Analytica Chimica Acta 678 (2010) 183–188
Contents lists available at ScienceDirect
Analytica Chimica Acta
journal homepage: www.elsevier.com/locate/aca
Enrichment of polycyclic aromatic hydrocarbons in seawater with magnesium
oxide microspheres as a solid-phase extraction sorbent
Jing Jin a,b , Zhiping Zhang c , Yun Li a , Peipei Qi a,b , Xianbo Lu a , Jincheng Wang a , Jiping Chen a,∗ , Fan Su a
a
b
c
Dalian Institute of Chemical Physics, Chinese Academy of Sciences, No. 457 Zhongshan Road, Dalian 116023, China
Graduate University of Chinese Academy of Science, Beijing 100039, China
Department of Chemistry and Biochemistry, Brigham Young University, Provo 84602, USA
a r t i c l e
i n f o
Article history:
Received 2 June 2010
Received in revised form 18 August 2010
Accepted 20 August 2010
Available online 21 September 2010
Keywords:
Polycyclic aromatic hydrocarbons
Solid-phase extraction
Magnesium oxide
Enrichment
Water
a b s t r a c t
The enrichment of polycyclic aromatic hydrocarbons (PAHs) in water samples with magnesium oxide
(MgO) microspheres was evaluated, and four 3–5-ring PAHs were used as probes to validate the adsorption capacity of the material. Factors affecting the recovery of PAHs were investigated in detail, including
the type and concentration of organic modiﬁers, elution solvents, particle size of the adsorbent, volume
and ﬂow rate of the samples, and the lifetime of MgO cartridges. The recoveries of four PAHs extracted
from 20 mL of seawater spiked with standard PAHs ranged from 85.8% to 102.0% under the optimised
conditions. The limits of detection varied from 1.83 ng L−1 to 16.03 ng L−1 , indicating that the analytical
method was highly sensitive. Additionally, the proposed method was successfully used to enrich PAHs
in seawater. Compared to conventional methods, the proposed method consumed less organic modiﬁer
(5% acetone), and cheaper sorbents with comparable extraction efﬁciency were employed.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
In recent years, polycyclic aromatic hydrocarbons (PAHs), a class
of organic compounds with two or more fused aromatic rings, have
drawn a considerable amount of attention due to their carcinogenicity and mutagenicity [1,2]. They are formed mainly as a result
of pyrolytic processes, especially the incomplete combustion of
organic materials during industrial activities. For instance, PAHs
are formed from the processing of coal and crude oil, the combustion of natural gas and refuse, and vehicular trafﬁc. Moreover, they
can be formed from natural processes such as carbonisation. The
United States Environmental Protection Agency had designated 16
PAHs as priority pollutants.
However, due to their non-polar structure and extremely low
water solubility, the concentration of PAHs in natural water is
extremely low. To improve the sensitivity of analysis, solid-phase
extraction (SPE) techniques are often used to enrich PAHs at trace
levels in various environmental matrices such as soil [3,4], water
[5–7], and air [8]. Sorbents are the most important component
Abbreviations: PAHs, polycyclic aromatic hydrocarbons; MgO, magnesium
oxide; SPE, solid-phase extraction; PHEN, phenanthrene; ANTH, anthracene;
PYR, pyrene; BaP, benzo[a]pyrene; ACN, acetonitrile; MeOH, methanol; DCM,
dichloromethane; HPLC, high-performance liquid chromatography; LOD, limits of
detection; LOQ, limits of quantiﬁcation; RSD, relative standard deviation; PF, preconcentration factor.
∗ Corresponding author. Tel.: +86 411 84379562; fax: +86 411 84379562.
E-mail address: [email protected] (J. Chen).
0003-2670/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.aca.2010.08.028
of SPE; thus, these materials have received a signiﬁcant amount
of attention from various research groups. A variety of different
SPE sorbents have been developed, including silica-based matrices
[9–12], multiwalled carbon nanotubes [13], octadecyl functional
magnetic ferrite microspheres [7], molecularly imprinted polymers
[6], and tetradecanoate hemimicelles [14]. In the past few years,
C18 bonded silica has become the most important SPE adsorbent.
To improve the recoveries of 2–6-ring PAHs, optimal parameters for
various C18 SPE cartridges have been determined [4,10–12,15]. For
example, Busetti and coworkers [4] tested three different sorbents
(StrataE, StrataM, and Supelclean Envi-18) and demonstrated that
StrataE exhibited superior recovery (83–102%) and reproducibility
for a variety of different PAHs; however, poor results (83–87%) were
obtained for 5–6-ring PAHs. Interactions between organic analytes
and C30, a hydrophobic SPE sorbent, are superior to that of other
materials; thus, Li et al. [9] developed a method for the determination of PAHs in airborne particulates and obtained satisfactory
results for high-ring PAHs using C30 bonded silica. Recently, Krupadam et al. [6] successfully synthesised a molecularly imprinted
polymer using standard PAHs as a template, and obtained excellent recoveries of nearly 100% for PAHs standard and 71–98%
for environmental samples using ﬂuorescence spectrophotometry,
including coastal sediments, atmospheric particulates, and industrial efﬂuents. Additionally, Liu et al. [7] and Ballesteros-Gómez
and Rubio [14] prepared C18 magnetic ferrite microspheres and
tetradecanoate hemimicelles on Fe3 O4 nanoparticles to determine
the concentration of PAHs in aqueous samples, respectively. The
results of the aforementioned study indicated that the proposed
184
J. Jin et al. / Analytica Chimica Acta 678 (2010) 183–188
extraction method was an effective procedure for sample pretreatment. Although these novel materials could improve recovery rates,
multiple steps must be conducted to properly prepare samples.
Therefore, new, simple, and alternative SPE methods for the enrichment of PAHs in water samples must be developed.
In a previous report, MgO microspheres were successfully prepared by seed-induced precipitation, and excellent efﬁciency in the
separation of PAHs was obtained [16]. In contrast to conventional
silica, MgO displays unique selectivity for the separation of basic
compounds in normal-phase liquid chromatography, including aniline, quinoline, and pyridine derivatives [17]. The high adsorption
capacity for electron-rich compounds is attributed to interactions
between oxygen vacancies at the surface or sub-surface of MgO and
lone pairs of electrons on the analyte [17]. In addition, MgO microspheres were successfully used to preconcentrate benzo[a]pyrene
(BaP) in environmental water, and recovery rates of 94–101% were
obtained [18]. However, the performance of MgO as a SPE sorbent
for enriching multifold PAHs has not yet been investigated. In this
study, the extraction efﬁciencies of MgO for four 3–5-ring PAHs
were studied systematically. Parameters affecting the extraction
efﬁciency of the material were investigated in detail, including the
type and concentration of organic modiﬁers, elution solvents, particle size of the adsorbent, volume and ﬂow rate of the sample, and
the lifetime of MgO cartridges. In addition, the optimised method
was used to enrich PAHs in seawater.
2. Experimental
2.1. Chemicals and materials
Magnesium nitrate (Mg(NO3 )2 ) and potassium carbonate (K2 CO3 ) were purchased from Kermel Chemical
Reagent Co., Ltd. (Tianjin, China). Sodium polyphosphate and
Mg5 (CO3 )4 (OH)2 ·4H2 O seeds were obtained from Beijing Chemical
Reagent Co., Ltd. (Beijing, China) and Beijing Shuanghuan Chemical Reagent Factory (Beijing, China), respectively. HPLC-grade
acetonitrile (ACN) and methanol (MeOH) were purchased from
Shandong Yuwang Industrial Co., Ltd. Chemical Branch (Yuwang,
China), and HPLC-grade dichloromethane (DCM), isopropanol
and n-hexane were purchased from Tianjin Chemical Reagents
Institute (Tianjin, China), Shenyang Chemical Reagents Institute
(Shenyang, China), and Tianjin Kermel Chemical Reagents Co.,
Ltd. (Tianjin, China), respectively. Reagent grade ethyl acetate
and acetone were purchased from Shenyang Chemical Reagents
Factory (Shenyang, China). Phenanthrene (PHEN), anthracene
(ANTH), pyrene (PRY), and BaP (>98%) were obtained from Fluka
(Buchs, Switzerland), Beijing Chemworks (Beijing, China), BDH
Laboratory Chemicals Group (Poole, England), and Fluka (Buchs,
Switzerland), respectively. Ultrapure water was obtained from
a Millipore system, and Sep-Pak cartridges (Vac C18 3cc) were
purchased from Waters (Milford, MA, USA).
2.2. Preparation of standard PAH solutions
20.52 g of Mg(NO3 )2 ·6H2 O was dissolved in 100 mL of ultrapure
water. The resulting solution was transferred to a 500-mL threenecked ﬂask and heated to 70 ◦ C. In a separate ﬂask, 0.3 g of sodium
polyphosphate was added to 200 mL of a 0.4 M solution of K2 CO3 ,
and the mixture was heated to 70 ◦ C. Under vigorous stirring, the
K2 CO3 solution was added into the solution of Mg(NO3 )2 over 4–5 s.
Next, two drops of 0.4 mg mL−1 of Mg5 (CO3 )4 (OH)2 ·6H2 O, which
had been dispersed in water by ultrasonication, was rapidly added
to the reaction mixture, and the resulting solution was stirred for
1 min. The mixture was maintained at 70 ◦ C for 1 h under static condition, and the temperature of the reaction was increased to 100 ◦ C
for 1 h. Finally, the product was collected, ﬁltered, and washed with
ultrapure water and ethanol. The material was calcined in air from
room temperature to 300 ◦ C at a rate of 2 ◦ C min−1 for 2 h. Upon
completion, the product was calcined from room temperature to
550 ◦ C in a mufﬂe furnace for 7 h to obtain MgO microspheres.
2.4. Chromatographic measurements of PAHs on MgO column
The SPE sorbent was packed into stainless-steel tubes by applying a slurry of MgO microspheres (100 mm × 4.6 mm i.d.). Standard
solutions of PHEN, ANTH, PYR, and BaP were analysed with a Waters
high-performance liquid chromatography (HPLC) system. The ﬂow
rate of the mobile phase (100% of MeOH) was set to 0.8 mL min−1 .
2.5. SPE procedure
2.5.1. Preparation of SPE columns
The SPE columns were prepared by packing 0.2 g of spherical
MgO into an empty cartridge (6 mL, polypropylene). Upper and
lower frits were placed at each end of the cartridge to hold the
packing material in place. Next, the cartridge was conditioned by
sequentially eluting with DCM (5 mL), organic modiﬁers (4 mL, e.g.,
acetone, MeOH, and isopropanol), and a solution of ultrapure water
and organic modiﬁers (4 mL of 5:95 (v/v); 15:85 (v/v); 25:75 (v/v),
respectively).
2.5.2. The analysis of synthetic water samples via SPE
To obtain synthetic water samples, aliquots of organic modiﬁers
(e.g., 0, 5, 15, and 25% (v/v) of MeOH, acetone, and isopropanol,
respectively) and 1 mL of PAH standard #1 were added to ultrapure water, and the resulting suspension was ultrasonicated. The
synthetic water samples were passed through each cartridge at a
speciﬁc ﬂow rate (e.g., 1.0, 2.5, and 5.0 mL min−1 ). After loading the
samples, the cartridges were washed with 3 mL of ultrapure water
and were dried thoroughly with a vacuum pump. Subsequently,
the target compounds were eluted with different elution solvents
(2 × 2.5 mL of DCM, n-hexane, or ethyl acetate) and the volume
of the elutant was evaporated to <100 ␮L with a gentle stream of
nitrogen. Finally, the volume of the elutant was reconstituted to
2 mL with 80% (v/v) ACN–H2 O, and the solution was stored in 2 mL
Teﬂon-lined screw capped glass vials at 4 ◦ C.
Stock solutions of PHEN
ANTH
PYR
(72 mg L−1 ), and BaP (16 mg L−1 ) were prepared in MeOH, respectively. PAHs standard #1 (0.28 mg L−1 PHEN, 0.22 mg L−1 ANTH,
0.63 mg L−1 PYR, and 0.33 mg L−1 BaP) and PAHs standard #2
(54 ␮g L−1 PHEN, 17 ␮g L−1 ANTH, 144 ␮g L−1 PYR, and 9.5 ␮g L−1
BaP) were prepared by diluting the standard stock solution with
ACN. The solutions were stored in a refrigerator at 4 ◦ C to reduce
losses by evaporation.
2.5.3. The analysis of spiked seawater via SPE
Prior to SPE, seawater was ﬁltered through a solvent ﬁltration
apparatus to remove suspended sediments and solid materials.
To prepare 20 mL of spiked seawater, a procedure similar to the
method described in Section 2.5.2 was employed; however, ultrapure water and PAHs standard #1 was replaced with seawater
and PAHs standard #2, respectively. SPE was carried out under the
optimised conditions, and the elutant was treated as described in
Section 2.5.2.
2.3. Preparation of MgO microspheres
2.6. Instrumental analysis
MgO microspheres were prepared according to the method
described in our previous report [16]. To synthesise MgO spheres,
Pretreated samples were analysed with a Waters HPLC system
consisting of a Waters 2695 separations module and a Waters
(72 mg L−1 ),
(68 mg L−1 ),
J. Jin et al. / Analytica Chimica Acta 678 (2010) 183–188
185
3.2.1. Optimisation of the type and concentration of organic
modiﬁers
In SPE, the non-reversible adsorption of hydrophobic PAHs on
sorbents often leads to low recoveries. Thus, to improve the efﬁciency of sample preparation, organic modiﬁers are often added
to water samples. In agreement with the results of previous studies, in the present investigation, the type and concentration of
organic modiﬁers had a signiﬁcant effect on the extraction efﬁciency of PAHs [4,9,10,13,15]. In general, many organic solvents
have been used as organic modiﬁers, and high recovery rates have
been obtained. Namely, isopropanol [10], MeOH [10,13], acetone
[9], and ACN [15] have been used as organic modiﬁers. Moreover, in
previous studies, 25% (v/v) isopropanol or 30% (v/v) ACN provided
the greatest enrichment of PAHs on Sep-Pak C18 [10,15]. Alternatively, for C30 bonded silica, the best results were obtained with
20% acetone [9]. Furthermore, on multiwalled carbon nanotubes,
maximum recovery rates were obtained with 15% MeOH [13].
Therefore, the type and amount of organic modiﬁers are important
factors affecting the enrichment of PAHs in water samples.
In the present study, to improve the extraction efﬁciencies of
PAHs on MgO, different concentrations (e.g., 0%, 5%, 15%, and 25%)
of three organic modiﬁers (MeOH, acetone, and isopropanol) were
investigated. As shown in Fig. 1A, the recovery rates of the PAHs
in the absence of organic modiﬁers ranged from 73% to 96%, which
is signiﬁcantly higher than those obtained from silica-based SPE
sorbents (54–84%) [19]. In general, as the MeOH content increased
from 0% to 5%, the recovery of PAHs increased. However, a further increase in the MeOH concentration led to variable results,
depending on the type of PAHs. For example, as the MeOH concentration increased from 5% to 25%, the recoveries of PHEN, ANTH,
and PYR gradually decreased from 99.6%, 87.4%, and 93.4% to 72.3%,
67.8%, and 75.1%, respectively. Alternatively, as the MeOH concentration increased from 5% to 15%, the recovery of BaP increased from
75.7% to 87.0%. However, the recovery of BaP gradually decreased
to 81.3% as the MeOH content increased from 15% to 25%. Similar trends were observed for other organic modiﬁers (e.g., acetone
and isopropanol), as shown in Fig. 1B and C. However, the addition
of acetone and isopropanol led to highly variable results. Organic
modiﬁers may enhance the recovery rates of PAHs by improving the
solubility of PAHs in water and increasing the eluotropic strength
of the sample. As a consequence, the breakthrough volume of PAHs
996 photodiode array detector/a Waters 474 scanning ﬂuorescence detector. The data obtained from the photodiode array
detector were collected at a wavelength of 254 nm. The ﬂuorescence excitation and emission wavelengths were modiﬁed
during the chromatographic separation to obtain better sensitivity. The following excitation/emission wavelengths were employed
throughout the chromatographic run: 275/350 nm at 0 min,
260/420 nm at 4.3 min, 270/440 nm at 4.9 min, and 290/430 nm
from 7.5 min to 20 min. PAHs were separated on a SUPELCOSILTM
LC-PAH HPLC column (150 × 4.6 mm i.d., 5 ␮m) at 30 ◦ C using a gradient elution of ACN [0–2 min: ACN–H2 O (80:20), 2–7 min: linear
gradient to 100% ACN; 7–17 min: 100% ACN; 17–18 min: linear gradient to ACN–H2 O (80:20); 18–20 min: ACN–H2 O (80:20)]. During
the separation process, 100 ␮L of sample was injected onto the
HPLC with an auto sampler. A personal computer equipped with
Waters Empower software was used to acquire and process chromatographic data, and the peak area was used to determine the
concentration of the analyte.
3. Results and discussion
3.1. Retention of PAHs on MgO
As shown in Table 1, the retention capacity of the target compounds on MgO increased from 0.28 to 3.48 as the number of
aromatic rings increased from 3 to 5. The relationship between the
number of rings and the retention capacity may be related to the
interaction between oxygen vacancies at the surface or sub-surface
of MgO and the ␲ electrons in the PAHs, which become stronger [17]
as the number of aromatic rings increases. As a result, the retention
capacity of PAHs increases with an increase in the number of rings.
Thus, the results suggested that MgO has the potential to provide
excellent recoveries of high-ring PAHs [4]. Moreover, the results
obtained in the present study are consistent with those of previous
investigations [18].
3.2. Optimisation of the parameters affecting SPE
In SPE, recovery is affected by various parameters, including the
elution solvent, particle size of the adsorbent, ﬂow rate, sample
volume and the type and concentration of organic modiﬁers.
Table 1
Comparisons of capacity factors for PAHs on MgO column.
3-Ring
Structure
Degree of unsaturation
Capacity factor
4-Ring
5-Ring
PHEN
ANTH
PYR
BaP
10
0.30
10
0.28
12
0.53
15
3.48
Fig. 1. The effect of the type and concentration (e.g., 0, 5%, 15%, and 25%) of organic modiﬁer (A. MeOH, B. acetone, C. isopropanol) on extraction efﬁciency. Conditions:
loading volume = 20 mL, ﬂow rate = 1.0 mL min−1 , elution solvent = 2 × 2.5 mL DCM.
186
J. Jin et al. / Analytica Chimica Acta 678 (2010) 183–188
Table 2
Parameters of MgO and effects of particle size on extraction efﬁciency (n = 3).
2
−1
Surface area (m g )
Pore diameter (nm)
Pore volume (cm3 g−1 )
Average recovery ± SD (PHEN)
Average recovery ± SD (ANTH)
Average recovery ± SD (PYR)
Average recovery ± SD (BaP)
Fig. 2. The effect of elution solvents on extraction efﬁciency. Conditions: loading
volume = 20 mL, ﬂow rate = 1.0 mL min−1 , elution solvents = 2 × 2.5 mL.
is reduced, which is conducive to lower recoveries [10]. Alternatively, interactions between MgO and PAHs may be dependent on
the number of aromatic rings. Thus, as the number of aromatic rings
increases, the strength of the interaction between MgO and PAHs
increases [17,18,20]. As a result, the addition of an organic modiﬁer
has a stronger effect on the breakthrough volume of 3–4-ring PAHs
than that of 5-ring PAHs. In other words, the recoveries of PHEN,
ANTH, and PYR were more sensitive to the concentration of organic
modiﬁer than that of BaP.
The addition of 5% acetone (v/v) provided the greatest recovery
rates, and the maximum recoveries of PHEN, ANTH, PYR, and BaP
were 110.7%, 96.1%, 110.1%, and 102.7%, respectively. Compared
to the concentration of organic modiﬁers used in previous studies [9,10,13,15], signiﬁcantly less organic solvent was used in the
present investigation, and high recoveries were achieved. Thus, the
proposed method is more environmentally friendly than traditional
SPE methodologies.
3.2.2. Optimisation of elution solvents
PAHs are non-polar contaminants, and their hydrophobicity
increases with an increase in molecular weight. Therefore, the elution solvent has a signiﬁcant effect on the recovery rate. In the
present study, three organic solvents (2 × 2.5 mL of n-hexane, ethyl
acetate, and DCM) were evaluated to investigate the effects of elution solvent on the recovery rates of PAHs. As shown in Fig. 2, DCM
yielded the highest recoveries (96–111%), and n-hexane provided
acceptable results (59–80%). Alternatively, when ethyl acetate
was used as the elution solvent, poor recoveries were obtained
(36–56%). Thus, the results indicated that DCM was the most appropriate solvent for the elution of 3–5-ring PAHs from MgO cartridges.
However, DCM, n-hexane, and toluene provided comparable results
for low-ring PAHs. Alternatively, with C30 bonded silica, DCM provided superior results for high-ring PAHs [9]. In contrast, n-hexane
and DCM provided comparable results for the elution of 4-ring PAHs
from Sep-Pak C18 cartridges [15]. In our opinion, solvents may have
different effects on various sorbents because the mechanism of
retention and elution is dependent on the sorbent [18].
3.2.3. Effect of sorbent particle size on extraction efﬁciency
In general, interactions between sorbents and analytes can be
enhanced by increasing the number of interaction sites and the
surface area of the sorbent. Li et al. [9] demonstrated that the
recovery of analytes from C30 was more sensitive to the ﬂow rate
than C18 because the average particle size of C30 is 7–8 times
larger than that of C18. To investigate the effect of particle size
on the recoveries of PAHs, two different particle sizes were evaluated. To control the particle size of the sorbent, the material was
sieved with 200 and 320-mesh sieves. The surface area and pore
parameter of the sorbents were measured with a N2 porosime-
200-Mesh
320-Mesh
128.8
16.1
51.8
90.2%
85.9%
94.6%
83.6%
133.1
16.9
56.2
110.7%
96.1%
110.1%
102.0%
±
±
±
±
9.5%
8.1%
1.3%
1.8%
±
±
±
±
17.9%
16.0%
0.1%
14.7%
ter (NOVA 4000). Prior to the measurements, the samples were
degassed at 300 ◦ C under a ﬂow of nitrogen for approximately
5 h. Nitrogen adsorption–desorption data were recorded at the
temperature of liquid nitrogen (77 K), and the results are provided in Table 2. The speciﬁc surface area was calculated using
the Brunauer–Emmett–Teller equation, and the size distribution of
meso-pores was determined via microspore analysis. As expected,
the surface area of 320-mesh particles (133.1 m2 g−1 ) was slightly
larger than that of 200-mesh particles (128.8 m2 g−1 ). The pore
diameters of the particles were 16.1 nm and 16.9 nm, and the pore
volumes were 51.8 cm3 g−1 and 56.2 cm3 g−1 . The two sorbents
were used to enrich PAHs in water samples, and the results are
provided in Table 2. In summary, the recovery rates of PAHs from
smaller particles were greater (96–111%) than those (84–95%) of
larger particles. Thus, 320-mesh MgO was employed in subsequent
experiments.
3.2.4. Optimisation of loading volumes
In previous studies [6,7,9–11], various loading volumes (20 mL,
50 mL, 100 mL, and 1 L) have been evaluated to optimise the recovery rate of an analyte or to determine the concentration of an
analyte in spiked samples. To study the effect of the loading volume on the recoveries of PAHs on MgO, the volume of synthetic
water samples was varied, and the recovery rates of PAHs were
determined. However, all of the samples contained equal amounts
of PAHs.
Fig. 3 shows the effect of the loading volume on the recoveries of
PAHs. For 3-ring PAHs (PHEN and ANTH), the recovery of PHEN and
ANTH increased from 84% to 111% and 96% as the loading volume
increased to 20 mL. However, as the loading volume increased from
20 mL to 100 mL, the recovery of PHEN and ANTH decreased to 82%
and 81%, respectively. Alternatively, the recoveries of 4-ring and
5-ring PAHs (PYR and BaP) displayed signiﬁcantly different results.
Namely, under the same conditions, the recoveries of 4-ring and
5-ring PAHs were greater than those of 3-ring PAHs. Speciﬁcally, as
the loading volume increased from 100 mL to 250 mL, the recover-
Fig. 3. The effect of loading volume on extraction efﬁciency. Conditions: different volumes of synthetic water samples containing equal amounts of PAHs, ﬂow
rate = 1.0 mL min−1 , elution solvent = 2 × 2.5 mL DCM.
J. Jin et al. / Analytica Chimica Acta 678 (2010) 183–188
Fig. 4. The effect of sample ﬂow rate on extraction efﬁciency. Conditions: loading
volume = 20 mL, elution solvent = 2 × 2.5 mL DCM.
ies of 4- and 5-ring PAHs were greater than 90%. However, under
identical conditions, the recoveries of 3-ring PAHs were approximately 80%. Although the recoveries of 4-ring and 5-ring PAHs were
greater than those of 3-ring PAHs, the general trends were identical.
Moreover, as the loading volume increased from 10 mL to 250 mL,
the preconcentration factor (PF), which is an important indicator
of a successful SPE, increased from 2 to 50. After evaporation and
dissolution of the residue in 2 mL of ACN–H2 O (80/20, v/v), the PF
increased from 10 to 125, and satisfactory recoveries were obtained
(80.0–110.7%). The results of the present study were comparable to
those (PF: 20 or 120, recovery: 69–97%) obtained in previous investigations [9,10]. However, the PF obtained in the present work was
lower than that (PF: 1000) obtained by Busetti et al. [4]. To achieve
high recovery rates and to reduce the analysis time, the loading
volume was set to 20 mL in subsequent experiments.
3.2.5. Optimisation of the sample ﬂow rate
In general, the complete adsorption of the target compound can
be achieved at low ﬂow rates; however, a decrease in the ﬂow rate
increases the analysis time. Moreover, a high ﬂow rate may result
in the loss of analytes due to incomplete adsorption. Therefore,
the sample ﬂow rate is a crucial factor in SPE, and a suitable ﬂow
rate must be selected to achieve high recovery rates and to reduce
loading times.
To determine the optimal conditions, the sample ﬂow rate was
varied between 1.0 and 5.0 mL min−1 , and the recovery rates of
PAHs were evaluated. As shown in Fig. 4, the recoveries of 3-ring
PAHs (PHEN and ANTH) were 91–106%, and the recoveries of 4–5rings PAHs (PYR and BaP) were 98–109%. The results indicated that
a sample ﬂow rate of 1.0–5.0 mL min−1 did not have a signiﬁcant
effect on the recoveries of PAHs. Similarly, in a previous report,
Junk and Richard [11] used C18 bonded silica as a SPE sorbent and
demonstrated that the ﬂow rate did not affect the recoveries of
PAHs. Even at a very high ﬂow rates (e.g., 25 mL min−1 ), signiﬁcant
differences in recovery rates were not observed. Alternatively, as
the ﬂow rate increased from 1.5 mL min−1 to 13.3 mL min−1 , the
recovery rates of 16 PAHs on C30 bonded silica increased to a
maximum value and then subsequently decreased [9]. The authors
187
Fig. 5. The lifetime of MgO cartridges. Conditions: loading volume = 20 mL, ﬂow
rate = 5.0 mL min−1 , elution solvents = 2 × 2.5 mL DCM.
suggested that the recovery rates of PAHs were dependent on the
ﬂow rate due to the relatively large particle size and compactness
of C30 bonded silica. In the present study, the weak effect of ﬂow
rate on the recovery of BaP may also be related to the small size of
MgO.
3.3. The lifetime of MgO SPE columns
The adsorption stability of MgO was tested by consecutive contact with synthetic water samples. As shown in Fig. 5, the best
recovery rates were obtained during the ﬁrst use of the MgO cartridge. Nevertheless, reasonable recovery rates were obtained after
three enrichment and desorption cycles (80–87%). However, upon
further use of the MgO cartridge, signiﬁcantly reduced performance
was observed. Thus, to maintain high recovery rates and to prevent interferences from impurities, the MgO cartridge should be
only used once. Despite the limited number of uses, MgO has great
potential as a SPE sorbent because the material can be prepared by
a simple synthetic method. Moreover, compared to traditional procedures, the proposed method consumes less organic solvent (5%
acetone) and is based on a cheaper SPE sorbent.
3.4. Determination of PAHs in spiked seawater and real seawater
To evaluate the proposed method in practical applications, seawater spiked with trace amounts of PAHs was enriched. Using
the optimised conditions, 20 mL of spiked seawater containing 5%
acetone (v/v) was pretreated with MgO. As shown in Fig. 6A, the
HPLC chromatogram of seawater after pretreatment was so clean
that the sensitivity of the analysis was improved. Although different concentrations of standard PAHs (#1 and #2) were employed,
the recoveries of PAHs from seawater ranged from 85.8% to
94.9% (Table 3), indicating that MgO exhibited excellent extraction efﬁciencies for PAHs over a broad concentration range. The
repeatability of the proposed method was satisfactory, and the relative standard deviation (RSD) of three runs was less than 12.5%.
In addition, a calibration curve was constructed by measuring
different concentrations of standard solutions, and the result-
Table 3
Recovery of PAHs in spiked seawater (n = 3), linear range, regression coefﬁcient, limits of quantiﬁcation (LOQ) and detection (LOD) of the method.
Linear range (␮g L−1 )
Recovery (%) ± RSD (%)
#1a
PHEN
ANTH
PYR
BaP
a
b
91.8
88.3
94.9
85.8
Regression coefﬁcient (r2 )
LOD (ng L−1 )
LOQ (ng L−1 )
0.9961
0.9999
1.0000
0.9999
16.03
2.62
28.8
1.83
53.43
8.73
96.0
6.08
#2b
±
±
±
±
3.0
0.9
1.8
1.9
102.0
99.7
89.6
86.2
±
±
±
±
12.5
6.1
0.8
5.0
2.7–108
0.68–13.6
0.36–144
0.0024–9.50
Spiked seawater using PAHs standard #1, detected by Waters 996 photodiode array detector.
Spiked seawater using PAHs standard #2, detected by Waters 474 scanning ﬂuorescence detector.
188
J. Jin et al. / Analytica Chimica Acta 678 (2010) 183–188
ANTH were 0.5 and 2.5 times greater when MgO was employed,
respectively. Moreover, the peak areas of PYR and BaP remained
constant and decreased by one-half when MgO was used to enrich
the PAHs, respectively. Thus, the extraction efﬁciencies of MgO
and Sep-Pak C18 for 4- and 5-ring PAHs were comparable. Moreover, the proposed method was successfully used to determine the
concentration of PAHs in seawater. The results shown in Table 4
indicated that MgO could be used to enrich PAHs and to determine
the concentration of PAHs in environmental water.
4. Conclusions
High extraction efﬁciencies for PAHs (85.8–102.0%) were
obtained with the proposed method. In the absence of organic
modiﬁers, the recoveries (73–96%) of the proposed method were
signiﬁcantly greater than that of a commercial C18 cartridge
(54–84%). Under the optimised conditions, the proposed method
was successfully used to enrich PAHs in seawater. In addition, the
results revealed that the method could be used to preconcentrate
PAHs in environmental water samples. In particular, the method
could be used to enrich high-ring PAHs. In general, the proposed
method provided reasonable extraction efﬁciencies (85.8–102.0%),
consumed less organic modiﬁers (5% acetone), and employed
cheaper sorbents than conventional methods. After SPE, overlapping peaks were observed in the PHEN region (e.g., silica and MgO);
thus, elution conditions must be optimised to remove interferences
from the matrix.
Fig. 6. (A) HPLC chromatograms of standard PAHs, seawater, and spiked seawater
pretreated with MgO (the equivalent concentration of the PAHs spiked in seawater
was: 2.7 ␮g L−1 PHEN, 0.85 ␮g L−1 ANTH, 7.2 ␮g L−1 PYR, and 0.475 ␮g L−1 BaP). (B)
HPLC chromatogram of seawater pretreated with MgO and Sep-Pak C18 cartridges
under the optimised conditions.
Table 4
Determination of PAHs in seawater (n = 3).
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (Grant Nos. 20775081, 20907051) and the
Chinese Ministry of Science and Technology 863 Project (No.
2007AA061601).
References
Seawater
PHEN
ANTH
PYR
BaP
Concentration (␮g L−1 )
RSD (%)
6.82
0.241
0.525
0.023
4.06
5.70
7.33
3.35
ing regression coefﬁcients (r2 ) were greater than 0.996 (PHEN:
2.7–108 ␮g L−1 ; ANTH: 0.68–13.6 ␮g L−1 ; PYR: 0.36–144 ␮g L−1 ;
BaP: 0.0026–9.5 ␮g L−1 ). The limit of detection (LOD) and the limit
of quantiﬁcation (LOQ), which was deﬁned as a signal to noise ratio
of 3:1 and 10:1, respectively, were obtained by measuring the minimum concentration of the standard solutions of the calibration
curve.
Under the optimised conditions, 20 mL of seawater containing 5% (v/v) acetone was directly loaded onto the MgO cartridge.
To compare the extraction efﬁciencies of PAHs on MgO and SepPak C18, 20 mL of seawater containing 25% (v/v) isopropanol was
loaded onto a Sep-Pak C18 cartridge, and optimal conditions were
employed [10]. Fig. 6B shows the HPLC chromatogram of seawater pretreated with MgO and a Sep-Pak C18 cartridge. Compared to
the results obtained with Sep-Pak C18, the peak areas of PHEN and
[1] R.M. Santella, M. Stefanidis, Chemical Analysis of Polycyclic Aromatic Compounds, Wiley, New York, 1989.
[2] T. Toyooka, Y. Ibuki, Environ. Toxicol. Pharmacol. 23 (2007) 256–263.
[3] T. Reemtsma, J. Mehrtens, Chemosphere 35 (1997) 2491–2501.
[4] F. Busetti, A. Heitz, M. Cuomo, S. Badoer, P. Traverso, J. Chromatogr. A 1102
(2006) 104–115.
[5] N. Li, H.K. Lee, J. Chromatogr. A 921 (2001) 255–263.
[6] R.J. Krupadam, B. Bhagat, S.R. Wate, G.L. Bodhe, B. Sellergren, Y. Anjaneyulu,
Environ. Sci. Technol. 43 (2009) 2871–2877.
[7] Y. Liu, H.F. Li, J.M. Lin, Talanta 77 (2009) 1037–1042.
[8] G. Carrera, P. Fernandez, R. Vilanova, J.O. Grimalt, J. Chromatogr. A 823 (1998)
189–196.
[9] K. Li, H.F. Li, L.B. Liu, Y. Hashi, T. Maeda, J.M. Lin, J. Chromatogr. A 1154 (2007)
74–80.
[10] G. Kiss, Z. Varga-Puchony, J. Hlavay, J. Chromatogr. A 725 (1996) 261–272.
[11] G.A. Junk, J.J. Richard, Anal. Chem. 60 (1988) 451–454.
[12] H. Kicinski, S. Adamek, A. Kettrup, Chromatographia 28 (1989) 203–208.
[13] W.D. Wang, Y.M. Huang, W.Q. Shu, J. Cao, J. Chromatogr. A 1173 (2007) 27–36.
[14] A. Ballesteros-Gómez, S. Rubio, Anal. Chem. 81 (2009) 9012–9020.
[15] M.S. García-Falcón, M. Pérez-Lamela, J. Simal-Gándara, J. Agric. Food Chem. 52
(2004) 6897–6903.
[16] Z.P. Zhang, Y.J. Zheng, J.P. Chen, Q. Zhang, Y.W. Ni, X.M. Liang, Adv. Funct. Mater.
17 (2007) 2447–2454.
[17] Z. Zhang, Y. Zheng, J. Zhang, J. Chen, X. Liang, J. Chromatogr. A 1165 (2007)
116–121.
[18] J. Jin, Z.P. Zhang, J.C. Wang, P.P. Qi, J.P. Chen, J. Sep. Sci. 33 (2010) 1836–1841.
[19] R.M. Marcé, F. Borrull, J. Chromatogr. A 885 (2000) 273–290.
[20] R. Sadowski, W. Wójcik, J. Chromatogr. 262 (1983) 455–459.