1
J. Sep. Sci. 2014, 00, 1–8
Longhui Liu1 Lijun He1 Xiuming Jiang1 Wenjie Zhao1 Guoqiang Xiang1 Jared L. Anderson2 1 School
of Chemistry and Chemical Engineering, Henan University of Technology, Zhengzhou, China 2 Department of Chemistry, School of Green Chemistry and Engineering, The University of Toledo, Toledo, OH, USA Received September 24, 2013 Revised January 27, 2014 Accepted January 27, 2014
Research Article
Macrocyclic polyamine-functionalized silica as a solid-phase extraction material coupled with ionic liquid dispersive liquid–liquid extraction for the enrichment of polycyclic aromatic hydrocarbons In this study, silica modified with a 30-membered macrocyclic polyamine was synthesized and first used as an adsorbent material in SPE. The SPE was further combined with ionic liquid (IL) dispersive liquid–liquid microextraction (DLLME). Five polycyclic aromatic hydrocarbons were employed as model analytes to evaluate the extraction procedure and were determined by HPLC combined with UV/Vis detection. Acetone was used as the elution solvent in SPE as well as the dispersive solvent in DLLME. The enrichment of analytes was achieved using the 1,3-dibutylimidazolium bis[(trifluoromethyl)sulfonyl]imide IL/acetone/water system. Experimental conditions for the overall macrocycle-SPE–IL-DLLME method, such as the amount of adsorbent, sample solution volume, sample solution pH, type of elution solvent as well as addition of salt, were studied and optimized. The developed method could be successfully applied to the analysis of four real water samples. The macrocyclic polyamine offered higher extraction efficiency for analytes compared with commercially available C18 cartridge, and the developed method provided higher enrichment factors (2768–5409) for model analytes compared with the single DLLME. Good linearity with the correlation coefficients ranging from 0.9983 to 0.9999 and LODs as low as 0.002 g/L were obtained in the proposed method. Keywords: Dispersive liquid–liquid microextraction / Ionic liquids / Macrocyclic polyamines / Polycyclic aromatic hydrocarbons / Solid-phase extraction DOI 10.1002/jssc.201301062
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Nowadays, SPE has become one of the most common sample preparation techniques due to its ease of operation and lower consumption of organic solvent compared with liquid–liquid extraction. The technique has been successfully applied in many areas of analytical chemistry including environmental, biological, and food analysis [1, 2]. The composition of adsorCorrespondence: Dr. Lijun He, School of Chemistry and Chemical Engineering, Henan University of Technology, 450001 Zhengzhou, China E-mail: [email protected], [email protected] Fax: +86-371-67756718
Abbreviations: [BBIM][Tf2 N], 1,3-dibutylimidazolium bis [(trifluoromethyl)sulfonyl]imide; DLLME, dispersive liquid– liquid microextraction; EF, enrichment factor; ER, extraction recovery; IL, ionic liquid; 30-MemMP, 30-membered macrocyclic polyamine; PAH, polycyclic aromatic hydrocarbon; RR, relative recovery
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
bent materials in SPE is a key factor in determining satisfactory extraction efficiency. Therefore, many efforts have been focused on the development of new adsorbent materials in SPE [3–6], such as carbon nanotubes, graphene, ionic liquids (ILs), and so on. Among these efforts, a generation of adsorbents based on the macrocyclic supramolecular systems has gained intensive interest for analytical extraction and preconcentration schemes. In these approaches, the extraction ability can be tailored according to the variety of recognition interactions between the macrocycles and target analytes. The use of cyclodextrin and its derivatives or modified crown ethers as adsorbents in SPE have been described in many studies for the extraction of trace metal ions and organic compounds in water [7–9]. Macrocyclic polyamine, an azamacrocyclic compound, is a supramolecule with a structure similar to crown ether [10, 11]. A variety of functionalized groups on the macrocyclic polyamine ligand can provide multiple interactions with guest molecules, such as hydrophobic, –, and hydrogen bonding interactions as well as dipole-induced dipole
www.jss-journal.com
2
J. Sep. Sci. 2014, 00, 1–8
L. Liu et al.
interactions resulting from alkyl, phenyl, and amino group substituents. The cavity size of macrocyclic polyamines can be easily adjusted by attaching different lengths of alkyl chains, and different numbers of the benzene rings and amino groups [12, 13]. The recognition ability for the specific target analyte through multiple interactions was demonstrated in their application as HPLC stationary phases [12,14] and capillary electrochromatography coatings [15, 16]. The chromatographic separations of various organic analytes, including phenols, polycyclic aromatic hydrocarbons (PAHs), and pesticides using the macrocyclic polyamine-based separation materials, demonstrated the various types of and strength of interactions between macrocyclic polyamines ligand and guest molecules, resulting in high separation selectivity. It is expected that the selective recognition of macrocyclic polyamines toward specific analytes can be exploited in SPE. It was reported that the preconcentration of palladium could be accomplished with a C18 cartridge containing the 15membered triolefinic macrocyclic polyamine [17]. However, to the best of our knowledge, the application of SPE material only consisting of macrocyclic polyamine has not been explored. Since it was introduced by the Assadi group in 2006, dispersive liquid–liquid microextraction (DLLME) has been recognized as an attractive sample preparation technique in environmental analysis, biological analysis, and food analysis [18–20] due to its some advantages including simple operation, quick extraction equilibrium, and good extraction performance. DLLME could be developed to couple with SPE to obtain higher enrichment factor (EF) and detection sensitivity. Fattahi et al. [21] successfully combined SPE with DLLME for the preconcentration of ultra-trace chlorophenols in aqueous samples with various matrices. SPE–DLLME was also reported for the trace enrichment of short-chained alcohol and macrocyclic lactone compounds [22, 23]. Instead of conventional organic solvents, IL has been exploited as extraction solvent in DLLME. A number of papers have proposed IL-DLLME as a successful alternative to traditional DLLME for extraction and enrichment of organic contaminants, DNA, and inorganic ions [24–26], as recently reviewed [27, 28]. In spite of some prominent characteristics of SPE– DLLME, no work was reported on the combination of SPE and IL-DLLME. In this paper, a 30-membered macrocyclic polyamine (30MemMP), namely, 1,4,11,14,21,21-hexaaza-(2,3:12,13:22,23)tributano-(6,9:16,19:26,29)-trietheno-(1H,2H,3H,4H,5H,10 H,11H,12H,13H,14H,15H,20H,21H,22H,23H,24H,25H,30 H)-octadecahydro-(30)-annulene, was bonded onto silica and used for the first time as an SPE adsorbent. 30-MemMP-SPE was further combined with IL-DLLME for the enrichment and determination of targets using HPLC. The influence of different extraction conditions on the extraction efficiency was studied and optimized using five PAHs as model compounds. The method performance was evaluated by examining the linearity, LOD, and repeatability, as well as the relative recovery (RR) of real water samples. The extraction performance of 30-MemMP and 30-MemMP C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
SPE–IL-DLLME were compared with commercial C18 and only IL-DLLME, respectively.
2 Materials and methods 2.1 Chemicals and reagents Silica (particle diameter: 40–63 m, pore size: 60 Å, surface area: 500 m2 /g) was purchased from Aladdin Reagent (Shanghai, China). 30-MemMP, 1,3-dibutylimidazolium bis[(trifluoromethyl)sulfonyl]imide ([BBIM][Tf2 N]) were synthesized in our own lab [14, 29]. Methanol, acetone, acetonitrile, and ethanol were purchased from Merck (Darmstadt, Germany). Other solvents for synthesis were obtained from different origins, and dried by usual means prior to use. All reagents used were of HPLC grade or at least analytical reagent grade. Naphthalene, acenaphthene, anthracene, chrysene, and perylene were obtained from Tianjin Kermel Chemical Reagent (Tianjing, China). Individual stock solutions of five analytes were prepared at a concentration of 1.0 g/L by dissolving the proper quantity of each compound in methanol (naphthalene and acenaphthene) or tetrahydrofuran (anthracene, chrysene, and perylene), and were stored at 4⬚C. The working solutions were obtained by appropriately diluting the stock solution prior to analysis.
2.2 Instrumentation The infrared spectra were recorded on a Prestige-21 spectrometer (Shimadzu, Kyoto, Japan) at 4000–400 cm−1 . A Flash EA 1112 elemental analyzer (Thermo, Waltham, USA) was utilized. Chromatographic analysis was performed using a Shimadzu HPLC system (Kyoto, Japan) consisting of a LC-10AT pump and an SPD-10A UV/Vis detector. A 7725i six-port injection valve with 20 L sample loop (Rheodyne, Rohnert Park, CA, USA) was employed. The chromatographic separations were carried out on a Shimadzu ODS-SP column (250 × 4.6 mm id, 5m). The mobile phase was composed of acetonitrile and water (80:20, v/v) at a flow rate of 1.0 mL/min. The empty SPE tubes were obtained from Shanghai Derian Instrument (Shanghai, China).
2.3 Preparation of 30-MemMP-modified silica The synthesis procedure of 30-MemMP-modified silica was similar to our previous work [14] except silica with different diameter and surface area was employed. Two main procedures were performed, including the synthesis of ␥-chloropropyltrimethoxy-sililated silica by reacting activated silica and ␥-chloropropyltrimethoxysilane using triethylamine as catalyst in toluene for 24 h under an argon atmosphere, and the preparation of 30-MemMP-modified silica by www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 00, 1–8
3
Figure 1. Structural representation of 30-MemMP-modified silica.
reacting 30-MemMP and ␥-chloropropyltrimethoxy-sililated silica in toluene for 12 h at 112⬚C under an argon atmosphere. A structural representation of 30-MemMP-modified silica is shown in Fig. 1.
2.4 Extraction procedure The 30-MemMP-SPE–IL-DLLME procedure was carried out as follows. The 30-MemMP-modified adsorbent (200 mg) was packed into a 3 mL SPE tube. Prior to each SPE run, the 30MemMP cartridge was activated by washing successively with methanol and water (each for 10 mL). Then, 150 mL of sample solution (pH 8) passed through the SPE column at a flow rate of about 10 mL/min with the aid of a vacuum pump. The desired compounds were eluted with 1.0 mL acetone and collected into 10 mL centrifuge tube. The [BBIM][Tf2 N] (25 L) and distilled water (5 mL) were rapidly injected into the above-mentioned acetone, and a cloudy solution was instantly formed. The mixture was centrifuged for 5 min at 4000 rpm, and 5 L of the sediment phase was injected into HPLC for analysis. The commercial Waters Sep-pak C18 cartridges (Waters, Milford, MA, USA) were used to compare. The C18 -SPE–IL-DLLME procedure was the same as the 30MemMP-SPE–IL-DLLME. In order to evaluate the extraction efficiency of overall process, EF and extraction recovery (ER) were used and can be expressed as follows. ER =
Csed Vsed nsed × 100% = × 100% n0 C0 V0
= EF ×
Vsed × 100% V0
(1)
Accordingly, nsed and n0 are the extracted amount of analytes and initial amount of analytes, respectively, while Vsed and V0 are the volume of sediment phase and sample solution, respectively. The volume of sample solution and sediment phase were kept constant in the optimization experiments; therefore, the change of EF followed the same trend as ER according to the above equation. Thus, only EF was used as an indicator of extraction efficiency for all analytes and shown in all subsequent studies unless otherwise specified. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Effect of the amount of adsorbent on the extraction efficiency of PAHs. Extraction conditions: water sample volume, 150 mL; flow rate, 10 mL/min; sample solution pH 8; elution solvent, 1.0 mL of acetone; extraction solvent, 25.0 L of [BBIM][Tf2 N]; room temperature; the concentration of analytes, 20 g/L for naphthalene and acenaphthene, 5 g/L for anthracene, chrysene, and perylene.
3 Results and discussion 3.1 Characterization of the 30-MemMP-based adsorbent 30-MemMP-bonded silica was characterized by infrared spectroscopy and elemental analysis. A comparison between the infrared spectra of silica and 30-MemMP-bonded silica (Supporting Information Fig. S1) showed that a new absorption of the methylene group appeared at 2930 and 2860 cm−1 on 30-MemMP-bonded silica. Compared with ␥-chloropropyltrimethoxy-sililated silica, the rise of nitrogen and carbon content on 30-MemMP-bonded silica was demonstrated by the elemental analysis. The content of carbon and nitrogen on 30-MemMP-bonded silica and ␥chloropropyltrimethoxy-sililated silica was 12.24 and 1.42%, and 8.20 and 0%, respectively. The data qualitatively and quantitatively verified the successful covalent immobilization of 30-MemMP onto silica.
3.2 Optimization of extraction conditions 3.2.1 Effect of the amount of 30-MemMP-based adsorbent The amount of adsorbent plays an important role in acquiring satisfactory extraction efficiency. In order to estimate the effect of 30-MemMP-based adsorbent amount on the overall extraction efficiency of 30-MemMP-SPE–IL-DLLME, different amounts ranging from 100 to 300 mg were examined. Figure 2 illustrates the effect of amount of adsorbent on www.jss-journal.com
4
L. Liu et al.
J. Sep. Sci. 2014, 00, 1–8
Figure 3. Effect of the volume of sample solution on the extraction efficiency of PAHs. Extraction conditions: amount of adsorbent, 200 mg; other conditions are the same as in Fig. 2.
Figure 4. Effect of sample solution pH on the extraction efficiency of PAHs. Extraction conditions are the same as in Fig. 2.
extraction efficiency. The results indicated that the EFs of all analytes increased when the adsorbent amount was increased from 100 to 200 mg, and then decreased when the amount was further increased to 300 mg. Using the lower amount of adsorbent, analytes could not be fully adsorbed onto the extraction materials, resulting in a low extraction efficiency. However, at the higher loading amount of adsorbent, the elution solvent with constant volume and elution strength could not completely elute the analytes from the adsorbent, leading to a low extraction efficiency. From the results, 200 mg of 30-MemMP-based adsorbent was used in the further experiments.
only the extraction efficiency but also the analysis time. The experiments were performed using 150 mL of sample solution at different flow rates ranging from 2 to 13 mL/min. The results showed the EF gradually increased with the increase of flow rate and then decreased up to 10 mL/min. Thus, 10 mL/min of flow rate of sample solution was used in the following experiments.
3.2.2 Effect of the volume and flow rate of sample solution The volume of sample solution will have an influence on both ER and EF of 30-MemMP-SPE–IL-DLLME. The concentration of analytes in the elution solvent would increase with the increase of sample solution volume, resulting in an increased ER and EF. However, further increasing the volume would lead to a saturation adsorption, subsequently decreasing ER. In the case of EF, the amount of eluted analytes is kept constant, so the EF will remain stable with further increases of sample volume. The effects of the sample solution volume on the ER and EF of five PAHs were investigated in the range of 50– 250 mL. The results shown in Fig. 3 indicate that ERs of the five PAHs increased with the increase of sample solution volume from 50 to 150 mL, but then decreased distinctly with further increase of volume. The EFs were observed to first increase and remain nearly constant at higher volume (Supporting Information Fig. S2). Thus, 150 mL of sample solution was selected in all subsequent experiments. In the 30-MemMP-SPE–IL-DLLME procedure, the flow rate of sample solution through the solid phase affects not C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.2.3 Effect of pH of sample solution The structure of 30-MemMP grafted onto silica plays a crucial role in enriching and extracting the target analytes in the proposed 30-MemMP-SPE–IL-DLLME procedure. The various functional groups on 30-MemMP including phenyl, alkyl, and amino groups, provide –, hydrophobic, hydrogen bonding, and dipole-induced dipole interactions, which aid in extracting the analytes from the aqueous solution. Among these groups, the form of the amino group is influenced by the pH of the sample solution. Figure 4 shows the dependence of the extraction efficiency on the sample solution pH. An increase in the extraction efficiency for five PAHs was observed when the pH of the aqueous solution was changed from pH 2 to 8. At low pH, the cationic amino groups interacted weakly with the hydrophobic analytes, while at higher pH, the neutral amino groups enhanced the hydrophobicity of the macrocycle and enlarged the conjugated system of the macrocycle, resulting in stronger interactions between 30-MemMP and PAHs. It is worth noting that the pH had a more obvious effect on naphthalene than on other analytes, which can be due to the fact that it has a higher polarity compared with the other analytes and is capable of undergoing dipole-induced dipole interactions. The above results indicate that the hydrophobic, –, and dipole-induced dipole interactions between 30-MemMP and PAHs may all play a role in the extraction performance. www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 00, 1–8
5
Table 1. Comparison results of different elution solvents in 30-MemMP-SPEa)
Elution solvent
Acetone Acetonitrile Ethanol Methanol Isopropanol
Naphthalene
Acenaphthene
Anthracene
Chrysene
Perylene
EF
ER (%)
EF
ER (%)
EF
ER (%)
EF
ER (%)
EF
ER (%)
95.4 94.5 94.3 81.0 87.4
63.6 63.0 62.8 54.0 58.3
80.3 86.2 83.6 73.5 78.6
53.5 57.5 55.7 49.0 52.4
53.7 57.1 60.5 36.2 39.0
35.8 38.1 40.3 24.1 26.0
70.2 93.6 80.8 59.5 57.2
46.8 62.4 53.9 39.7 38.1
62.7 77.9 64.8 52.7 45.8
41.8 51.9 43.2 35.1 30.5
a) The data were obtained in the following 30-MemMP-SPE conditions: volume of sample solution, 150 mL; elution rate, 10 mL/min; elution solvent volume, 1 mL.
3.2.4 Selection of the elution solvent (disperser solvent) The elution solvent has dual functions in the developed 30-MemMP-SPE–IL-DLLME method. It elutes the analytes from the 30-MemMP-based adsorbent in SPE procedure and simultaneously disperses the extraction solvent into fine droplets in DLLME procedure. Thus, the selection of elution solvent is not only dependent on the elution ability in SPE but also determined by the dispersive capacity in DLLME. Five solvents (acetone, acetonitrile, ethanol, methanol, and isopropanol) were first evaluated as eluting solvents (volume of 1 mL) in 30-MemMP-SPE to test their elution abilities. The results, shown in Table 1, indicate that acetone, acetonitrile, and ethanol provided higher EF and ER than the other two solvents. Thus, the former three solvents, along with methanol, which is often used as dispersive solvent in DLLME, were further studied in 30-MemMP-SPE– IL-DLLME. From the experiments, the highest EFs and ERs were obtained using acetone as the dispersive solvent and elution solvent, which is ascribed to the stronger elution ability of acetone compared with methanol. In addition, acetone could form a more obvious turbid solution with extraction solvent and water than acetonitrile and ethanol. On the basis of the results, acetone was chosen as the elution and dispersive solvent for the proposed 30-MemMP-SPE–IL-DLLME method.
phase was not influenced by the addition of salt, subsequently no obvious change in extraction efficiency was observed with the addition of salt. It can be concluded that the proposed method would have little matrix effect when being applied in sample with a high concentration of salt. So the experiments were carried out without salt.
3.3 Performance of the proposed 30-MemMP-SPE–[BBIM][Tf2 N]-DLLME method Table 2 summarizes the parameters of the calibration curves for the PAH analysis. The developed method indicated a good linearity with the correlation coefficient ranging from 0.9983 to 0.9999. The repeatability was evaluated by five replicate extractions of two different spiked levels, and the RSDs were calculated to be in the range of 3.3–5.4%. The LODs, based on S/N of 3, for naphthalene, acenaphthene, anthracene, chrysene, and perylene were 0.1, 0.1, 0.002, 0.02, and 0.02 g/L, respectively. Under the optimized experimental conditions, the EFs of analytes ranged between 2768 and 5409. The high EF verifies the fact that 30-MemMP-SPE–[BBIM][Tf2 N]DLLME can provide satisfactory extraction efficiency for the target compounds.
3.4 Real-sample analysis 3.2.5 Effect of salt concentration The influence of added salt was evaluated at 0–4% m/v NaCl levels while other experimental parameters were kept constant. The results showed that the salt concentration had no significant effect on the extraction efficiency of PAHs, which was unlike the case in single IL-DLLME procedure [29, 30], where the addition of salt mainly affected the volume of sediment phase and subsequently influenced the EF and ER. In the proposed 30-MemMP-SPE–IL-DLLME procedure, salts were added to the sample solution and allowed to flow through the SPE cartridge along with sample solution. In fact, no salt was present in the acetone/IL/water system in the subsequent DLLME procedure. Thus, the volume of sediment C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The applicability of the developed method was assessed by determining PAHs in four different real water samples including well water, tap water, surface water, and river water. The river water was collected from Yellow River in Zhengzhou city (China), while other water samples were obtained from Henan University of Technology (Zhengzhou, China). All samples were filtered prior to use and analyzed according to the procedures described in Section 2.4. From the results shown in Table 3, all analytes were below the LODs, except that 0.1 and 4.3 g/L anthracene was detected in well river and surface water, respectively, and 0.1 g/L1 chrysene was detected in well water. Moreover, the spiked recoveries and RSDs of all PAHs at two concentration levels were determined in the four samples for testing the accuracy www.jss-journal.com
6
J. Sep. Sci. 2014, 00, 1–8
L. Liu et al.
Table 2. Quantitative results of the developed 30-MemMP-SPE–IL-DLLME-HPLC method in the determination of PAHs
Compounds
Linear equation
LRa) (g/L)
R2
LOD (g/L)
EF
RSD (%, n = 5)
Naphthalene Acenaphthene Anthracene Chrysene Perylene
y = (15 784 ± 348)x + (32 649 ± 6269) y = (9339 ± 520)x + (38 108 ± 5178) y = (362 157 ± 7621)x + (303 229 ± 11 069) y = (262 624 ± 8063)x + (38 344 ± 2696) y = (173 948 ± 1993)x + (62 537 ± 7662)
1.0–500 1.0–500 0.02–20 0.2–50 0.2–50
0.9998 0.9999 0.9983 0.9999 0.9999
0.1 0.1 0.002 0.02 0.02
2768 3640 4847 4837 5409
4.0–6.8b) 0.4–6.2b) 8.4–1.0b) 6.5–3.5b) 6.9–2.3b)
a) LR, linear range. b) The former RSD obtained at lower spiked levels (10 g/L for naphthalene and acenaphthene; 0.2 g/L for anthracene, chrysene, and perylene), and the latter RSD obtained at higher spiked levels (200 g/L for naphthalene and acenaphthene; 10 g/L for anthracene; 20 g/L for chrysene and perylene). Table 3. Results for determination of analytes in spiked tap, well, river, and surface waters
Compounds
Naphthalene
Acenaphthene
Anthracene
Chrysene
Perylene
Added (g/L)
0 20.0 100.0 0 20.0 100.0 0 5.0 10.0 0 5.0 10.0 0 5.0 10.0
Tap water (n = 3)
Well water (n = 3)
River water (n = 3)
Surface water (n = 3)
Found (g/L)
RR (%)a)
RSD (%)
Found (g/L)
RR (%)
RSD (%)
Found (g/L)
RR (%)
RSD (%)
Found (g/L)
RR (%)
RSD (%)
– 17.1 87.9 – 17.4 98.0 – 5.7 10.5 – 5.6 10.1 – 5.7 8.1
– 85.6 87.9 – 86.9 98.0 – 114.1 105.4 – 111.8 101.0 – 114.2 80.9
– 3.5 5.3 – 1.7 4.6 – 5.9 10.1 – 9.2 8.9 – 3.2 2.4
– 21.6 111.7 – 20.0 104.8 0.1 5.2 12.0 0.1 5.2 10.5 – 5.0 9.7
– 108.1 111.7 – 100.2 104.8 – 102.3 119.1 – 102.4 104.3 – 100.4 97.1
– 4.2 8.8 – 2.1 6.8 2.6 7.9 1.4 9.1 6.2 1.9 – 1.6 4.6
– 22.6 110.6 – 20.5 108.8 – 5.2 11.6 – 5.4 9.8 – 5.6 9.99
– 112.7 110.6 – 102.5 108.8 – 104.7 116.1 – 108.2 97.1 – 111.2 99.9
– 4.5 4.2 – 5.8 8.9 – 5.4 5.2 – 2.1 2.5 – 1.4 2.5
– 19.2 104.3 – 17.4 106.5 4.3 8.6 16.3 – 6.1 10.2 – 5.4 8.3
– 95.8 104.3 – 86.9 106.5 – 86.7 120.1 – 103.2 93.2 – 107.0 83.2
– 7.5 2.8 – 2.1 4.1 5.3 1.9 9.9 – 5.6 4.1 – 4.6 1.7
and precision of the method. The data indicated that the developed 30-MemMP-SPE–[BBIM][Tf2 N]-DLLLE–HPLC method can be applied to analyze PAHs in water samples with satisfactory RRs (80.9–120.1%) and RSDs (1.4–10.1%).
3.5 Comparison with C18 -SPE and [BBIM][Tf2 N]-DLLME In order to elucidate the advantages of new SPE material and acquire a good understanding of the retention mechanism of 30-MemMP-silica adsorbent, a commercial C18 cartridge was performed under the same extraction conditions and other four PAHs (biphenyl, phenanthrene, fluoranthene, and pyrene) were extracted along with the above mentioned five analytes. Figure 5 shows the chromatogram of nine analytes after SPE with 30-MemMP-silica and C18 . It was observed evidently that 30-MemMP-SPE provided higher extraction efficiency for all analytes than C18 -SPE, and the longer the retention time of analytes is, the more the extraction efficiency on 30-MemMP-SPE is high. The extraction efficiency
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
is mainly dependent on the hydrophobic interaction between adsorbent and analytes using C18 as extraction material, and the intensity of acetone (1 mL) was not enough to elute the analytes with strong hydrophobility from C18 adsorbent, thus the ER decreased with the increase of retention time, especially for chrysene and perylene, whose ERs are
J. Sep. Sci. 2014, 00, 1–8
Longhui Liu1 Lijun He1 Xiuming Jiang1 Wenjie Zhao1 Guoqiang Xiang1 Jared L. Anderson2 1 School
of Chemistry and Chemical Engineering, Henan University of Technology, Zhengzhou, China 2 Department of Chemistry, School of Green Chemistry and Engineering, The University of Toledo, Toledo, OH, USA Received September 24, 2013 Revised January 27, 2014 Accepted January 27, 2014
Research Article
Macrocyclic polyamine-functionalized silica as a solid-phase extraction material coupled with ionic liquid dispersive liquid–liquid extraction for the enrichment of polycyclic aromatic hydrocarbons In this study, silica modified with a 30-membered macrocyclic polyamine was synthesized and first used as an adsorbent material in SPE. The SPE was further combined with ionic liquid (IL) dispersive liquid–liquid microextraction (DLLME). Five polycyclic aromatic hydrocarbons were employed as model analytes to evaluate the extraction procedure and were determined by HPLC combined with UV/Vis detection. Acetone was used as the elution solvent in SPE as well as the dispersive solvent in DLLME. The enrichment of analytes was achieved using the 1,3-dibutylimidazolium bis[(trifluoromethyl)sulfonyl]imide IL/acetone/water system. Experimental conditions for the overall macrocycle-SPE–IL-DLLME method, such as the amount of adsorbent, sample solution volume, sample solution pH, type of elution solvent as well as addition of salt, were studied and optimized. The developed method could be successfully applied to the analysis of four real water samples. The macrocyclic polyamine offered higher extraction efficiency for analytes compared with commercially available C18 cartridge, and the developed method provided higher enrichment factors (2768–5409) for model analytes compared with the single DLLME. Good linearity with the correlation coefficients ranging from 0.9983 to 0.9999 and LODs as low as 0.002 g/L were obtained in the proposed method. Keywords: Dispersive liquid–liquid microextraction / Ionic liquids / Macrocyclic polyamines / Polycyclic aromatic hydrocarbons / Solid-phase extraction DOI 10.1002/jssc.201301062
Additional supporting information may be found in the online version of this article at the publisher’s web-site
1 Introduction Nowadays, SPE has become one of the most common sample preparation techniques due to its ease of operation and lower consumption of organic solvent compared with liquid–liquid extraction. The technique has been successfully applied in many areas of analytical chemistry including environmental, biological, and food analysis [1, 2]. The composition of adsorCorrespondence: Dr. Lijun He, School of Chemistry and Chemical Engineering, Henan University of Technology, 450001 Zhengzhou, China E-mail: [email protected], [email protected] Fax: +86-371-67756718
Abbreviations: [BBIM][Tf2 N], 1,3-dibutylimidazolium bis [(trifluoromethyl)sulfonyl]imide; DLLME, dispersive liquid– liquid microextraction; EF, enrichment factor; ER, extraction recovery; IL, ionic liquid; 30-MemMP, 30-membered macrocyclic polyamine; PAH, polycyclic aromatic hydrocarbon; RR, relative recovery
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
bent materials in SPE is a key factor in determining satisfactory extraction efficiency. Therefore, many efforts have been focused on the development of new adsorbent materials in SPE [3–6], such as carbon nanotubes, graphene, ionic liquids (ILs), and so on. Among these efforts, a generation of adsorbents based on the macrocyclic supramolecular systems has gained intensive interest for analytical extraction and preconcentration schemes. In these approaches, the extraction ability can be tailored according to the variety of recognition interactions between the macrocycles and target analytes. The use of cyclodextrin and its derivatives or modified crown ethers as adsorbents in SPE have been described in many studies for the extraction of trace metal ions and organic compounds in water [7–9]. Macrocyclic polyamine, an azamacrocyclic compound, is a supramolecule with a structure similar to crown ether [10, 11]. A variety of functionalized groups on the macrocyclic polyamine ligand can provide multiple interactions with guest molecules, such as hydrophobic, –, and hydrogen bonding interactions as well as dipole-induced dipole
www.jss-journal.com
2
J. Sep. Sci. 2014, 00, 1–8
L. Liu et al.
interactions resulting from alkyl, phenyl, and amino group substituents. The cavity size of macrocyclic polyamines can be easily adjusted by attaching different lengths of alkyl chains, and different numbers of the benzene rings and amino groups [12, 13]. The recognition ability for the specific target analyte through multiple interactions was demonstrated in their application as HPLC stationary phases [12,14] and capillary electrochromatography coatings [15, 16]. The chromatographic separations of various organic analytes, including phenols, polycyclic aromatic hydrocarbons (PAHs), and pesticides using the macrocyclic polyamine-based separation materials, demonstrated the various types of and strength of interactions between macrocyclic polyamines ligand and guest molecules, resulting in high separation selectivity. It is expected that the selective recognition of macrocyclic polyamines toward specific analytes can be exploited in SPE. It was reported that the preconcentration of palladium could be accomplished with a C18 cartridge containing the 15membered triolefinic macrocyclic polyamine [17]. However, to the best of our knowledge, the application of SPE material only consisting of macrocyclic polyamine has not been explored. Since it was introduced by the Assadi group in 2006, dispersive liquid–liquid microextraction (DLLME) has been recognized as an attractive sample preparation technique in environmental analysis, biological analysis, and food analysis [18–20] due to its some advantages including simple operation, quick extraction equilibrium, and good extraction performance. DLLME could be developed to couple with SPE to obtain higher enrichment factor (EF) and detection sensitivity. Fattahi et al. [21] successfully combined SPE with DLLME for the preconcentration of ultra-trace chlorophenols in aqueous samples with various matrices. SPE–DLLME was also reported for the trace enrichment of short-chained alcohol and macrocyclic lactone compounds [22, 23]. Instead of conventional organic solvents, IL has been exploited as extraction solvent in DLLME. A number of papers have proposed IL-DLLME as a successful alternative to traditional DLLME for extraction and enrichment of organic contaminants, DNA, and inorganic ions [24–26], as recently reviewed [27, 28]. In spite of some prominent characteristics of SPE– DLLME, no work was reported on the combination of SPE and IL-DLLME. In this paper, a 30-membered macrocyclic polyamine (30MemMP), namely, 1,4,11,14,21,21-hexaaza-(2,3:12,13:22,23)tributano-(6,9:16,19:26,29)-trietheno-(1H,2H,3H,4H,5H,10 H,11H,12H,13H,14H,15H,20H,21H,22H,23H,24H,25H,30 H)-octadecahydro-(30)-annulene, was bonded onto silica and used for the first time as an SPE adsorbent. 30-MemMP-SPE was further combined with IL-DLLME for the enrichment and determination of targets using HPLC. The influence of different extraction conditions on the extraction efficiency was studied and optimized using five PAHs as model compounds. The method performance was evaluated by examining the linearity, LOD, and repeatability, as well as the relative recovery (RR) of real water samples. The extraction performance of 30-MemMP and 30-MemMP C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
SPE–IL-DLLME were compared with commercial C18 and only IL-DLLME, respectively.
2 Materials and methods 2.1 Chemicals and reagents Silica (particle diameter: 40–63 m, pore size: 60 Å, surface area: 500 m2 /g) was purchased from Aladdin Reagent (Shanghai, China). 30-MemMP, 1,3-dibutylimidazolium bis[(trifluoromethyl)sulfonyl]imide ([BBIM][Tf2 N]) were synthesized in our own lab [14, 29]. Methanol, acetone, acetonitrile, and ethanol were purchased from Merck (Darmstadt, Germany). Other solvents for synthesis were obtained from different origins, and dried by usual means prior to use. All reagents used were of HPLC grade or at least analytical reagent grade. Naphthalene, acenaphthene, anthracene, chrysene, and perylene were obtained from Tianjin Kermel Chemical Reagent (Tianjing, China). Individual stock solutions of five analytes were prepared at a concentration of 1.0 g/L by dissolving the proper quantity of each compound in methanol (naphthalene and acenaphthene) or tetrahydrofuran (anthracene, chrysene, and perylene), and were stored at 4⬚C. The working solutions were obtained by appropriately diluting the stock solution prior to analysis.
2.2 Instrumentation The infrared spectra were recorded on a Prestige-21 spectrometer (Shimadzu, Kyoto, Japan) at 4000–400 cm−1 . A Flash EA 1112 elemental analyzer (Thermo, Waltham, USA) was utilized. Chromatographic analysis was performed using a Shimadzu HPLC system (Kyoto, Japan) consisting of a LC-10AT pump and an SPD-10A UV/Vis detector. A 7725i six-port injection valve with 20 L sample loop (Rheodyne, Rohnert Park, CA, USA) was employed. The chromatographic separations were carried out on a Shimadzu ODS-SP column (250 × 4.6 mm id, 5m). The mobile phase was composed of acetonitrile and water (80:20, v/v) at a flow rate of 1.0 mL/min. The empty SPE tubes were obtained from Shanghai Derian Instrument (Shanghai, China).
2.3 Preparation of 30-MemMP-modified silica The synthesis procedure of 30-MemMP-modified silica was similar to our previous work [14] except silica with different diameter and surface area was employed. Two main procedures were performed, including the synthesis of ␥-chloropropyltrimethoxy-sililated silica by reacting activated silica and ␥-chloropropyltrimethoxysilane using triethylamine as catalyst in toluene for 24 h under an argon atmosphere, and the preparation of 30-MemMP-modified silica by www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 00, 1–8
3
Figure 1. Structural representation of 30-MemMP-modified silica.
reacting 30-MemMP and ␥-chloropropyltrimethoxy-sililated silica in toluene for 12 h at 112⬚C under an argon atmosphere. A structural representation of 30-MemMP-modified silica is shown in Fig. 1.
2.4 Extraction procedure The 30-MemMP-SPE–IL-DLLME procedure was carried out as follows. The 30-MemMP-modified adsorbent (200 mg) was packed into a 3 mL SPE tube. Prior to each SPE run, the 30MemMP cartridge was activated by washing successively with methanol and water (each for 10 mL). Then, 150 mL of sample solution (pH 8) passed through the SPE column at a flow rate of about 10 mL/min with the aid of a vacuum pump. The desired compounds were eluted with 1.0 mL acetone and collected into 10 mL centrifuge tube. The [BBIM][Tf2 N] (25 L) and distilled water (5 mL) were rapidly injected into the above-mentioned acetone, and a cloudy solution was instantly formed. The mixture was centrifuged for 5 min at 4000 rpm, and 5 L of the sediment phase was injected into HPLC for analysis. The commercial Waters Sep-pak C18 cartridges (Waters, Milford, MA, USA) were used to compare. The C18 -SPE–IL-DLLME procedure was the same as the 30MemMP-SPE–IL-DLLME. In order to evaluate the extraction efficiency of overall process, EF and extraction recovery (ER) were used and can be expressed as follows. ER =
Csed Vsed nsed × 100% = × 100% n0 C0 V0
= EF ×
Vsed × 100% V0
(1)
Accordingly, nsed and n0 are the extracted amount of analytes and initial amount of analytes, respectively, while Vsed and V0 are the volume of sediment phase and sample solution, respectively. The volume of sample solution and sediment phase were kept constant in the optimization experiments; therefore, the change of EF followed the same trend as ER according to the above equation. Thus, only EF was used as an indicator of extraction efficiency for all analytes and shown in all subsequent studies unless otherwise specified. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 2. Effect of the amount of adsorbent on the extraction efficiency of PAHs. Extraction conditions: water sample volume, 150 mL; flow rate, 10 mL/min; sample solution pH 8; elution solvent, 1.0 mL of acetone; extraction solvent, 25.0 L of [BBIM][Tf2 N]; room temperature; the concentration of analytes, 20 g/L for naphthalene and acenaphthene, 5 g/L for anthracene, chrysene, and perylene.
3 Results and discussion 3.1 Characterization of the 30-MemMP-based adsorbent 30-MemMP-bonded silica was characterized by infrared spectroscopy and elemental analysis. A comparison between the infrared spectra of silica and 30-MemMP-bonded silica (Supporting Information Fig. S1) showed that a new absorption of the methylene group appeared at 2930 and 2860 cm−1 on 30-MemMP-bonded silica. Compared with ␥-chloropropyltrimethoxy-sililated silica, the rise of nitrogen and carbon content on 30-MemMP-bonded silica was demonstrated by the elemental analysis. The content of carbon and nitrogen on 30-MemMP-bonded silica and ␥chloropropyltrimethoxy-sililated silica was 12.24 and 1.42%, and 8.20 and 0%, respectively. The data qualitatively and quantitatively verified the successful covalent immobilization of 30-MemMP onto silica.
3.2 Optimization of extraction conditions 3.2.1 Effect of the amount of 30-MemMP-based adsorbent The amount of adsorbent plays an important role in acquiring satisfactory extraction efficiency. In order to estimate the effect of 30-MemMP-based adsorbent amount on the overall extraction efficiency of 30-MemMP-SPE–IL-DLLME, different amounts ranging from 100 to 300 mg were examined. Figure 2 illustrates the effect of amount of adsorbent on www.jss-journal.com
4
L. Liu et al.
J. Sep. Sci. 2014, 00, 1–8
Figure 3. Effect of the volume of sample solution on the extraction efficiency of PAHs. Extraction conditions: amount of adsorbent, 200 mg; other conditions are the same as in Fig. 2.
Figure 4. Effect of sample solution pH on the extraction efficiency of PAHs. Extraction conditions are the same as in Fig. 2.
extraction efficiency. The results indicated that the EFs of all analytes increased when the adsorbent amount was increased from 100 to 200 mg, and then decreased when the amount was further increased to 300 mg. Using the lower amount of adsorbent, analytes could not be fully adsorbed onto the extraction materials, resulting in a low extraction efficiency. However, at the higher loading amount of adsorbent, the elution solvent with constant volume and elution strength could not completely elute the analytes from the adsorbent, leading to a low extraction efficiency. From the results, 200 mg of 30-MemMP-based adsorbent was used in the further experiments.
only the extraction efficiency but also the analysis time. The experiments were performed using 150 mL of sample solution at different flow rates ranging from 2 to 13 mL/min. The results showed the EF gradually increased with the increase of flow rate and then decreased up to 10 mL/min. Thus, 10 mL/min of flow rate of sample solution was used in the following experiments.
3.2.2 Effect of the volume and flow rate of sample solution The volume of sample solution will have an influence on both ER and EF of 30-MemMP-SPE–IL-DLLME. The concentration of analytes in the elution solvent would increase with the increase of sample solution volume, resulting in an increased ER and EF. However, further increasing the volume would lead to a saturation adsorption, subsequently decreasing ER. In the case of EF, the amount of eluted analytes is kept constant, so the EF will remain stable with further increases of sample volume. The effects of the sample solution volume on the ER and EF of five PAHs were investigated in the range of 50– 250 mL. The results shown in Fig. 3 indicate that ERs of the five PAHs increased with the increase of sample solution volume from 50 to 150 mL, but then decreased distinctly with further increase of volume. The EFs were observed to first increase and remain nearly constant at higher volume (Supporting Information Fig. S2). Thus, 150 mL of sample solution was selected in all subsequent experiments. In the 30-MemMP-SPE–IL-DLLME procedure, the flow rate of sample solution through the solid phase affects not C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
3.2.3 Effect of pH of sample solution The structure of 30-MemMP grafted onto silica plays a crucial role in enriching and extracting the target analytes in the proposed 30-MemMP-SPE–IL-DLLME procedure. The various functional groups on 30-MemMP including phenyl, alkyl, and amino groups, provide –, hydrophobic, hydrogen bonding, and dipole-induced dipole interactions, which aid in extracting the analytes from the aqueous solution. Among these groups, the form of the amino group is influenced by the pH of the sample solution. Figure 4 shows the dependence of the extraction efficiency on the sample solution pH. An increase in the extraction efficiency for five PAHs was observed when the pH of the aqueous solution was changed from pH 2 to 8. At low pH, the cationic amino groups interacted weakly with the hydrophobic analytes, while at higher pH, the neutral amino groups enhanced the hydrophobicity of the macrocycle and enlarged the conjugated system of the macrocycle, resulting in stronger interactions between 30-MemMP and PAHs. It is worth noting that the pH had a more obvious effect on naphthalene than on other analytes, which can be due to the fact that it has a higher polarity compared with the other analytes and is capable of undergoing dipole-induced dipole interactions. The above results indicate that the hydrophobic, –, and dipole-induced dipole interactions between 30-MemMP and PAHs may all play a role in the extraction performance. www.jss-journal.com
Sample Preparation
J. Sep. Sci. 2014, 00, 1–8
5
Table 1. Comparison results of different elution solvents in 30-MemMP-SPEa)
Elution solvent
Acetone Acetonitrile Ethanol Methanol Isopropanol
Naphthalene
Acenaphthene
Anthracene
Chrysene
Perylene
EF
ER (%)
EF
ER (%)
EF
ER (%)
EF
ER (%)
EF
ER (%)
95.4 94.5 94.3 81.0 87.4
63.6 63.0 62.8 54.0 58.3
80.3 86.2 83.6 73.5 78.6
53.5 57.5 55.7 49.0 52.4
53.7 57.1 60.5 36.2 39.0
35.8 38.1 40.3 24.1 26.0
70.2 93.6 80.8 59.5 57.2
46.8 62.4 53.9 39.7 38.1
62.7 77.9 64.8 52.7 45.8
41.8 51.9 43.2 35.1 30.5
a) The data were obtained in the following 30-MemMP-SPE conditions: volume of sample solution, 150 mL; elution rate, 10 mL/min; elution solvent volume, 1 mL.
3.2.4 Selection of the elution solvent (disperser solvent) The elution solvent has dual functions in the developed 30-MemMP-SPE–IL-DLLME method. It elutes the analytes from the 30-MemMP-based adsorbent in SPE procedure and simultaneously disperses the extraction solvent into fine droplets in DLLME procedure. Thus, the selection of elution solvent is not only dependent on the elution ability in SPE but also determined by the dispersive capacity in DLLME. Five solvents (acetone, acetonitrile, ethanol, methanol, and isopropanol) were first evaluated as eluting solvents (volume of 1 mL) in 30-MemMP-SPE to test their elution abilities. The results, shown in Table 1, indicate that acetone, acetonitrile, and ethanol provided higher EF and ER than the other two solvents. Thus, the former three solvents, along with methanol, which is often used as dispersive solvent in DLLME, were further studied in 30-MemMP-SPE– IL-DLLME. From the experiments, the highest EFs and ERs were obtained using acetone as the dispersive solvent and elution solvent, which is ascribed to the stronger elution ability of acetone compared with methanol. In addition, acetone could form a more obvious turbid solution with extraction solvent and water than acetonitrile and ethanol. On the basis of the results, acetone was chosen as the elution and dispersive solvent for the proposed 30-MemMP-SPE–IL-DLLME method.
phase was not influenced by the addition of salt, subsequently no obvious change in extraction efficiency was observed with the addition of salt. It can be concluded that the proposed method would have little matrix effect when being applied in sample with a high concentration of salt. So the experiments were carried out without salt.
3.3 Performance of the proposed 30-MemMP-SPE–[BBIM][Tf2 N]-DLLME method Table 2 summarizes the parameters of the calibration curves for the PAH analysis. The developed method indicated a good linearity with the correlation coefficient ranging from 0.9983 to 0.9999. The repeatability was evaluated by five replicate extractions of two different spiked levels, and the RSDs were calculated to be in the range of 3.3–5.4%. The LODs, based on S/N of 3, for naphthalene, acenaphthene, anthracene, chrysene, and perylene were 0.1, 0.1, 0.002, 0.02, and 0.02 g/L, respectively. Under the optimized experimental conditions, the EFs of analytes ranged between 2768 and 5409. The high EF verifies the fact that 30-MemMP-SPE–[BBIM][Tf2 N]DLLME can provide satisfactory extraction efficiency for the target compounds.
3.4 Real-sample analysis 3.2.5 Effect of salt concentration The influence of added salt was evaluated at 0–4% m/v NaCl levels while other experimental parameters were kept constant. The results showed that the salt concentration had no significant effect on the extraction efficiency of PAHs, which was unlike the case in single IL-DLLME procedure [29, 30], where the addition of salt mainly affected the volume of sediment phase and subsequently influenced the EF and ER. In the proposed 30-MemMP-SPE–IL-DLLME procedure, salts were added to the sample solution and allowed to flow through the SPE cartridge along with sample solution. In fact, no salt was present in the acetone/IL/water system in the subsequent DLLME procedure. Thus, the volume of sediment C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The applicability of the developed method was assessed by determining PAHs in four different real water samples including well water, tap water, surface water, and river water. The river water was collected from Yellow River in Zhengzhou city (China), while other water samples were obtained from Henan University of Technology (Zhengzhou, China). All samples were filtered prior to use and analyzed according to the procedures described in Section 2.4. From the results shown in Table 3, all analytes were below the LODs, except that 0.1 and 4.3 g/L anthracene was detected in well river and surface water, respectively, and 0.1 g/L1 chrysene was detected in well water. Moreover, the spiked recoveries and RSDs of all PAHs at two concentration levels were determined in the four samples for testing the accuracy www.jss-journal.com
6
J. Sep. Sci. 2014, 00, 1–8
L. Liu et al.
Table 2. Quantitative results of the developed 30-MemMP-SPE–IL-DLLME-HPLC method in the determination of PAHs
Compounds
Linear equation
LRa) (g/L)
R2
LOD (g/L)
EF
RSD (%, n = 5)
Naphthalene Acenaphthene Anthracene Chrysene Perylene
y = (15 784 ± 348)x + (32 649 ± 6269) y = (9339 ± 520)x + (38 108 ± 5178) y = (362 157 ± 7621)x + (303 229 ± 11 069) y = (262 624 ± 8063)x + (38 344 ± 2696) y = (173 948 ± 1993)x + (62 537 ± 7662)
1.0–500 1.0–500 0.02–20 0.2–50 0.2–50
0.9998 0.9999 0.9983 0.9999 0.9999
0.1 0.1 0.002 0.02 0.02
2768 3640 4847 4837 5409
4.0–6.8b) 0.4–6.2b) 8.4–1.0b) 6.5–3.5b) 6.9–2.3b)
a) LR, linear range. b) The former RSD obtained at lower spiked levels (10 g/L for naphthalene and acenaphthene; 0.2 g/L for anthracene, chrysene, and perylene), and the latter RSD obtained at higher spiked levels (200 g/L for naphthalene and acenaphthene; 10 g/L for anthracene; 20 g/L for chrysene and perylene). Table 3. Results for determination of analytes in spiked tap, well, river, and surface waters
Compounds
Naphthalene
Acenaphthene
Anthracene
Chrysene
Perylene
Added (g/L)
0 20.0 100.0 0 20.0 100.0 0 5.0 10.0 0 5.0 10.0 0 5.0 10.0
Tap water (n = 3)
Well water (n = 3)
River water (n = 3)
Surface water (n = 3)
Found (g/L)
RR (%)a)
RSD (%)
Found (g/L)
RR (%)
RSD (%)
Found (g/L)
RR (%)
RSD (%)
Found (g/L)
RR (%)
RSD (%)
– 17.1 87.9 – 17.4 98.0 – 5.7 10.5 – 5.6 10.1 – 5.7 8.1
– 85.6 87.9 – 86.9 98.0 – 114.1 105.4 – 111.8 101.0 – 114.2 80.9
– 3.5 5.3 – 1.7 4.6 – 5.9 10.1 – 9.2 8.9 – 3.2 2.4
– 21.6 111.7 – 20.0 104.8 0.1 5.2 12.0 0.1 5.2 10.5 – 5.0 9.7
– 108.1 111.7 – 100.2 104.8 – 102.3 119.1 – 102.4 104.3 – 100.4 97.1
– 4.2 8.8 – 2.1 6.8 2.6 7.9 1.4 9.1 6.2 1.9 – 1.6 4.6
– 22.6 110.6 – 20.5 108.8 – 5.2 11.6 – 5.4 9.8 – 5.6 9.99
– 112.7 110.6 – 102.5 108.8 – 104.7 116.1 – 108.2 97.1 – 111.2 99.9
– 4.5 4.2 – 5.8 8.9 – 5.4 5.2 – 2.1 2.5 – 1.4 2.5
– 19.2 104.3 – 17.4 106.5 4.3 8.6 16.3 – 6.1 10.2 – 5.4 8.3
– 95.8 104.3 – 86.9 106.5 – 86.7 120.1 – 103.2 93.2 – 107.0 83.2
– 7.5 2.8 – 2.1 4.1 5.3 1.9 9.9 – 5.6 4.1 – 4.6 1.7
and precision of the method. The data indicated that the developed 30-MemMP-SPE–[BBIM][Tf2 N]-DLLLE–HPLC method can be applied to analyze PAHs in water samples with satisfactory RRs (80.9–120.1%) and RSDs (1.4–10.1%).
3.5 Comparison with C18 -SPE and [BBIM][Tf2 N]-DLLME In order to elucidate the advantages of new SPE material and acquire a good understanding of the retention mechanism of 30-MemMP-silica adsorbent, a commercial C18 cartridge was performed under the same extraction conditions and other four PAHs (biphenyl, phenanthrene, fluoranthene, and pyrene) were extracted along with the above mentioned five analytes. Figure 5 shows the chromatogram of nine analytes after SPE with 30-MemMP-silica and C18 . It was observed evidently that 30-MemMP-SPE provided higher extraction efficiency for all analytes than C18 -SPE, and the longer the retention time of analytes is, the more the extraction efficiency on 30-MemMP-SPE is high. The extraction efficiency
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
is mainly dependent on the hydrophobic interaction between adsorbent and analytes using C18 as extraction material, and the intensity of acetone (1 mL) was not enough to elute the analytes with strong hydrophobility from C18 adsorbent, thus the ER decreased with the increase of retention time, especially for chrysene and perylene, whose ERs are
