Ke-Jing
Huang
*,
Yu-Jie
Liu
,
Jing
Li
,
Tian
Gan
and
Yan-Ming
Liu
College of Chemistry and Chemical Engineering, Xinyang Normal University, Xinyang 464000, P.R. China. E-mail: kejinghuang@163.com; Tel: +86-376-6390611
First published on 30th October 2013
A graphene (Gr) functionalized silica gel (Gr–SiO2) sorbent was prepared and used as a new sorbent for solid-phase extraction (SPE) of polycyclic aromatic hydrocarbons (PAHs) including naphthalene, acenaphthylene, fluorene, phenanthrene, fluoranthene, and pyrene. The analytes were separated and detected by high-performance liquid chromatography (HPLC) with a UV detector. Factors affecting the extraction efficiency including the eluent type and its volume, adsorbent amount, sample volume, sample pH and flow rate were optimized. Under the optimized conditions, the present method showed a wide linear range (0.01–600 μg L−1), acceptable repeatability of the extraction (RSDs of 1.7–8.2%, n = 3) and satisfactory detection limits (0.0029–0.052 μg L−1). The recovery studies using this new sorbent were performed by three consecutive extraction steps of water and milk samples at three spiked levels. The average recoveries of the analytes in environmental water and milk samples were 89.0–115.4%, which demonstrated the applicability of the developed method.
Some analytical techniques have been developed for the determination of PAH compounds in different samples, such as synchronous fluorescence spectroscopy,4 gas chromatography,5 capillary electrophoresis,6 and high-performance liquid chromatography (HPLC).7–10 However, clean-up and enrichment procedures are usually required because of the complexity of the matrix and the low content of PAHs in environmental samples. Because of its advantages of high enrichment factor, high recovery, rapid phase separation, low cost, and low consumption of organic solvents, solid-phase extraction (SPE) has been the most common technique used for preconcentration of analytes in various samples.11 The choice of appropriate adsorbent is a critical factor to obtain good recovery and a high enrichment factor in the SPE procedure.12 For sample pretreatment, carbon nanoparticles represent some exciting prospects, such as the ease of material processing, low cost of synthesis, high surface area, good chemical stability, and strong mechanical strength,13,14 and they have been widely applied in sample pretreatment procedures and good analytical performance is obtained.15,16
As one of the carbon nanomaterials, graphene (Gr) attracts an enormous amount of interest from scientists.17–20 The two-dimensional structure of Gr provides numerous unexpected properties for sample pretreatment: both sides of the planar sheets are available for molecule adsorption, suggesting high sorption capacity;21 the high specific surface area (close to 2630 m2 g−1) means high sorption capacities; easy synthesis and processing imply low cost. These unusual and robust properties are useful in sample treatment procedures. Most recently, we reported a high-performance liquid chromatography (HPLC) method for sensitive determination of γ-aminobutyric acid, taurine and glycine in rat brain using Gr as a sorbent for SPE.22 The results revealed the potential of Gr as a sorbent in the analysis of biological samples. Wu et al. reported a Gr-based magnetic nanocomposite used as an adsorbent for the preconcentration of the five carbamate pesticides (metolcarb, carbofuran, pirimicarb, isoprocarb and diethofencarb) in environmental water samples prior to high performance liquid chromatography-diode array detection.23 Zhang et al. reported gas chromatography-mass spectrometry coupled with micro-solid-phase extraction for the determination of polycyclic aromatic hydrocarbons (PAHs) in water based on sulfonated Gr sheets.24 However, some problems still need to be solved when using Gr as an adsorbent in sample pretreatment. Gr usually congregates when packed into SPE cartridges, which leads to blockage in the extraction column. Therefore, many interesting and unique properties of Gr can only be realized after it is integrated into more complex assemblies. A useful technique to incorporate Gr into such assemblies is through chemical modification, which enables chemical covalent bonding between the Gr and the material of interest.25
In this work, a new sorbent of SPE was prepared based on Gr functionalized silica gel (Gr–SiO2). Six PAHs including naphthalene (Nap), acenaphthylene (Acy), fluorene (Flo), phenanthrene (Ph), fluoranthene (Fl) and pyrene (Py) were selected to test the extraction performances of Gr–SiO2 sorbents. The six PAHs were separated and detected by the HPLC-UV technique after extracted by Gr–SiO2 packed SPE cartridges. The results obtained with the Gr–SiO2 were compared with those of a commercial C18 sorbent. Finally, the Gr–SiO2 was tested in the environmental water and milk samples.
The preparation of Gr functionalized-silica gel included four steps: (a) 5 g silica gel was added to HCl solution (6 mol L−1) under stirring, and then incubated at 97 °C for 4 h. After cooling naturally, the white precipitate was collected and washed with water and ethanol, respectively, and then dried in a vacuum oven at 80 °C; (b) the resultant composite was dispersed in 75 mL toluene. 5 mL APS was then added into the above mixture and stirred at 97 °C for 12 h. The preliminary activation of silica gel was performed after washing with water and ethanol; (c) 1.0 g preliminary activated silica gel was added into 150 mL GO solution. Subsequently, 2.5 g DCC was added immediately. The mixture was stirred and kept at 97 °C until white precipitates turned to gray; (d) 0.8 g sodium borohydride was added to the above mixture and incubated at 70 °C for 8 h. The precipitate was collected and washed thoroughly with water and methanol, respectively. A Gr–SiO2 composite was obtained after being dried at 60 °C under vacuum.
EDX was used to determine the composition of the Gr–SiO2 composites. The results showed that the atomic ratio percentage C/Si was about 3 in the samples, indicating the successful anchoring of Gr onto the surface of the silica gel. The specific surface areas calculated by the BET method of silica gel and Gr–SiO2 composites are 481.5 m2 g−1 and 578.2 m2 g−1, respectively. Obviously, the BET specific surface area of Gr–SiO2 composites is much higher than that of silica gel, which means that the Gr sheets anchoring onto the silica gel can effectively increase the specific surface areas of the silica gel.
As far as the SPE method is concerned, the selection of organic solvent used as an eluent plays the most important role in the desorption efficiency. Five eluents including acetonitrile, methanol, acetone, ethanol and cyclohexane were tested. The experimental results demonstrated that acetonitrile, methanol and acetone give the higher elution efficiency than ethanol and cyclohexane (Fig. 2A). However, when acetone was used as an eluent, some impurity peaks appeared in the chromatogram, which interfered the separation of the PAHs (Fig. 2B). So acetonitrile was chosen as the eluent for further experiments.
The amount of eluent affects the desorption of PAHs. Herein, the volume of acetonitrile was optimized in the range of 0.15–0.5 mL. The results are shown in Fig. 3A. It is obvious that the extraction performance of the Gr–SiO2 packed cartridge is affected by the amount of the eluent. The peak areas of 6 objectives increase when the volume of acetonitrile varies from 0.15 to 0.2 mL, and then decrease when the eluent volume exceeds 0.2 mL. Therefore, 0.2 mL acetonitrile was selected in the further experiments.
Fig. 3 Effect of eluent volume (A), sample flow rate (B), sample pH (C), amount of NaCl (D), and sample volume (E) on the peak area of PAHs. The other conditions are same as in Fig. 4. |
For SPE, mass transfer of analytes from the sample to the sorbent materials is time-dependent and equilibrium based rather than an exhaustive extraction process. The sample flow rate is a key factor for extraction efficiency. In this work, the sample flow rate is estimated within the range of 1–5 mL min−1 (Fig. 3B). The experimental results show that the peak areas of PAHs increase with the sample flow rate up to 1.5 mL min−1 with no significant improvement thereafter. A sample flow rate of 1.5 mL min−1 was therefore selected for further work.
Sample pH plays an important role in the SPE procedure, because the pH value of the solution determines the present state of analytes in solution as an ionic or a molecular form, and thus determines the extraction efficiency of the target analytes. Herein, a range of sample pH values from 3 to 10 is evaluated (Fig. 3C). No significant effect is observed with sample pH change, suggesting that a non-electrostatic interaction such as pi–pi interaction plays an important role in the adsorption of PAHs to the Gr–SiO2 sorbent. Hence, all the environmental water samples can be preconcentrated directly without adjusting acidity because the pH value of environmental water hardly exceeds the range of pH 3–10.
The effect of salt addition has been used in SPE to decrease the solubility of analytes in the aqueous sample and enhance their partitioning into the organic solvent. For this investigation, various amounts of sodium chloride (ranging from 0 to 30%, w/v) are added to the sample to study the phenomenon (Fig. 3D). 100 mL of sample solution was used. No significant variation in the extraction efficiency is observed with the change of the NaCl concentration indicating that the ionic strength is negligible for the next analytical procedure.
The amount of adsorbent is another important parameter that affects the extraction efficiency. A quantitative retention is not obtained when the amount of Gr–SiO2 is less than optimum. On the other hand, an excess amount of adsorbent prevents the quantitative elution of the retained PAHs by a small volume of eluent. In order to examine the optimum amounts of solid phase, 10–50 mg of Gr–SiO2 was packed into the columns and the same volume of sample solution (100 mL) was treated by applying the general procedure. The results showed that the biggest peak areas of 6 PAHs were obtained when 30 mg Gr–SiO2 was used.
The adsorption capacity of a sorbent is an important parameter in assessing its ability to retain selected analytes. Breakthrough volume and optimal sample volume were investigated in order to obtain a higher extraction efficiency. In this study, different sample volumes (25, 50, 100, 125, 150, 175 mL) were employed and the peak areas of the target compounds were determined under the following conditions: 30 mg Gr–SiO2 sorbent, a flow rate of 1.5 mL min−1, and 0.2 mL acetonitrile as the eluent. The results are shown in Fig. 3E. For all analytes, the peak areas increase greatly in the sample volume range of 25–100 mL. However, a small change of the peak area is observed when the sample volume is over 100 mL. In practice, the sample volume is chosen according to the required sensitivity and the time acceptable for a whole analysis. Generally, further increasing the sample volume is not desirable for routine analysis since the total time needed for one analysis would be longer. So a sample volume of 100 mL was selected for subsequent analysis.
Analytes | Regression equation | Linear range (μg L−1) | R | RSD | LOD (μg L−1) | LOQ (μg L−1) | |
---|---|---|---|---|---|---|---|
Intraday (n = 3) | Interday (n = 3) | ||||||
Nap | y = 0.486x + 13.723 | 0.10–300 | 0.9933 | 3.6 | 7.3 | 0.052 | 0.17 |
Acy | y = 0.545x + 29.113 | 0.10–300 | 0.9958 | 5.2 | 8.2 | 0.024 | 0.079 |
Flo | y = 5.164x + 274.471 | 0.01–400 | 0.9976 | 1.7 | 3.3 | 0.0029 | 0.0096 |
Ph | y = 31.206x + 988.506 | 0.01–400 | 0.9998 | 1.9 | 4.1 | 0.0082 | 0.027 |
Fl | y = 28.007x + 118.494 | 1–600 | 0.9996 | 3.4 | 8.0 | 0.014 | 0.046 |
Py | y = 23.291x + 103.660 | 1–600 | 0.9998 | 4.5 | 7.1 | 0.012 | 0.040 |
Adsorbents | LOD (μg L−1) | Linear range (μg L−1) | RSD% | References |
---|---|---|---|---|
Multi-walled carbon nanotubes | Nap: 0.021, Acy: 0.058, Flo: 0.026, Ph: 0.009, Fl: 0.034, Py: 0.036 | Nap, Acy: 0.4–100, Flo: 0.1–25, Ph: 0.04–10, Fl, Py: 0.2–50 | Nap: 4.8, Acy: 4.0, Flo: 4.6, Ph: 3.0, Fl: 4.6, Py: 4.7 | 27 |
Magnetic C18 microspheres | Nap: 0.8, Acy: 4.1, Flo: 3.9, Ph: 3.4, Py: 5.1 | Nap, Acy, Flo, Ph, Py: 10–800 | Nap: 4.1, Acy: 5.0, Flo: 4.9, Ph: 5.2, Py: 4.0 | 28 |
TiO2 nanotubes | Acy: 0.031, Flo: 0.026, Ph: 0.053, Fl: 0.017, Py: 0.059 | Acy: 10–500, Fl, Flo: 50–400, Ph: 50–800, Py: 10–800 | — | 29 |
Zeolite imidazolate framework 8 | Nap: 0.083, Acy: 0.050, Flo: 0.067, Ph: 0.050, Fl: 0.029, Py: 0.044 | Nap, Acy, Ph, Py: 0.1–50, Flo: 0.5–50, Fl: 0.05–50 | Nap: 3.1, Acy: 4.1, Flo: 8.5, Ph: 2.2, Fl: 2.7, Py: 2.1 | 30 |
Titanate nanotube array modified by cetyltrimethylammonium bromide | Nap: 0.27, Acy: 0.19, Ph: 0.013, Fl: 0.048, Py: 0.069 | Nap, Acy, Ph, Fl, Py: 0.2–100 | Nap: 8.3, Acy: 9.2, Ph: 8.4,Fl: 1.7, Py: 2.8 | 31 |
Graphene functionalized silica gel | Nap: 0.052, Acy: 0.024, Flo: 0.0029, Ph: 0.0082, Fl: 0.014, Py: 0.012 | Nap, Acy: 0.10–300; Flo, Ph: 0.01–400; Fl, Py: 1–600 | Nap: 3.6, Acy: 5.2, Flo: 1.7, Ph: 1.9, Fl: 3.4, Py: 4.5 | This work |
The recoveries of PAHs in the spiked environmental water samples obtained from the Gr–SiO2 cartridge and C18 cartridge were 89.2–114.0% and 80.2–123.4%, respectively (Table 3), indicating that Gr–SiO2 can be an excellent SPE sorbent for PAH pretreatment and enrichment from environmental water samples. The developed method was also used to detect PAHs in milk samples. No PAHs was found in these milk samples. The recoveries of PAHs in the spiked milk samples obtained from the Gr–SiO2 cartridge and C18 cartridge were 89.0–115.4% and 82.5–125.6%, respectively (Table 4).
SPE cartridge | Analyte | Added (μg L−1) | Pond water samples | River water samples | ||
---|---|---|---|---|---|---|
Recovery (%) | RSD (%) | Recovery (%) | RSD (%) | |||
Gr–SiO2 cartridge | Nap | 5 | 94.0 | 2.1 | 114.0 | 2.6 |
50 | 109.2 | 2.7 | 108.2 | 3.7 | ||
Acy | 5 | 98.0 | 3.9 | 98.0 | 3.1 | |
50 | 109.8 | 1.7 | 98.6 | 2.2 | ||
Flo | 5 | 110.0 | 2.3 | 102.0 | 4.2 | |
50 | 102.2 | 3.6 | 101.6 | 3.6 | ||
Ph | 5 | 110.0 | 4.4 | 106.0 | 3.0 | |
50 | 113.0 | 4.3 | 94.0 | 2.5 | ||
Fl | 5 | 100.0 | 3.9 | 112.0 | 2.0 | |
50 | 101.6 | 3.7 | 89.2 | 3.7 | ||
Py | 5 | 92.0 | 3.3 | 110.0 | 3.1 | |
50 | 103.2 | 2.9 | 101.4 | 3.9 | ||
C18 cartridge | Nap | 5 | 92.4 | 3.5 | 110.3 | 4.0 |
50 | 85.6 | 3.6 | 123.4 | 3.5 | ||
Acy | 5 | 88.3 | 2.9 | 109.4 | 3.2 | |
50 | 109.3 | 4.1 | 96.4 | 2.2 | ||
Flo | 5 | 115.6 | 2.5 | 92.5 | 2.9 | |
50 | 87.2 | 3.3 | 88.4 | 3.5 | ||
Ph | 5 | 80.2 | 3.9 | 107.2 | 3.7 | |
50 | 107.5 | 4.3 | 105.3 | 2.9 | ||
Fl | 5 | 120.7 | 2.2 | 87.6 | 2.5 | |
50 | 110.2 | 3.6 | 85.3 | 3.1 | ||
Py | 5 | 96.3 | 3.8 | 91.4 | 4.2 | |
50 | 90.1 | 3.0 | 88.9 | 3.2 |
SPE cartridge | Analyte | Added (μg L−1) | Whole milk samples | Skimmed milk samples | ||
---|---|---|---|---|---|---|
Recovery (%) | RSD (%) | Recovery (%) | RSD (%) | |||
Gr–SiO2 cartridge | Nap | 5 | 89.0 | 3.2 | 98.0 | 3.6 |
50 | 102.2 | 1.9 | 95.2 | 4.2 | ||
Acy | 5 | 96.0 | 3.4 | 106.0 | 1.1 | |
50 | 104.8 | 1.8 | 115.4 | 2.2 | ||
Flo | 5 | 112.0 | 2.4 | 106.0 | 3.2 | |
50 | 107.4 | 2.3 | 94.2 | 4.0 | ||
Ph | 5 | 112.0 | 3.4 | 94.0 | 3.1 | |
50 | 102.6 | 2.4 | 97.4 | 2.3 | ||
Fl | 5 | 91.6 | 3.4 | 102.0 | 2.6 | |
50 | 96.2 | 4.2 | 90.8 | 3.7 | ||
Py | 5 | 98.0 | 2.4 | 104.0 | 3.2 | |
50 | 102.4 | 4.2 | 105.6 | 1.8 | ||
C18 cartridge | Nap | 5 | 85.3 | 4.5 | 88.2 | 3.8 |
50 | 90.4 | 3.2 | 91.6 | 2.5 | ||
Acy | 5 | 87.6 | 3.6 | 125.6 | 3.5 | |
50 | 92.5 | 3.8 | 111.2 | 2.9 | ||
Flo | 5 | 82.5 | 2.1 | 119.3 | 4.1 | |
50 | 115.3 | 2.9 | 87.2 | 4.8 | ||
Ph | 5 | 110.6 | 3.9 | 85.6 | 4.5 | |
50 | 84.6 | 3.4 | 86.2 | 3.7 | ||
Fl | 5 | 109.2 | 4.3 | 94.3 | 3.1 | |
50 | 89.3 | 2.2 | 108.2 | 2.7 | ||
Py | 5 | 96.5 | 3.7 | 114.5 | 3.8 | |
50 | 120.4 | 3.3 | 87.8 | 3.5 |
This journal is © The Royal Society of Chemistry 2014 |