Application of a Zn(II) based metal–organic framework as an efficient solid-phase extraction sorbent for preconcentration of plasticizer compounds

Abstract

A solid-phase extraction (SPE) sorbent, a Zn(II) based metal–organic framework, was prepared via a simple, solventless, green and a low-cost mechanosynthesis process. This sorbent was applied to the extraction of three plasticizer compounds (Di-n-butyl phthalate (DBP), di-2-(ethylhexyl) phthalate (DEHP) and dioctyladipate (DOA)) prior to analysis by gas chromatography with a flame ionization detector (GC-FID). The applied sorbent had the advantages of large surface area, simple and low cost preparation process and good stability. Analytical performances of the proposed method were investigated under the optimum extraction conditions and compared with other SPE methods reported for the same analytes. Validation experiments showed that the optimized method presents good linearity (R² > 0.994), satisfactory precision (RSD ≤ 6.3%), and suitable pre-concentration factors (92–295). The results indicated that solid phase extraction with the proposed sorbent had advantages of convenience, good sensitivity, high efficiency, and it could also be applied successfully to analyze plasticizers in real water samples.

---

Introduction

Metal–organic frameworks (MOFs) as crystalline porous organic–inorganic hybrid materials are composed of metal cations or metal ion clusters interconnected with organic ligands as linkers.^1–3 MOFs impart unique structural properties such as a large range of pore sizes, high surface areas, specific adsorption affinities, and controllable particle dimensions and morphology, which make them as very inspiring materials for a wide range of potential research fields.^4,5 In recent years, many different applications have been reported for these intriguing materials in the fields of gas storage,^6,7 catalysis,^8,9 sensing,¹⁰ separation¹¹ and drug delivery.¹² In all of these applications, stability of the MOFs in applied conditions is a critical parameter. However, water stability of the most MOFs is questionable, which limits their further applications, especially in sample preparation techniques which deal with aqueous samples. MOFs offer a unique platform for the development of adsorbents for sample preparation procedures.^13–15

Phthalic acid esters (PAEs) known as phthalates, and adipic acid esters (adipates) are used ubiquitously as plasticizers worldwide, particularly in the preparation of flexible polyvinyl chloride (PVC) plastics.^16,17 Plasticizers are not chemically bound to plastic and thus they have become ubiquitous environmental contaminants due to volatilization and leaching from the plastic container to the product or various environmental matrices.¹⁸ PAEs are of concern because of their ability to alter the developmental and reproductive system of animals and wildlife as well as humans.^19,20 Several studies have shown that PAEs can produce effects similar to those of estrogen, causing feminization of male infants and disturbances in genital development and testes maturation.^21–23 Certain phthalates and/or their metabolites are suspected human cancer-causing agents, and estrogenic chemicals or endocrine disruptors.^21–23 According to Section 307 of the US Clean Water Act, di-2-(ethylhexyl) phthalate (DEHP) and di-n-butyl phthalate (DBP) should be considered priority toxic pollutants.²⁴ The World Health Organization (WHO) has established a guideline value of 8 ng mL⁻¹ for DEHP for fresh and drinking water,²⁵ which is similar to the maximum contaminant level (MCL) for DEHP set by the Environmental Protection Agency (EPA) (6 ng mL⁻¹).

In order to prevent uncontrolled effects of these compounds on the human health, and the deleterious effects in the aquatic environment, their control is of great importance and the development of the new methods for their determination in diverse environmental matrices is highly significant. Gas chromatography (GC) and liquid chromatography (LC) coupled with mass spectrometry (MS) are the most widely used techniques for the trace analysis of plasticizers.^26–30 Most of the mentioned methods are sophisticated and expensive, thus limiting their applicability.

In spite of substantial technological advances in the analytical field, direct evaluation of trace different analytes content in the complex matrices with low concentrations (near the instrumentation detection limit) generally has low accuracy. Therefore, it is often essential to establish simple, fast, low-cost, sensitive, and selective analytical methods for extraction and pre-concentration before their analysis using GC or LC analysis or other techniques and often plays a vital role in the overall analytical scheme. In most cases, extraction and pre-concentration steps are required for analysis of environmental samples, because the analysis of these compounds is influenced by the complexity of the environmental matrix.

Among these methods, solid-phase extraction (SPE) has become a more popular and well-established sample preparation method to preconcentrate desired components from sample matrix due to its high concentration factor, simplicity, rapidity, minimal cost, and easy automation.^31,32 The selection of appropriate adsorbents in SPE is of vital importance to improve analytical performance (such as analytical sensitivity, selectivity and precision) of SPE techniques.^33,34 In the dispersive SPE procedure, the adsorbent does not need to be packed into the SPE cartridge; instead, the micro-/nano-sorbents are dispersed in the sample solution. So it is capable of being exposed completely with the entire sample that can improve extraction time.^35–38

In our previous study, applicability of three Zn(II) based metal–organic frameworks (TMU-4, TMU-5 and TMU-6) as sorbent for extraction and removal of some heavy metal ions were demonstrated. Lewis acid–base interaction of metal ions with free electrons of nitrogen atoms on the pillars of the MOFs was found that has important role in adsorption of metal ions on the MOFs.³⁹ These MOFs showed good water stability which becomes them as suitable sorbent in SPE procedures. The objective of this study was to investigate the suitability of one of the MOFs previously studied (TMU-6) as a sorbent for SPE procedure for extraction and preconcentration of trace levels of plasticizers in water samples and followed by determination by gas chromatography with flame ionization detector (GC-FID). The factors affecting the extraction were investigated in details, and the applicability of the method was evaluated by the analysis of trace analytes in real water samples.

Experimental

Materials and physical techniques

All chemicals were of analytical reagent grade. HPLC-grade acetonitrile, methanol, 1-propanol, ethyl acetate were acquired from Merck (Darmstadt, Germany). Di-n-butyl phthalate (DBP), di-(2-ethylhexyl) phthalate (DEHP), dioctyladipate (DOA) (Fig. 1) and DMF (dimethylformamide) were obtained from Sigma-Aldrich (Milwaukee, WI, USA). Zinc(II) acetate dihydrate, and 4,4′-oxybis(benzoic acid) (H₂oba) were purchased from Aldrich and Merck, respectively and used as received. The mixed stock solution containing the analytes was prepared monthly by dissolving appropriate amount of plasticizers in methanol and stored in glass bottles at 4 °C. Working standard solutions were prepared by appropriate dilution of the corresponding stock solution with methanol. Ultra-pure quality water was used throughout which was produced by a model Aqua Max-Ultra Youngling ultra-pure water purification system (Dongan-gu, South Korea). Mineral water samples were purchased from a local supermarket.

Fig. 1 Chemical structure of three plasticizers: di-n-butyl phthalate (DBP), di-(2-ethylhexyl) phthalate (DEHP), and dioctyladipate (DOA).

X-ray powder diffraction (XRD) measurements were performed using a Philips X'pert diffractometer with mono chromated Cu-Kα radiation. The sample was characterized with a field emission scanning electron microscope (FE-SEM) ZEISS SIGMA VP (Germany) with gold coating.

An Agilent 7890A (Agilent, Palo Alto, CA, USA) GC instrument equipped with a split-splitless injector and a flame ionization detector (FID) was used for separation and determination of the analytes. Chromatographic separation of target analytes was performed by HP-5 Agilent fused-silica capillary column (30 m × 0.32 mm i.d. × 0.25 μm film thickness). Helium (99.999%) was used as the carrier gas with a flow rate of 4.0 mL min⁻¹. Detector and injector temperatures were 300 and 270 °C, respectively. The GC oven temperature program was: 150 °C for 1 min, increased to 215 °C at 15 °C min⁻¹, then increased to 260 °C at the rate of 8 °C min⁻¹; after that increased to 280 °C at 20 °C min⁻¹ and then held at 280 °C for 3 min.

Preparation of metal–organic framework

The ligand N¹,N⁴-bis((pyridin-4-yl)methylene)-benzene-1,4-diamine (4-bpmb) was synthesized according to previously reported methods.⁴⁰ In a typical synthesis, TMU-6 [Zn(oba)(4-bpmb)_0.5]_n·(DMF)_z, H₂oba = 4,4′-oxybisbenzoic acid, and 4-bpmb = N¹,N⁴-bis((pyridin-4-yl)methylene)-benzene-1,4-diamine was prepared by the mechanochemical procedure. TMU-6 was synthesized by grinding Zn(OAc)₂·2H₂O (1 mmol), H₂oba (1 mmol) and 4-bpmb (0.5 mmol) by hand for 15 minutes. The resulting powder was washed with small amounts of DMF in order to remove any unreacted starting material. The product was heated at 100 °C for 24 h before usage.³⁹

Solid-phase extraction procedure

Seven milligrams of MOF was activated with 20 μL of methanol, and then added into 75 mL of aqueous sample solution in a glass beaker containing 15% w/v of NaCl. The mixture was mixed for 10 min by magnetic stirrer. Subsequently, the solution was transferred into a conical tube and centrifuged. After discarding the supernatant, the sorbent was rinsed with 140 μL of ethyl acetate as eluent with gently mixing by vortex for 2 min. Finally, the eluent was separated from the sorbent by centrifuging and then 2 μL of this solution was injected into the GC system for analysis.

Results and discussion

Characterization

Mechanosynthesized TMU-6 was prepared according to the previously reported paper³⁹ (Fig. S1 and S2, ESI†). In the structure of [Zn(oba)(4-bpmb)_0.5]_n·(DMF)_z (TMU-6), there are 2D sheets composed of three consecutive Zn(II) centers from two different units which are connected to each other by dicarboxylate oba ligand. Three dimensional MOF structure with large 1D pore channels running along the [101] direction (aperture size: 9.1 × 8.9 Å, taking into account the van der Waals radii; 34.2% void space per unit cell) has been obtained by connecting these 2D sheets through the linear 4-bpmb (Fig. 2). The pores of TMU-6 walls are composed of phenyl rings of oba and 4-bpmb ligands, the oxygen atoms of ether group of oba ligand and also decorated with nitrogen atoms of imine groups which make TMU-6 favorable for adsorption of analytes by π–π stacking interactions, hydrophobic interactions and also hydrogen bonds (Fig. 2b and c and S3, ESI†). TMU-6 is porous to CO₂ at 195 K and 1 bar with BET surface area of 456 m² g⁻¹.³⁹

Fig. 2 (a) Representation of the pores of TMU-6 along the [101] direction, (b) highlighting the imine groups (in blue) and (c) highlighting the ether groups (in red). Hydrogen atoms and DMF molecules are omitted for clarity.

Optimization of SPE parameters

Taking into consideration the sensitivity of the determination of target analytes, as well as the selectivity and the precision of the method, the effect of the most important parameters such as extraction time, quantity of the sorbent, sample volume, ionic strength and desorption conditions on the analytical signal was examined in details and optimized. First, effects of type of eluent, extraction time and ionic strength were investigated via one-at-a-time method. Subsequently, three remaining parameters including the quantity of the adsorbent, sample volume, and eluent volume were evaluated by experimental design.

One of the important factors which affects the preconcentration procedure efficiency is the type of the eluent used for the stripping the retained analytes on the sorbent. For this purpose, different organic eluents were tested: methanol, acetonitrile, acetone, 1-propanol and ethyl acetate. The obtained results (Fig. 3) showed that the best desorption efficiency can be achieved while using ethyl acetate as desorption solvent. So, it was used as desorption solvent in the following experiments. It is also notable that TMU-6 is stable to all above mentioned solvents (Fig. 4a).⁴¹

Fig. 3 Effect of eluent type on extraction efficiency. Extraction conditions: sample solution, 30 mL of 100 μg L⁻¹ of target analytes; MOF, 4 mg; eluent, 100 μL; extraction time, 5 min; desorption time, 1 min.

Fig. 4 Comparison of PXRD patterns for testing stability of TMU-6 (a) in ethyl acetate, acetone and 1-propanol and (b) before and after 5 successive runs using ethyl acetate as eluent.

In order to achieve good extraction efficiency towards the target compounds, the effect of ionic strength in the solution was investigated in the range from 0 to 20% (w/v) NaCl. When the amount of salt increased, the recovery of all analytes increased sharply in the initial stage up to 15% (w/v) NaCl, and decreased after that for DOA and DEHP, while recovery of DBP was increased. Based on the above results (Fig. S4, ESI†), the addition of 15% (w/v) of NaCl to samples was employed for the following studies. This phenomenon could be explained in this way that increase of extraction efficiencies may be due to the salting-out effect since addition of salt can decrease the solubility of the target analytes in solution, which will benefit the adsorption of plasticizers to the sorbent. On the other hand, salt addition at higher concentrations can increase viscosity of solution which result in more difficult mass transfer of analytes to the adsorbent surface and it would reduce the interaction of analytes with adsorbent. These effects decrease extraction efficiencies of some analytes at high concentration of salt.

The contact time between adsorbate and adsorbent is one of the most important parameter that affects the performance of adsorption processes. The effect of extraction time on the adsorption of analytes by the proposed sorbent was examined and the results are presented in Fig. S5.† Peak areas of analytes almost remained unchanged with prolonging extraction time over 10 min, which indicates that the extraction equilibrium can be achieved in 10 min. The shortened adsorption time is a significant advantage of nano sorbents because of their high surface area to-volume ratio and short diffusion route. Therefore, an extraction time of 10 min was selected. The experimental results also indicated that 2 min is sufficient to desorb analytes from the sorbent surface and no significant effect was observed when the time of desorption was greater than 2 min.

Additionally, the powder X-ray diffraction (PXRD) patterns of the TMU-6 sorbent before and after five repeated experiments clearly show that the structure remains intact (Fig. 4b).

Multivariate optimization strategies such as factorial designs allow the simultaneous variation of all the factors affecting the experiment and evaluation of interactions among them. The factorial experimental design reduces the time needed for the optimization of an investigated procedure and overall costs.^42,43

In this work, a rotatable and orthogonal central composite design (CCD) was utilized to achieve optimum conditions for the three remaining parameters including adsorbent amount as well as volumes of sample and eluent. Rotatability provides constant variance of the estimated response corresponding to all new observation points that are at the same distance from the center point of the design.^42,43 The experimental design generation and statistical analyses were performed by means of the software package Design-Expert version 8.0.6 trial for Windows (Stat-Ease Inc., MN, USA).

This design consists of two-level full factorial design (2k) augmented with star points (2^k), where k is the number of variables to be optimized, and with the points at the center of the experimental region, which can run n times. In this way, the interactions of parameters and the curvature among experimental variables were studied and, therefore, a real optimum was achieved. A CCD is made orthogonal and rotatable by the choice of a suitable axial point, “α”, for the star design. The value of “α” needed to be ensure orthogonality and rotatability can be calculated from eqn (1).

where N_f = 2^f is the number of factorial points. Using eqn (1) the axial spacing was ±1.682. Then, N₀ was obtained using eqn (2) equal to 9.where N_a (=2f) is the number of axial points.

The total number of experimental runs (N), suggested by CCD, is obtained by the equation N = 2^f + 2f + N₀, wherein f is the number of variables. Therefore, with 3 factors and 9 center points, totally 23 experiments had to be run for the CCD.

Table S1† presents the factor levels used in the CCD, the corresponding design matrix in natural units and responses. The 23 experiments shown in Table S1† were run in a random manner, in order to minimize the effect of uncontrolled variables on the response. Normalized peak area for each run was selected as response objective for the study.⁴⁴

For normalization of the response, peak areas of three analytes in each experimental run in Table S1,† were obtained. Then, the peak area of each analyte in any run was divided by its smallest peak area among all of the experimental runs performed on the basis of experimental design (Table S1†). Summation of normalized peak areas of all three analytes in each run was used as total normalized peak area and used as the response for each run. Second-order polynomial model, consisted of three main effects, three two-factor interaction effects and three curvature effects, showed good agreement with the experimental data provided by the central composite design, as indicated in the following equation:

Response = +31.71 + 9.30A − 12.36B − 8.43C − 1.82AB + 1.75AC + 5.46BC + 0.030A² − 3.35B² + 2.83C² (3)

An analysis of variance (ANOVA) test was used to evaluate the significance of the main effects and their interactions. The strength of the influence of a factor is indicated by the magnitude of the F-value (factors with F-values over 2.71 have a significant influence at the 5% significance level), while the direction of this influence is shown by the sign of the effect. On the other hand, on the basis of p test, with a p-value of less than 0.05, the factor has a statistically significant effect at the 95% confidence level.

According to the ANOVA (Table S2, ESI†), the F-value of the model regression (F_model = 25.6135) was much greater than the tabular F-value with the same number of degrees of freedom of two sources of variance (F_0.05(9, 3) = 2.71), indicating that the treatment differences are highly significant.

The analysis of the results showed that all the three main effects (A, B and C), one cross-product interaction (BC) and two curvature effects (B² and C²) are statistically significant and sample volume (B) is the most significant effect. Whilst, the rest of factors and interactions are not statistically significant.

According to the results, sample volume was the most significant variable having a positive effect. Additionally, the interaction between the sample volume and eluent volume (BC) was significant with a positive effect on the efficiency of the extraction. However, interactions between adsorbent amount and sample volume (AB) and between adsorbent amount and eluent volume (AC) were not-significant. ANOVA results also reveal that two curvature effects of B² and C² appeared as significant effects showing negative and positive effects, respectively. Whilst, the curvature effect of adsorbent amount (A²), revealed a not-significant effect on the extraction efficiency.

In eqn (3), the coefficients for sample volume and eluent volume are negative, which indicates the chromatographic peak areas decrease with increasing these variables. Whilst, the adsorbent amount effect appears with a positive sign in model equation, that indicates its positive effect on extraction efficiency.

The quality of the second-order polynomial model was expressed by the coefficient of determinations (R² and adjusted-R²). The R² and the adjusted-R² values were 0.9466 and 0.9097, respectively, that implies the model obtained adequately correlates the experimental data and it can explain 94.66% of the variability in the response. Moreover, according to the F-value of “lack of fit” (2.6788), the model fitted the data and that lack of fit is not significant relative to the pure error. The large adjusted-R² value indicates a good relationship between the experimental data and the fitted model.

The three-dimensional plots shown in Fig. 5 are useful to interpret the variation of the response as a function of each pair of independent variables graphically. These response surfaces were obtained against two experimental factors while the third is held constant at its optimum level. The plots, shown in Fig. 5, were obtained by plotting adsorbent amount vs. sample volume with the eluent volume fixed at 140 μL (Fig. 5A), adsorbent amount vs. eluent volume, whilst keeping the sample volume of 75 mL (Fig. 5B), and finally, sample volume as a function of eluent volume, for a constant 7.0 mg of adsorbent (Fig. 5C).

Fig. 5 Three-dimensional representation of the response surfaces where (A) adsorbent amount vs. sample volume; (B) adsorbent amount vs. eluent volume; and (C) sample volume vs. eluent volume.

According to the results of the optimization study, the following experimental conditions were chosen: adsorbent amount of 7.0 mg; sample volume of 75 mL and eluent volume of 140 μL.

Method validation

Evaluation of the method performance. The optimized SPE-GC-FID procedure was characterized in terms of linearity (linear range and coefficient of determination), precision (expressed as relative standard deviation), sensitivity (limit of detection) and extraction efficiency (extraction recovery and pre-concentration factor). The results of these assays are reported in Table 1. In this sense, calibration curve was established for each analyte at eight different concentrations by SPE combined with GC-FID under the optimum extraction conditions in the range of 0.5 to 100 μg L⁻¹. The value of limit of detection (LOD) was calculated as the analyte concentration which gives signal to noise whose height is 3 times the baseline noise and was obtained in the range of 0.2–0.7 μg L⁻¹. The pre-concentration factors for all the analytes, which were obtained by comparing the slops of the calibration curves before and after the preconcentration, were in the range from 92 (for DBP) to 295 (for DEHP). The repeatability, expressed as relative standard deviation (RSD) for four replicate analyses, was evaluated by spiking samples at a concentration level of 50 μg L⁻¹ that were calculated to be between 4.8 and 6.3%.

Table 1 Analytical performance characteristics of the proposed method for the plasticizers determination under the optimum conditions

Analyte	Linearity		LOD (μg L^−1)		PF	ER^b	Precision^c (RSD%, n = 4)
	LDR^a (μg L^−1)	R^2
a Linear dynamic range.b Extraction recovery.c Data were calculated based on the extraction of 50 μg L^−1 of each analyte.
DBP	1.5–100	0.998		0.7	92	17	6.3
DOA	0.5–100	0.995		0.2	175	33	4.8
DEHP	0.5–100	0.994		0.2	295	55	5.4

---

Table 2 provides a comparison between the characteristics of the proposed method with other SPE methods that were recently reported in the literature for determination of plasticizers. It is shown that along with its simplicity, this method demonstrated good linearity range, satisfactory sensitivity and an acceptable reproducibility which is comparable with existing methods. However, the LOD of the proposed method is higher than those of other works presented in Table 2, but using a more sensitive analytical instrument such as GC-MS it is possible to achieve better LODs. Meanwhile, the consumption of organic solvents in this method is minimized compared with conventional SPE. Also, the amount of sorbent used in this method with sample volume of 75 mL is comparable with those used in other SPE methods reported. Moreover, the proposed sorbent is successfully synthesized via mechanosynthesis as a convenient, rapid, low cost, solventless and green process showing high stability in experimental conditions.^7,39

Table 2 Comparison of the current method characteristics by metal–organic framework sorbent with those of the other sorbents presented for solid phase extraction of plasticizers

Analyte	Adsorbent/instrument	LOD (μg L^−1)	LDR (μg L^−1)	RSD (%)	Sample volume (mL)	Adsorbent amount (mg)	Eluent volume (mL)	Ref.
DBP	Barium alginate caged Fe3O4@C18 magnetic nanoparticles/HPLC-UV	0.059	0.1–20	<9	500	100	0.50	45
DBP	Cetyltrimethylammonium bromide-coated titanate nanotubes/HPLC-UV	0.019	0.1–20	<10	250	100	2.00	46
DBP-DEHP	Magnetic grapheme/GC-MS	0.010	0.1–200	5.9–6.6	10	20	0.40	47
DBP-DEHP	Multi-walled carbon nanotubes/GC-MS-MS	0.010–0.017	50–10 000	<8.2	200	20	1.00	48
DBP-DEHP	Fe3O4@PPy NPs/GC-MS	0.014–0.018	0.5–100	8.9–11.7	10	30	0.80	49
DBP-DEHP-DOA	Fe3O4@PTh NPs/GC-FID	0.2–0.4	0.4–100	4.0–12.3	45	100	0.10	50
DBP	Magnetic dummy molecularly imprinted/GC-FID	0.6	4–400	3.1	10	100	7.00	51
DBP-DEHP-DOA	Zn-based MOF (TMU-6)/GC-FID	0.2–0.7	0.5–100	4.8–6.3	75	7	0.14	This work

---

Analysis of real samples. In order to evaluate the influence of the sample matrix on the performance of the proposed method, three water samples (two bottled mineral water samples and one boiling water sample exposed to polyethylene vial) were analyzed using the proposed method under optimized conditions. DEHP was detected in the water samples whereas the levels of DBP and DOA were below their LODs. The accuracy of the proposed method was estimated by spiking the samples at some concentration levels.

Satisfactory relative recoveries were obtained in all cases (88–110%), revealing that the matrices of the analyzed water samples had little effect on the performance of the proposed SPE method. According to the obtained results in Table 3, it can be considered that the present method provides acceptable recoveries and precision for the determination of trace amounts of plasticizers in real water samples. Fig. 6 shows the typical chromatograms of the extracted analytes from the mineral water sample using proposed method before and after spiking with 2.0 and 5.0 μg L⁻¹ of each analytes.

Table 3 Results from determination of target analytes by the proposed method in different real water samples

Sample	Analyte	Cinitial (μg L^−1)	Cadded (μg L^−1)	Cfound (μg L^−1)	Relative recovery%	RSD%
Bottled mineral water 1	DBP	<LOD	2.0	2.2	110	8.3
			5.0	5.2	104	7.5
			20.0	19.2	96	7.1
			80.0	78.4	98	6.5
	DOA	<LOD	2.0	1.8	89	5.2
			5.0	4.5	90	5.8
			20.0	21.2	106	5.5
			80.0	76.5	96	4.8
	DEHP	1.4	2.0	3.5	88	6.3
			5.0	5.8	88	5.7
			20.0	19.4	90	6.9
			80.0	77.2	95	5.7
	DBP	<LOD	2.0	2.2	108	7.5
			5.0	5.3	107	5.9
			20.0	18.7	94	7.3
			80.0	83.6	105	6.5
Boiling water exposed to polyethylene vial	DOA	<LOD	2.0	1.8	91	3.8
			5.0	5.1	101	5.5
			20.0	22.1	110	5.3
			80.0	84.5	106	5.9
	DEHP	1.9	2.0	3.7	90	6.1
			5.0	6.3	88	4.2
			20.0	20.3	92	6.8
			80.0	76.5	93	5.8
	DBP	<LOD	2.0	2.0	102	8.4
			5.0	4.8	96	5.9
			20.0	21.2	106	7.8
			80.0	79.1	99	6.2
Bottled mineral water 2	DOA	<LOD	2.0	2.2	110	6.1
			5.0	5.3	106	4.7
			20.0	17.9	90	6.3
			80.0	78.5	98	5.2
	DEHP	0.7	2.0	2.5	92	6.9
			5.0	6.1	109	5.2
			20.0	18.9	91	6.4
			80.0	78.6	97	5.7

---

Fig. 6 GC-FID chromatograms of the plasticizers after extraction using proposed SPE method from bottled mineral water sample; (a) non-spiked (b); spiked with 2.0 μg L⁻¹ of each analytes; and (c) spiked with 5.0 μg L⁻¹ of each analytes; (1) DBP, (2) DOA, and (3) DEHP.

Conclusions

To test the feasibility of the Zn(II) based MOF (TMU-6) as a sorbent in SPE; it was used as an extraction sorbent for pre-concentration of trace amounts of three plasticizer compounds from water samples. The proposed procedure was based on low consumption sorbent and organic solvents, compared with others solid-phase extraction methods. The adsorption of analytes was mainly based on π–π stacking interaction (between the π bonds of MOF pillars and DBP and DEHP) and hydrophobic interaction (between the all three analytes and hydrophobic parts of MOF).

The advantages of the proposed method based on the new sorbent include rather easy, simple, fast, and inexpensive synthesis method; rapid and convenient extraction operation, good sensitivity, and precision and accuracy in pre-concentration and determination of environmental pollutants such as plasticizers. Therefore, it is believed that the proposed procedure can easily be expanded to determination of other pollutants in environmental samples.

Acknowledgements

Support of this investigation by Tarbiat Modares University is gratefully acknowledged.

Notes and references

1. A. K. Cheetham, C. N. R. Rao and R. K. Feller, Chem. Commun., 2006, 4780–4795.
2. C. Janiak, Dalton Trans., 2003, 2781–2804.
3. M. Y. Masoomi, A. Morsali and P. C. Junk, CrystEngComm, 2015, 17, 686–692.
4. J. R. Long and O. M. Yaghi, Chem. Soc. Rev., 2009, 38, 1213–1214.
5. M. O'Keeffe, Chem. Soc. Rev., 2009, 38, 1215–1217.
6. M. Dincă and J. R. Long, Angew. Chem., Int. Ed., 2008, 47, 6766–6779.
7. M. Y. Masoomi, K. C. Stylianou, A. Morsali, P. Retailleau and D. Maspoch, Cryst. Growth Des., 2014, 14, 2092–2096.
8. J. Lee, O. K. Farha, J. Roberts, K. A. Scheidt, S. T. Nguyen and J. T. Hupp, Chem. Soc. Rev., 2009, 38, 1450–1459.
9. D. Farrusseng, S. Aguado and C. Pinel, Angew. Chem., Int. Ed., 2009, 48, 7502–7513.
10. G. J. Halder, C. J. Kepert, B. Moubaraki, K. S. Murray and J. D. Cashion, Science, 2002, 298, 1762–1765.
11. S. Ma, D. Sun, M. Ambrogio, J. A. Fillinger, S. Parkin and H.-C. Zhou, J. Am. Chem. Soc., 2007, 129, 1858–1859.
12. P. Horcajada, C. Serre, M. Vallet-Regí, M. Sebban, F. Taulelle and G. Férey, Angew. Chem., Int. Ed., 2006, 45, 5974–5978.
13. Z.-Y. Gu, C.-X. Yang, N. Chang and X.-P. Yan, Acc. Chem. Res., 2012, 45, 734–745.
14. X. Chen, N. Ding, H. Zang, H. Yeung, R.-S. Zhao, C. Cheng, J. Liu and T. W. D. Chan, J. Chromatogr. A, 2013, 1304, 241–245.
15. N. Chang, Z.-Y. Gu, H.-F. Wang and X.-P. Yan, Anal. Chem., 2011, 83, 7094–7101.
16. U. Heudorf, V. Mersch-Sundermann and J. Angerer, Int. J. Hyg. Environ. Health, 2007, 210, 623–634.
17. X.-y. Hu, B. Wen and X.-q. Shan, J. Environ. Monit., 2003, 5, 649–653.
18. K.-X. Wang and H.-T. Cai, in Practical Industrial Additives, Chemical Industry Press, Beijing, 2001, p. 133.
19. X.-F. Zhang, L.-J. Zhang, L. Li, Y.-N. Feng, B. Chen, J.-M. Ma, E. Huynh, Q.-H. Shi, M. De Felici and W. Shen, Environ. Mol. Mutagen., 2013, 54, 354–361.
20. S. Carbone, O. J. Ponzo, N. Gobetto, Y. A. Samaniego, R. Reynoso, P. Scacchi, J. A. Moguilevsky and R. Cutrera, Horm. Behav., 2013, 63, 692–699.
21. G. Cinelli, P. Avino, I. Notardonato, A. Centola and M. V. Russo, Anal. Chim. Acta, 2013, 769, 72–78.
22. P. Liang, J. Xu and Q. Li, Anal. Chim. Acta, 2008, 609, 53–58.
23. H. Zhang, X. Chen and X. Jiang, Anal. Chim. Acta, 2011, 689, 137–142.
24. US Environmental Protection Agency, Introduction to Water Policy Standards, Office of Water, Washington, DC, 1999.
25. World Health Organization (WHO), Chemical Aspects, Guidelines for Drinking-Water Quality, Draft, 3rd edn, World Health Organization, Geneva, 2003, ch. 8.
26. Z. Guo, S. Wang, D. Wei, M. Wang, H. Zhang, P. Gai and J. Duan, Meat Sci., 2010, 84, 484–490.
27. D. Koniecki, R. Wang, R. P. Moody and J. Zhu, Environ. Res., 2011, 111, 329–336.
28. R. Starink, Results of Proficiency Test Phthalates in PVC, Institute for Inter laboratory Studies, Netherlands, 2005.
29. P. Gimeno, A.-F. Maggio, C. Bousquet, A. Quoirez, C. Civade and P.-A. Bonnet, J. Chromatogr. A, 2012, 1253, 144–153.
30. A. D. LaFleur and K. A. Schug, Anal. Chim. Acta, 2011, 696, 6–26.
31. C. F. Poole, TrAC, Trends Anal. Chem., 2003, 22, 362–373.
32. Y. Wen, L. Chen, J. Li, D. Liu and L. Chen, TrAC, Trends Anal. Chem., 2014, 59, 26–41.
33. J. Fu, L. Chen, J. Li and Z. Zhang, J. Mater. Chem. A, 2015, 3, 13598–13627.
34. J. Ma, R. Xiao, J. Li, J. Yu, Y. Zhang and L. Chen, J. Chromatogr. A, 2010, 1217, 5462–5469.
35. X. Jiang, K. Huang, D. Deng, H. Xia, X. Hou and C. Zheng, TrAC, Trends Anal. Chem., 2012, 39, 38–59.
36. L. Chen, T. Wang and J. Tong, TrAC, Trends Anal. Chem., 2011, 30, 1095–1108.
37. E. Tahmasebi and Y. Yamini, Microchim. Acta, 2013, 181, 543–551.
38. E. Tahmasebi, Y. Yamini, S. Seidi and M. Rezazadeh, J. Chromatogr. A, 2013, 1314, 15–23.
39. E. Tahmasebi, M. Y. Masoomi, Y. Yamini and A. Morsali, Inorg. Chem., 2015, 54, 425–433.
40. D. M. Ciurtin, Y.-B. Dong, M. D. Smith, T. Barclay and H.-C. zur Loye, Inorg. Chem., 2001, 40, 2825–2834.
41. M. Y. Masoomi, S. Beheshti and A. Morsali, J. Mater. Chem. A, 2014, 2, 16863–16866.
42. D. C. Montgomery, Design and Analysis of Experiments, John Wiley, 2004.
43. G. D. Morgan, Chemometrics: Experimental Design (Analytical Chemistry by Open Learning), John Wiley, London, 1991.
44. M. Rezazadeh, Y. Yamini, S. Seidi, E. Tahmasebi and F. Rezaei, J. Agric. Food Chem., 2014, 62, 3134–3142.
45. S. Zhang, H. Niu, Y. Cai and Y. Shi, Anal. Chim. Acta, 2010, 665, 167–175.
46. H. Niu, Y. Cai, Y. Shi, F. Wei, S. Mou and G. Jiang, J. Chromatogr. A, 2007, 1172, 113–120.
47. Q. Ye, L. Liu, Z. Chen and L. Hong, J. Chromatogr. A, 2014, 1329, 24–29.
48. Y. Jiao, S. Fu, L. Ding, Q. Gong, S. Zhu, L. Wang and H. Li, Anal. Methods, 2012, 4, 2729–2734.
49. J. Meng, J. Bu, C. Deng and X. Zhang, J. Chromatogr. A, 2011, 1218, 1585–1591.
50. E. Tahmasebi, Y. Yamini, M. Moradi and A. Esrafili, Anal. Chim. Acta, 2013, 770, 68–74.
51. J. Qiao, M. Wang, H. Yan and G. Yang, J. Agric. Food Chem., 2014, 62, 2782–2789.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra06560k

---