated to the quantification of PCP in
wood processing wastewaters, employing gas chroma
tography with mass spectrometric detection (GC
MS) after SPME.
EXPERIMENTAL
Chemicals. Pentachlorophenol was obtained from Su
pelco (Bellefonte, PA, USA). Sodium sulfate anhydrous
p.a. and sulfuric acid 95–97% p.a. were purchased to
Merck (Darmstadt, Germany). An aqueous effluent with
unknown concentration of PCP was collected in a wood
processing factory.
Preparation of standard solutions. PCP solutions from
0.8 to 40 μg/L were made from a 2400 mg/L stock solu
tion, which was prepared by accurately weighing 0.06 g of
JOURNAL OF ANALYTICAL CHEMISTRY Vol. 66
757
PCP and diluting in a 25 mL volumetric flask with 0.1 M
sodium hydroxide. An intermediate stock solution was
prepared with 100 μL of the stock solution diluted in a
50 mL volumetric flask with the same solvent. The work
ing standards were then made with different aliquots of the
intermediate stock solution in 50 mL volumetric flasks,
using a saturated solution (pH 2) as solvent. This solution
was prepared by adding sodium sulfate (Na2SO4) until
saturation of the deionised water, then controlling the pH
with 1 M sulfuric acid.
Experimental procedure. The quantification of PCP
was attained with an 6890 Series gas chromatograph from
Agilent (Santa Clara, CA, USA) equipped with a 5973 N
Series mass spectrophotometer selective detector in scan
mode, ranging from 15.0 to 500 amu (atomic mass unit).
The interface temperature was 160°C and the ion source
(electron ionization) was set at 230°C with electron ener
gy of 70 eV, whilst the quadrupole mass filter was kept at
150°C. The capillary column was a HewlettPackard
5MS (30 m × 0.25 mm × 0.25 μm) (Avondale, PA, USA).
Helium (99.9999% purity) was the carrier gas, at a con
stant 1 mL/min flow through the column. The oven was
initially set at 80°C, and then raised to 260°C at
15°C/min. The injector was in splitless mode at 250°C,
closed for 3 min before purging with helium at
20 mL/min.
For sample preparation, a 85μm PA fiber (Supelco)
and the respective SPME sampling manual holder (Su
pelco) were employed. The fiber was preconditioned at
300°C for 2 h in the GC injection port. For the extraction
of PCP, 2 mL of the standard solution or sample were
measured into a 4 mL amber vial and fiber was immersed
in the solution for 30 min at room temperature
(25 ± 2°C), with rapid and constant stirring. After this pe
riod, the fiber was removed and desorbed in the GC injec
tion port for 3 min.
Global uncertainty calculation. Equation 1 can calcu
late the estimation of global uncertainty U, expressed as a
relative standard deviation [34]:
U = U 21 + U 22 + U 32 + U 42 . (1)
The quantification of the partial relative uncertainties
U1, U2, U3 and U4 is henceforth explained. The uncertain
ty associated with the preparation of standard solutions,
U1, is equal to the highest individual uncertainty of prepa
ration of each standard, ust, which is given by
2
⎛ Δz i ⎞
ust = ∑ ⎜ z ⎟,
⎝ i ⎠
(2)
i
where Δzi is the inaccuracy associated to the measurement
equipment and zi is the value measured by the respective
equipment. In other terms, the uncertainty associated to
preparation of the standards is the sum of the uncertainties
of each measurement necessary to prepare the standards.
Uncertainty is expected to rise with decreasing quantities
measured.
No. 8 2011
758 BRÁS et al.
Abundance
1500000
1000000
500000
6 8 10
Fig. 1. Chromatogram of 40 µg/L PCP after SPME in GCMS scan mode.
The uncertainty associated with the calibration
curve (U2) is the ratio between the standard deviation
for a given standard concentration, sxo, and the respec
tive standard concentration, calculated by the calibra
tion curve, x0, where sxo is determined by Eq. 3:
12
s ⎧⎪ ( y − y ) 2 ⎪⎫ ,
sxo= y / x ⎨ 1 + 1 + 2 0 ⎬ (3)
∑
b ⎪⎩m n b ( xi − x ) 2⎪⎭
where sy/x is the standard deviation of the regression anal
ysis, b is the slope of the linear regression fit, m is the num
ber of experimental values of y0 obtained for each value of
x0, n is the number of standards used to build the calibra
tion curve, y0 is the experimental value of y obtained by
the linear regression fit for x0, y is the mean of the experi
mental values of y, xi are the values of the standard solu
tions used to obtain the calibration curve, and x is the
mean of the xi values.
The uncertainty associated to the method precision,
U3, is the worst estimate of the relative standard deviation
of replicated measurements of the standards. The preci
sion of a given methodology may be evaluated by three
kinds of assays, namely, the repeatability (replicates mea
sured in the same conditions, same equipment, same op
erator, in the shortest time), the intermediate precision
(the same as previous, but with one variable changed, for
example, evaluation in different days) and the reproduc
ibility (obtained by interlaboratorial assays). Further
more, the precision evaluation is usually done at different
concentration levels, with a minimum of six replicates.
The individual relative standard deviation up is calculated
by Eq. (4):
up = s . (4)
x0 n
In this equation, s is the standard deviation of the
data obtained for the precision study and n is the num
ber of replicates.
The last parameter taken into account for the estima
tion of U is the uncertainty associated to the method ac
curacy, U4 (= ue), which derives from recovery studies with
spiked samples and is calculated by Eq. (5):
JOURNAL OF ANALYTICAL CHEMISTRY
Abundance 266
300000
250000
200000
150000
100000 165
50000 60 7995115130145 179202 230 281
0
50 100 150 200 250
m/z
12
Time
ue = s ( η ),
(5)
n
where s(η) is the relative standard deviation of the per
centage of recovery (standard deviation/mean recov
ery) and n is the number of samples. Recovery is
obtained by the ratio between the concentration of the
spiked sample and the expected one.
Finally, it is customary to add a coverage factor of 2 to
the uncertainty U, to obtain the expanded uncertainty.
Results are to be reported associating the measurement
(M) with the global uncertainty, in the following form:
M ± U (units).
RESULTS AND DISCUSSION
The analysis of pentachlorophenol in environmen
tal samples is normally performed for either pollution
evaluation purposes or to assess the conformity with
legal limits. Both cases need very low detection ranges
and a convenient preconcentration/extraction meth
odology associated to a sensible detection is required.
SPME is within the acceptable extraction methodolo
gies, but careful validation must be previously
obtained. Also GCMS has advantages in this deter
mination because it allows the structural identification
of the detected analytes, thus avoiding misinterpreta
tion of the results. The chromatographic analysis of
PCP was performed in 12 min (Fig. 1), with clear peak
resolution.
To estimate the uncertainty associated to standard so
lution preparation (U1), the errors associated to the mea
suring material (inaccuracy) are referred in Table 1, as well
as the scheme which helps to understand the determina
tion of ust (=U1) for the preparation of the 40 μg/L PCP
standard solution. Such preparation comprises different
tasks, each associated to a partial uncertainty. Thus, the
uncertainty associated with the preparation of that solu
tion is:
u st
–3 –3 –3
= ( 0.2785 × 10 ) + ( 2.5014 × 10 ) ( 0.1466 + 10 )
–2
= 5.4098 × 10 .
Vol. 66 No. 8 2011
 UNCERTAINTY IN THE QUANTIFICATION OF PENTACHLOROPHENOL 759

Table 1. Errors (inaccuracy, Δz) associated with measuring apparatus and the different tasks for the calculation of ust (=U1),
for the 40 μg/L PCP standard
Apparatus Reference Δz
25 μL volumetric flask V1 0.04 mL (20°C)
50 μL volumetric flask V2 0.06 mL (20°C)
100–1000 μL pipette P1 0.005 mL (20°C)
1000–10000 μL pipette P2 0.005 mL (20°C)
Balance B 0.001 g
Task Measure Apparatus ust (partial)

(0.0602 ) ( 25 )
m = 0.0602 g B 2 2
0.001 + 0.04 = 0.2785 ×10 −3
A. Stock solution preparation
V = 25 mL V1

(0.100 ) ( 50 )
V = 0.100 mL P1 2 2
B. Intermediate stock solution preparation 0.005 + 0.06 = 2.5014 × 10 −3
V = 50 mL V2

(0.005
0.415) ( 50 )
V = 0.415 mL P1 2 2
−3
C. Dilute stock solution preparation + 0.06 = 0.1460 × 10
V = 50 mL V2

For the other standards, results are: 5.4335 × 10–2
(0.8 μg/L); 5.4114 × 10–2 (4.0 μg/L); 5.4107 × 10–2
(8.0 μg/L); 5.3358 × 10–2 (12.0 μg/L); and 5.7864 × 10–2
(20.0 μg/L).
Quantification of the PCP concentration in samples
was done by building an external calibration curve, with
standards extracted by SPME in the same experimental
conditions as samples. Three extractions were performed
for each standard and PCP was quantified by the GCMS
respective chromatogram peak area. The range of lineari
ty attained was between 0.8 and 40 μg/L, with a limit of
detection (LOD) for this method of 0.75 μg/L, obtained
by signaltonoise (S/N) ratio. This is an improvement to
previously reported values for PCP analysis in aqueous
matrices. For instance, Estevinho et al. [11] obtained a
level of 0.96 μg/L.
The uncertainty associated to the calibration curve
(U2) was performed using equation 3, according to the da
ta in Fig. 2, which shows the trend line for PCP standards.
Table 2 lists the values of the experimental results (yi), sxo,
x0 and, consequently, U2, for each of the six calibration
standards. It is clear that the uncertainty associated with
the calibration curve rises exponentially as we approach
the least concentrated standards and of course, the LOD
for the given method.
The precision parameter is obtained using repeatabili
ty and intermediate precision assays. The former repre
sents shorttime variability and was obtained by doing
consecutive extractions of 6 different standard solutions
with the same concentration in the same day. The inter
mediate precision intends to analyze time as factor of
method variation, by repeating 6 extractions of different
standard solutions with the same concentration in differ
ent days. For both contributions, the parameter value is
calculated dividing the standard deviation by the mean
concentration, as given by the calibration curve of all the
determinations and is known as relative standard error
(RSD%) or coefficient of variation (CV%). Repeatability
and intermediate precision were tested with extractions of
a 4 μg/L PCP standard, to assess the uncertainty associat
ed to precision (U3 = up/x0). Results are given in Table 3.
In this case, we have two possibilities to perform the calcu
lation, and it was decided to use the intermediate preci
sion values, once they would give the worst case scenario
results in terms of “partial” uncertainty:
up = 1.1098 = 0.3347, and then, for the 4 μg/L
6
standard (see Table 2), U3 = 0.4531/4.9550 = 0.0914.
Repeatability and intermediate precision were already
referred by some authors in similar studies, in terms of co
efficient of variation (CV). The first with better values than
the current study: 12% [14] and 11.3% [18]. The second
with worse performances: 21.3% [28], 26.6% [41] and
10%. Overall, this study obtained common results.
Peak area
2.5E+07 r2 = 0.9953
2.0E+07
1.5E+07
1.0E+07
5.0E+06
0.0E+00
0 5 10 15 20 25 30 35 40
Standard solutions concentration, µg/L
Fig. 2. Calibration curve for PCP standard solutions, ex
traction by SPME.

JOURNAL OF ANALYTICAL CHEMISTRY Vol. 66 No. 8 2011

760 BRÁS et al.

Table 2. Values of experimental results (yi), sxo, x0 and the un sents the mean between the SPME responses of the
certainty associated to the calibration curve, U2 (m = 2; n = 6) blank and the standard solution.
Standard, The uncertainty attributed to accuracy (U4 = ue) was
yi sxo x0 U2
μg/L known through four assays of recovery using a PCP stan
dard of 4 μg/L, blanks and spiked (1 : 1) solutions. Table 4
0.8 659482 0.9988 1.5769 0.6334 shows the results for all solutions in all assays and the re
4 2651487 0.9542 4.9550 0.1926 spective percentage of recovery. With equation 5, the result
8 4280931 0.9271 7.7183 0.1201 of U4 is:
12 5930119 0.9090 10.5151 0.0864
ue = 9.5 106.6 = 0.045.
20 11103340 0.9177 19.2881 0.0476 4
40 23757001 1.2846 40.7468 0.0315 Helaleh et al. [18] reported a similar recovery for
PCP analysis under similar conditions: 108.8%.
With the data gathered above, and using equations 2 to
Table 3. Results for repeatability and intermediate precision 5, the global uncertainty (U) could be calculated for each
of the six standards used in the calibration curve (Eq. 1).
Repeatability Intermediate precision As U3 and U4 were only determined for a given standard,
(CV%), n = 6 (CV%), n = 6
this estimate only suffered the influence of U1 and pre
4.2531 3.9240 dominantly U2 as a function of standards concentration.
4.6070 4.5855 Taking into account the aforementioned uncertainty
5.8953 4.1846 parameters, the most important contribution to U is, in
5.6801 6.6385 fact, U2 (calibration curve). Combining all the partial
terms, the global uncertainty was plotted as a function of
3.5968 3.6520
the standard concentrations (Fig. 3).
3.9333 5.3518 As expected, the global uncertainty is stable or with a
x = 4.6610 x = 4.7227 slight increase throughout most of the calibration range
s = 0.9374 s = 1.1098 (between 10 and 20%), and rises exponentially for the
lower part of that same range, reaching 64.4% for the
CV (%) = 20.11 CV (%) = 23.50 0.8 μg/L standard. In practical terms, any result near this
x —mean; s—standard deviation; CV (%)—coefficient of varia area can be affected of a considerable doubt, even more
tion (s/ x *100). evident in terms of LODs. In this PCP analysis, the meth
od LOD is 0.75 μg/L, which corresponds to an uncertain
ty close to 65%. In the vicinity of the legal limit for PCP in
Table 4. Results of the four assays for the study of recovery waters (2.0 μg/L), U is about 50%, which means that a
2.0 μg/L result could in fact be any result between 1.0 and
Assay Blank Standard (St), Bl + St Recovery, 3.0 μg/L and, consequently, be measurable or not detect
number (Bl) 4 mg/L (1 : 1) % ed. Hence, the authors stress the importance of this par
ticular parameter in the validation of analytical methodol
1 0.64 4.26 2.87 117.1 ogies and their consequent data, particularly in the case of
2 0.63 4.73 2.95 110.1 trace contaminants, which appear in concentrations close
to the LODs. The establishment of regulatory measures in
3 0.77 3.72 2.35 104.7 terms of admissible limits of pollutants must take these par
4 0.74 4.12 2.30 94.7 ticular aspects into account. There are some published works
on this subject, although only few refer SPME/GCMS
Mean 106.6 methodology [35, 42–46]. Díaz et al. [35] obtained similar
Std dev 9.5 results for the determination of nonylphenol in water us
ing DVBCARPDMS fibers for the same concentration
levels: 42.8% for 2.5 μg/L and 56.0% for 0.25 μg/L. To
Accuracy of a given method can be determined by our knowledge, only one study focuses on PCP using
recovery assays. This procedure consists in quantifying SPME/GCMS [20]. In that study, the authors report a
27% global uncertainty for PCP, but only for one concen
SPME extractions in a blank solution, in a 1 : 1 spiking tration level (0.71 μg/L). Furthermore, in this case the
of blank solution with a standard solution, and in the matrix was raw or pretreated drinking water, which is in
same standard solution. The determination is principle cleaner than industrial wastewaters. Then, no
obtained by the ratio of the concentration obtained for further comparison is possible as far as the global uncer
the 1 : 1 spiking and half the concentration of the stan tainty is concerned.
dard solution (observed value divided by the expected This study explains how to calculate the global
value). For an ideal 100% recovery, the result repre uncertainty associated to the quantification of pen

JOURNAL OF ANALYTICAL CHEMISTRY Vol. 66 No. 8 2011

 UNCERTAINTY IN THE QUANTIFICATION OF PENTACHLOROPHENOL 761

Global uncertainty, %
70
60
50
40
30
20
10
0 5 10 15 20 25 30 35 40
Standard concentration, µg/L
Fig. 3. Variation of global uncertainty as a function of stan
dard concentration.
tachlorophenol in waters, by solid phase micro
extraction and gas chromatography with mass spec
trometry.
This calculation refers to the sum of the individual
uncertainty contributions associated to the prepara
tion standard solutions, calibration curve, precision
and accuracy, for the given method. Plotting the global
uncertainty as a function of the calibration standards,
it was shown that, for the higher and medium concen
tration ranges, its value went from 10 to 20%, yet rap
idly rising when approaching the lower ranges, to a
maximum of 64.4% for the lowest standard tested
(0.8 μg/L).
These results show the importance of estimating
uncertainty in trace compounds analysis, crucial in
the characterization and quantification of environ
mental samples.
ACKNOWLEDGMENTS
The authors wish to thank Fundação Calouste Gul
benkian (Portugal) for the financial support.
REFERENCES
1. Council Directive 76/464/EEC. Official Journal
L0464, 4 May 1976, relating to the pollution caused by
certain discharges into the aquatic environment of the
Community.
2. Keith, L.H. and Telliard, W.A., Environ. Sci. Technol.,
1979, vol. 13, no. 4, p. 416.
3. Mäenpää, K., Sorsa, K., Lyytikäinen, M., Leppänen, M.T.,
and Kukkonen, J.V.K., Ecotoxicol. Environ. Safe.,
2008, vol. 69, no. 1, p. 121.
4. Fisher, B., J. Pesticide Reform, 1991, vol. 11, no. 1, p. 2.
5. Orton, F., Lutz, I., Kloas, W., and Routledge, E.J., En
viron. Sci. Technol., 2009, vol. 43, no. 6, p. 2144.
6. Zhu, B.Z. and Shan, G.Q., Chem. Res. Toxicol.,
2009, vol. 22, no. 6, p. 969.
7. Eduljee, G., Sci. Total. Environ., 1999, vol. 232,
no. 3, p. 193.
8. Council Directive 86/280/EEC. Official Journal
L0280, 12 June 1986, on limit values and quality objec
tives for discharges of certain dangerous substances in
cluded in List I of the Annex to Directive 76/464/EEC.
9. Deichmann, W. and Schafer, L.J., Ind. Eng. Chem.
Anal. Ed., 1942, vol. 14, no. 4, p. 310.
10. Zheng, M.H., Zhang, B., Bao, Z.C., Yang, H., and
Xu, X.B., Bull. Environ. Contam. Toxicol., 2000,
vol. 64, no. 1, p. 16.
11. Estevinho, B.N., Ratola, N., Alves, A., and Santos, L.,
J. Hazard. Mater., 2006, vol. 137, no. 2, p. 1175.
12. Estevinho, B.N., Martins, I., Ratola, N., Alves, A., and
Santos, L., J. Hazard. Mater., 2007, vol. 143, no. 1⎯2,
p. 535.
13. Patel, U.D. and Suresh, S., Separ. Purif. Tech., 2008,
vol. 61, no. 2, p. 115.
14. Buchholz, K.D. and Pawliszyn, J., Anal. Chem., 1994,
vol. 66, no.1, p. 160.
15. Engwall, M.A., Pignatello, J.J., and Grasso, D., Water
Res., 1999, vol. 33, no. 5, p. 1151.
16. Li, X., Zeng, Z., and Zhou, J., Anal. Chim. Acta, 2004,
vol. 509, no. 1, p. 27.
17. Besner, A., Gilbert, R., Tétreault, P., Lépine, L., and
Archambault, J.F., Anal. Chem., 1995, vol. 67, no. 2,
p. 442.
18. Helaleh, M.I.H., Fujii, S., and Korenaga, T., Talanta,
2001, vol. 54, no. 6, p. 1039.
19. Düring, R.A., Zhang, X., Hummel, H.E., Czynski, J.,
and Gäth, S., Anal. Bioanal. Chem., 2003, vol. 375,
no. 4, p. 584.
20. Simões, N.G., Cardoso, V.V., Ferreira, E., Benoliel, M.J.,
and Almeida, C.M.M., Chemosphere, 2007, vol. 68,
no. 3, p. 501.
21. Gao, J., Liu, L., Liu, X., Zhou, H., Huang, S.,
Wang, Z., Chemosphere, 2008, vol. 71, no. 6, p. 1181.
22. Sarrión, M.N., Santos, F.J., and Galcerán, M.T.,
J. Chromatogr. A, 2002, vol. 947, no. 2, p. 155.
23. Brossa, L., Pocurull, E., Borrull, F., and Marcé, R.M.,
Chromatographia, 2004, vol. 59, no. 7–8, p. 419.
24. Praus, P., Anal. Chim. Acta, 1995, vol. 302, no. 1, p. 39.
25. Jáuregui, O., Moyano, E., and Galcerán, M.T.,
J. Chromatogr. A, 2000, vol. 896, no. 1–2, p. 125.
26. Bernal, J.L., Jiménez, J.J., Rivera, J.M., Toribio, L.,
and del Nozal, M.J., J. Chromatogr. A, 1996, vol. 754,
no. 1–2, p. 145.
27. RodríguezAlcala, M., YañezSedeño, P., and Polo
Díez, L.M., Talanta, 1988, vol. 35, no. 8, p. 601.
28. Supelco, Solid Phase Microextraction: Theory and Opti
mization of Conditions, Bulletin 923, SigmaAldrich Co,
1998.
29. Pawliszyn, J., SolidPhase Microextraction—Theory
and Practice, New York: WileyVCH, 1997.
30. Lee, M.R., Yeh, Y.C., Hsiang, W.S., and Hwang, B.H.,
J. Chromatogr. A, 1998, vol. 806, no. 2, p. 317.
31. Brás, I., Lemos, L., Alves, A., and Pereira, M.F.R.,
Chemosphere, 2005, vol. 60, no. 8, p. 1095.

JOURNAL OF ANALYTICAL CHEMISTRY Vol. 66 No. 8 2011

762 BRÁS et al.
32. Kawaguchi, M., Inoue, K., Yoshimura, M., Sakui, N.,
Okanouchi, N., Ito, R., Yoshimura, Y., and Nakazawa, H.,
J. Chromatogr. A, 2004, vol. 1041, no. 1–2, p. 19.
33. LópezJiménez, F.J., Rubio, S., and PérezBendito, D.,
J. Chromatogr. A, 2008, vol. 1195, no. 1–2, p. 25.
34. EURACHEM/CITAC Guide, Quantifying Uncertainty
in Analytical Measurement, 2nd Edition. UK: Ellison
SLR, Rosslein M., Williams A. Editors, 2000.
35. Díaz, A., Vázquez, L., Ventura, F., and Galcerán, M.T.,
Anal. Chim. Acta, 2004, vol. 506, no. 1, p. 71.
36. Ratola, N., Martins, L., and Alves, A., Anal. Chim. Ac
ta, 2004, vol. 513, no. 1, p. 319.
37. Ratola, N., Santos, L., Herbert, P., and Alves, A., Anal.
Chim. Acta, 2006, vol. 573⎯574, p. 202.
38. Abreu, S.M., Caboni, P., Cabras, P., Alves, A., and Ga
rau, V.L., J. Chromatogr. A, 2006, vol. 1103, no. 2,
p. 362.
39. Moreira, J.L., Polidoro, A.M., Heitor, C., Nunes, O.C.,
Alves, A., and Santos, L., Fresenius Environ. Bull.,
2006, vol. 15, no. 4, p. 289.
JOURNAL OF ANALYTICAL CHEMISTRY
40. Hund, E., Massart, D.L., and SmeyersVerbeke, J.,
Trends Anal. Chem., 2001, vol. 20, no. 8, p. 394.
41. Ribeiro, A., Neves, M.H., Almeida, M.F., Alves, A.,
and Santos, L., J. Chromatogr. A, 2002, vol. 975,
no. 2, p. 267.
42. Maroto, A., Boqué, R., Riu, J., and Rius, F.X., Anal.
Chim. Acta, 2001, vol. 440, no. 2, p. 171.
43. Maroto, A., Boqué, R., Riu, J., and Rius, F.X., Anal.
Chim. Acta, 2001, vol. 446, no. 1–2, p. 133.
44. Prado, C., Garrido, J., and Periago, J.F., J. Chro
matogr. B, 2004, vol. 804, no. 2, p. 255.
45. Arrebola, F.J., GonzálezRodríguez, M.J., Garrido
Frenich, A., MarínJuan, A., and Martínez Vidal, J.L.,
Anal. Chim. Acta, 2005, vol. 552, no. 1–2, p. 60.
46. BeceiroGonzález, E., Guimarães, A., and Alpendura
da, M.F., J. Chromatogr. A, 2009, vol. 1216, no. 29,
p. 5563.
Vol. 66 No. 8 2011