Ionic liquid polymer functionalized carbon nanotubes-coated polyaniline for the solid-phase microextraction of benzene derivatives

Abstract

Multiwalled carbon nanotubes (MWCNTs) were non-covalently functionalized with poly(imidazolium ionic liquids) (PILs), and the resulting PIL/MWCNTs composite suspension was coated on electrodeposited polyaniline (PANI) film to fabricate a novel solid-phase microextraction coating (PANI-PIL/MWCNTs). The coating showed porous structure and had large specific surface area (231 m² g⁻¹, determined by the Brunauer–Emmett–Teller N₂ adsorption method). Its adsorption properties were explored by preconcentrating benzene derivatives from water samples prior to gas chromatography-flame ionization detection. The results showed that the PANI-PIL/MWCNTs composite had high enrichment capacity for the analytes and high stability for repeatable use, due to the synergistic effect of PANI and PIL/MWCNTs. Hence, a detection method was developed for them, and good linearity (correlation coefficients higher than 0.9953), low limits of detection (17.7–32.6 ng L⁻¹) and high precision (relative standard deviations less than 6.5% (n = 5)) were achieved. The method was applied to the determination of the benzene derivatives in real samples with recoveries of 84.0–106.9%.

---

1. Introduction

Nanostructured materials have received increasing attention because of their extensive application in many fields such as sorbents, catalyst, electric capacitors and fuel cells.^1,2 Among various nanomaterials, carbon nanotubes (CNTs) show large surface area, high electrical conductivity, rich stacking π electrons, high mechanical strength, remarkable chemical and thermal stability.^3–5 These unique characteristics make CNTs promising for diverse applications in analytical chemistry, such as in sensors,⁶ sampling,⁷ solid-phase extraction,^8–10 solid-phase microextraction (SPME)^11–14 and chromatography.^15,16

Recently, CNTs have gained more and more attention in SPME. For example, they were used as adsorption materials for the separation of polybrominated diphenyl ethers¹² and polar aromatic compounds.¹³ They were also used as adsorbents for the determination of phenolic compounds in environmental samples.¹⁴ In those works, CNTs were anchored on proper substrate surface either through physical deposition^12,13 or through chemical-bonding process.¹⁴ The chemical-bonding process is usually complicated. Furthermore, the extended π conjugation in CNTs may be disrupted and the electrical and/or mechanical properties may be altered, influencing their application. The physical deposition is relatively simple to handle. However, CNTs tend to agglomerate in the processing solvent due to strong tube–tube interactions, making physical depositing difficult. To improve the compatibility and dispersion of CNTs, they were modified by two important approaches including covalent and non-covalent functionalization.¹⁷ The covalent functionalization approach had its limitations such as long functionalization time and harsh chemical treatment, leading to destabilization of nanostructure, surface defects and tube shortening.¹⁸ On the contrary, non-covalent wrapping had gained recognition by virtue of simplicity, versatility, speed and green chemistry aspects.¹⁷ Physical interactions such as van der Waals, π–π stacking or electrostatic forces between CNTs and functional groups provide a non-covalent way to prepare functionalized CNTs.

The non-covalent functionalization of CNTs with polymerized ionic liquids (PILs) was proved to be effective. Tunckol et al.¹⁹ reported the non-covalent functionalization of CNTs with PILs, the hybrids showed improved conductivity, compatibility and stability owing to the excellent physicochemical properties of PILs. Moreover, PILs are favorable adsorbent materials in SPME.^20,21 Feng et al.²² fabricated CNTs doped-PILs fiber for multiple SPME of 2-naphthol in fruit samples. The adsorption capacity and stability of the composite were significantly improved owing to the synergistic effect of CNTs and PILs. Nevertheless, the pretreatment of stainless steel wire substrate was time-consuming.

In this work, PIL functionalized multiwalled CNTs (PIL/MWCNTs) were prepared. The approach was based on the thermal-initiation free radical polymerization of IL monomer 1-vinyl-3-hexyl-imidazolium bis(trifluoromethanesulfonyl)imide ([VHIM]NTf₂) to form PIL on MWCNTs surface. The PIL film could prevent the aggregation of MWCNTs and improved the dispersion of MWCNTs in water and organic solvent.^23,24 On the other hand, the structure was less damaged by this way because of the mild polymerization of the IL monomer. The PIL/MWCNTs was coated on an electrodeposited polyaniline (PANI) fiber to give PANI-PIL/MWCNTs composite coating. The illustration of the preparation of PANI-PIL/MWCNTs coated fiber was presented in Scheme 1. In this case, PIL not only played the role of a functional reagent, but also of a sorbent. The obtained coating was applied to the separation and enrichment of benzene derivatives from water samples. Benzene derivatives are important environmental contaminants, they come from industrial production, automobile exhaust and human activity etc., and their sensitive detection is very required. Owing to the combined effect of porous PANI and rich π electron stacking PIL/MWCNTs, the composite exhibited high affinity to the benzene derivatives, thus it could effectively adsorb and enrich them.

Scheme 1 Schematic demonstration of the preparation of PANI-PIL/MWCNTs coated fiber.

2. Experimental

2.1. Apparatus

A CHI 617A electrochemical workstation (CH Instrument Corp., Shanghai, China) was employed for preparing SPME fibers. Conventional three-electrode system was adopted, including a stainless steel wire (2 cm × 250 μm O.D.) as working electrode, a Pt wire (2.5 cm × 0.1 cm O.D.) as counter electrode and a saturated calomel electrode (SCE) as reference electrode. The commercial fibers polydimethylsiloxane (PDMS, 100 μm) and polydimethylsiloxane/divinylbenzene (PDMS/DVB, 65 μm) were purchased from Supelco (Bellefonte, PA, USA), which were used for the comparison experiment.

A Model GC-2010 gas chromatography system equipped with a flame ionization detector (FID) (Shimadzu Corporation, Japan) was used for GC experiments. A GC solution chromatographic workstation program (Shimadzu Corporation) was used to process chromatographic data. The separation of benzene derivatives was carried out on the capillary column HP-5 (USA, 30 m long × 0.25 mm I.D., 0.25 μm film thickness). The column temperature was programmed as follows: initial temperature was held at 50 °C for 3 min, programmed at 10 °C min⁻¹ to 110 °C, and 2 °C min⁻¹ to 120 °C, then 10 °C min⁻¹ to 160 °C and kept at this temperature for 2 min. The total run time was 20 min. The injector mode was splitless, the temperatures of injector and detector were 250 °C. Ultrapure nitrogen was used as carrier gas at a constant flow rate of 1 mL min⁻¹. Hydrogen and air flow rates were maintained at 40 and 400 mL min⁻¹, respectively. The SPME device was laboratory-made. FT-IR spectra were recorded with a Nexus-670 Fourier transform infrared spectrometer (Nicolet, USA). The scanning electron microscopy (SEM) images were obtained using an LEO 1530 field emission SEM (Carl Zeiss NTS GmbH, Germany). Transmission electron microscopy (TEM) analysis was performed with a JEM-2100 (200 kV) electron microscope. Surface area arrangements were performed on an ASAP-2020 M gas adsorption instrument (Micromeritics, Atlanta) (77 K) based on the Brunauer, Emmet, and Teller (BET) method.

2.2. Reagents

All chemicals and reagents were analytical grade. Aniline (ANI), nitric acid, 2,2′-azobis(isobutyronitrile) (AIBN, 98%) and N,N-dimethylformamide (DMF) were obtained from the Reagent Factory of Shanghai (China). ANI was further purified by distillation before use. 1-Vinyl-3-hexyl-imidazolium bis(trifluoromethanesulfonyl)imide ([VHIM]NTf₂), 1-vinyl-3-ethyl imidazole bromine ([VEIM]Br) and 1-vinyl-3-ethyl imidazole bis(trifluoromethanesulfonyl)imide ([VEIM]NTf₂) were supplied by Reagent Factory of Lanzhou (Lanzhou, China). Pristine MWCNT (average number of walls: 5–15, length: 0.1–10 μm, external diameter: 10–15 nm, purity > 95%) was purchased from XF NANO, INC (Nanjing, China). Chlorobenzene (CB), bromobenzene (BB), 4-bromine toluene (4-BT), 1,4-dichlorobenzene (1,4-DCB) and 1,2,4-trichlorobenzene (1,2,4-TCB) were purchased from the Reagent Factory of Shanghai. The stock solutions (100 mg mL⁻¹) of these benzene derivatives were prepared with methanol. The mixture solution contained 1 mg mL⁻¹ of each benzene derivative. The working solutions (0.05–500 μg L⁻¹) were prepared by diluting the mixture solution with saturated NaCl aqueous solution.

The water samples were lake water (Sample 1, from East Lake, Wuhan, China), waste water (Sample 2, from a local petrochemical factory, Wuhan) and tap water (Sample 3, from the laboratory, Wuhan). These samples were analyzed immediately after sampling without any pretreatment process.

2.3. In situ synthesis of PIL/MWCNTs

PIL/MWCNTs was prepared according to literature with minor modifications.¹⁹ In situ polymerization of IL monomer was carried out in ethanol in the presence of pristine MWCNTs (5 wt% with respect to the monomer) and AIBN (2 wt% with respect to the monomer). Prior to reaction, the MWCNTs were dispersed in the medium by sonication. Then the mixture was refluxed for 22 h at 65 °C under vigorous stirring and N₂ atmosphere. At the end of the reaction, it was observed that the viscosity of the medium increased. The product was filtered (with 0.45 μm filter funnel), washed with ethanol for several times to thoroughly remove weakly adsorbed polymer and unreacted monomer from the surface of MWCNTs. The final product, referred to as PIL/MWCNTs, was dried for 24 h in an oven at 40 °C to remove the residual solvent. PILs were prepared for comparing study.

2.4. Preparation of PANI-PIL/MWCNTs coated fiber

The PANI coating was directly electrodeposited on a stainless steel wire surface from a 10 mL 1 M HNO₃ aqueous solution containing 0.1 M ANI by using cyclic voltammetry (CV) technique. Prior to electrochemical polymerization, the stainless steel wire was ultrasonicated in 0.5 M H₂SO₄, 1 M NaOH and distilled water each for 15 min and then dried at room temperature. The CV was performed between −0.2 V and 1.5 V at a scan rate of 50 mV s⁻¹, and the number of scan cycle was set at 50. The obtained fiber was washed with distilled water to remove unwanted chemicals such as monomer and supporting electrolyte, and subsequently kept in a desiccator for 10 h at room temperature. Afterwards, it was aged in an electric furnace at 100 °C for 30 min and then at 250 °C for 2 h under a gentle stream of N₂.

The SPME fiber was coated with PIL/MWCNTs composite through the following procedures. Briefly, the composite was dispersed in DMF by sonication to prepare a 2 mg mL⁻¹ suspension. The PANI fiber was dipped vertically into the coating solution and slowly removed from it and kept in air for 10 min for the DMF to evaporate. This dipping and evaporating process was repeated for three times to achieve relatively thickness and better repeatability of the coating. After that, the coated fiber was placed in a desiccator at room temperature for 12 h. Finally, it was mounted on a GC microsyringe device for SPME. Before use, the PANI-PIL/MWCNTs coated fiber was conditioned in the GC injector at 250 °C under nitrogen until a stable GC baseline was obtained. For comparing study, a PANI-PIL coating was fabricated by the same way. The surface area of the coating was determined by BET N₂ adsorption method. The process was as follows: the PANI-PIL/MWCNTs coating was scraped from the fiber after aging, and then gathered together, and finally ground into powder.

2.5. Headspace SPME (HS-SPME) procedure and GC determination

In view of the abundant π electron stacking in PIL/MWCNTs, benzene derivatives were used as model analytes. A 10 mL aqueous solution of benzene derivatives (0.05–500 μg L⁻¹) was transferred in a 15 mL vial capped with PTFE-coated septum. To perform extraction, the needle of the SPME device was pierced through the septum, and the PANI-PIL/MWCNTs coated fiber was exposed to the headspace above solution for a certain time. In this work, extraction was performed at 40 °C for 20 min under magnetic agitation at 500 rotation per min. After extraction, the needle was removed from the vial and immediately transferred to the GC injection port for thermal desorption at 250 °C for 3 min and for subsequent analysis. Every day, before the measurements, the fibers were conditioned in the GC injection chamber (at 250 °C) for 30 min. In addition, the column blank and fiber blank were recorded to determine the extent of any laboratory contamination.

3. Results and discussion

3.1. Characterization of PANI-PIL/MWCNTs coating

3.1.1. Characterization of PIL/MWCNTs. The resulting PIL/MWCNTs is verified by FT-IR (Fig. 1). According to the spectrum of MWCNTs (curve a), the characteristic peaks at 1724 and 1580 cm⁻¹ can be assigned to the carboxylic C=O band²⁵ and the asymmetric C=C stretching vibration, respectively.²⁶ The FT-IR spectrum of PIL/MWCNTs (curve b) shows the characteristic vibration bands corresponding to the imidazolium cation. The peak at 1470 cm⁻¹ is ascribed to the stretching vibrations of the C–N bonds in the aromatic system of imidazole and 1190 cm⁻¹ is ascribed to imidazole ring in-plane asymmetric stretching vibration. These absorption bands are consistent with those of PIL (curve c), indicating that PIL is successfully introduced onto MWCNTs.

Fig. 1 FT-IR spectra of MWCNTs (a), PIL/MWCNTs (b) and PIL (c).

The microstructures of MWCNTs and PIL/MWCNTs composite are investigated by TEM (Fig. 2). Fig. 2b demonstrates the presence of a thin continuous layer of PIL on the surface of MWCNTs. The thickness of the PIL layer varies between 1 and 3 nm.

Fig. 2 TEM images of MWCNTs (a) and PIL/MWCNTs (b).

The PIL/MWCNTs showed good dispersibility in DMF. When they were dispersed in DMF they could readily form homogeneous black solution (Fig. S1a†), the suspension was fairly stable and no coagulation was observed even after storing for more than one month. The reason was that when MWCNTs were decorated with PIL, a large number of functional groups were introduced on the surface of MWCNTs. On the other hand, the PIL film on the MWCNTs created a distribution of ionic species with positive charge that prevented the aggregation of the MWCNTs. In contrast, MWCNTs couldn't be well dispersed in DMF, and they easily precipitated to the bottom of the vial after ultrasonication (Fig. S1b†).

3.1.2. Morphology of PANI-PIL/MWCNTs coated fiber. As shown in Fig. 3, the SEM images clearly demonstrate that the PANI-PIL/MWCNTs coating is uniform, with a thickness of about 35 μm. From the higher magnification image, it can be learnt that it is porous with scattered pasty covering. Compared with previously reported net-like PANI coating, the composite coating has large effective surface area (231 m² g⁻¹), which is favorable for the adsorption of analytes. Through the cross section, it can be observed that the outer layer (PIL/MWCNTs) has a close contact with the inner layer (PANI), the thickness of the outer layer is approximately 30 μm. The strong interaction between these two layers enhances the mechanical stability and durability of the composite coating.

Fig. 3 SEM images of PANI-PIL/MWCNTs fiber (a), the cross-section image of PANI-PIL/MWCNTs fiber (b). Inset: magnified image of PANI-PIL/MWCNTs.

3.1.3. Thickness control of the coating. It was necessary to optimize the coating method (spraying or dipping) and the overall parameters such as the number of coating cycle. The spraying method resulted in highly irregular coatings, thus, dip-coating was adopted. The optimized procedure was that presented in the Experimental section. The thickness of the PIL/MWCNTs coating was controlled by changing the coating times. The coating thickness was observed to increase linearly with the number of coating cycles (Fig. S2†). In general, the enrichment efficiency increased with the coating thickness increasing. Considering the time consumption and enrichment efficiency, here the number of coating cycle was set at three for fabricating the PIL/MWCNTs coating for further application.

3.2. Adsorptive selectivity of the PANI-PIL/MWCNTs coating

The research about the adsorption characteristics of PANI-PIL/MWCNTs composite for organic analytes is critical for its potential applications. Considering the rich stacking π electrons of PIL/MWCNTs, a series of organic compounds, including nonpolar analytes (i.e. halogenated aromatics), polar analytes (i.e. benzaldehyde, aniline, phemethylol, phenol and benzoic acid, n-octanol) and aliphatic hydrocarbons (i.e. n-octane) were selected as models, and their adsorption affinity on PANI-PIL/MWCNTs coating was evaluated by the enhancement factors (EFs). The EF was defined as the ratio of the chromatographic peak area of an analyte after HS-SPME to that before extraction (i.e., by directly injecting 1 μL standard solution).²⁷ The results were shown in Table 1. The PANI-PIL/MWCNTs coating showed much higher affinity to halogenated aromatics (EFs, 430–629) than to other analytes (EFs ≤ 265) studied. As nonpolar compounds, the adsorption of halogenated aromatics on the PANI-PIL/MWCNTs coating mainly relied on hydrophobic effect (K_OW), so hydrophobicity of analytes was considered an important factor.

Table 1 EFs for different compounds on PANI-PIL/MWCNTs coated fiber

Compounds	Structure	Molecular weight	log KOW^a	EFs (mean ± SD, n = 3)
a KOW: n-octanol/water partition coefficients. Data taken from ref. 27 and a databank at http://www.syrres.com/what-we-do/databaseforms.aspx?id=386.
Chlorobenzene		112.5	2.72	430.2 ± 24.0
Bromobenzene		157	3.02	628.7 ± 12.3
1,4-Dichlorobenzene		147	3.38	575.0 ± 28.2
4-Bromotoluene		171	3.45	516.7 ± 23.4
1,2,4-Trichlorobenzene		181.5	4.44	431.5 ± 17.0
Benzaldehyde		106	1.48	126.8 ± 3.6
Phenylamine		93	0.90	132.9 ± 6.8
Phemethylol		108	1.10	175.6 ± 5.2
Phenol		94	1.46	152.3 ± 5.9
Benzoic acid		122	1.87	265.2 ± 10.2
n-Octanol		130	3.07	118.3 ± 4.5
n-Octane		114	5.18	65.8 ± 5.7

---

The presence of a delocalized π system in the target molecules played a significant role in the selectivity of the PANI-PIL/MWCNTs coating. For example, the hydrophobicity of n-octane was higher than that of benzene derivatives, but it exhibited much lower EFs. Although the hydrophobicity of bromobenzene and n-octanol was similar, bromobenzene showed much higher EF than n-octanol. This indicated that the π–π interaction between analytes and PANI-PIL/MWCNTs coating was another potential factor.

It was reported that hydrogen bonding also contributed to the interaction between the polar analytes and the adsorbents. The EFs of polar analytes were higher than n-octane, probably because their stronger ability to form hydrogen bonding with oxygen atoms in [VHIM]NTf₂.

According to that mentioned above, we concluded that the high extraction ability of the PANI-PIL/MWCNTs coating was due to the combined interaction of hydrophobic effect, π–π interaction and hydrogen bonding interaction. Therefore, five benzene derivatives (i.e. chlorobenzene, bromobenzene, 4-bromine toluene, 1,4-dichlorobenzene and 1,2,4-trichlorobenzene) were selected as target analytes for investigating the extraction properties of the PANI-PIL/MWCNTs composite. The adsorption isotherms of the PANI-PIL/MWCNTs coating for these benzene derivatives were preliminarily investigated (Fig. 4). The high adsorption affinity to benzene derivatives makes the composite an excellent adsorbent in real sample analysis.

Fig. 4 Adsorption yield curves of the PANI-PIL/MWCNTs coating for benzene derivatives in a concentration range of 0.05–500 μg L⁻¹. Error bars show the standard deviation (n = 3).

3.3. Influence of experimental conditions on extraction efficiency of the PANI-PIL/MWCNTs fiber

Different experimental parameters that affected extraction efficiency, including extraction time, extraction temperature and ion strength, were investigated and optimized in sequential order using 10 mL aqueous solution spiked with 50 μg L⁻¹ of benzene derivatives (Fig. S3†).

The effect of extraction time on the extraction efficiency was examined from 10 to 50 min. Results indicated that the extraction efficiency reached the maximum around 20 min and then decreased a little. This could be ascribed to the competitive adsorption of water molecules on PANI-PIL/MWCNTs coating. The equilibrium was reached in short time due to the high diffusion coefficients and fast adsorption kinetics of the analytes on PANI-PIL/MWCNTs coating. Hence, 20 min was adopted for further experiments.

Temperature is another major parameter affecting extraction efficiency. In general, increasing extraction temperature the extraction rate increases, but the distribution constant decreases. As shown in Fig. S3,† the highest extraction efficiency was observed at 40 °C for five analytes. Accordingly, 40 °C was selected for subsequent experiments.

Ionic strength usually influences the solubility of some analytes in aqueous phase due to the salt-out effect, hence their concentrations in the headspace change. In this work, the influence of ionic strength was investigated by varying NaCl concentration from 0 to 0.35 g mL⁻¹. Results showed that the highest peak area could be obtained at 0.35 g mL⁻¹ (i.e. saturated NaCl solution).

In regards to desorption time, 3 min was sufficient to desorb the analytes from the PANI-PIL/MWCNTs fiber at 250 °C. To study the carryover effect, blank tests were run with the fiber after desorption of the extracted compounds. No signal of analytes was detected after desorption at 250 °C for 3 min.

3.4. Comparison of PANI-PIL/MWCNTs fiber with PANI-PIL and PANI fibers

The developed PANI-PIL/MWCNTs fiber was compared with PANI-PIL and PANI fibers for SPME of the benzene derivatives under the same conditions. The comparison was accomplished by calculating the adsorption amounts of analytes, and the results were shown in Fig. 5A. The PANI-PIL/MWCNTs fiber (coating-thickness: 35 μm) displayed much higher enrichment efficiency (minimum 3-fold for CB and maximum 5-fold for BB) than PANI fiber (with same coating thickness) for SPME of the benzene derivatives. The improved performance of the PANI-PIL/MWCNTs coating likely resulted from enhanced π–π interaction and the hydrophobic effect. When compared with PANI-PIL fiber, the enrichment efficiency increased by 2–3 folds. It deserved to mention that [VHIM]NTf₂ was selected in the experiments. With [VHIM]NTf₂, the obtained PANI-PIL/MWCNTs coating demonstrated higher extraction efficiency than that with the other two (i.e. [VEIM]Br and [VEIM]NTf₂) (Fig. S4†).

Fig. 5 (A) Comparison of the PANI-PIL/MWCNTs fiber with PANI-PIL and PANI fibers for the SPME of the benzene derivatives at 50 μg L⁻¹. (B) Comparison of the PANI-PIL/MWCNTs fiber with commercial PDMS and PDMS/DVB fibers for the SPME of the benzene derivatives at 50 μg L⁻¹. Sample volume, 10 mL; extraction time, 20 min; extraction temperature, 40 °C; desorption time, 3 min; desorption temperature, 250 °C. Error bar shows the standard deviation for triplicate extractions.

3.5. Comparison of PANI-PIL/MWCNTs fiber with commercial PDMS and PDMS/DVB fibers

The PANI-PIL/MWCNTs coating was also compared with commercial PDMS (coating-thickness: 100 μm) and PDMS/DVB (coating-thickness: 65 μm) fibers for SPME of the benzene derivatives under the same conditions. Since PANI-PIL/MWCNTs and commercial coatings have different thickness and length, EFs of these coatings for various target compounds were finally calculated and compared based on per cubic millimeter of fiber coatings in this study. The PDMS/DVB fiber worked better than the PDMS fiber but worse than the PANI-PIL/MWCNTs fiber for the SPME of benzene derivatives (Fig. 5B). The high adsorption affinity of the PANI-PIL/MWCNTs coating was related to the large surface area and unique porous structure of PANI and PIL/MWCNTs, hydrophobic effect, π–π interactions of the aromatic rings of the analytes with PIL/MWCNTs.

3.6. Reusability of the coating

The stability of a coating over time is a practical parameter of top importance in the SPME technique. To evaluate the endurance and reusability of the PANI-PIL/MWCNTs coating, it was subjected to a series of 200 successive headspace SPME cycles. As presented in Fig. S5,† the fiber endurance was measured as the amount of analytes extracted along with the extraction times. Responses were normalized by taking extraction efficiency of the first extraction for each benzene derivative as 100%. Extraction efficiency decreased by 13–18% in comparison with the maximum measured value after 200 adsorption/desorption cycles. This indicated that the PANI-PIL/MWCNTs fiber was very stable and reproducible for SPME.

3.7. Method evaluation

The characteristic parameters of the analytical method were investigated and listed in Table 2. Linearity was tested by extracting from aqueous solution containing these benzene derivatives at concentration changing from 0.05 μg L⁻¹ to 500 μg L⁻¹. Satisfactory linearity was obtained with correlation coefficients above 0.9953. The limits of detection (LODs), defined as three times of the standard deviation of the obtained peak areas at the lowest concentration divided by the slope of the calibration curve, were 17.7 to 32.6 ng L⁻¹.²⁸ The limits of quantitation (S/N = 10) were found to be 53.6 to 103.7 ng L⁻¹. The repeatability was measured, and the relative standard deviations (RSDs) were found to be <5.2% for one fiber (n = 5) and <6.5% for fiber-to-fiber (five fibers, 3 replicates each).

Table 2 Analytical parameters for benzene derivatives measured with PANI-PIL/MWCNTs coated fiber based HS-SPME-GC method

Analytes	Regression equation	R	Linear range (μg L^−1)	LOD (ng L^−1)	LOQ (ng L^−1)	RSD (%)
						One fiber (n = 5)	Fiber to fiber (n = 5)
CB	y = 2653.6x + 13 225	0.9975	0.1–250	32.6	101.1	3.4	5.1
BB	y = 3009.1x + 18 418	0.9983	0.1–250	32.4	103.7	4.5	5.7
DCB	y = 3594.6x + 26 751	0.9953	0.05–250	17.7	53.6	3.8	6.1
BT	y = 3203.5x + 21 456	0.9956	0.05–250	18.8	57.8	5.2	6.5
TCB	y = 3002.6x + 15 245	0.9967	0.05–250	21.6	71.3	3.9	5.3

---

3.8. Sample analysis

The method was applied to the analysis of benzene derivatives in three water samples, including lake water, petrochemical waste water and tap water. As shown in Table 3 and Fig. 6, CB and DCB were detected at the concentration of 2.6 and 3.2 μg L⁻¹ for Sample 2, respectively, while the concentration of other benzene derivatives were below the LODs. No analytes were found in Sample 1 and Sample 3. Recoveries were obtained by spiking with 5 μg L⁻¹ of benzene derivatives and ranged from 84.0 to 106.9%. Fig. 6 presents the GC chromatograms of benzene derivatives extracted with the PANI-PIL/MWCNTs fiber from the standard solution (a) and Sample 2 (b). According to the peak areas and linear regression equations, the concentrations of the analytes were calculated (Table 3). For comparison, the chromatograms of benzene derivatives in Sample 2 after extracted with the PDMS and PDMS/DVB fibers were presented in Fig. S6.† They were not so good as Fig. 6. All the characteristic parameters of the method validated that this method was reliable and sensitive for the analysis of benzene derivatives in water samples.

Table 3 GC-FID analysis of benzene derivatives in water samples after HS-SPME with PANI-PIL/MWCNTs coating

Analytes	Sample 1		Sample 2		Sample 3
	Detected concentration (μg L^−1)	Recovery^a (%)	Detected concentration (μg L^−1)	Recovery^a (%)	Detected concentration (μg L^−1)	Recovery^a (%)
a Spiked: 5 μg L^−1 level of benzene derivatives.b nd, not detected.c Mean value ± standard deviation (n = 3).
CB	nd^b	106.5 ± 8.4^c	3.8 ± 0.17	105.9 ± 1.1	nd	104.0 ± 3.4
BB	nd	104.8 ± 2.1	nd	101.4 ± 1.7	nd	95.3 ± 1.2
DCB	nd	91.7 ± 3.5	2.1 ± 0.09	96.8 ± 7.4	nd	84.0 ± 3.8
BT	nd	100.5 ± 8.1	nd	89.4 ± 9.8	nd	96.8 ± 3.9
TCB	nd	86.5 ± 6.1	nd	106.9 ± 3.1	nd	87.8 ± 4.7

---

Fig. 6 Chromatograms of benzene derivatives in 5 μg L⁻¹ standard solution (a) and Sample 2 (b) after extracted with the PANI-PIL/MWCNTs coating. Peaks: (1) CB, (2) BB, (3) DCB, (4) BT, (5) TCB. Extraction temperature, 40 °C; extraction time, 20 min; saturated NaCl; desorption temperature, 250 °C; desorption time, 3 min.

4. Conclusions

In summary, we report the first example of the preparation of PANI-PIL/MWCNTs composite coating by non-covalent functionalizing MWCNTs with PILs and then immobilizing them on electrodeposited PANI fiber. The coating presents porous structure and has large specific surface area. Owing to the synergetic effect of porous PANI and rich π electron stacking PIL/MWCNTs, the PANI-PIL/MWCNTs coating offers high enrichment efficiency for the benzene derivatives. Coupled with GC-FID analysis, wide linear ranges, low LODs, and good repeatability are achieved. The PANI-PIL/MWCNTs composite is robust enough for 200 repeatable uses without damage of adsorption performance. In general, the PANI-PIL/MWCNTs coated fiber has potential application as an effective and useful extraction tool.

Acknowledgements

The authors appreciate the support of the National Natural Science Foundation of China (Grant No. 21275112).

References

1. C. G. Wu and T. Bein, Science, 1994, 266, 1013–1015.
2. S. H. Joo, S. J. Choi, I. Oh, J. Kwak, Z. Liu, O. Terasaki and R. Ryoo, Nature, 2001, 412, 169–172.
3. R. H. Baughman, A. A. Zakhidov and W. A. de Heer, Science, 2002, 297, 787–792.
4. P. M. Ajayan and O. Z. Zhou, Top. Appl. Phys., 2001, 80, 391–425.
5. L. T. Qu and L. M. Dai, J. Am. Chem. Soc., 2005, 127, 10806–10807.
6. J. Wang, G. Chen, M. P. Chatrathi and M. Musameh, Anal. Chem., 2004, 76, 298–302.
7. A. A. Ensafi, M. Jokar and M. Ghiaci, J. Iran. Chem. Soc., 2015, 12, 457–467.
8. Q. Zhou, J. Xiao and Y. Ding, Anal. Chim. Acta, 2007, 602, 223–228.
9. I. Marquez-Sillero, E. Aguilera-Herrador, S. Cardenas and M. Valcarcel, J. Chromatogr. A, 2010, 1217, 1–6.
10. Z. Guan, Y. Huang and W. Wang, Anal. Chim. Acta, 2008, 627, 225–231.
11. W. Du, F. Q. Zhao and B. Z. Zeng, J. Chromatogr. A, 2009, 1216, 3751–3757.
12. J. X. Wang, D. Q. Jiang, Z. Y. Gu and X. P. Yan, J. Chromatogr. A, 2006, 113, 8–14.
13. W. F. Chen, J. B. Zeng, J. M. Chen, X. L. Huang, Y. Q. Jiang, Y. R. Wang and X. Chen, J. Chromatogr. A, 2009, 1216, 9143–9148.
14. H. M. Liu, J. B. Li, X. Liu and S. X. Jiang, Talanta, 2009, 7, 929–935.
15. P. W. Stege, A. V. Lapierre and L. D. Martinez, Talanta, 2011, 86, 278–283.
16. C. André, T. Gharbi and Y. C. Guillaume, J. Sep. Sci., 2009, 32, 1757–1764.
17. N. Karousis and N. Tagmatarchis, Chem. Rev., 2010, 110, 5366–5397.
18. P. X. Hou, C. Liu and H. M. Cheng, Carbon, 2008, 46, 2003–2025.
19. M. Tunckol, S. Fantini, F. Malbosc, J. Durand and P. Serp, Carbon, 2013, 57, 209–216.
20. T. D. Ho, H. L. Yu, W. T. S. Cole and J. L. Anderson, Anal. Chem., 2012, 84, 9520–9528.
21. J. J. Feng, M. Sun, J. B. Li, X. Liu and S. X. Jiang, J. Chromatogr. A, 2012, 1227, 54–59.
22. J. J. Feng, M. Sun, L. L. Li, X. J. Wang, H. M. Duan and C. N. Luo, Talanta, 2014, 123, 18–24.
23. B. H. Wu, D. Hu, Y. J. Kuang, B. Liu, X. H. Zhang and J. H. Chen, Angew. Chem., Int. Ed., 2009, 48, 4751–4754.
24. B. H. Wu, D. Hu, Y. M. Yu, Y. J. Kuang, X. H. Zhang and J. H. Chen, Chem. Commun., 2010, 46, 7954–7956.
25. M. J. Park, J. K. Lee, B. S. Lee, Y. W. Lee, I. S. Choi and S. G. Lee, Chem. Mater., 2006, 18, 1546–1551.
26. H. J. Jiang, L. B. Zhu, K. S. Moon and C. P. Wong, Carbon, 2007, 45, 655–661.
27. H. L. Xu, Y. Li, D. Q. Jiang and X. P. Yan, Anal. Chem., 2009, 81, 4971–4977.
28. S. L. Zhang, Z. Du and G. K. Li, Anal. Chem., 2011, 83, 7531–7541.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra19027d

---