In situ derivatization/solid-phase microextraction for the determination of haloacetic acids in water

Article abstract of DOI:10.1021/ac000479d

An in situ derivatization solid-phase microextraction method has been developed for the determination of haloacetic acids (HAAs) in water. The analytical procedure involves derivatization of HAAs to their methyl esters with dimethyl sulfate, headspace sampling using solid-phase microextraction (SPME), and gas chromatography-ion trap mass spectrometry (GC/ITMS) determination. Parameters affecting both derivatization efficiency and headspace SPME procedure, such as the selection of the SPME coating, derivatization-extraction time and temperature, and ionic strength, were optimized. The commercially available Carboxen-poly(dimethylsiloxane) (CAR-PDMS) fiber appears to be the most suitable for the determination of HAAs. Moreover, the formation of HAA methyl esters was dramatically improved (up to 90-fold) by the addition of tetrabutylammonium hydrogen sulfate (4.7 μmol) to the sample as ion-pairing agent in the derivatization step. The precision of the in situ derivatization/HS-SPME/GC/ITMS method evaluated using an internal standard gave relative standard deviations (RSDs) between 6.3 and 11.4%. The method was linear over 2 orders of magnitude, and detection limits were compound-dependent, but ranged from 10 to 450 ng/L. The method was compared with the EPA method 552.2 for the analysis of HAAs in various water samples, and good agreement was obtained. Consequently, in situ derivatization/HS-SPME/GC/ITMS is proposed for the analysis of HAAs in water.

Full text of DOI:10.1021/ac000479d

Anal. Chem. 2000, 72, 4865-4873
In S it u De riva t iza t io n /S o lid -P h a s e Mic ro e x t ra c t io n
fo r t h e De t e rm in a t io n o f Ha lo a c e t ic Ac id s in Wa t e r
M. N. Sa rri o´ n, F. J . Sa ntos , a nd M. T. Ga lc e ra n*
Departament de Qu ´ı mica Anal ´ı tica, Facultat de Qu ´ı mica, Universitat de Barcelona, Mart ´ı i Franqu e` s,
1-11, 08028-Barcelona, Spain
1
An in situ derivatization solid-phase microextraction
method has been developed for the determination of
haloacetic acids (HAAs) in water. The analytical procedure
involves derivatization of HAAs to their methyl esters with
dimethyl sulfate, headspace sampling using solid-phase
microextraction (SP ME), and gas chromatography-ion
trap mass spectrometry (GC/ ITMS) determination. P a-
rameters affecting both derivatization efficiency and head-
space SP ME procedure, such as the selection of the
SP ME coating, derivatization-extraction time and tem-
perature, and ionic strength, were optimized. The com-
mercially available Carboxen-poly(dimethylsiloxane) (CAR-
P DMS) fiber appears to be the most suitable for the
determination of HAAs. Moreover, the formation of HAA
methyl esters was dramatically improved (up to 9 0 -fold)
by the addition of tetrabutylammonium hydrogen sulfate
(such as dose of chlorine and free chlorine contact time). In the
early days of DBP research, trihalomethanes (THMs) received
almost exclusive attention, because chloroform was shown to be
an animal carcinogen. However, awareness that HAAs present
serious human health hazards has increased. The U.S. Environ-
mental Protection Agency (USEPA) has classified dichloroacetic
acid as a group B2, probable human carcinogen, on the basis of
positive carcinogenic findings in the liver of two animal species,
2
3
1
male B6C3F mouse and male rats, and trichloroacetic acid as a
group C, possible human carcinogen, on the basis of limited
evidence of carcinogenicity to the liver in animals.2,4 Accordingly,
in the first stage of the D/ DBP rule, the USEPA has established
a maximum contamination level (MCL) of 60 µg/ L for the sum
of five HAAs: monochloroacetic acid, dichloroacetic acid, trichlo-
roacetic acid, monobromoacetic acid, and dibromoacetic acid.5
Under stage II of this rule, this MCL is expected to be reduced
to 30 µg/ L. Consequently, efforts must be made to develop fast
and accurate analytical methods of monitoring concentration,
behavior, and distribution of HAAs in water.
Most of the methods used to determine HAAs involve gas
chromatography (GC) with electron capture detection (ECD) or
coupled with mass spectrometry (GC/ MS). For the analysis of
these compounds by GC, a prior derivatization step is necessary
because of their low volatility and high polarity. Liquid-solid or
liquid-liquid extraction (LLE) is commonly used for the isolation
of HAAs. After an extraction step, the derivatization of HAAs to
(
4 .7 µmol) to the sample as ion-pairing agent in the
derivatization step. The precision of the in situ derivati-
zation/ HS-SP ME/ GC/ ITMS method evaluated using an
internal standard gave relative standard deviations (RSDs)
between 6 .3 and 1 1 .4 %. The method was linear over 2
orders of magnitude, and detection limits were compound-
dependent, but ranged from 1 0 to 4 5 0 ng/ L. The method
was compared with the EP A method 5 5 2 .2 for the
analysis of HAAs in various water samples, and good
agreement was obtained. Consequently, in situ derivati-
zation/ HS-SP ME/ GC/ ITMS is proposed for the analysis
of HAAs in water.
6
short-chain esters using different reagents, such as diazomethane,
acid-alcohol,7,8 or BF
-methanol, is often applied. Other authors
9
3
(
1) Minear, R. A.; Amy, G. L. Water Disinfection and Natural Organic Matter:
Characterization and Control; ACS Symposium Series 649; American Chemi-
cal Society: Washington, DC, 1996.
2) Bull, J. R.; Sanchez, I. M.; Nelson, M. A.; Larson, J. L.; Lansing, A. J.
Toxicology 1 9 9 0 , 63, 341-359.
3) DeAngelo, A. B.; Daniel, F. B.; Most, B. M.; Olson, G. R. Toxicology 1 9 9 6 ,
Chemical oxidants, such as chlorine, ozone, chloramines,
chlorine dioxide, etc., are used in the treatment of drinking water
in many parts of the world, mostly for disinfection and removal
of obnoxious chemical compounds or potential toxicants. However,
water treatment practices also generate disinfection byproducts
(
(
(
(
1
14, 207-221.
4) DeAngelo, A. B.; Daniel, F. B.; Most, B. M.; Olson, G. R. J. Toxicol. Environ.
Health 1 9 9 7 , 52, 425-445.
5) U.S. Environmental Protection Agency (USEPA); Disinfectants and Disinfec-
tion Byproducts: Proposed Rule; Fed. Reg. 59:38668-38829; Washington, DC,
(
DBPs) due to reactions of these oxidants with natural organic
matter present in water. Nowadays, the most common drinking
water disinfectant worldwide is chlorine, which produces triha-
lomethanes (THMs) and haloacetic acids (HAAs) as the most
prevalent groups of chlorination byproducts. The levels and
speciation of these compounds depend on the water quality
conditions (such as total organic content, bromide concentration,
temperature, and pH), as well as on the disinfection conditions
1
994.
(6) Heller-Grossman, L.; Manka, J.; Limoni-Relis, B.; Rebhun, M. Water Res.
9 9 3 , 27, 1323-1331.
1
(
7) Methods for the determination of Organic Compounds in Drinking Water.
Supplement III. Determination of Organic Compounds in Drinking Water by
Liquid-Liquid Extraction, Derivatization and Gas Chromatography with
Electron Capture Detection; Method 552.2, EPA/ 600/ R-95/ 131; U.S. Envi-
ronmental Protection Agency, Environmental Monitoring Systems Labora-
tory: Cincinnati, OH, 1995.
*
Corresponding author: (e-mail) galceran@zeus.qui.ub.es; (fax) +34 93 402
(8) Reimann, S.; Grob, K.; Frank, H. Environ. Sci. Technol. 1 9 9 6 , 30, 2340-
1
2 33.
2344.
1
0.1021/ac000479d CCC: $19.00 © 2000 American Chemical Society
Analytical Chemistry, Vol. 72, No. 20, October 15, 2000 4865
Published on Web 09/14/2000

1
0
proposed simultaneous extraction-derivatization or a derivati-
zation step prior to the extraction.
described the analysis of HAAs in drinking water by in situ acidic
derivatization to the methyl esters with HCl/ methanol followed
by headspace SPME (HS-SPME) and GC-ECD determination
using a poly(dimethylsiloxane) fiber. Although this method is
rapid, a high limit of detection was obtained for monochloroacetic
acid (400 µg/ L) and only three chlorinated acetic acids were
studied. More recently, a new method for the analysis of the six
HAAs included in EPA method 552 using HS-SPME/ GC coupled
to ion-trap mass spectrometry (ITMS) was developed in our
laboratory.32 Acid-catalyzed ethylation instead of methylation was
used to form volatile esters with high partition constants on the
fiber in order to obtain low detection limits. The method showed
good sensitivity (detection limits from 10 to 200 ng/ L), but was
relatively time- and labor-intensive.
In this paper, a new method for analysis of HAAs in water using
HS-SPME/ GC/ ITMS is proposed. Direct derivatization of HAAs
in water by dimethyl sulfate or diethyl sulfate was tested in order
to obtain low detection limits, to avoid tedious preconcentration
steps, and to reduce the analysis time. Experimental conditions
to obtain high efficiency in the derivatization step were established
and HS-SPME parameters were optimized to achieve good
sensitivity in the GC. The optimized procedure was applied to the
analysis of nine HAAs (EPA method 552.2) in tap water and
swimming-pool water.
8
,11,12
The derivatization is
performed either in dry conditions by evaporation of the matrix
sample or directly in water, and in both cases the derivatives are
then extracted by organic solvents8,11 or directly analyzed by the
headspace technique.12 On the other hand, the Grob closed-loop
stripping analysis (CLSA) technique has recently been employed
for the determination of the most common nonpolar and polar
halogenated DBPs using GC-ECD analysis, but only results for
dichloroacetic acid and dibromoacetic acid have been reported.13
Determination of HAAs without derivatization is possible by using
liquid chromatography,6,14-16 especially ion chromatography15,16
or capillary zone electrophoresis.17 Some of these methods are
able to achieve detection limits for HAAs similar to or even better
than GC methods, but multistep procedures, involving trace
enrichment processes, are also necessary, which is tedious and
time-consuming.
Solid-phase microextraction (SPME), pioneered by Pawliszyn
and co-workers,1
8-20
is a rapidly growing technique. It involves
the partitioning of organic analytes from the aqueous or gaseous
medium onto the stationary-phase coating of a SPME fiber. The
analytes can be determined by GC via thermal desorption at the
GC injector port, or by HPLC via a special interface. SPME has
been used successfully for the analysis of a wide range of organic
compounds in water samples,21,22 and recently, it has been applied
for the determination of various DBPs in water samples, such as
trihalomethanes and halogenated solvents,23-25 iodinated halo-
methanes,26,27 carbonyl compounds,28 cyanogen halides,29 and also
for compounds causing taste and odor in water supplies, such as
geosmin and 2-methylisoborneol.30 To our knowledge, only two
studies report HAAs analysis by SPME. Aikawa and co-workers31
EXPERIMENTAL SECTION
Chemicals and Materials. Monochloroacetic acid (MCAA,
9
9%), monobromoacetic acid (MBAA, 99%), dichloroacetic acid
(
DCAA, 99%), dibromoacetic acid (DBAA, 98%), trichloroacetic
acid (TCAA, 99.5%), and tribromoacetic acid (TBAA, 99%) were
obtained from Fluka (Buchs, Switzerland); bromodichloroacetic
acid (BDCAA, 99%) and chlorodibromoacetic acid (CDBAA, 99%)
were purchased from Supelco (Bellefonte, PA), and finally,
bromochloroacetic acid (BCAA, 98%) was supplied by Chem
Service (West Chester, PA). All standards were used as received.
A commercially available EPA 552.2 esters calibration mixture in
methyl tert-butyl ether (MtBE), containing the methyl esters of
the 9 HAAs at a purity higher than 97% and concentrations
between 200 and 2000 µg/ mL, was obtained from Supelco
(
9) Boucharat, C.; Desauziers, V.; Le Cloirec, P. Talanta 1 9 9 8 , 47, 311-323.
(
(
(
10) Benanou, D.; Acobas, F.; Sztajnbok, P. Water Res. 1 9 9 8 , 32, 2798-2806.
11) Scott, B. F.; Alaee, M. Water Qual. Res. J. Can. 1 9 9 8 , 33, 279-293.
12) Neitzel, P. L.; Walther, W.; Nestler, W. Fresenius’ J. Anal. Chem. 1 9 9 8 ,
3
61, 318-323.
(
(
(
13) Kampioti, A. A.; Stephanou, E. G. J. Chromatogr., A 1 9 9 9 , 857, 217-229.
14) Hashimoto, S.; Otsuki, J. J. High Resolut. Chromatogr. 1 9 9 8 , 21, 55-58.
15) Sarzanini, C.; Bruzzoniti, M. C.; Mentasti, E. J. Chromatogr., A 1 9 9 9 , 850,
1
97-211
(
(
(
16) Lopez-Avila, V.; Liu, Y.; Charan, C. J. AOAC Int. 1 9 9 9 , 82, 689-704.
17) Mart ´ı nez, D.; Borrull, F.; Calull, M. J. Chromatogr., A 1 9 9 9 , 835, 187-196.
18) Arthur, C. L.; Killam, L. M.; Buchholz, K. D.; Pawliszyn, J. Anal. Chem.
(
(
Bellefonte, PA). The derivatization reagents dimethyl sulfate
DMS) and diethyl sulfate (DES), as well as the ion-pairing agent,
1
9 9 2 , 64, 1960-1966.
(
(
19) Zhang, Z.; Pawliszyn, J. Anal. Chem. 1 9 9 3 , 65, 1843-1852.
20) Pawliszyn, J. Solid-Phase Microextraction: Theory and Practice; Wiley-VCH:
New York, 1997.
21) Eisert, R.; Levsen, K. J. Chromatogr., A 1 9 9 6 , 733, 143-157.
22) Pawliszyn, J. Applications of Solid-Phase Microextraction; RSC Chromatog-
raphy Monographs: Cambridge, 1999.
4
tetrabutylammonium hydrogen sulfate (TBA-HSO ), and the
dechlorinating agent, ammonium chloride, were obtained from
Fluka at a high purity (g99%). The derivatization reagents, as well
as some HAAs are carcinogenic or toxic and were handled in
accordance with the most current material safety data sheets. The
compounds, 2,3-dibromopropionic acid (98%) and 1,2-dibromopro-
pane (97%), used as the surrogate standard and internal standard
were purchased from Fluka and Sigma-Aldrich (Milwaukee, WI).
Methanol of residue analysis grade and sulfuric acid for analysis
were supplied by Merck (Darmstadt, Germany), whereas MtBE
of residue analysis grade was obtained from Fluka. Anhydrous
sodium sulfate and copper (II) sulfate pentahydrate were pur-
chased from Panreac (Barcelona, Spain) and Probus (Badalona,
Spain), respectively. Water from the Milli-Q water purification
system (Millipore Corp., Bedford, MA) was used.
(
(
(
23) Nilsson, T.; Pelusio F.; Montanarella L.; Larsen B.; Facchetti, S.; Madsen,
J. O. J. High Resolut. Chromatogr. 1 9 9 5 , 18, 617-624.
(
(
24) Popp, P.; Paschke, A. Chromatographia 1 9 9 7 , 46, 419-424.
25) Apfalter, S.; Krska R.; Linsinger, T.; Oberhauser, A.; Kandler, W.; Grasser-
bauer, M. Fresenius’ J. Anal. Chem. 1 9 9 9 , 364, 660-665.
26) Frazey, P. A.; Barkley, R. M.; Sievers, R. E. Anal. Chem. 1 9 9 8 , 70, 638-
44.
27) Cancho, B.; Ventura, F.; Galceran, M. T. J. Chromatogr., A 1 9 9 9 , 841, 197-
06.
(
(
(
(
6
2
28) Bao, M.; Pantani, F.; Griffini, O.; Burrini, D.; Santianni, D.; Barbieri, K. J.
Chromatogr., A 1 9 9 8 , 809, 75-87.
29) Cancho, B.; Ventura, F.; Galceran, M. T. submitted to J. Chromatogr., A.
Departament of Organic Analytical Chemistry. Societat General d’Aig u¨ es
de Barcelona, S. A. Barcelona, 1999.
(
(
30) McCallum, R.; Pendleton, P.; Schumann, R.; Trinh, M.-U. Analyst (Cambridge,
U. K.) 1 9 9 8 , 123, 2155-2160.
31) Aikawa, B.; Burk, R. C. Int. J. Environ. Anal. Chem. 1 9 9 7 , 66, 215-224.
(32) Sarri o´ n, M. N.; Santos, F. J.; Galceran, M. T. J. Chromatogr., A 1 9 9 9 , 859,
159-171.
4866 Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

SPME experiments were performed with a manual fiber holder
supplied from Supelco (Bellefonte, PA). Five commercially avail-
able fibers, Poly(dimethylsiloxane), PDMS, 100 µm; Polyacrylate,
PA, 85 µm; Carboxen-Poly(dimethylsiloxane), CAR-PDMS, 75 µm;
Poly(dimethylsiloxane)-divinylbenzene, PDMS-DVB, 65 µm; and
StableFlex Divinylbenzene-Carboxen-Poly(dimethylsiloxane), DVB-
CAR-PDMS, 50/ 30 µm were purchased from Supelco. Before use,
each fiber was conditioned in a heated GC split/ splitless injection
port under helium flow according to the manufacturer’s instruc-
tions. Screw-capped vials (10, 30, and 40 mL), sealed with a Teflon-
lined silicon septum and used for storing the standard solutions
as well as for sample derivatization and extraction in both HS-
SPME and LLE procedures, were obtained from Wheaton (Millville,
NJ). The vials were cleaned with AP-13 Extran alkaline soap
reference. The transfer line and the ion trap manifold temperatures
were maintained at 270 and 220 °C, respectively. For ethylation
experiments, the instrument was operated in full-scan mode and
the mass range was from m/ z 27 to 325 at 0.8 s/ scan. For
quantification, two selected ions of each ethyl haloacetate were
monitored: m/ z 77/ 94 for MCAA, 83/ 85 for DCAA, 117/ 82 for
TCAA, 121/ 138 for MBAA, 174/ 120 for DBAA, 251/ 172 for TBAA,
129/ 109 for BCAA, 163/ 161 for BDCAA, and 209/ 205 for CDBAA.
For methylation experiments, the mass spectrometer was operated
in full-scan mode between m/ z 29 and 260 at 0.8 s/ scan, with an
ionization time of 100 ms. After establishing the optimized
conditions, different acquisition segments during each chromato-
graphic run were applied using a narrow mass range and different
scan rates. To enhance the response, ionization times of 200 or
400 ms, depending on the compound, were used. The ions of the
methyl haloacetates used for quantification were as follows: m/ z
59/ 108 for MCAA, 83/ 85 for DCAA, 117/ 119 for TCAA, 93/ 95
for MBAA, 173/ 171 for DBAA, 251/ 253 for TBAA, 129/ 127 for
BCAA, 163/ 161 for BDCAA, and 209/ 205 for CDBAA. For MCAA
m/ z 59 instead of 77 was used to prevent interference from the
fiber. Saturn version 5.2 software was used for data acquisition.
Analyses of HAAs by EPA method 552.2 were performed on a
Carlo Erba 5300 Mega Series gas chromatograph (Milan, Italy),
equipped with a 63Ni electron capture detector (ECD). A 60 m ×
0.25 mm i.d. × 0.25 µm DB-1701 fused-silica capillary column
(J&W Scientific, Folsom, CA) was used. The column temperature
program was 36 °C, held for 21 min, then increased at 10 °C/ min
to 140 °C, held for 3 min, up to 240 °C at 20 °C/ min, held for 5
min, and finally increased at a rate of 20 °C/ min to 280 °C and
held for 10 min. Carrier gas was hydrogen (32 cm/ s at 36 °C),
and nitrogen was used as makeup gas (50 mL/ min). A 30-m DB-
17 (50% phenyl, 50% methylpolysiloxane) fused-silica capillary
column with 0.25-mm i.d. and 0.25-µm film thickness (J&W) was
used for confirmation. Injector and detector temperatures were
kept at 200 and 330 °C, respectively, and splitless injection mode
(1 min) was used in all analyses. ChromCard version 1.3 software
(Fisons Instruments, Spain) was used for data acquisition.
Confirmation of analytes in the more complex sample, swimming-
pool water, was performed by GC/ MS.
(
(
Merck) for 24 h; rinsed consecutively with (i) deionized water,
ii) 1:10 HCl/ water, (iii) again with deionized water, and (iv) finally
with Milli-Q water; and baked at 110 °C overnight. Volumetric
glassware was washed as described above, but was air-dried.
Anhydrous sodium sulfate was heated to 400 °C overnight to
remove phthalates and other interfering organic substances and
then stored at 110 °C until use.
Stock standard solutions of each HAA (2000 µg/ mL) were
prepared by weight in Milli-Q water. Standard mixtures were
prepared weekly or daily, depending on their concentration, except
for TBAA, which was prepared daily because it decomposes
6
spontaneously at 25 °C. All solutions were stored frozen in the
dark at -17 °C until use. For evaluating the SPME procedure,
water standards containing 200 µg/ L of each HAA were prepared
by adding 50 µL of a standard mixture of 45 µg/ mL into 10 mL of
Milli-Q water and then sealing them in a 30-mL screw-capped vial.
Barcelona tap water and swimming-pool water samples were
analyzed using the proposed HS-SPME protocol as well as the
EPA method 552.2. The samples were collected in 250-mL amber
glass bottles with PTFE-faced septa and polypropylene screw caps
containing ammonium chloride, avoiding the presence of head-
space at the top of the bottles. All analyses were performed within
2
days of sampling. Ammonium chlorine was added as a dechlo-
rinating agent to preserve the samples by converting the highly
reactive free chlorine to the less reactive monochloramine. Thus,
the free chlorine was prevented from reacting with precursor
organic matter to form additional DBPs.
HS-SP ME P rocedure. In situ derivatization/ HS-SPME was
optimized using different derivatization reagents such as DES and
DMS in order to obtain the ethyl and methyl haloacetates prior
to the analysis by HS-SPME. Briefly, 10 mL of Milli-Q water
containing 200 µg/ L of each compound (total amount of HAAs,
0.12 µmol) was placed in a 30-mL screw-cap glass vial containing
a 10 × 5 mm Teflon-coated stir bar and 5 g (3.5 M) of anhydrous
Instrumentation. All analyses using SPME procedures were
carried out with a Varian 3400 CX GC capillary gas chromatograph
coupled to a Saturn 3 GC/ MS ion trap mass spectrometer (Sugar
Land, TX). Separations were conducted on a DB-5 MS fused-silica
capillary column, 30 m × 0.25 mm i.d. × 0.25 µm (J&W Scientific,
Folsom, CA), with helium as carrier gas, at a linear velocity of 34
cm/ s. The column was held at 40 °C for 1 min, ramped at 20
4
sodium sulfate. After addition of an ion-pairing agent (TBA-HSO ,
2.3 µmol, as aqueous solution), the vial was closed. One hundred
microliters of derivatization reagent (DES, 0.73 mmol, or DMS,
1.05 mmol, i.e., high excess) was then injected through the
septum, and the vial was clamped inside a water-thermostatized
bath, which was placed on a hot plate/ stirrer. After 5 min at 55
°C, the fiber was exposed to the headspace above the aqueous
solution for the desired extraction time. Magnetic stirring at 1200
rpm was applied during both stabilization and extraction. Finally,
the fiber was desorbed in the injection port of the gas chromato-
graph for 2 min, at different temperatures depending on the fiber
coating (splitless injection mode). Several parameters affecting
°
C/ min to 60 °C, then up to 120 °C at 5 °C/ min, held for 3 min,
and finally ramped at 25 °C/ min to 280 °C and held for 10 min.
Desorption time and injection port temperature were set at the
optimum values.
The ion trap mass spectrometer (ITMS) was operated in
electron ionization (EI) positive-mode using automatic gain control
(
AGC). For EI experiments, the instrumental parameters were
set at the following values: a filament emission current of 75 µA,
an electron multiplier voltage of 1850 V, and a modulation
amplitude of 2.5 V, using perfluorotributylamine (FC-43) as
Analytical Chemistry, Vol. 72, No. 20, October 15, 2000 4867

both derivatization and HS-SPME were then studied: derivatiza-
tion reagent (DMS or DES) and fiber stationary phase type,
derivatization-extraction temperature (from 35 to 70 °C), deriva-
tization-extraction time (up to 60 min), volume of derivatization
reagent (between 10 and 160 µL of DMS) and amount of ion-
pairing agent (up to 6.6 µmol). Other parameters affecting the
HS-SPME procedure, such as desorption temperature (280 and
pairing agent. In addition, ethylation was initially used instead of
methylation in order to form relatively high molecular weight
volatile esters, with expected high partition constants on the fiber.
Optimization of the SP ME Conditions. (a) Derivatization
Reagent and Fiber Selection. Ethylation of HAAs with DES was
initially tested, and the analysis of the derivatives formed using
HS-SPME with different stationary phases was evaluated to obtain
high sensitivity and selectivity. Five fibers were tested: PDMS,
100 µm; PA, 85 µm; PDMS-DVB, 65 µm; CAR-PDMS, 75 µm;
PDMS-DVB, 65 µm; and StableFlex DVB-CAR-PDMS, 50/ 30 µm.
SPME conditions are described in the Experimental Section. A
long extraction time (60 min) was applied to select the fiber
coating, to ensure that a large amount of ethyl derivatives was
formed and extracted. The desorption temperatures were within
the recommended operating range for each fiber: 250 °C for
PDMS, 290 °C for PA, 260 °C for PDMS-DVB and DVB-CAR-
PDMS, and 280 °C for CAR-PDMS. No carryover on second
desorptions was found for any of the fibers, indicating complete
removal of analytes at these temperatures. The results showed
that these fibers were not suitable for all the analytes because
only small amounts of MBAA, DCAA, and DBAA ethyl haloac-
etates were detected. Moreover, all fibers showed an extremely
broad peak close to those of the TCAA and BCAA ethyl esters,
which interfered with the determination of these compounds.
Since these problems were hard to overcome, methylation was
performed using DMS, and the fiber coating was optimized. The
relative responses obtained for methyl haloacetates using the
fibers are shown in Figure 1. Homogeneous polymer coatings,
such as PDMS and PA fibers, did not extract HAAs methyl esters;
only small amounts of di- and trihalogenated HAAs derivatives
3
00 °C), desorption time (up to 2 min), and the effect of ionic
strength in the aqueous solution (from 0 to 5 M of anhydrous
sodium sulfate) were also optimized. Possible carryover was
prevented by keeping the fiber in the injector for an additional
time with the injector in the split mode (purge on). Moreover,
blanks were run periodically during the analysis to confirm the
absence of contaminants. For optimization, all determinations were
performed in duplicate, and the average values are reported.
Liquid-Liquid Extraction Procedure (EPA Method 552.2).
Liquid-liquid extraction for the determination of HAAs in tap
water and swimming-pool water was performed in triplicate
7
following EPA method 552.2 with minor modifications. Briefly,
1
1 µL of a MtBE solution of 2,3-dibromopropionic acid 22 µg/
mL, as surrogate standard, 3 mL of concentrated sulfuric acid (to
obtain pH < 0.5), 12 g of anhydrous sodium sulfate, 3 g of copper
(
II) sulfate pentahydrate, and 2 mL of MtBE were added to a 30-
mL water sample placed in a 40-mL vial. The vials were then sealed
with Teflon-faced septa, shaken for 15 min in a rotary mixer, and
allowed to stand for 5 min. To derivatize the HAAs, 1 mL of the
MtBE extract and 2 mL of methanol/ sulfuric acid (9:1) were
transferred to a 10-mL vial, which was placed in a thermostatic
water bath at 50 °C for 1 h. After cooling to 4 °C, 5 mL of a CuSO
4
/
Na SO solution was added and the mixture was shaken by hand
2
4
for 2 min. An aliquot of 300 µL of MtBE extract was transferred
to a 2-mL vial, and 3 µL of a MtBE solution of 1,2-dibromopropane,
of 10 mg/ L, was added as internal standard. Finally, 1 µL of the
MtBE extract was injected into the GC.
(
from 0.03 to 3.7% respect to maximum response area for DBAA
with the CAR-PDMS fiber) were extracted. For coatings formed
by porous polymeric phases such as PDMS-DVB, CAR-PDMS,
and the dual-coated DVB-CAR-PDMS, the HAAs methyl esters
were successfully extracted (Figure 1). This may be due to the
strong retention of the analytes into the pores. Generally, the high
molecular weight derivatives showed best extraction efficiencies
for these three fibers, except TBAA, which showed low sensitivity.
This could be explained by the fact that TBAA undergoes
RESULTS AND DISCUSSION
The objective was to develop and optimize both the in situ
derivatization and the extraction conditions for the quantification
of HAAs from water. To derivatize carboxylic acids, an esterifi-
2 3
cation reaction is frequently used. For instance, NaOH or K CO
6
spontaneous decomposition to bromoform in aqueous solution
catalyzes in situ methylation of diols, phenols, or even aliphatic
acids with low molecular weight, but cannot be used for direct
analysis of halogenated acids in water, such as HAAs, because
they decompose to halogenated hydrocarbons in acidic or alkaline
media. Another possibility of in-matrix derivatization of organic
acids involves the alkylation with dimethyl sulfate12,33 to the
corresponding methyl ester. The addition of modifiers, such as
ion-pairing agents, which activate the analytes during derivatiza-
tion, increases esterification yields and thus improves the sensitiv-
ity of the procedure.12 Ammonium quaternary salts, such as
tetrabutylammonium bromide or tetrabutylammonium chloride,
can be used as ion-pairing agents for the analysis of HAAs in water.
However, they react with the derivatization reagents, dimethyl
sulfate or diethyl sulfate, to form halogenated hydrocarbons, which
could compete with HAAs derivatives for the fiber. So, tetrabu-
even at 25 °C, or could have undergone thermal degradation
during esterification with DMS.12 The presence of bromoform was
confirmed by identification in the chromatogram. Moreover, peaks
corresponding to the decomposition of BDCAA, CDBAA, and
TCAA to the respective halogenated hydrocarbons were also
observed in the chromatogram, but the yield of formation was
low. On the other hand, only the dual-coated DVB-CAR-PDMS
and CAR-PDMS fibers extracted all HAAs, including the mono-
halogenated species; CAR-PDMS showed the highest extraction
efficiency for all the compounds (responses were 1.7-8 times
higher than those obtained with dual-coated DVB-CAR-PDMS
fiber). The mean micropore diameter (10 Å) of Carboxen is lower
than that of divinylbenzene (17 Å), so it would be ideal for SPME
analyses of small molecules in the C
2
-C
6
range,22 such as HAAs
methyl esters. In terms of selectivity, it should be noted that some
additional peaks appeared at the beginning of the chromatogram
with CAR-PDMS fiber in comparison with those with PDMS-DVB
and DVB-CAR-PDMS fibers. However, these peaks did not
4
tylammonium hydrogen sulfate (TBA-HSO ) was used as the ion-
(
33) Neu, H.-J.; Ziemer, W.; Merz, W. Fresenius’ J. Anal. Chem. 1 9 9 1 , 340, 65-
0.
7
4868 Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

Fig u re 1 . Extraction efficiency of five commercial SPME fibers. All recoveries are normalized to the maximum area response obtained. Milli-Q
water containing 200 µg/L of each HAA; DMS, 100 µL; TBA-HSO4, 2.3 µmol; sodium sulfate, 3.5 M; extraction time, 60 min; extraction temperature,
55 °C.
interfere with the detection of our derivatives, so CAR-PDMS
coating was selected for all subsequent experiments. As mentioned
above, 60 min was the time used to compare the behavior of the
fibers. To reduce extraction time, the efficiency of extractions
conducted for 25 min was compared with those conducted for 60
min with the CAR-PDMS fiber. Mean responses were similar for
all the derivatives, except that for DBAA methyl ester, which
showed an increase of 12% in the area after the 60-min extraction.
Therefore, a sampling time of 25 min was used for optimization
purposes.
Derivatization-Extraction Temperature. The effect of sample
temperature on the derivatization/ HS-SPME was examined from
3
5 to 65 °C (Figure 2). An increase in temperature enhanced the
derivatization reaction yield. Moreover, the mass transfer process
was also favored by increasing the vapor pressure of the analytes
in the headspace. On the other hand, the distribution constant of
the methyl haloacetates between the headspace and the fiber
coating was also temperature-dependent. At high temperatures
(
above 55 °C), the affinity of the analytes for the fiber coating
diminished and the relative responses decreased (Figure 2). Fifty-
five degrees Celsius was the optimum temperature for all the
methyl haloacetates.
Derivatization Reagent and Ion-P airing Agent Amount.
The concentration of the derivatization reagent also affects the
reaction yield of HAAs methyl esters. Amounts of DMS ranging
from 10 (0.10 mmol) to 160 µL (1.89 mmol) were tested at 55°
Fig u re 2 . Temperature profiles for in situ methylation HS-SPME
of the nine HAAs. Conditions as in Figure 1, with a 25-min extraction
time and the CAR-PDMS fiber. Compound identification: ([) MCAA,
(9) DCAA, (2) TCAA, (4) MBAA, (f) DBAA, (O) TBAA, (]) BCAA,
0) BDCAA, (b) CDBAA.
(
(
Figure 3). In general, volumes of DMS between 60 (0.63 mmol)
and 100 µL (1.05 mmol) gave satisfactory reaction yields for most
of the compounds. The relative responses for some HAAs
decreased at high amounts of DMS. The addition of 60 µL (0.63
mmol) of DMS ensured the maximum response for MBAA methyl
ester.
The effect of the amount of ion-pairing agent on the derivati-
4
zation of HAAs was studied by adding TBA-HSO to the water up
to 6.6 µmol and using 60 µL of DMS. The responses relative to
the maximum value obtained for DBAA methyl ester are given in
Table 1. The best results for most compounds were obtained using
Analytical Chemistry, Vol. 72, No. 20, October 15, 2000 4869

Fig u re 3 . Effect of dimethyl sulfate on the reaction yield of HAAs. Conditions as in Figure 2, extraction temperature 55 °C.
Ta b le 1 . Effe c t o f Io n -P a irin g Ag e n t (TBA-HS O4 ) o n
fiber augmented significantly when sodium sulfate concentration
increased to 3.5 M. At higher concentrations, constant responses
were obtained for some compounds (MCAA, TBAA, BDCAA, and
CDBAA) whereas a decrease was observed for others (DCAA,
TCAA, MBAA, DBAA, and BCAA). So a concentration of 3.5 M
was maintained for subsequent studies.
HAA Re a c t io n Yie ld s t o t h e Co rre s p o n d in g Me t h yl
a
Ha lo a c e t a t e s
area response
b
ratio with
amt of TBA-HSO
µmol)
4
without the use
(
of TBA-HSO
4
compd
0
0.23 0.47 2.36 4.72 5.66 6.61
Derivatization-Extraction Time and Stirring Rate. The
extraction time profiles of the methyl haloacetates were then
studied up to 60 min (Figure 4). Different equilibration times were
obtained for the analytes, depending on the headspace/ aqueous
relative responses with respect to
the maximum area value (%)
MCAA
DCAA
TCAA
MBAA
DBAA
TBAA
BCAA
0.9
1.2
1.4
4.3
5.7
2.5
1.4
66.8
60.3
4.6
6
23
90
3
41
9
27
80
44
3.6 24.2 55.4 84.4 82.8 70.3
0.8 29.6 63.0 72.0 69.3 69.8
1.7
1.5
1.6
3.5
5.7
5.1
sample distribution constant, Khs, and the fiber coating/ headspace
distribution constant, Kfh.19 Some derivatives (MCAA, BDCAA, and
2.3 15.1 34.8 95.5 95.4 89.6 100
0.3
1.4
1.5
2.9
2.9
2.2
2.8
71.0
31.1
20.1
CDBAA methyl esters) achieved equilibration in 5-15 min but
other compounds needed 25 or 35 min. Consequently, an exposure
time of 35 min was chosen as optimal for all the haloacetates.
The effect of stirring rate on the responses was tested between
2.7 23.6 46.3 75.3 72.3 64.6
BDCAA 0.4
CDBAA 0.5
8.9 16.1 29.9 31.1 26.5
5.3 7.0 19.2 20.8 17.4
a
Milli-Q water containing 200 µg/ L of each HAA (∑ HAAs ) 0.12
µmol). Amount of TBA-HSO4, 4.72 µmol.
b
9
00 and 1200 rpm. On the basis of extraction efficiency, similar
responses were obtained for MCAA, BDCAA, CDBAA, and TBAA
methyl esters at all the rates studied. However, for DCAA, TCAA,
MBAA, DBAA, and BCAA methyl esters, equilibrium was not
reached for rates below 1200 rpm, indicating that diffusion through
the water was the rate-controlling step in the adsorption. In terms
of precision, RSDs (n ) 3) lower than 7% were obtained for five
HAAs at low stirring rate; however, values that were too high were
observed for the remaining compounds (up to 15%). So, 1200 rpm
was maintained for further studies, taking into account the shorter
equilibration time assessed for all the compounds at this rate.
Optimization of Desorption Conditions. Two desorption
temperatures, 280 and 300 °C, were evaluated for a desorption
time of 2 min. Results showed no differences in the GC responses
between the two temperatures. In addition, carryover was not
detected at either temperature; therefore, the lower temperature
was chosen for further experiments to avoid degradation of the
fiber with temperature. All compounds were quantitatively des-
orbed from the CAR-PDMS fiber in 1 min at 280 °C (Figure 4).
In summary, for optimum in situ derivatization and HS-SPME
4
.7 µmol (1.6 mg) of TBA-HSO
on the yield of the derivatization is shown by the increase in the
peak area obtained with the use of TBA-HSO at the optimized
4
. The effect of the ion-pairing agent
4
conditions. The responses increased up to 90-fold for some of the
compounds (Table 1).
Effect of Ionic Strength. For many organic analytes, aqueous
solubility decreases with increasing ionic strength, and thus, the
partitioning from the aqueous solution to the headspace is
improved.20 To raise the ionic strength, an inorganic salt is often
added to the aqueous matrix. The most common salts used in
SPME are sodium chloride and sodium sulfate; however, sodium
chloride is not recommended when analyzing DBPs due to the
presence of bromide as an impurity which can enhance the
amount of bromide-HAAs in the sample.34 So, sodium sulfate was
chosen. As expected, the amount of analyte adsorbed onto the
(
34) Xie, Y. Effect of sodium chloride on DBP analytical results, Division of
Environmental Chemistry, American Chemical Society Annual Conference;
Aug 21-26, 1995, Chicago, IL; Extended Abstract.
2 4
sampling of HAAs from water, 3.5 M Na SO , 4.7 µmol of TBA-
4870 Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

replicates of a standard mixture containing the nine esters at
concentrations between 10.8 and 108 µg/ mL on one day and on
three different days, respectively. Good precision was achieved
with relative standard deviations (RSDs) for run-to-run precision
between 2.5 and 6.6% and for day-to-day precision from 4.0 to 8.4%.
Parameters of the in situ derivatization/ HS-SPME/ GC/ ITMS
method were evaluated using the optimized conditions. In the first
step, quantification with external calibration without internal
standard, as is frequently used in SPME, was tested. The precision
was determined for a set of five replicates of Milli-Q water spiked
with HAAs and analyzed consecutively on one day and on three
different days (Table 2). RSDs for run-to-run precision ranged
between 9.8 and 13.9% and for day-to-day precision between 10.1
and 15.6%. The run-to-run precision of the optimized SPME
method (expressed as RSDs) was not as good as that reported
7
for the EPA method 552.2 for water spiked at similar concentra-
tions of HAAs (Table 2).
To increase the precision, the effect of addition of an internal
standard was evaluated. In the first step, the use of 1,2-dibro-
mopropane, recommended as an internal standard by the Standard
Method 6251B was tested. This compound achieved equilibrium
in 30 min, and quantitative desorption from the fiber occurred in
1
min. However, the response depended on the amount of HAAs
in the aqueous sample. For instance, a decrease in the 1,2-
dibromopropane area of 34% was observed when the concentration
of HAAs rose from 18 to 405 µg/ L. To eliminate these problems,
a compound similar to the HAAs, 2,3-dibromopropionic acid, was
then chosen as an internal standard. This compound achieved
equilibrium in 35-40 min, and quantitative desorption from the
fiber was observed in 1 min. Moreover, satisfactory precision for
five determinations (RSD lower than 6%) was obtained. In addition,
an invariable response of this compound was observed at different
concentrations of HAAs in the water sample. Using this internal
standard, all HAAs showed RSD values for run-to-run precision
of less than 10%, except for MCAA (10.9%) and day-to-day
precisions (RSDs) lower than 11.4%. The linearity of the optimized
HS-SPME/ GC/ ITMS method was examined over the range 0.1-
300 µg/ L, expressed as the initial concentration of HAAs in water.
The linear range was established from the curves obtained by
plotting the relative area of each methyl haloacetate to that of the
internal standard (A/ Ais) versus the concentration of each HAA
Fig u re 4 . (a) Extraction time and (b) desorption time profiles of
methyl haloacetates by in situ methylation HS-SPME/GC/ITMS at 55
°
C. Fiber and conditions as in Figure 2, with 60 µL of DMS and 4.7
µmol of TBA-HSO4 as ion-pairing agent. In (b) extraction time was
5 min. Compound identification: ([) MCAA, (9) DCAA, (2) TCAA,
4) MBAA, (f) DBAA, (O) TBAA, (]) BCAA, (0) BDCAA, (b) CDBAA.
3
(
(
Table 2). Most methyl haloacetates showed good linearity and
2
For MCAA, MBAA, and TBAA methyl esters, the scale is shown on
the right.
correlations (r g 0.995). Detection limits were calculated using
Milli-Q water spiked at low levels of HAAs and analyzed using
the optimized procedure. In these experimental conditions, LODs
were from 0.01 to 0.45 µg/ L, which are 1.8- to 25-fold lower than
those obtained with the EPA method 552.2. Moreover, LODs using
this procedure were 71 to 2000 times lower than those reported
by Aikawa and co-workers31 for MCAA, DCAA, and TCAA using
in situ derivatization with methanol/ HCl and GC/ ECD determi-
nation.
4
HSO , and 60 µL of DMS were added to 10 mL of water. The
sample was maintained at 55 °C stirred at 1200 rpm and a CAR-
PDMS fiber was exposed to the headspace for 35 min.
Linearity, P recision, and Sensitivity Study. Calibration
parameters of the GC/ ITMS were determined by conventional
injection. Linear dynamic ranges were from 0.45 to 450 ng injected,
depending on the compound. Limits of detection (LOD), defined
as the concentration of the analyte that produces a chromato-
graphic peak with a signal-to-noise ratio (S/ N) of greater than 3,
were evaluated in full-scan mode at optimized conditions and
ranged from 17 pg for DBAA and BCAA methyl ester to 156 pg
for TBAA methyl ester. The run-to-run and day-to-day precision
of the system was assessed by consecutively analyzing 10
Analysis of Water Samples. To examine the feasibility of the
HS-SPME method, two water samples with different amounts of
HAAs, one from Barcelona’s water distribution system (<10 µg/
L) and the other from a swimming pool (10-150 µg/ L), were
analyzed. HAAs were determined in triplicate using the optimized
HS-SPME method as well as LLE (EPA method 552.2). Internal
standard calibration was used in both methods. HS-SPME/ GC/
Analytical Chemistry, Vol. 72, No. 20, October 15, 2000 4871

Ta b le 2 . Lin e a r Ra n g e , De t e c t io n Lim it s , a n d P re c is io n fo r In S it u Me t h yla t io n /HS -S P ME/GC/ITMS Me t h o d
precisionb
EPA method
552.2
run-to-runc
day-to-dayd
linear range
(µg/ L)
corr coeff
LODa
(µg/ L)
target value
(µg/ L)
with int std
(without)
with int std
(without)
LODa,f
(µg/ L)
2
run-to-runb,e
compd
(r )
MCAA
DCAA
TCAA
MBAA
DBAA
TBAA
BCAA
BDCAA
CDBAA
1.00-140
0.50-75
0.20-100
1.30-150
0.50-60
1.35-180
0.50-120
1.00-135
1.50-250
0.998
0.998
0.997
0.999
0.995
0.997
0.999
0.996
0.998
0.20
0.07
0.02
0.40
0.01
0.45
0.01
0.15
0.40
13.5
6.83
6.80
13.5
6.92
13.5
6.66
6.68
13.3
10.9 (13.9)
8.0 (12.2)
9.0 (9.8)
9.8 (12.1)
7.2 (11.1)
9.3 (12.9)
6.3 (11.5)
7.8 (11.8)
7.0 (11.1)
11.4 (14.4)
8.4 (12.9)
9.3 (10.1)
10.7 (12.5)
8.6 (13.4)
10.0 (15.6)
6.8 (13.4)
8.1 (13.8)
7.3 (11.8)
13
11
8.3
11
6.0
7.6
9.3
9.6
7.6
0.60
0.24
0.20
0.20
0.20
1.5
0.25
0.40
0.75
a
LOD, limit of detection. Precisions expressed as RSDs (%). c n ) 5. d n ) 5 replicates × 3 days. n ) 7 (reagent water spiked between 2.00
b
e
f
and 20.0 µg/ L). The LOD is defined as a level of a compound in a sample yielding a peak in the final extract with a signal-to-noise (S/ N) ratio
of approximately five, whichever is greater.
ITMS was found to be highly selective for the analysis of HAAs
in both drinking water and swimming-pool water. No interferences
from other compounds that may have been in the sample matrix
were detected in these conditions, see Figure 5 for swimming-
pool water. The results obtained for the water samples using HS-
SPME and LLE are given in Table 3. The total HAAs concentration
found in the tap water (∼30 µg/ L) is lower than the MCL (60
µg/ L) established for the USEPA for the sum of 5 HAAs, whereas
for swimming pool water the concentration rose to ∼300 µg/ L,
more than 10-fold that of tap water. As for HAA speciation, several
factors may influence the results: chlorine dose, reaction time,
natural organic matter content, and bromide concentration.
Trihalogenated acetic acids, especially TCAA, constituted the
greatest fraction of the total HAA concentration in swimming-pool
water (81%), whereas for tap water the tri- and dihalogenated
species were similar (45 and 50%, respectively) (Table 3). On the
other hand, the fraction of monohalogenated species was practi-
cally negligible (from 1.2 to 5%) in both samples. To compare the
results of HS-SPME with those of LLE, the significance of the
mean values was studied statistically using the Student’s t-test.
When unequal variances were obtained (F-test), Cochran’s test
was applied. The significance values (p) obtained are given in
Table 3. Generally, the results with HS-SPME using internal
standard agree with those obtained with LLE in both cases (P <
0.05). The reproducibility of both HS-SPME and LLE was generally
high for swimming-pool water. The coefficients of variation (up
to 17%) can be considered acceptable, considering matrix com-
plexity. In situ methylation/ HS-SPME/ GC/ ITMS showed some
advantages over LLE and GC/ ECD method, such as the avoidance
of organic solvents and labor-intensive sample manipulation steps.
Consequently, losses of analytes and analysis time are minimized.
In addition, lower detection limits were obtained with the proposed
method.
Fig u re 5 . HS-SPME/GC/ITMS total-ion chromatogram and single-
ion chromatograms of methyl haloacetates from swimming-pool water.
Compound identifications: 1, MCAA; 2, DCAA; 3, TCAA; 4, BCAA;
5, DBAA; 6, BDCAA; 7, CDBAA; and 8, TBAA methyl esters; IS, 2,3-
dibromopropionic acid methyl ester.
CONCLUSIONS
The feasibility of HS-SPME/ GC/ ITMS for the analysis of HAAs
in water after in situ derivatization with dimethyl sulfate has been
demonstrated. The CAR-PDMS fiber was found to be the most
effective coating for the analysis of HAAs methyl esters, especially
for monohalogenated ones. Maximum responses were obtained
using 10-mL water samples salted with sodium sulfate and set at
an equilibration time of 35 min at 55 °C. A large increase in the
4872 Analytical Chemistry, Vol. 72, No. 20, October 15, 2000

Ta b le 3 . Qu a n t it a t io n Re s u lt s fo r HAAs in Ba rc e lo n a Ta p Wa t e r a n d S w im m in g P o o l Wa t e r b y S P ME a n d EP A
Me t h o d 5 5 2 .2
tap water
headspace SPMEa,b
swimming pool water
headspace SPMEa,b EPA method 552.2a
EPA method 552.2a
mean
significance level (P-value)c
mean
mean
mean
(µg/ L)
swimming-pool
compd
MCAA
DCAA
TCAA
MBAA
DBAA
TBAA
(µg/ L)
RSD (%)
(µg/ L)
RSD (%)
(µg/ L)
RSD (%)
RSD (%)
tap water
water
1.43
8.27
5.64
n.d.
1.64
n.d.
4.29
5.28
1.90
28.4
8.3
12.8
7.1
n.d.
5.5
n.d.
8.9
13.9
5.9
1.43
9.10
6.34
n.d.
1.71
n.d.
4.18
6.32
2.22
31.3
8.6
7.6
9.9
n.d.
3.0
n.d.
3.9
4.22
45.2
155
14.8
16.7
17.0
n.d.
13.9
15.1
15.9
10.9
7.1
3.73
49.4
129
15.7
13.6
11.6
n.d.
10.9
17.4
9.8
0.993
0.981
0.203
0.377
0.520
0.155
n.d.
n.d.
2.76
18.9
10.5
60.6
32.8
330
2.22
17.5
11.9
61.2
32.5
307
0.312
0.109
0.643
0.352
0.617
0.903
BCAA
0.678
0.101
0.161
BDCAA
CDBAA
total HAAs
6.8
13.5
5.3
8.9
a
n ) 3. b Int std is 2,3-dibromopropionic acid. c Significant differences between procedures for P < 0.05 (at the 95% confidence level); n.d., not
detected.
ACKNOWLEDGMENT
reaction yield was obtained using TBA-HSO
4
as modifier for the
This study was supported by CICYT-Spain project number
AMB97-0405 and the Pla de Recerca de Catalunya (Generalitat
de Catalunya), project number 1999SGR00049. M.N. Sarri o´ n
acknowledges a Ph.D. fellowship from the University of Barcelona.
in situ methylation with dimethyl sulfate. HS-SPME in conjunction
with GC/ ITMS gave good precision; it was linear over 2 orders
of magnitude and the detection limits were at the low ppb level.
The method is proposed as an alternative to the liquid-liquid
extraction (EPA method 552.2) for the analysis of HAAs in
aqueous matrixes containing either low or high HAAs levels.
Received for review April 26, 2000. Accepted July 19,
2000.
AC000479D
Analytical Chemistry, Vol. 72, No. 20, October 15, 2000 4873