Abstract
Aniline derivates are widely used in rubber products as anti-aging agents, but their biological toxicity causes harm in the use and recycling of rubber products. Therefore, in this study, a high-performance poly(3,4,5,6-tetra-fluorophthalonitrile-β-cyclodextrin)(P-TFPN-β-CD) solid-phase extraction (SPE) adsorbent was synthesized and applied to extract and detect aniline derivatives in rubber specimens. After crosslinking with 3,4,5,6-tetrafluorophthalonitrile (TFPN), the polymer exhibited mesopores (7.88 nm) and a large specific area (55.2 m2/g). The porous structure significantly improved the extraction efficiency (recovery was between 90.1% and 110.5%) and rate (60% in 10 s) of 4 aniline derivatives (aniline, N-ethyl aniline, 1,2-phenylenediamine, and p-phenylenediamine). With a combined gas chromatographic-flame ionization detector (GC-FID), the extracted aniline derivatives were accurately detected (RSD = 1.60% to 9.90%). Due to the weak interaction between P-TFPN-β-CD and analyte, the prepared poly-TFPN-β-CD SPE adsorbent was regenerated via mild washing with methanol. The high absorption capacities remained after 4 extracting-washing cycles. The novel porous SPE adsorbent showed good extracting and recycling performance for the analysis and detection of aniline compounds in rubber. Thus, it has good prospects for application to improve detection efficiency and reduce cost.
Download PDF
Full Article
Novel Poly(β-cyclodextrin) Porous Material as Solid Phase Extraction Sorbent for Aniline Derivatives in Rubber Samples
Jinlan Dai,a,b Honglei Yin,b Lingyun Yu,c Tang Hong,b Lei Zhou,d and Minghua Liu e,*
Aniline derivates are widely used in rubber products as anti-aging agents, but their biological toxicity causes harm in the use and recycling of rubber products. Therefore, in this study, a high-performance poly(3,4,5,6-tetra-fluorophthalonitrile-β-cyclodextrin)(P-TFPN-β-CD) solid-phase extraction (SPE) adsorbent was synthesized and applied to extract and detect aniline derivatives in rubber specimens. After crosslinking with 3,4,5,6-tetrafluorophthalonitrile (TFPN), the polymer exhibited mesopores (7.88 nm) and a large specific area (55.2 m2/g). The porous structure significantly improved the extraction efficiency (recovery was between 90.1% and 110.5%) and rate (60% in 10 s) of 4 aniline derivatives (aniline, N-ethyl aniline, 1,2-phenylenediamine, and p-phenylenediamine). With a combined gas chromatographic-flame ionization detector (GC-FID), the extracted aniline derivatives were accurately detected (RSD = 1.60% to 9.90%). Due to the weak interaction between P-TFPN-β-CD and analyte, the prepared poly-TFPN-β-CD SPE adsorbent was regenerated via mild washing with methanol. The high absorption capacities remained after 4 extracting-washing cycles. The novel porous SPE adsorbent showed good extracting and recycling performance for the analysis and detection of aniline compounds in rubber. Thus, it has good prospects for application to improve detection efficiency and reduce cost.
Keywords: Poly (β-cyclodextrin), Adsorbent; Aniline derivatives; Solid-phase extraction; GC-MS
Contact information: a: Technical Center of Fuzhou Customs District of P.R. China , Fuzhou 350015, P.R. China; b: Fujian Key Laboratory for Technology Research of Inspection and Quarantine, Fuzhou 350116, P.R. China; c: Technical Center of Chengdu Customs District of P.R. China, Chengdu610041, P.R. China; d: Technical Center for Safety of Industrial Products of Tianjin Customs District of P.R. China, Tianjin 300308, P.R. China; e: College of Environment and Resource, Fuzhou University, Fuzhou 350116, P.R. China; *Corresponding author: mhliu2000@fzu.edu.cn
INTRODUCTION
Rubber is widely used in various products, such as gloves, toys, shoes, seals, and tires due to its excellent elasticity, wear-resistance, and electrical insulation. However, its poor aging resistance hinders its application in daily life; therefore, aniline derivatives, which are stable and effective anti-aging agents, are added to rubber to protect it from damage by free radical reactions (Harrison and Jollow 1987; Müller et al. 1997; Janzen et al. 2006; Sarafraz-Yazdi and Es’haghi 2006; Oughlis-Hammache et al. 2016). However, according to ISO/TR 16178 (2012), aniline derivatives in rubber are considered category 5 critical substances, which are suspected to have an effect on the wearer. Further, they are potentially hazardous materials that can be inhaled or absorbed into the human body through the skin, which can cause tissue hypoxia, met-hemoglobin, and damage to the central nervous system and cardiovascular system (Ryzhakov 2018; Bhuiyan et al. 2019; Chaturvedi and Katoch 2020). Moreover, waste rubber products that contain aniline derivative residues cause soil and water pollution and affect the growth and reproduction of aquatic organisms (Miskam et al. 2014). Therefore, accurately detecting the concentration of aniline derivatives in rubber and its waste is conducive to the standardization of rubber quality control. In addition, it provides guidance for improving rubber production methods, which can reduce the damage to the environment and human body caused by over standard samples. For example, Ulsaker and Teien (1979) used gas chromatography/mass spectrometry (GC/MS) to detect the N-ethylaniline in polyethylene-bottle-packed intravenous solutions from the rubber part of the two-component closure, and they found that the concentration of N-ethylaniline ranged from 0.1 ppm to 2 ppm. Although GC-MS has high selectivity and sensitivity, it is necessary to eliminate the matrix and interfering compounds via solid phase extraction (Guo and Lee 2013; Jamali et al. 2013; Miskam et al. 2014; Krawczyk and Stanisz 2016; Rajabi et al. 2016; Niu et al. 2018).
Solid-phase extraction is a useful analytical approach for sample pretreatment prior to gas chromatography analysis. The process is shown schematically in Fig. 1. The first step is to pretreat the column solution to activate it. Then the sample extraction solution containing target object and chaff interferent is loaded to the column. The target object and partial chaff interferent are retained. Then most of the chaff interferent is washed with special solution. The last step is to elute the target object for determination. Amiri et al. (2017) prepared magnetic Wells–Dawson heteropoly acid nanoparticles for magnetic solid-phase extraction of aromatic amines in water samples with detection limits in the range of 0.003 ng × mL−1 to 0.01 ng × mL−1. The enrichment factors (EFs) ranged from 75 to 113. Relative standard deviations (RSDs) were 4.8% to 8.3%, and the relative recovery values for the spiked water samples ranged from 90.7% to 99.8% (Amiri et al. 2017). Nazari et al. (2019) synthesized magnetite graphene oxide composites to extract and pre-concentrate aniline, and the limits of detection were 0.341, 0.110, and 0.167 g mL-1 for aniline, phenol, and naphthalene, respectively. The relative standard deviations ranged from 3.3% to 5.7% in 8 repeated extractions. However, these extractants cannot be reused (Nazari et al. 2019).
Fig. 1. Sketch diagram of solid-phase extraction
Cyclodextrins (CDs), common macrocyclic oligosaccharides with a hydrophobic inner cavity and hydrophilic exterior, can encapsulate small organic molecules into their cavities to form inclusion complexes. Due to their compatibility with various organic compounds and ability to be restored via a simple solvent treatment, CDs have been extensively studied and applied in the absorption field (Liu et al. 2017; Nojavan and Yazdanpanah 2017; Chen et al. 2018; 2019; Jiang et al. 2019; Takeo et al. 2019; Wu et al. 2019; Yazdanpanah and Nojavan 2019). Researchers have demonstrated that modified cyclodextrin has good reusability in the absorption of organic compounds (Zhou et al. 2012; Feng et al. 2016). Feng et al. (2016) prepared Fe3O4/hydroxypropyl-β-cyclodextrin polymer magnetic nanomaterials to absorb malachite green, and the absorption rate of the solid-phase extraction agent was 92% and could be used 5 times. After 5 uses, the removal efficiency was above 85% (Xie et al. 2017). Li et al. (2014) prepared anionic β-cyclodextrin/Fe3O4 microspheres to absorb Cu2+, and they could be reused 8 times. After 8 uses, the removal efficiency of Cu2+ decreased from 95% to 88.21% (Li et al. 2014). Cross-linked β-cyclodextrin polymers typically have low surface areas and poorer absorption performance than conventional extractants. Thus, in this study, 3,4,5,6-tetrafluoro-phthalonitrile (TFPN) was employed to crosslink CDs to create a porous and layered structure. The resulting cross-linked poly-TFPN-β-CD had a mesoporous structure with a high surface area. It effectively adsorbed and extracted 4 common aniline derivatives (aniline, N-ethyl aniline, 1,2-phenylenediamine, and p-phenylenediamine). In addition, the polymer could be regenerated at least 4 times without substantial reduction in absorption performance via a mild washing procedure. These findings demonstrated the promising application of porous cyclodextrin-based polymers for the accurate and repeatable detection of aniline derivatives in rubber.
EXPERIMENTAL
Materials
β-cyclodextrin, dichloromethane, and anhydrous potassium carbonate (K2CO3) were purchased from Sinopharm Group Chemical Reagent Co., Ltd. (Shanghai, China). 3,4,5,6-Tetrafluorophthalonitrile (TFPN) was purchased from Shanghai Macklin Reagent Co. Ltd (Shanghai, China). N,N-dimethylformamide (DMF) was supplied by Aladdin reagent Co. Ltd. (Shanghai, China). Anhydrous tetrafuran (THF) was purchased from Energy Chemical (Shanghai, China). Methanol of chromatographic purity, aniline, N-ethyl aniline, 1,2-phenylenediamine, and p-phenylenediamine (> 99.5%, standard (GC)) were supplied by CATO Research Chemical Inc. (Eugene, OR, USA). The working standard solutions of the four aniline derivatives were freshly prepared by dilution of the stock solutions with methanol to the desired concentrations. The rubber samples were purchased from the local market, positive rubber samples of aniline derivatives were prepared. The rubber samples of less than 3 mm × 3 mm × 5 mm in size were prepared by immersion in a mixture of 50 mL of standard solution with a concentration of 500 mg/L and 10 mL of isopropanol. Then, after oscillation for 24 h at room temperature under shading conditions and volatilization of the solution under shading conditions, 3 different positive rubber samples of aniline derivatives were obtained. Ultrapure water was prepared via the Milli-Q system (Aubagne, France). All other reagents were of analytical grade and used without further purification.
Characterization
Infrared spectra were determined with an AVATAR 360 transform infrared spectrometer (Nicolet, Madison, WI, USA). Morphology was studied with a scanning electron microscope (SU8010, Hitachi High-Technologies Corporation, Ishikawa, Japan). The nitrogen absorption-desorption measurements were performed on an ASAP2020-HD88 analyzer (Micromeritics Co., Ltd., Atlanta, GA, USA). The surface area of poly(β-cyclodextrin) was calculated by the Brunauer-Emmett-Teller (BET) method (Zhang and Lu 2019). The pore size distribution of the polymer was determined by modeling according to the Brett-Jouner-Halenda (BJH) theory. Solid nuclear magnetic resonance spectrum testing was performed on a Bruker AG 500 MHz AVANCEⅢ nuclear magnetic resonance spectrometer (Fällanden, Switzerland). Thermogravimetry (TG) was determined with a TGA 2 Mettler Toledo thermal analyzer (Greifensee, Switzerland). Differential scanning calorimetry (DSC) was determined with a DSC 3 Mettler Toledo thermal analyzer (Greifensee, Switzerland). X-ray photoelectron spectroscopy (XPS) was determined with an ESCALAB 250 (Thermo Fisher Scientific, West Sussex, United Kingdom).
Gas chromatographic analysis was performed on an Agilent 7890A equipped with a flame ionization detector (FID) and an HP-5 phenyl methyl siloxan column (30 m × 320 μm × 0.25 μm) (Agilent Company, Wilmington, DE, USA). The injection mode was split (5:1). N2 (99.999%) served as the carrier gas with a constant flow of 1.0 mL × min-1. Further, the column temperature was increased from 50 °C to 260 °C t 4.0 ℃ × min-1, raised to 280 ℃ at a rate of 40.0 ℃ × min-1, and then maintained for 2 min.
Synthesis of the Poly-TFPN-β-CD
The poly(β-cyclodextrin) was synthesized via one-pot nucleophilic substitution condensation copolymerization with β-CD and 3,4,5,6-tetrafluorophthalonitrile (TFPN) as the monomers (Fig. 2). β-cyclodextrin (0.52 g, 0.45 mmol), TFPN (0.30 g, 1.50 mmol), and K2CO3 were added into a reaction tube. Then, 20 mL of mixed solvent (THF / DMF, 9:1v/v) was added into the tube with dry nitrogen flushing. After 3 min of dry nitrogen flushing, the reaction tube was sealed, and the mixture was stirred at 85 °C for 48 h. The yellow products were precipitated and thoroughly washed with 1N HCl, water (25 mL × 2), THF (25 mL × 2), and CH2Cl2 (40 mL). The final product was frozen at -196 °C in a liquid nitrogen bath for 10 min and then dried in a freezer dryer. The yield of the polymer was 37.6% to 40.2%. The synthetic route is shown in Fig. 2.
Fig. 2. The synthetic route of P-TFPN-β-CD
Solid-Phase Extraction (SPE) and GC-FID Detection
The prepared positive rubber sample was extracted with 10 mL of ethanol via an ultrasonic method before testing. Next, 0.45 μm nyron-66 filter membrane was placed on the bottom of the 0.5 mL syringe, and 10 mg of poly(β-cyclodextrin) was added onto it. Then, another 0.5μm nyron-66 filter membrane was placed above the adsorbent poly(β-cyclodextrin). One mL of the analytes was then passed through this thin layer, and the adsorbent layer was washed with 5 mL of methanol. The resulting absorption solution was collected and evaporated to 1 mL, and the analytes were filtered using a 0.22 μm filter membrane. Then, the sample was injected into the GC-FID system for further analysis. Finally, the residual adsorbent was washed with an additional 10 mL of methanol to prepare for the next extraction process.
RESULTS AND DISCUSSION
Structural Analysis of Poly-TFPN-β-CD and β-CD
KBr discs samples were prepared. In the FT-IR spectra of the polymer and the crude materials (Fig. 3), the broad OH stretching band around 3450 cm-1 reflected the stretching vibration of multiple OH functional groups of β-CD. The peak at 2894 cm−1 was attributed to the C-H vibration of CH2, and the peak at 2252 cm−1 was the characteristic signal of the cyano group in 2, 3, 5, 6-tetrafluoroterephthalonitrile (TFPN). The characteristic signals of above functional groups were all observed in the spectrum of the polymer (P-CD), which confirmed the successful synthesis of β-CD polymer. Figures 3b and 3c show the solid-state 13C NMR and 19 F NMR of the β-CD polymer, respectively. In the 13C NMR spectra (Fig. 3b), the peaks from 140 ppm to 153 ppm were attributed to the signals of the carbon atom of C-F (a) and C-O (b) of the aromatic ring. The peaks at 109.9 ppm and 97.7 ppm were attributed to the signal of the carbon atom of CN (c) and aromatic ring (C-CN, d, Ar) (Fig. 3b). The signal of the carbon atom in the sugar ring appeared at 107.3 ppm and 72.8 ppm. In the 19 F NMR spectra, the peaks at -122.9 and -137.3 ppm belonged to the signal of the F atom of β-CD polymer (Fig. 3c).
Fig. 3. Structural analysis of poly-TFPN-β-CD and β-CD before and after crosslinking; (a) FT-IR spectra of β-CD, poly-TFPN-β-CD, and 3,4,5,6-tetra fluorophthalonitrile (TFPN); (b) The solid 13CNMR spectra of poly-TFPN-β-CD; (c) The 19F NMR spectra of poly-TFPN-β-CD
Thermal Properties of Poly-TFPN-β-CD
Figure 4 shows the TG and DSC curves of β-CD and poly-TFPN-β-CD. There were two decomposition stages of cyclodextrin (Fig. 4a).
Fig. 4. Thermo analysis of β-CD and poly-TFPN-β-CD; (a) TG curves of β-CD and poly-TFPN-β-CD; (b) DSC curves of β-CD and poly-TFPN-β-CD
The first stage may have been the removal of crystal water, and the second stage was the degradation of β-CD at 324 °C. For poly-TFPN-β-CD, there were also two decomposition sections. Similarly, the first stage was the removal of crystal water, although the proportion of decomposition was relatively small for the hydrophobicity increment of β-CD after modification. The decomposition temperature of poly-TFPN-β-CD was 285 °C. Moreover, because of the introduction of the aromatic compound TFPN, the residual carbon rate of cyclodextrin was greatly increased to 56.0%. The DSC curve is shown in Fig. 4b. An obvious exothermic peak was observed at 111.5 °C, which corresponded to the crystallization temperature of β-CD. After crosslinking with TFPN, the crystallization temperature of poly-TFPN-β-CD increased to 126.6 °C due to the formation of long chain polymers.
Porous Structure of Poly-TFPN-β-CD
The porous structure and specific surface area are important for absorption. Therefore, SEM and BET were carried out to evaluate the porous structure and specific surface area of poly-TFPN-β-CD. Figure 5a shows that the original β-CD had a minimally porous structure. After modification, the multi-layer polymer endowed the product with a more abundant pore structure that could effectively increase the specific surface area. The porous structures of β-CD before and after modification were further verified by BET analysis. The specific surface area of the modified β-CD was 55.2 m2 × g-1. In addition, the N2 nitrogen absorption curve of poly-TFPN-β-CD was a type II curve, which is indicative of a mesoporous structure. The treatment greatly improved the specific surface area of cyclodextrin; thus, its absorption performance was expected to improve. The total pore volume calculated by the single-point absorption method was 0.108 cm3 × g-1. The Barrett Joyner Halenda model (BJH) was applied to calculate the pore size distribution (Zhang and Lu 2019). The pore diameter was 7.88 nm, which indicated that a mesoporous structure formed (Miskam et al. 2014).
Fig. 5. Characterization of the porous structures of poly-TFPN-β-CD; (a) An SEM image of β-CD; (b) An SEM image of poly-TFPN-β-CD; (c) The N2 absorption (lower line) and desorption (upper line) isotherms of poly-TFPN-β-CD; (d) The cumulative pore volume of poly-TFPN-β-CD obtained by NLDFT analysis
Absorption Properties of Poly-TFPN-β-CD and the Accuracy of the Method
The absorption efficiency of β-CD and poly-TFPN-β-CD for aniline derivatives were determined, and the results are shown in Fig. 6. Poly-TFPN-β-CD adsorbed over 60% of aniline derivatives in 10 s, whereas the absorption capacity of β-CD was always below 40%, which indicated that the absorption efficiency of poly (β-CD) was three times higher than that of β-CD due to its high specific surface area.
To test the accuracy of the method, various concentrations of the standard solutions of 4 aniline derivatives between 1 mg × L-1 and 100 mg × L-1 were applied to plot the working curve (n = 3). The concentration of the analytes exhibited a linear relationship with peak area. The linear equation and relative parameters are listed in Table 1. The precision of this method was determined by relative standard deviation (RSD) values. The RSD of this method ranged from 1.61% to 7.64%, which illustrated that the developed method had good reproducibility.
The standard chromatographic curve under optimum conditions is shown in Fig. S1. In former research, when Fe3O4/hydroxypropyl-β-cyclodextrin polymer magnetic nanomaterials were used for solid-phase extraction, the LOD of malachite green was 5.6 mg/L (Feng et al. 2016). In this study, the LOD of aniline derivatives was better. Liu et al. (2012) determined the LOD of aniline, o-tolidine, o-toluidine, and 3-aminophenol in environmental water by capillary electrophoresis with on-line concentration, and they found that LOD ranged from 0.29 mg/L to 0.44 mg/L. In this research, the LOD of aniline was greatly increased.
Fig. 6. The absorption efficiency of 10 mg of poly-TFPN-β-CD or 10 mg β-CD with a concentration of 100 mg × L-1 four aniline derivatives
Table 1. Linear Equation and Relative Parameters of Four Aniline Derivatives
To evaluate the applicability of this method in practical matrix, 3 kinds of aniline derivatives in commercial rubber (purchased from a local market in Fuzhou, China) were analyzed.
Because no analyte was detected in the rubber samples tested, standard solutions were added to the commercial rubber samples to identify the recovery rates when using this method. The results are summarized in Tables 2, 3, and 4. Relative recovery ranged from 86.8% to 110.5%, and the RSD values of 1.61% to 9.91% were indicative of the high accuracy of this method.
The recovery and RSD results in Tables 2, 3, and 4 showed that this method had good precision in real sample testing. The results showed that the poly-TFPN-β-CD-based solid-phase extraction material was suitable for use in practical testing in various fields.
Table 2. Results of the Determination of Four Aniline Derivatives in Rubber Nipple (n = 6)
Table 3. Results of the Determination of Four Aniline Derivatives in Baby Car Tire (Recycled Rubber) (n = 6)
Table 4. Results of the Determination of Four Aniline Derivatives in Rubber Sole (n = 6)