Вы находитесь на странице: 1из 7

Talanta 168 (2017) 291–297

Contents lists available at ScienceDirect

Talanta
journal homepage: www.elsevier.com/locate/talanta

Determination of phthalates in bottled water by automated on-line solid MARK


phase extraction coupled to liquid chromatography with uv detection
Daniel Salazar-Beltrán, Laura Hinojosa-Reyes, Edgar Ruiz-Ruiz, Aracely Hernández-Ramírez,

Jorge Luis Guzmán-Mar
Universidad Autónoma de Nuevo León, UANL, Facultad de Ciencias Químicas, Cd. Universitaria, San Nicolás de los Garza, Nuevo León C.P. 66455, Mexico

A R T I C L E I N F O A BS T RAC T

Keywords: An on-line solid phase extraction coupled to liquid chromatography with UV detection (SPE/LC-UV) method
Bottled water was automated by the multisyringe flow-injection analysis (MSFIA) system for the determination of three
Phthalates phthalic acid esters (PAEs). The PAEs determined in drinking water stored in polyethylene terephthalate (PET)
On-line solid-phase extraction bottles of ten commercial brands were dimethyl phthalate (DMP), diethyl phthalate (DEP) and dibutyl phthalate
MSFIA automation
(DBP). C18-bonded silica membrane was used for isolation and enrichment of the PAEs in water samples. The
LC-UV detection
calibration range of the SPE/LC-UV method was 2.5–100 μg L−1 for DMP and DEP and 10–100 μg L−1 for DBP
Migration
PET containers with correlation coefficients (r) ranging from 0.9970 to 0.9975. Limits of detection (LODs) were between 0.7
and 2.4 μg L−1. Inter-day reproducibility performed at two concentration levels (10 and 100 μg L−1) expressed
as relative standard deviation (%RSD) were found in the range of 0.9–4.0%. The solvent volume was reduced to
18 mL with a total analysis time of 48 min per sample. The major species detected in bottled water samples was
DBP reaching concentrations between 20.5 and 82.8 μg L−1. The recovery percentages for the three analytes in
drinking water were 80–115%. The migration test showed a great variation in the sum of migrated PAEs level
(10.2–50.6 μg L−1) among the PET bottle brands analyzed indicating that the presence of these contaminants in
the plastic containers may depend on raw materials and the conditions used during their production process.

1. Introduction Due to the human health and environmental adverse effects caused
by this group of compounds, certain PAEs as DMP, DEP and DBP have
The dimethyl phthalate (DMP), diethyl phthalate (DEP) and dibutyl been identified as priority hazardous substances by the European
phthalate (DBP) belongs to the family of phthalates or phthalic acid Union (EU) and the United States Environmental Protection Agency
esters (PAEs). These compounds are extensively used as a plastic (EPA) controlling their use as plasticizers in articles intended to come
additives particularly in the poly(vinyl) chloride (PVC) and polyethy- into contact with food [11,12]. Thus, it is necessary to develop faster
lene terephthalate (PET) to improve properties such as flexibility, and efficient methodologies for determination of PAEs in bottled water
transparency and durability of products produced with these polymeric and for evaluation of their migration rate from plastic containers into
matrices [1–3]. Human exposure to phthalates can occur as a result of water. Several studies have been proven that phthalates could migrate
their direct contact and use or indirectly through leaching into other from the plastics used to storage drinks [13–15] indicating the
products, or general environmental contamination. The diet has been importance of use a standardized method to evaluate the migration
considered the major source of PAEs exposure [4,5]. PAEs are degree from plastic packaging material into foods. The European Union
recognized as potential endocrine disruptors, can bioaccumulate in (EU) 82/711/EEC and 85/572/EEC directives describe the basic rules
tissues, and are poorly biodegradable [6]. Because PAEs molecules are necessary for testing migration of the constituents of plastic materials
not chemically attached to polymer chains [7], these compounds can and articles intended to come into contact with foodstuffs specifying
easily migrate from the plastic packaging into food, drinks and water the use of food simulants, the contact time and exposure temperature.
[8]. The migration rate of PAEs are related to physicochemical factors The EU specific migration limit for DBP is 0.3 mg kg−1 of simulant
as temperature, radiation, solubility, diffusion coefficient, pressure, and [16,17].
presence of solvents and additives [9,10], resulting in the ubiquitous PAEs have been commonly analyzed by gas chromatography (GC)
presence of these plasticizers in drinking water. and liquid chromatography (LC) methods using mass spectrometric


Corresponding author.
E-mail address: jorge.guzmanmr@uanl.edu.mx (J. Luis Guzmán-Mar).

http://dx.doi.org/10.1016/j.talanta.2017.03.060
Received 31 January 2017; Accepted 17 March 2017
Available online 19 March 2017
0039-9140/ © 2017 Elsevier B.V. All rights reserved.
D. Salazar-Beltrán et al. Talanta 168 (2017) 291–297

Fig. 1. On-line SPE/LC-UV system used for phthalates determination exploiting MSFIA system for automated SPE procedure.

(MS) and UV detection [8,18–22]. Despite of the fact that GC-MS 2. Experimental
presents higher resolution and offers good separation for the environ-
mental analysis of PAEs in several matrices including drinking water, 2.1. Chemicals and material
LC-UV is a good alternative to analyze PAEs since it is a more
affordable technique that ensure good performance and comparable Double-distilled deionized water was used to prepare the mobile
limits of detection (LODs) than those performed by GC-MS and LC-MS phases (HPLC grade, TEDIA, Fairfield, OH, USA). All standards used
methods [8,23,24]. A sample preconcentration step is usually required were analytical grade quality. Dimethyl phthalate (DMP, purity 99.5%),
prior to chromatographic analysis in order to obtain low LODs and diethyl phthalate (DEP, purity 99.5%) and dibutyl phthalate (DBP,
good selectivity of PAEs when their concentration levels are low in purity 99.5%) standards were obtained from Sigma Aldrich (St. Louis,
environmental matrices. LC-UV has been widely combined with solid MO, USA). Acetonitrile and methanol were high-performance liquid
phase extraction (SPE) for analyte enrichment and sample clean-up. chromatography (HPLC) grade and were purchased from TEDIA
However, most of the reported SPE methods are used off-line [25–27]. (Fairfield, OH, USA). Solid phase extractions were performed with
The major problems of off-line SPE are risks of sample loss and 47 mm diameter×0.5 mm thickness×10 μm (nominal) particle size 3M
contamination. Thus, implementation of an on-line SPE step followed Empore membrane disks containing C18 (octadecyl-bonded silica),
by LC-UV detection brings advantages for PAEs analysis. On-line SPE/ SDB-RPS (poly(styrenedivinylbenzene) copolymer modified with sul-
LC-UV is based on the selective extraction of the analyte on an fonic acid groups) or SDB-XC (poly(styrenedivinylbenzene) copoly-
extraction column and its transfer to an analytical column by using a mer). The membrane disks were purchased from 3M Empore (St. Paul,
two-position column switching valve in a closed and usually automated MN, USA). Stock solutions of each phthalate were prepared in water
system. In this way, the use of flow analysis techniques as multisyringe with a certain percentage of acetonitrile, DMP (0%), DEP (6%) and
flow-injection analysis (MSFIA) for automation allows high sample DBP (12%). A series of standard solutions were prepared by mixing an
throughput and injection reproducibility, reduces analytical costs and appropriate amount of the stock solutions with water in a 50 mL
waste production, and minimize human intervention [28]. The on-line volumetric flask. All the standard solutions were stored at 4 °C and
SPE/LC-UV through MSFIA technique has been described for the protected from light in amber glass bottles. All glassware was thor-
screening of phenolic pollutants in water and environmental samples oughly washed, rinsed and sonicated with methanol to remove
using for preconcentration step polystyrene-divinylbenzene sorbent ubiquitous pollutants.
[29]. The MSFIA system has been applied by Boonjob et al. for
determination of chlorotriazine herbicides in environmental waters 2.2. Samples
and crude soil extracts by on-line SPE/LC method using dispersed
carbon nanotubes (CNTs) as the sorptive material [30]. Ten samples of bottled water in plastic bottles from ten commercial
Thus, the aim of this work was to develop an automated precon- brands were randomly collected from a market in Monterrey City,
centration stage by a multisyringe flow injection analysis (MSFIA) Mexico. The requirements for them to be eligible were that they were
coupled to LC-UV detection for the determination of DMP, DEP and bottles made with PET, that the volume they contained was in the
DBP in drinking water stored in PET bottles using reversed-phase C18- range of 400–750 mL and that their date of expiration was not over. All
bonded silica extraction disks. The procedure was also applied to study samples were stored in amber glass bottles in a refrigerator at 4 °C until
the migration of the three PAEs from PET bottles into water using the analysis.
migration test conditions described by EU Directives 82/711/EEC and
85/572 EEC [16,17]. To the best of our knowledge, no on-line SPE 2.3. MSFIA manifold and on-line SPE /LC-UV procedure
method with membrane disks through MSFIA for PAEs assay in
drinking water and for evaluation of PAEs migration from PET The MSFIA manifold used for the automatization of SPE is shown
containers into water has been reported to date. schematically in Fig. 1, consisted of a multisyringe piston pump with
programmable speed (MicroBu 4 S, Crison Instruments, Alella,
Barcelona, Spain) was used as a liquid driver. It was equipped with
one high-precision 5 mL bidirectional syringe (S) (Hamilton, Bonaduz,
Switzerland). The S is connected to the central port of an eight-port

292
D. Salazar-Beltrán et al. Talanta 168 (2017) 291–297

Table 1
Protocol sequence for the automated on-line solid phase extraction of PAEs from water.

Step Description SV operation

SV operation LC system SV position Volume (mL) Flow (mL min−1)

1 Syringe is filled with acetonitrile 4 5 10


2 Acetonitrile is sent through the membrane for 5 5 0.5
conditioning
3 Syringe is filled with sample 7 18 10
4 Sample is loaded to the preconcentration unit where 5 18 0.5
analytes retention take place
5 Syringe is filled with acetonitrile 4 1 10
6 Acetonitrile is sent through the membrane eluting The loop of the injection valve is filled with the eluent 5 1 0.5
the retained analytes
7 Syringe is filled The segment of eluent trapped in the loop is injected and 1 5 10
the chromatographic run starts.

multiposition selection valve (SV) (Sciware Systems, Bunyola, Spain), not detected (N.D.). The recovery of PAEs from water samples was
which is responsible for changing positions to pick up or dispense any determined by fortifying sample with 15 μg L−1 of each PAE followed
of the reagents used during the preconcentration step. Position 5 of the by on-line SPE extraction and LC–UV analysis under the optimized
SV is connected to the preconcentration unit that consists of a operational condition. All experiments were done in triplicate.
cylindrical PEEK (polyether ether ketone) block with a central conical
cavity that was 1.1 cm in diameter. A 1 cm i.d. extraction disk (C18) 2.4. Migration tests
was placed in the central conical cavity to enable the fluids used in SPE
to pass through. A 20 µm average pore diameter polyethylene frit A migration study based on EU 82/711/EEC and 85/572 EEC
(Sigma Aldrich, St. Louis, MO, USA) was attached to the conical section directives [16,17] was carried out on PET plastic bottles used to store
of the gap to prevent distortion of the membrane. water from different brands. The bottles were cut into small pieces of
The protocol sequence adopted for the automation of solid-phase approximately 1 cm2; 5g of plastic were weighed and placed in glass
extraction process is described in Table 1. The complete sequence vials to which 50 mL of HPLC grade water was added. The samples
included seven steps. 5 mL of acetonitrile were loaded into the holding were shaken at 100 rpm in a thermostatic bath with reciprocal stirring,
loop (step 1) and dispensed through the membrane in order to which allowed an adequate contact between the plastic and water,
condition it (step 2). Afterwards, 18 mL of sample were picked up into during 10 days at 40 °C. After 10 days, the plastic was separated from
the holding loop (step 3) and then dispensed through the preconcen- the water, and water samples were stored in amber flasks at 4 °C for
trator where analytes retention takes place (step 4). Subsequently, further analysis.
1 mL of acetonitrile was loaded into the holding loop (step 5) and then
dispensed through the membrane in order to remove the analytes 3. Results and discussion
retained in the membrane filling the loop of the HPLC injection valve
(step 6). The steps 1–7 were repeated three times for each analyzed A simple, environmentally friendly and fully automated method
sample. The reusability of each membrane was of 10 cycles. was developed for the separation, preconcentration and quantification
Then, 300 μL of the eluent were manually injected into the of three phthalates (DMP, DEP and DBP) in drinking water samples
chromatographic system beginning the chromatographic run and with an analysis time of 48 min per sample.
finally, the syringe was refilled (step 7).
Chromatographic separation of PAEs after on-line elution from SPE 3.1. Chromatographic analysis
membrane was performed on a reversed phase C18 column (5 µm,
4.6×250 mm) from Agela Technologies protected by a Phenomenex Reversed-phase liquid chromatography with UV detection is an
Gemini C18 guard column (5 µm, 4×3 mm). The mobile phase was analytical tool widely used in the determination of PAEs employing
acetonitrile: water (CH3CN:H2O) with the following gradient elution solvents as methanol and acetonitrile as organic modifiers of the
program: 0.0–5.5 min (70:30, v/v); 5.5–12.0 min (100:0, v/v); and mobile phase [25,32]. The absorption spectrum of the studied analytes
finally 12.0–16.0 min (70:30, v/v); the last step allowed to stabilize the showed that the maximum absorption wavelength was 230 nm and was
system to the initial conditions. The flow rate of the mobile phase was due to the π-π* transition of the C-C aromatic rings and n-π* of the
1 mL min−1. The UV monitoring wavelength was chosen at 230 nm. C=O bond [33]. Thus, detection was performed at 230 nm in order to
The sample injection volume was 300 μL and the separation was obtain best sensitivity and precision. Due to the different polarity of the
carried out at room temperature. Autoanalysis v5.0 software analytes a gradient elution was adopted. To determine the composition
(Sciware, Palma de Mallorca, Spain) was used for instrument control, of the mobile phase, two organic modifiers (methanol and acetonitrile)
data acquisition, and data processing. Raw data were saved and were evaluated in gradient elution mode to observe their effect on
processed by use of the Origin v8 Professional software package retention time (tR) and signal intensity of each of the analytes.
(OriginLab Corporation, Northampton, MA, USA). Quantitative analy- A lower efficiency was observed in the chromatographic separation,
sis was based on peak area. External standard calibration was used to when using methanol, due to this solvent has a lower elution power
quantitate PAEs. The amount of analyte was determined using a ten- (ε°=0.95) than acetonitrile (ε°=0.65) [33]; whereby acetonitrile was
point calibration curve injected by triplicate in the range of 2.5– selected as organic modifier of the mobile phase for PAEs separation.
100 µg L−1 for DMP and DEP, and from 10 to 100 µg L−1 for DBP. The working conditions of the LC-UV system are described in the
According to the definition of IUPAC, the limits of detection (LOD) and experimental part. Under these conditions, chromatogram for standard
quantification (LOQ) were calculated as three and ten times, respec- species of DMP, DEP and DBP was with sharp and symmetric peaks
tively, the standard deviation of six blank samples divided by the slope and well separated (as seen in Fig. 2A). The retention times for DMP,
of the calibration curve [31]. Values less than the LOQ were reported as DEP and DBP were 5.0, 6.6 and 10.8 min respectively, with total

293
D. Salazar-Beltrán et al. Talanta 168 (2017) 291–297

0,05 study, a volume of 2 mL of standard solution containing the three


(C)
0,04 analytes at a level of 0.5 mg L−1 was pumped through the system at a
3 flow rate of 0.5 mL min−1. The analytes were eluted by 1 mL of the
0,03 selected solvent at an elution rate of 0.5 mL min−1. As can be observed
0,02 2 in Fig. 3A, a C18-bonded silica membrane was more suited for
1 extracting PAEs from water samples allowing recoveries ranged from
0,01 100% to 102% than polymeric membranes. The lower recoveries
0,00 percentages obtained using SDB-RPS (44% for DMP, 40% for DEP
and 50% for DBP) and SDB-XC (62% for DMP, 33% for DEP and 81%
-0,01 for DBP) can be due to the mixed-mode of the polymer membranes
0,05 which have more capacity for polar compounds, therefore C18 mem-
(B)
0,04 brane was selected for further experiments.
Then the volume of solvent required for the elution of the retained
0,03
Absorbance

analytes was evaluated using 0.5, 0.75 and 1.0 mL of 100% acetonitrile
0,02 as eluent while keeping all the other experimental variables as
3 described previously. In Fig. 3B can be observed a comparison of the
0,01
normalized peak areas of the analytes using different eluent volumes.
0,00 The extraction efficiency for the three analytes increased while
increasing the eluent volume, selecting 1.0 mL. The extraction effi-
-0,01
ciencies were 106%, 105% and 106% for DMP, DEP and DBP,
0,05 3
(A) respectively.
0,04 For the preconcentration extraction flow rate; three different flow
rates (0.5, 0.75 and 1.0 mL min−1) were evaluated using 2 mL contain-
0,03
ing the three PAEs at 0.5 mg L−1, in Fig. 3C can be observed that as the
1 2
0,02 flow increases, the normalized peak areas decreases, which may be due
to the low contact time of the analytes with the membrane making a
0,01
potential detrimental effect on the efficiency of the SPE process.
0,00 Whereby 0.5 mL min−1 was selected as the extraction flow rate allow-
ing the highest recovery percentages (99–102%).
-0,01
0 3 6 9 12 15 The last parameter evaluated in the preconcentration stage was the
maximum sample volume that can be preconcentrated by the mem-
Time (min)
brane using 10, 15, 18 and 20 mL containing the three analytes at
Fig. 2. Chromatogram of the on-line SPE/LC-UV of the samples: (A) standard solution 100 μg L−1 (Fig. 3D). The use of a larger sample volume increased the
of PAEs mixture at 15 µg L−1, (B) water before and (C) water after spiking with 15 µg L−1 preconcentration factor of the method. Good recoveries between 81%
of each analyte [DMP(1), DEP (2). DBP (3)]. and 112% were obtained for preconcentrating the three analytes from
10 up to 18 mL of sample volume. Nevertheless when sample volume
analysis time of 16 min including column stabilization. These results increased to 20 mL the recovery decreases (55–83%) which indicates
were comparable with some of the reported HPLC methods (analysis that the membrane is close to saturation and the analyte breakthrough
time ranged from 8 to 41 min) [8,9,23,34,35]. is high. A further increase on the sample volume will just increase the
analysis time, so that 18 mL was selected as the maximum sample
3.2. Implementation of the preconcentration stage by MSFIA system volume.
In Table 2 are presented the final experimental conditions of the
The development of highly sensitive analysis techniques for the SPE/LC-UV system for the preconcentration and determination of
quantification of PAEs in food matrices is necessary because they are phthalates.
present at very low concentrations [9,27]. Thus, a pretreatment stage is
needed to extract and concentrate the PAEs and to improve the 3.3. Analytical performance
sensitivity of the analysis.
The use of membranes was explored as an alternative to the packed The method for PAEs determination in water was validated under
columns for the implementation of the preconcentration stage through the selected experimental conditions for linearity, limits of detection
the automated SPE/LC-UV method. The application of membranes (LODs), limits of quantification (LOQs), inter-day repeatability (RSD),
decreases the solvent consumption and does not present undesirable and accuracy (recovery). The analytical features are summarized in
overpressures allowing high flow rates and reducing risk of plugging. Table 3. Linearity was evaluated by injection of 300 μL of the ten
The result is a fast, efficient and reproducible extraction of the analytes concentration level standards analyzed using the on-line SPE/LC-UV
that facilitates the extraction of large sample volumes and enables system. In Table 3 can be observed that the linearity obtained for each
automation [36,37]. During this stage were evaluated different para- of the calibration curves was satisfactory with correlation coefficients
meters such as membrane type, eluent volume, sample volume and (r) ranging from 0.9970 to 0.9975. LODs were between 0.7 and
extraction flow rate; the results are shown in Fig. 3 using normalized 2.4 μg L−1 which were comparable with previous published works
peak area as a response. A factor by factor method was used for the [8,23,25,32]. Inter-day reproducibility studies were performed at two
optimization of the parameters affecting the extraction efficiency. concentration levels (10 and 100 μg L−1) and the experiments were
Three types of apolar membranes were assessed (C18, SDB-RPS carried out on triplicate for three continues days. The values of the
and SDB-XC): C18 is completely apolar membrane due to the hydro- relative standard deviation (%RSD) were found in the range of 0.9–
carbon chain bonded to the silica surface, SDB-XC is poly(styrenedi- 4.0%, indicating that the method shows good reproducibility. The
vinylbenzene) copolymer (PS-DVB) which shows moderate polarity enrichment factors (EF), defined as the ratio of the slopes of the
with affinity for moderately polar water soluble analytes and SDB-RPS samples before and after the preconcentration stage [38], were between
is PS-DVB modified with sulfonic acid groups showing hydrophilic 50.8 and 67.4 showing the high performance of the developed
character with higher selectivity for analytes of polar nature. For this automated SPE procedure using C18 membrane disks.

294
D. Salazar-Beltrán et al. Talanta 168 (2017) 291–297

(A) (B)

1.0 1.0

0.8 0.8
DMP

Peak area
Peak area

DMP
0.6 0.6
DEP
DEP
0.4 DBP 0.4
DBP
0.2 0.2

0.0 0.0
C18 SDB-RPS SDB-XC 0.50 0.75 1.00
Membrane type Eluent volume (mL)

(C) (D)
1.0 1.0

0.8 0.8
Peak area

Peak area
0.6 DMP DMP
0.6
DEP DEP
0.4 0.4
DBP DBP
0.2 0.2

0.0 0.0
0.50 0.75 1.00 10 15 18 20
Extraction flow rate (mL/min) Sample volume (mL)
Fig. 3. Study of the effect of (A) membrane type, (B) eluent volume, (C) extraction flow rate, (D) Sample volume on the extraction of PAEs by using automated SPE/LC-UV system.

Table 2 with the mixture of phthalate standards at the concentration level of


Operating conditions of the SPE/LC-UV system for the preconcentration of PAEs in 15 μg L−1 to evaluate matrix effects and to verify the accuracy of the
water. developed system. The results obtained from the analysis of original
and fortified water samples are shown in Table 4. In the first case,
Parameter Condition
among the ten analyzed drinking water samples, DBP was detected in
Extraction disk 3 M Empore C18 (1 cm I.D.) all samples at concentration between 20.5 and 82.8 μg L−1. DMP was
Eluent Acetonitrile (1 mL) detected only in one of the ten analyzed samples at a concentration of
Sample volume 18 mL
2.9 μg L−1 and DEP was not detected in the analyzed samples. DBP has
Volume of eluent 1 mL
Flow rate 0.5 mL min−1 been detected in seven commercial brands of Jordanian mineral water
reaching concentrations between 1.2 and 13.5 μg L−1 [23], DBP has
also detected in bottled water from Ireland at concentrations under
0.07 μg L−1 [8].
Table 3
Analytical performance data for the PAEs by the SPE/LC-UV system.
The recoveries of each analyte were investigated fortifying each
water sample at 15 μg L−1 concentration level, and injected in tripli-
Parameter DMP DEP DBP cate. In this spiked level, the recoveries for the three analytes ranged
−1
from 80% to 115% and %RSD values were below 5.2%, demostrating
Calibration range (μg L ) 2.5–100 2.5–100 10–100
Regression equation y=0.0062x y=0.0061x y=0.0031x
that the accuracy of the measurements and precisions for the determi-
+0.0184 +0.0149 −0.0283 nation of PAEs in real samples is acceptable for trace analysis. An
r 0.9970 0.9973 0.9975 example of chromatograms of bottled water sample before and after
spiking is shown in Fig. 2B and C.
%RSDa 10 μg L−1 0.9 3.5 1.9
The proposed method was compared with other reported studies for
100 μg L−1 4.0 1.8 3.4
LOD (μg L−1) 0.7 1.2 2.4 the determination of PAEs. This method uses reduced solvent volume
LOQ (μg L−1) 2.3 3.9 7.9 (18 mL) compared to other works that apply on-line and off-line SPE
Enrichment factor 67.4 53.2 50.8 methods (5–80 mL) which indicates that proposed on-line SPE/LC-UV
a
system was environmentally friendly [20,23,25,27]. Although the LODs
n=3 replicates.
of the proposed method (0.7–2.4 μg L−1) were slightly poorer than GC
method for PAEs analysis in bottled water with LODs of 0.20 μg L−1 for
3.4. Accuracy of the method and analysis of bottled water samples DMP, 0.13 μg L−1 for DEP and 0.12 μg L−1 for DBP and RSD values
from 3.03% to 3.05% [23], were higher than the maximum limit
In order to validate the applicability of the developed on-line SPE/ established by the EPA (6 µg L−1) [11] but similar to other LC methods.
LC-UV method, it was applied to determine the analytes in bottled PAEs has been determined by Zhiyong (2010) in orange juice using
water samples. Because of the lack of certified reference materials, SPE and LC-UV reaching LODs of 2.6, 4.4 and 7.6 μg L−1 for DMP,
bottled water samples were enriched before the preconcentration stage DEP and DBP, respectively with RSD values from 2% to 4.7% [25]. The

295
D. Salazar-Beltrán et al. Talanta 168 (2017) 291–297

Table 4
Results obtained from the analysis of PAEs in bottled water samples and recoveries of these plasticizers from spiked water samples (n=3).

Sample Amount added (μg L−1) DMP DEP DBP

Found (μg L−1) ± S.D Recovery (%) Found (μg L−1) ± S.D Recovery (%) Found (μg L−1) ± S.D Recovery (%)

1 0 N.D 89 N.D 97 33.3 ± 0.7 115


15 13.3 ± 0.9 14.6 ± 0.8 50.6 ± 0.5

2 0 N.D 109 N.D 101 41.1 ± 1.7 103


15 16.3 ± 0.4 15.2 ± 0.9 56.6 ± 0.9

3 0 N.D. 102 N.D 97 43.6 ± 0.7 82


15 15.3 ± 0.02 14.6 ± 0.01 55.9 ± 0.01

4 0 N.D 107 N.D 105 82.8 ± 1.4 100


15 16.1 ± 1.7 15.7 ± 1.0 97.8 ± 1.3

5 0 N.D 86 N.D 102 64.5 ± 0.1 90


15 13.0 ± 0.5 15.3 ± 0.4 78.1 ± 0.5

6 0 2.9 ± 0.1 89 N.D 111 59.9 ± 1.5 113


15 16.3 ± 0.5 16.7 ± 0.8 76.9 ± 0.3

7 0 N.D 105 N.D 102 20.5 ± 1.0 108


15 15.7 ± 0.8 15.3 ± 0.3 36.8 ± 0.8

8 0 N.D 113 N.D 95 27.6 ± 0.5 100


15 16.9 ± 0.5 14.2 ± 0.5 42.5 ± 0.5

9 0 N.D 80 N.D 91 29.3 ± 0.4 108


15 12.0 ± 0.2 13.7 ± 1.0 45.5 ± 1.3

10 0 N.D 103 N.D 105 48.8 ± 1.2 82


15 15.4 ± 0.5 15.8 ± 0.4 61.1 ± 0.8

S.D. Standard deviation


N.D. Not detected.

LPME-LC-UV has been used for the determination of DBP in mineral Table 5
water by Sun (2013) reaching a LOD of 2.4 μg L−1 and a RSD of 6.2% Migration of PAEs (mean ± S.D.) in μg L−1 into bottled water at 40 °C for 10 daysa.
[32]. By comparison, the LODs achieved here were better that those by
Sample DMP DEP DBP
LLE-GC-MS for PAEs in bottled water (27.4 μg L−1 for DMP, 23 μg L−1
for DEP and 16.1 μg L−1 for DBP) with RSD values between 0.5% and 1 50.6 ± 0.3 N.D. N.D.
1.8% [8]. The preconcentration time of the proposed system was 2 N.D. N.D. 21.2 ± 0.2
10 min for membrane activation, 36 min for preconcentration and 3 4.1 ± 0.2 N.D. 14.5 ± 0.3
4 10.9 ± 0.001 N.D. 19.8 ± 0.1
2 min for elution of the analytes from the membrane, with a total time 5 25.2 ± 0.3 N.D. 24.6 ± 1.4
of 48 min per sample; thus, this method has the advantage of being 6 N.D. N.D. N.D.
simple, sensitive, and fully automated so that the analyst is not in 7 N.D. N.D. 25.6 ± 2.0
contact with the sample which helps to reduce contamination and 8 N.D. N.D. N.D.
9 N.D. N.D N.D
manipulation of the sample. Compared with other reported methods,
10 10.2 ± 0.5 N.D N.D
the repeatability and recovery of the method are satisfactory, and the
limits of detection could meet the requirement of specific migration N.D. Not detected.
a
limits of determination. S.D. for n=3 determinations.

present in five of the ten analyzed samples, with concentrations from


3.5. Migration of phthalates from PET bottles 4.1 μg L−1 to 50.6 μg L−1, whereas DEP was not detected in any of the
analyzed samples. Compared to the literature data, DMP and DEP at
The migration tests were performed on ten different commercial concentrations of 0.8 μg L−1 and 1.2 μg L−1, respectively were detected
brands PET bottles used for drinking water storage according to the EU in a migration study performed on PET bottles used for drinking water
directives [16,17]. The migrated PAEs content were determined by the storage [40]. DMP, DEP and DBP concentrations were found for in
on-line SPE/LC-UV method. The repeatability of the migration test (% bottles for keeping mineral water: 87, 85 and 312 μg L−1 in a migration
RSD) was determined by analyzing the migrated amount of the study reported by Ustun (2015). Keresztes (2013) reported a 0.6 μg L−1
analytes in water reaching values between 0.6% and 7.9%. migrated concentration for DBP in PET-bottled used for non-carbo-
Prior to the migration tests, the PAEs levels in water samples were nated mineral water [13].
measured and used as the baseline values for blank controls in
subsequent experiments. DBP was the only detected compound. The
results obtained from the PAEs migration test are summarized in 4. Conclusions
Table 5, where it can be observed that DBP is present in five of the ten
analyzed samples reaching concentrations from 14.5 μg L−1 up to A SPE/LC-UV system using C18-bonded silica membrane was
25.6 μg L−1. Any sample was over the DBP migration limit of implemented successfully. It was applied for the determination of
0.3 mg kg−1 of simulant stipulated by the EU [39]. DMP was also three phthalates (DMP, DEP, and DBP) in bottled water of ten

296
D. Salazar-Beltrán et al. Talanta 168 (2017) 291–297

commercial brands. The system presents not only low limits of come into contact with foodstuffs, in: E.E. Community (Ed.) Off. J. Eur. Union,
1982, pp. 1–10.
detection 0.7–2.4 μg L−1), but also high recoveries for the three [17] E. Union, Council Directive 85/572/EEC Laying down the list of simulants to be
analytes (80–115%). Additionally the %RSD values were below 5.2%, used for testing migration of constituents of plastic materials and articles intended
using reduced solvent volume (18 mL) in a total analysis time of to come into contact with foodstuffs, in: E.E. Community (Ed.) Off. J. Eur. Union,
1985, pp. 14–21.
48 min per sample. The advantages of this method in comparison with [18] L. Chang, P. Bi, X. Li, Y. Wei, Study of solvent sublation for concentration of trace
off-line SPE procedures are that it is a greener method, reducing not phthalate esters in plastic beverage packaging and analysis by gas chromatography-
only the use of solvents but also the interaction with the analyst. mass spectrometry, Food Chem. 177 (2015) 127–133.
[19] J. Xu, P. Liang, T. Zhang, Dynamic liquid-phase microextraction of three phthalate
Analysis of PAEs in bottled water samples indicated that DBP was the esters from water samples and determination by gas chromatography, Anal. Chim.
major species detected in all the analyzed samples at concentrations Acta 597 (2007) 1–5.
between 20.5 and 82.8 μg L−1. The migration test according to [20] C. Cheng, K.-C. Chang, Sensitive analysis of phthalate esters in plastic bottled water
via on-line capillary solid-phase microextraction liquid chromatography electro-
normative EU 82/711/EEC was performed for DMP, DEP and DBP
spray ionization-ion trap mass spectrometry, Anal. Methods 8 (19) (2016)
in PET bottles used for containing drinking water reaching total 3910–3919.
migrated PAEs concentrations between 10.2 and 50.6 μg L−1 indicating [21] W. Jia, X. Chu, Y. Ling, J. Huang, J. Chang, Analysis of phthalates in milk and milk
that the presence of these contaminants in the plastic containers may products by liquid chromatography coupled to quadrupole orbitrap high-resolution
mass spectrometry, J. Chromatogr. A 1362 (2014) 110–118.
depend on raw materials and the conditions used during their [22] Y. Fan, S. Liu, Q. Xie, Rapid determination of phthalate esters in alcoholic
production process. beverages by conventional ionic liquid dispersive liquid-liquid microextraction
coupled with high performance liquid chromatography, Talanta 119 (2014)
291–298.
Acknowledgement [23] M.F. Zaater, Y.R. Tahboub, A.N. Al Sayyed, Determination of phthalates in
Jordanian bottled water using GC-MS and HPLC-UV: environmental study, J.
The authors thank the financial support from the Facultad de Chromatogr. Sci. 52 (2014) 447–452.
[24] G. Tranfo, L. Caporossi, E. Paci, C. Aragon, D. Romanzi, C. De Carolis, M. De Rosa,
Ciencias Químicas of the Universidad Autónoma de Nuevo León, and S. Capanna, B. Papaleo, A. Pera, Urinary phthalate monoesters concentration in
the National Council of Science and Technology of México (CONACyT couples with infertility problems, Toxicol. Lett. 213 (2012) 15–20.
Ciencia Básica, project number 177990). D. Salazar-Beltrán is also [25] G. Zhiyong, W. Danyi, W. Meili, W. Sui, Determination of six phthalic acid esters in
orange juice packaged by PVC bottle using SPE and HPLC–UV: application to the
acknowledges CONACyT for a doctoral fellowship.
migration study, J. Chromatogr. Sci. 48 (2010) 760–765.
[26] X. Gao, B. Yang, Z. Tang, X. Luo, F. Wang, H. Xu, X. Cai, Determination of
References phthalates released from paper packaging materials by solid-phase extraction-high-
performance liquid chromatography, J. Chromatogr. Sci. 52 (2014) 383–392.
[27] J. Li, Y. Cai, Y. Shi, S. Mou, G. Jiang, Analysis of phthalates via HPLC-UV in
[1] K. Khosravi, G.W. Price, Determination of phthalates in soils and biosolids using environmental water samples after concentration by solid-phase extraction using
accelerated solvent extraction coupled with SPE cleanup and GC–MS quantifica- ionic liquid mixed hemimicelles, Talanta 74 (2008) 498–504.
tion, Microchem. J. 121 (2015) 205–212. [28] S. Clavijo, J. Avivar, R. Suárez, V. Cerdà, Analytical strategies for coupling
[2] M.J. Silva, A.R. Slakman, J.A. Reidy, J.L. Preau, A.R. Herbert, E. Samandar, separation and flow-injection techniques, TrAC Trends Anal. Chem. 67 (2015)
L.L. Needham, A.M. Calafat, Analysis of human urine for fifteen phthalate 26–33.
metabolites using automated solid-phase extraction, J. Chromatogr. B 805 (2004) [29] H.M. Oliveira, M.A. Segundo, J.L. Lima, V. Cerda, Multisyringe flow injection
161–167. system for solid-phase extraction coupled to liquid chromatography using mono-
[3] W.J. Peijnenburg, J. Struijs, Occurrence of phthalate esters in the environment of lithic column for screening of phenolic pollutants, Talanta 77 (4) (2009)
the Netherlands, Ecotoxicol. Environ. Saf. 63 (2006) 204–215. 1466–1472.
[4] Y. Guo, Q. Wu, K. Kannan, Phthalate metabolites in urine from China, and [30] W. Boonjob, M. Miro, M.A. Segundo, V. Cerda, Flow-through dispersed carbon
implications for human exposures, Environ. Int. 37 (2011) 893–898. nanofiber-based microsolid-phase extraction coupled to liquid chromatography for
[5] H. Fromme, L. Gruber, M. Schlummer, G. Wolz, S. Bohmer, J. Angerer, R. Mayer, automatic determination of trace levels of priority environmental pollutants, Anal.
B. Liebl, G. Bolte, Intake of phthalates and di(2-ethylhexyl)adipate: results of the Chem. 83 (13) (2011) 5237–5244.
integrated exposure assessment survey based on duplicate diet samples and [31] G. Chen, H. Hu, T. Wu, P. Tong, B. Liu, B. Zhu, Y. Du, Rapid and sensitive
biomonitoring data, Environ. Int. 33 (2007) 1012–1020. determination of plasticizer diethylhexyl phthalate in drink by diffuse reflectance
[6] Y. Gao, T. An, Y. Ji, G. Li, C. Zhao, Eco-toxicity and human estrogenic exposure UV spectroscopy coupled with membrane filtration, Food Control 35 (2014)
risks from OH-initiated photochemical transformation of four phthalates in water: 218–222.
a computational study, Environ. Pollut. 206 (2015) 510–517. [32] M. Sun, R. Tang, Q. Wu, C. Wang, Z. Wang, Graphene reinforced hollow fiber
[7] M. Rahman, C. Brazel, The plasticizer market: an assessment of traditional liquid-phase microextraction for the determination of phthalates in water, juice and
plasticizers and research trends to meet new challenges, Prog. Polym. Sci. 29 milk samples by HPLC, Anal. Methods 5 (2013) 5694–5700.
(2004) 1223–1248. [33] D.A. Skoog, F.J. Holler, T.A. Nieman, Principios de analisis instrumental, in: M.G.
[8] P. Otero, S.K. Saha, S. Moane, J. Barron, G. Clancy, P. Murray, Improved method Hill (Ed.), 2001.
for rapid detection of phthalates in bottled water by gas chromatography-mass [34] X. Liu, Z. Sun, G. Chen, W. Zhang, Y. Cai, R. Kong, X. Wang, Y. Suo, J. You,
spectrometry, J. Chromatogr. B 997 (2015) 229–235. Determination of phthalate esters in environmental water by magnetic Zeolitic
[9] P. Serodio, J.M. Nogueira, Considerations on ultra-trace analysis of phthalates in Imidazolate Framework-8 solid-phase extraction coupled with high-performance
drinking water, Water Res. 40 (2006) 2572–2582. liquid chromatography, J. Chromatogr. A 1409 (2015) 46–52.
[10] X. Ni, X. Xing, Y. Cao, G. Cao, Determination of phthalates in food packing [35] I. Ustun, S. Sungur, R. Okur, A. Sumbul, S. Oktar, N. Yilmaz, C. Gokce,
materials by electrokinetic chromatography with polymeric pseudostationary Determination of phthalates migrating from plastic containers into beverages, Food
phase, Food Chem. 190 (2016) 386–391. Anal. Methods 8 (2014) 222–228.
[11] D. Gao, Z. Li, Z. Wen, N. Ren, Occurrence and fate of phthalate esters in full-scale [36] V. Cerdà, L. Ferrer, J. Avivar, A. Cerdà, Flow Analysis: A practical guide, in:
domestic wastewater treatment plants and their impact on receiving waters along Elsevier (Ed.) 2014, p. 278.
the Songhua River in China, Chemosphere 95 (2014) 24–32. [37] C. Erger, T.C. Schmidt, Disk-based solid-phase extraction analysis of organic
[12] C. Pérez-Feás, M.C. Barciela-Alonso, P. Bermejo-Barrera, Presence of phthalates in substances in water, TrAC Trend Anal. Chem. 61 (2014) 74–82.
contact lens and cleaning solutions, Microchem. J. 99 (2011) 108–113. [38] J.M. Castor, L. Portugal, L. Ferrer, L. Hinojosa-Reyes, J.L. Guzman-Mar,
[13] S. Keresztes, E. Tatar E, Z. Czegeny, G. Zaray, V.G. Mihucz, Study on the leaching of A. Hernandez-Ramirez, V. Cerda, An evaluation of the bioaccessibility of arsenic in
phthalates from polyethylene terephthalate bottles into mineral water, Sci. Total corn and rice samples based on cloud point extraction and hydride generation
Environ. 458–460 (2013) 451–458. coupled to atomic fluorescence spectrometry, Food Chem. 204 (2016) 475–482.
[14] Q. Xu, X. Yin, M. Wang, H. Wang, N. Zhang, Y. Shen, S. Xu, L. Zhang, Z. Gu, [39] E. Union, Council Directive 85/572/EEC laying down the list of simulants to be
Analysis of phthalate migration from plastic containers to packaged cooking oil and used for testing migration of constituents of plastic materials and articles intended
mineral water, J. Agric. Food Chem. 58 (2010) 11311–11318. to come into contact with foodstuffs, in: E.E. Community (Ed.), Off. J. Eur. Union,
[15] M. Jeddi, N. Rastkari, R. Ahmadkhaniha, M. Yunesian, Concentrations of phtha- 2007, pp. 17–36.
lates in bottled water under common storage conditions: do they pose a health risk [40] I. Al-Saleh, N. Shinwari, A. Alsabbaheen, Phthalates residues in plastic bottled
to children?, Food Res. Int. 69 (2015) 256–265. waters, J. Toxicol. Sci. 36 (4) (2011) 469–478.
[16] E. Union, Council Directive 82/711/EEC Laying down the basic rules necessary for
testing migration of the constituents of plastic materials and articles intended to

297

Вам также может понравиться