Coupling Solid-Phase Microextraction with Ambient Mass

Oct 16, 2014 - Guangdong Provincial Public Laboratory of Analysis and Testing Technology, China National Analytical Center Guangzhou, 100 Xianlie Midd...
0 downloads 35 Views 4MB Size
Download PDF
Article pubs.acs.org/ac

Coupling Solid-Phase Microextraction with Ambient Mass Spectrometry Using Surface Coated Wooden-Tip Probe for Rapid Analysis of Ultra Trace Perfluorinated Compounds in Complex Samples Jiewei Deng,† Yunyun Yang,‡ Ling Fang,§ Li Lin,† Haiyun Zhou,§ and Tiangang Luan*,† †

MOE Key Laboratory of Aquatic Product Safety, School of Life Sciences, Sun Yat-Sen University, 135 Xingangxi Road, Guangzhou 510275, China ‡ Guangdong Provincial Public Laboratory of Analysis and Testing Technology, China National Analytical Center Guangzhou, 100 Xianlie Middle Road, Guangzhou 510070, China § Instrumental Analysis & Research Center, Sun Yat-Sen University, 135 Xingangxi Road, Guangzhou 510275, China S Supporting Information *

ABSTRACT: Coupling solid-phase microextraction (SPME) with ambient mass spectrometry using surface coated woodentip probe was achieved for the first time and applied in the analysis of ultra trace perfluorinated compounds (PFCs) in complex environmental and biological samples. We modified n-octadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium chloride on the surface of sharp wooden tip via silanization to form a novel SPME probe, which was then used for highly selective enrichment of PFCs from complex matrices and applied as a solid substrate to induce electrospray ionization for mass spectrometric analysis. The porous structural surface together with the dual extraction mechanisms (reversed phase adsorption and ion exchange adsorption) demonstrated that the SPME probe has an outstanding enrichment capacity, enhancing sensitivity by approximately 4000−8000 folds for the detection in aqueous samples, and 100−500-fold in whole blood and milk samples. The method showed good linearity, with correlation coefficient values (r2) of no less than 0.9931 for eight target PFCs. The limits of detection and qualification of the eight PFCs were 0.06−0.59 and 0.21−1.98 ng/L, respectively. Quantification of real samples was achieved by isotope internal standard calibration curve method or isotope dilution method, and ultratrace levels of PFCs present in lake water, river water, whole blood, and milk samples had been successfully detected and qualified.

A

for rapid, direct, in situ, and in vivo analysis of PFCs are highly desirable. Ambient mass spectrometry (AMS)5−7 is an excellent choice for rapid and direct analysis of complicated samples because this technique operates under ambient and open-air condition and requires minimal or no sample pretreatment and no chromatographic separation. AMS developed rapidly since the introduction of desorption electrospray ionization (DESI)8−10 by Cooks and his co-workers of Purdue University in 2004.8 DESI is derived from electrospray ionization (ESI),11−13 a mass spectrometric ionization technique that is widely used today. Besides DESI, a series of ambient ionization techniques based on ESI mechanism, that is, direct electrospray probe (DEP),14 electrospray laser desorption/ionization (ELDI),15,16 extractive electrospray ionization (EESI),17−19 laser ablation electrospray ionization (LAESI), 20,21 probe electrospray ionization

ccurate and sensitive detection of ultra trace perfluorinated compounds (PFCs) in environmental and biological media has been recognized as one of the hot research topics in recent years, because it gives crucial information to understand the environmental distribution and ecotoxicological effect of these persistent organic pollutants. Liquid chromatography coupled with mass spectrometry (LC-MS) is the conventional method for PFC analysis.1,2 However, this method is laborintensive and time-consuming because it requires large-volume samples as well as multistep pretreatments for enrichment of the ultratrace level of target analytes and elimination of matrix interference.1−3 In addition, the extraction of polar PFCs from polar and complex matrices is rather difficult. The complicated pretreatment work may lead to great variations and uncontrolled errors. Moreover, PFCs are ubiquitous in fluorine-containing materials, such as tubes and seals, which may result in contamination. The adsorption of PFCs to glass vials may also occur unless the sample was analyzed immediately.4 All these issues make the quality control procedure of PFCs analysis a challenging task. Thus, techniques © 2014 American Chemical Society

Received: June 22, 2014 Accepted: October 16, 2014 Published: October 16, 2014 11159

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166

Analytical Chemistry

Article

Figure 1. Schematic diagrams of preparation of surface coated wooden-tip SPME probe and analysis procedure.

(PESI),22−24 paper spray,25,26 wooden-tip ESI,27−29 extraction spray,30,31 and touch spray,32 etc., have been developed for improving the sampling and ionization efficiency of conventional ESI. Meanwhile, ambient ionization techniques based on atmospheric pressure chemical ionization (APCI) mechanism, that is, direct analysis in real time (DART),33,34 desorption atmospheric pressure chemical ionization (DAPCI), 35,36 dielectric discharge barrier ionization (DBDI),37 low-temperature plasma (LTP),38,39 and desorption corona beam ionization (DCBI),40 etc., were also developed and implemented into analytical practices. All of these ambient ionization methods are rapid and straightforward for analyzing complex samples, but their sensitivities are often similar to or lower than those of conventional LC-MS methods. In addition, the introduction of real samples containing complex matrices into mass spectrometer may result in high matrix effect and low sensitivity for target analytes. Direct analysis of ultratrace level of compounds in complex environmental and biological samples remains a great challenge for AMS methods. Coupling solid-phase microextraction (SPME)41−44 to AMS is an effective strategy to improve the sensitivity and suppress the matrix effects for direct analysis of complicated samples.45 In the past several years, SPME has been successfully hyphenated with several ambient ionization techniques including DESI,46−48 DART,49−51 DCBI,40 and LTP.52 These hyphenated methods apply charged droplets/excited-state species to desorb the analytes enriched on a SPME fiber and generate secondary gaseous ions of analytes for mass spectrometric analysis. Although the matrix effects are greatly reduced, these hyphenations are complicated in equipment, and an accurate position of SPME fiber at ion source is extremely critical. Otherwise, it may cause low desorption/ionization efficiency, loss in sensitivity, and even error in the measurement. By inserting a SPME fiber into the space inside of a copper coil, Kuo and Shiea14 achieved the successful hyphenation of SPME with DEP. The sensitivity of SPME-DEP was greatly improved, because analytes were efficiently desorbed by direct application of a small amount of solvent and electrospray was induced under a high voltage. However, the copper coil makes the operation of SPME-DEP very complicated, and there

is no further report about the application of this hyphenated technique. Directly using a SPME fiber to induce electrospray in a similar way as “paper spray/wooden-tip ESI” seems simple for operation; however, the surface of commercial SPME fiber is typically hydrophobic, and without the support of copper coil, the spray solvent cannot be effectively adhered onto the fiber surface, resulting unstable and short signals. Recently, Zhang and his co-workers24 applied a fine-tungsten needle as a SPME probe for sampling single cells and enrichment of metabolites for PESI-MS analysis. However, without sorbent coated on the tungsten needle surface, their method only enhanced the signal intensity by about 30 folds when compared with conventional nanoESI. To overcome the aforementioned shortcomings of existing SPME-AMS methods, here we report a novel SPME and AMS hyphenated strategy and its application for directly analyzing ultra trace PFCs in complex environmental and biological samples. This hyphenated strategy is performed by developing a surface coated wooden tip as a SPME probe, which is functionalized with an efficient sorbent on the probe surface, for highly selective enrichment of PFCs from complex matrices via both reversed phase and ion exchanged adsorption mechanisms. The SPME probe serves also as a solid substrate to induce ESI for mass spectrometric analysis (Figure 1). Compared with the existing conventional methods, the proposed SPME-AMS method shows distinctive potentials in ultra trace analysis: (i) excellent sensitivity, improved sensitivity of 2−4 orders of magnitude via enrichment and direct desorption/ionization of analytes; (ii) high selectivity, selection of specific sorbents for target compounds; and (iii) speediness, integrating extraction, cleanup, enrichment, and detection into one step, without the need of tedious pretreatment steps and separation, and greatly simplifying the analytical procedures.

■

EXPERIMENTAL SECTION Materials and Reagents. Wooden toothpicks (birch wood) were purchased from a PARKnSHOP supermarket in Hong Kong. n-Octadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium chloride (50% in methanol) was from Fluorochem 11160

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166

Analytical Chemistry

Article

Detection. The loaded surface coated wooden-tip SPME probe was mounted onto a three-dimensional moving stage. The tip of the probe was placed pointing to the MS inlet and adjusted to a position of 10 mm away from the MS inlet. Then, a high voltage of −3.5 kV was applied to the probe. Subsequently, 5 μL of spray solvent (methanol) was added onto the probe surface by a micropipette. The enriched analytes were desorbed into the solvent, and then transported to the pointed end through the microchannels of probe surface, leading to the formation of a Taylor cone, which was then sprayed to generate charged ions for mass spectrometric analysis. The duration of signals is ∼20 s per analysis, and one loaded probe can be repeatedly desorbed several times, but the most abundant signals of analytes were obtained in the first time of desorption (Figure S-1 in the Supporting Information). Mass Spectrometry. The mass spectrometric analysis was performed on an Orbitrap Elite mass spectrometer or a TSQ Quantum Ultra triple quadrupole mass spectrometer (Thermo Fisher Scientific, San Jose, CA). The ionization source was set up using the corresponding ESI configuration. The Orbitrap MS was recorded over a mass-to-charge ratio (m/z) of 300− 650 with negative ion detection mode, and accurate mass measurement was conducted with a resolution of 60,000. The triple quadrupole MS was operated in multireaction monitoring (MRM) mode, with experimental conditions shown in Table S1 (Supporting Information). All of the experimental control and data acquisition were conducted via the Xcalibur 2.2 software (Thermo Fisher Scientific, San Jose, CA). Quantitative Calculation. Signals for the first time of desorption were used for quantitative calculation. For pure water, lake water, and river water samples, the quantitative analysis were performed using the isotope internal standard calibration curve method. Blank pure water solutions spiked with a series of concentrations of PFCs (0.5−100 ng/L) were used for construction of the calibration curves. Internal standard (IS) compounds were added into the spray solvent with a concentration of 10 ng/L. 13C4-PFOS was used as the IS of PFOS, and 13C4-PFOA was the IS of PFHxA, PFHpA, PFOA, PFNA, PFDA, PFUnDA, and PFDoDA. A triple quadrupole MS in negative ion MRM mode was used for detection, and the peak area ratios of fragment ions generated from PFCs to those from IS compounds were used for quantitative calculation. For whole blood and milk samples, the quantitative analysis was performed by Orbitrap MS with a high-resolution (60,000) scan mode to distinguish target analyte signals from the matrix signals. An isotope dilution calibration method was used for quantitative analysis, and 13C4-PFOS and 13C4-PFOA were used as the IS compounds to calculate the concentrations of PFOS and PFOA, respectively. First, a MS response factor (RF) was calculated by analyzing blank samples spiked with a series of varying concentrations of PFCs and ISs. Note that the concentration of PFC and its corresponding IS should be the same for each sample. The RF values were determined for each calibration level with the following equation (eq 1):

Ltd. (Graphite Way, Hadfield SK13 IQH, Germany). Octadecyltrimethoxysilane, bovine serum albumin (BSA), N,N-dimethylformamide (DMF, purified through calcium hydride desiccation and vacuum distillation before use), sodium chloride (NaCl), and calcium hydride were obtained from Sigma-Aldrich (St. Louis, MO, USA). Humic acid was obtained from Adamas-beta (Shanghai, China). HPLC grade methanol was purchased from Burdick & Jackson (Muskegon, MI, USA). Standard materials of perfluorooctanesulfonic acid (PFOS), perfluorohexanoic acid (PFHxA), perfluoroheptanoic acid (PFHpA), perfluorooctanoic acid (PFOA), perfluorononanoic acid (PFNA), perfluorodecanoic acid (PFDA), perfluoroundecanoic acid (PFUnDA), and perfluorododecanoic acid (PFDoDA) were purchased from Accustandard Co. Ltd. (NewHaven, CT, USA). 13C4-PFOS and 13C4-PFOA were purchased from Wellington Laboratories (Guelph, ON, Canada). Preparation of Surface-Coated Wooden-Tip SPME Probe. The wooden toothpicks were cut into a length of ∼2 cm using a cutter, and the tip-end was further sharpened until the tip-end o.d. was in the range of 150−200 μm. Then, the sharp wooden tips were dispersed in 100 mL of anhydrous DMF, followed by adding 5 mL of n-octadecyldimethyl[3(trimethoxysilyl)propyl]ammonium chloride under vigorous stirring. The mixtures were heated with reflux at 120 °C for silanization of 12 h under nitrogen atmosphere. After the reaction, the surfaces of wooden tips were modified with a layer of sorbent, and the SPME probes were obtained after the surface coated wooden tips were washed with methanol and dried. Characterization of Surface Coated Wooden-Tip SPME Probe. X-ray photoelectron spectroscopy (XPS) analysis was carried out using an ESCA LAB 250 XPS instrument (Thermo Fisher Scientific, San Jose, CA), with monochromated Al Kα radiation (hυ = 1486.6 eV), 45° photoelectron takeoff angle, and a 500 μm beam size. Scanning electron microscope (SEM) analysis was performed by a Quanta 400 FEG field emission SEM instrument (FEI, The Netherlands). Atomic force microscope (AFM) analysis was performed using a Dimension Fastscan Bio instrument (Bruker Daltonics, Bremen, Germany). Samples and Extraction. Pure water was purified by a Milli-Q water-purification system (Milford, MA, USA). Lake water samples were collected from the Westlake of Sun Yat-Sen University (Guangzhou, China). River water samples were obtained from Pearl River (Guangzhou, China). All the water samples were collected in 1000 mL precleaned polypropylene containers (methanol-rinsed and air-dried). Whole blood samples from healthy volunteers were donated by the first affiliated hospital of Southern Medical University, and stored at −20 °C prior to use. Fresh milk samples were purchased from the local store in Guangzhou. All samples were extracted by direct immersion mode using the probes. Before extraction, the SPME probes were rinsed with methanol for 30 s. The extraction conditions for different samples were as follows: (i) for aqueous samples (including pure water, lake water, and river water), 1000 mL of untreated sample was extracted for 60 min with stirring at 800 rpm; (ii) for whole blood samples, 1 mL of sample (without any pretreatment) was extracted under gentle agitation by a vortexer for 10 min; and (iii) for milk samples, 200 mL of sample was extracted at pH = 3, with stirring at 800 rpm for 20 min. After extraction, the SPME probes were rinsed quickly with pure water for 10 s, and then dried out for analysis.

RF = (C PFC/C IS)/(IPFC/IIS)

(1)

where CPFC and CIS are the concentrations of PFC and IS added into the blank sample, respectively; IPFC and IIS are the peak intensities of analyte and IS. Then, the known amount of IS solution was spiked into the investigated sample solution, and the concentrations of PFOA 11161

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166

Analytical Chemistry

Article

disposable because it is easy to prepare at a very low price, and thus cross-contamination and memory effects are much reduced. Surface coated wooden-tip SPME probe is suitable for analyzing environmental and biological samples with minimum amounts of 1 mL, especially for extremely complicated samples such as whole blood and milk. Subsequently, a highly efficient sorbent should be chosen. Octodecyl (C18) is a commonly used sorbent for enrichment of PFCs,53 and it bonds to the alkyl chain of PFCs via reversed phase adsorption mechanism. Considering PFCs are a group of compounds containing both a nonpolar alkyl chain and a polar sulfonic/carboxyl group, n-octadecyldimethyl[3-(trimethoxysiyl)propyl]ammonium chloride, a compound containing both nonpolar C18 group and polar positively charged quaternary ammomium moiety, is selected as the adsorbing material. This absorbent exhibits high extraction capacity toward PFCs with dual functions: the nonpolar alkyl chain of PFCs bonds to the hydrophobic C18 group with reversed phase adsorption, while the polar acidic group interacts with the quaternary ammomium ion via ion exchange adsorption (Figure 1). To demonstrate the capacity of the dual extraction mechanisms for enrichment of PFCs, here we compared two adsorbing m aterials, i.e ., n- o c t a d e c y l d i m e t h y l [ 3 (trimethoxysilyl)propyl]ammonium chloride and octadecyltrimethoxysilane, by extracting and analyzing pure water spiked with 0.05 μg/L of each target PFC. As illustrated in Figure 2,

and PFOS in the sample solution were calculated via the following equation Csample = (Ianalyte/IIS) × C IS × RF

(2)

where Ianalyte is the peak intensity of PFOA or PFOS generated from the sample analyzed; IIS is the peak intensity of IS; CIS is the concentration level of IS added into the sample; RF is the response factor. To reduce errors and obtain a more accurate quantitative result, the IS compound should be spiked into sample solution with a concentration similar to that of its corresponding target analyte. Enrichment Factor. The enrichment factor (EF) was defined as the ratio of the analyte mass spectral intensity obtained by SPME-AMS to direct AMS analysis. To investigate the EFs, PFCs were spiked into various sample matrices for SPME-AMS analysis. For direct AMS analysis, higher concentrations of spiked samples were used. IS compounds were added into the spray solvent, and the EFs were calculated by comparing the IPFC/IIS value of SPME-AMS to that of direct AMS analysis, and multiplying the concentration coefficient. More details for the calculation of EFs were provided in the Supporting Information.

■

RESULTS AND DISCUSSION Preparation and Characterization of Surface Coated Wooden-Tip SPME Probe. First, a solid substrate for preparation of the SPME probe was selected. Optical fiber,14 fine metal probe,22−24 paper triangle,25,26 and wooden tip,27−29 which are solid substrates of DEP, PESI, paper spray, and wooden-tip ESI, respectively, are all feasible candidates. Optical fiber is smooth and hydrophobic, which is made of insulation material (glass or plastic), thus it is difficult to induce ESI by applying a high voltage and spray solvent onto the fiber. Attempts have been made by inserting the optical fiber into a copper coil,14 but this extra step makes the operation very complicated. Paper triangle is flattened and soft, and the extraction procedure of SPME using such a solid substrate is not convenient for operation. SPME probe using a fine metal probe as the solid substrate for functionalization of sorbents is based on nanoESI mechanism because the diameter of the fine metal probe is usually at several-micrometer level. This type of SPME probe is low sample consumption, and it is suitable for direct sampling and extraction of target molecules from a small quantity of biological tissues and even single cell for mass spectrometric analysis. Different from fine metal probe, the tipend o.d. of wooden tip is 150−200 μm, which is similar to the diameter of the capillary of commercially available ESI source, and thus the application of surface coated wooden-tip SPME probe for mass spectrometric analysis is based on ESI mechanism. Wooden tip is a desirable solid substrate for sampling and induction of ESI because of its porous and hydrophilic properties. When some spray solvent is applied onto wooden tip, it becomes conductive, and electrospray is readily generated by applying a high voltage.27−29 In addition, wooden tip contains abundant hydroxyl groups on the surface, making it an excellent candidate for functionalization with various sorbents. The hydrophilicity of the wooden tip allows uniform distribution of solvent on its surface to desorb the analytes sufficiently. The narrow-stick shape of wooden tip also avoids rapid diffusion and vaporization of spray solvent, and thus allows a long signal duration time. Moreover, a SPME probe with wooden tip as the solid substrate is usually

Figure 2. Comparison of detection sensitivity of PFCs extracted by surface coated wooden-tip SPME probes using octadecyltrimethoxysilane and n-octadecyldimethyl-[3-(trimethoxysilyl)propyl]ammonium chloride as sorbents. Extraction of a 1000 mL pure water solution spiked with 0.05 μg/L of each PFC for 60 min, and 5 μL of methanol solution with 50 μg/L of each IS was applied as spray solvent for mass spectrometric analysis (n = 3).

when n-octadecyl- dimethyl[3-(trimethoxysilyl)propyl]ammonium chloride was used as the sorbent, the peak area ratios of each PFC to IS have increased by 9−14 folds as compared to those obtained with octadecyltrimethoxysilane that lacked the ion exchange capability of the former sorbent, indicating higher preconcentration ability of dual extraction mechanisms. n-Octadecyldimethyl[3(trimethoxysilyl)propyl]ammonium chloride was bonded to the hydroxyl groups on the wooden tip surface via silanization,53 which was confirmed by XPS analysis. Figure 3a shows that the carbon-, oxygen-, nitrogen-, and 11162

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166

Analytical Chemistry

Article

Figure 3. XPS spectra of (a) the n-octadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium chloride coated wooden-tip SPME probe surface, (b) Si 2p peak, and (c) N 1s peak.

Figure 4. SEM images at magnifications of (a) 200× , (b) 1500× , and (c) 3000× of n-octadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium chloride coated wooden-tip SPME probe.

probe surface was formed due to the expansion of wooden fiber when it was immersed into DMF solution. AFM was applied to further investigate the surface morphology and thickness of the sorbent coating. Because the surface of the wooden tip is rough and porous, it cannot be analyzed directly by AFM. Thus, we utilized planar sample plate (i.e., tungsten plate) instead of the wooden tip as substrate to functionalize with n-octadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium chloride (using the same experimental conditions) for AFM analysis, and the results revealed the morphologic differences between the surface coated tungsten plate and the uncoated tungsten plate. Formation of regular globules was observed from the AFM images of surface coated tungsten plate (Figure 5a), indicating a successful silanization. By contrast, the uncoated tungsten plate was relatively neat, with no evident globules observed on the surface (Figure 5b). The root-mean-square roughness, average roughness, and max roughness of the coated material were 6.70, 5.34, and 38.5 nm,

silicon-related peaks were observed in the XPS spectrum of the developed SPME probe surface. However, in the XPS spectrum of the original wooden tip surface (Figure S-2 in the Supporting Information), only carbon- and oxygen-related peaks were found. The content percentages of Si and N in the coated wooden tip surface are 2.14% and 2.35%, respectively. Deconvolution of the Si 2p peak shows the presence of both Si-CH2 at 102.8 eV and Si−O at 102.2 eV (Figure 3b), and deconvolution of the N 1s peak demonstrates the presence of N−CH2 at 399.6 eV and N−CH3 at 402.2 eV (Figure 3c). All these results reveal the introduction of n-octadecyldimethyl[3(trimethoxysilyl)propyl]ammonium chloride onto the SPME probe surface. Figure 4 shows the SEM images of an n-octadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium chloride coated wooden-tip SPME probe. Abundant holes were observed on the probe surface, although they were not completely uniform. The diameters of these holes were at several-μm level. Such porosity gives a high specific surface area for analytes in conjunction with the sorbents. It also offers microchannels for transportation of solvent to achieve ambient ionization for mass spectrometric analysis. To investigate the source of the porous structure, an original wooden tip without any treatment was used for SEM analysis. The results showed that its surface was winkled and loose, with the scale like fibers covered on the whole surface (Figure S-3 in the Supporting Information). This structure looked quite different from a surface coated woodentip SPME probe. To confirm the formation of the porous structure, a wooden tip was immersed in anhydrous DMF solution heated with reflux at 120 °C for 12 h without adding noctadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium chloride, and then analyzed by SEM. Similar porous structure was also observed (Figure S-4 in the Supporting Information), which indicated that the porous structure of wooden-tip SPME

Figure 5. AFM height images of (a) surface coated tungsten plate and (b) uncoated tungsten plate. 11163

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166

Analytical Chemistry

Article

Table 1. Enrichment Factor of Each PFC Extracted by n-Octadecyldimethyl[3-(trimethoxysilyl)propyl]ammonium Chloride Coated Wooden-Tip SPME Probe analyte PFOS PFHxA PFHpA PFOA PFNA PFDA PFUnDA PFDoDA

pure watera 8420 4570 4960 7354 6865 6380 6465 5781

± ± ± ± ± ± ± ±

338 128 149 256 210 232 241 170

lake watera 8025 4019 4470 6960 6425 6085 6130 5235

± ± ± ± ± ± ± ±

river watera

293 142 132 233 226 167 198 135

8242 4370 4721 7125 6467 6358 6370 5420

± ± ± ± ± ± ± ±

346 153 168 267 210 196 226 178

whole bloodb 245 107 102 105 141 160 150 124

± ± ± ± ± ± ± ±

12 4 3 5 8 11 8 5

milkc 502 352 363 466 439 431 419 494

± ± ± ± ± ± ± ±

28 18 19 22 26 23 21 17

a

1000 mL of sample, extracted 60 min, without pretreatment. b1 mL of sample, extracted 10 min, without pretreatment. c200 mL of sample, extracted 20 min, pH = 3.

Investigation of Anti-interference Ability. Humic acid and protein were selected to investigate the anti-interference ability of the developed surface coated wooden-tip SPME probe. They are two common natural organic matters present widely in environmental and biological matrices and can easily adhere to the SPME probe surface, contaminate the sorbent, and result in the reduction of extraction ability.53 Two pH values, that is, pH = 3 (the optimal pH value for PFCs enrichment) and pH = 7 (close to in situ and in vivo conditions), were investigated, and the obtained recoveries are shown in Table S-2 (Supporting Information). At pH = 3, the recoveries of each target PFC were no less than 65.8% in the presence of 50 mg/L humic acid (a rather high concentration in real samples), while the recoveries of all analytes were no less than 51.2% in the presence of 0.2 g/L BSA (a very high concentration in real samples). These recoveries are considered reasonable for analysis of ultra trace compounds in complex environmental and biological samples. At pH = 7, the recoveries of PFCs were no less than 36.1% with the addition of humic acid at a concentration of 50 mg/L, and no less than 17.2% with the addition of BSA at a concentration of 0.2 g/L. The results indicate that the developed SPME probe is still applicable for in situ and in vivo analysis. Considering the biological medium was salty, tolerance of high salty condition was also examined. The experimental results indicated that the developed surface coated wooden-tip SPME probe had excellent salt-tolerant ability, with recoveries greater than 85% observed under a condition of 10 mM NaCl solution at both pH = 3 and pH = 7. Extraction Capacity. The extraction capacity was investigated by determination of the EFs in different matrices. Note that all samples are investigated on the condition that most closely achieve direct, in situ, and in vivo analysis. For aqueous matrices, 1000 mL (five-fold of the optimal volume) of sample solution was extracted for 60 min (three folds of optimal time) without adjusting the pH value to obtain a condition close to in situ analysis. For whole blood, 1 mL of sample was extracted for 10 min without any pretreatment to study the EFs under similar conditions as those in in vivo analysis. For milk sample, an optimal laboratory analysis condition was used, that is, 200 mL of solution was extracted for 20 min at pH = 3. The obtained EFs are summarized in Table 1. The experimental results revealed that the developed SPME probe possessed excellent extraction capacity for analysis of aqueous samples, with EF values of approximately 4000−8000 for pure water, lake water, and river water samples. As for direct analysis of whole blood, the SPME probe also showed desirable enrichment abilities, possessing EFs of 100−300 even under a

respectively (Figure S-5 in the Supporting Information). In addition, the radii and heights of the globules were measured in the range of 20−50 and 7−10 nm, respectively (Figure S-6 in the Supporting Information). Previous studies have reported that a self-assembled monolayer of the material was formed on the surface of various substrates via silanization. 54−58 Considering that the maximum theoretical length of the siloxane molecules in all-trans conformation is of 2.62 nm,55,56 the thickness of the obtained coating is expected to be at several-nanometer level, and our experimental results are in agreement with reported values in literatures. Optimization of Extraction Conditions. The extraction conditions including solution pH, sample volume, and adsorption time were optimized by analyzing pure water samples spiked with 100 ng/L of each PFC. Different solution pH values (i.e., 3, 5, 7, 9, and 11) were investigated. The obtained results showed that the signal intensities, for most of the PFCs, decreased with increasing solution pH values (Figure S-7a in the Supporting Information). Because PFCs are acidic compounds, lowering the pH solution value would shift the acid/base equilibrium toward the molecular form, which would increase their affinity to the sorbent, thereby enhancing extraction efficiency. It should be noted that pH = 3 was observed to be the optimal pH value for PFCs enrichment, while the proposed SPME probe also showed desirable extraction ability at pH = 7 (which is close to in situ and in vivo conditions). Different sample volumes (i.e., 10, 100, 200, 500, and 1000 mL) were investigated. According to the SPME theory,43,44 the amount of analyte extracted onto sorbent (which is directly proportional to the signal intensity) depends on the distribution coefficient of analyte between sorbent and sample matrix, the initial concentration of analyte, the sorbent volume/ surface area, and the sample volume. Thus, significant upward trends of signal intensities were observed as the sample solution volume increased from 10 to 200 mL. When the sample volume is very large, the amount of analyte extracted is independent of sample volume.43,44 In this study, the extraction was observed to reach equilibrium when the sample volume increased to 200 mL, and the signal intensity plateaued (Figure S-7b in the Supporting Information). Different extraction time (i.e., 2, 5, 10, 20, 40, and 60 min) was also studied because the distribution of analytes between the sample solution and the SPME probe is a dynamic process of equilibrium. It was found that the signal intensities increased with increasing extraction time, and the equilibrium was reached at 20 min for most of the PFCs (Figure S-7c in the Supporting Information). 11164

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166

Analytical Chemistry

Article

Table 2. Validation of Surface Coated Wooden-Tip SPME-AMS Method repeatability (RSD, %)a

a

analyte

linear range (ng/L)

PFOS PFHxA PFHpA PFOA PFNA PFDA PFUnDA PFDoDA

0.5−100 1−100 0.5−100 0.5−100 0.5−100 0.5−100 1−100 2−100

regression equation y y y y y y y y

= = = = = = = =

0.1367x 0.0432x 0.0416x 0.0662x 0.0951x 0.1099x 0.0884x 0.0853x

+ 0.3424 + 0.0522 + 0.0928 + 0.1022 + 0.0258 − 0.1069 − 0.0086 + 0.1100

r2

LOD (ng/L)

LOQ (ng/L)

one probe (n = 12)

probe-to-probe (n = 6)

recovery (%)a

0.9992 0.9934 0.9974 0.9982 0.9942 0.9931 0.9939 0.9944

0.06 0.21 0.09 0.10 0.08 0.10 0.29 0.59

0.21 0.71 0.31 0.33 0.25 0.33 0.95 1.98

4.8 8.9 10.2 2.9 8.2 9.6 13.2 10.6

11.5 16.4 11.6 5.8 12.9 14.3 10.4 11.9

103 107 89 96 94 91 112 104

Analysis of pure water sample spiked with 10 ng/L of PFCs.

nonequilibrium extraction condition. For milk samples, the EF values of 350−500 were obtained at an optimized extraction condition. The extraction ability was greatly reduced when analyzing whole blood and milk samples, because the sorbent was contaminated by the complicated matrices. In spite of that, the proposed method still improved the analytical sensitivity by more than 2 orders of magnitude. Quantitative Ability. The quantitative ability of the proposed SPME-AMS method was evaluated by analyzing pure water samples spiked with a series of concentrations of PFCs in the range of 0.5−100 ng/L. 13C4-PFOS and 13C 4PFOA were used as the isotope IS compounds, adding into spray solvent at a concentration of 10 ng/L. In ambient ionization methods, IS compounds are commonly used to calibrate the variations of analytical procedure and improve the method reproducibility. Satisfactory linearity for each target PFC was obtained, with correlation coefficient values (r2) of no less than 0.9931 (Table 2). The limits of detection (LOD) and quantification (LOQ), which were calculated as the concentrations producing signal-to-noise ratios of 3 and 10, were 0.06−0.59 and 0.21−1.98 ng/L, respectively. For repeatability study, one SPME probe was used for 12 repeated extraction of pure water samples spiked with 10 ng/L PFCs under the same condition, and the obtained relative standard deviation (RSD) was less than 13.2% for each analyte. The results proved the reusability and stability of the surfacecoated wooden-tip SPME probe, although it is usually disposable upon use. Note that every time for reuse, the probe should be washed with 5 mL of methanol for 5 min. Probe-to-probe repeatability was also examined by analyzing six pure water samples spiked with 10 ng/L PFCs using six different SPME probes, with RSDs no higher than 16.4% observed. The method also exhibited good accuracy, with recoveries in the range of 89−112%. All these results demonstrate that the proposed SPME-AMS method is suitable for direct quantitative determination of ultratrace level of PFCs. Analysis of Environmental and Biological Samples. Analysis of lake water and river water samples were performed by using a triple quadrupole MS with isotope IS calibration curve method. Because the aqueous matrix did not obviously influence the extraction capacity of the SPME probe, the regression equations listed in Table 2 were used for quantitative calculation. In addition, recoveries were also examined by spiking PFCs into real samples at concentrations of 1 and 10 ng/L to evaluate the accuracy of detection. In one lake water sample, PFOS, PFHpA, PFOA, and PFDA were detected at the concentrations of 5.82, 1.91, 5.14, and 0.63 ng/L, respectively, with recoveries in the ranged of 90−108%. In one river water

sample, PFOS, PFOA, and PFDA were determined to be 9.21, 4.13, and 0.55 ng/L, respectively, with recoveries ranging from 93 to 110% (Table S-3 in the Supporting Information). The satisfactory results of recovery measurements confirmed that the sorbent material was not impacted by the aqueous matrices, which was important for ultra trace analysis. Isotope dilution calibration method was used for analysis of whole blood and milk samples. PFOS was detected in one whole blood sample at a concentration of 3.45 μg/L. PFOS and PFOA were spiked into the blood sample with concentrations at 1 and 10 μg/L, respectively, and the recoveries were in the range of 76−86%. Similarly, PFOS and PFOA were measured in a milk sample at concentrations of 157.3 and 68.6 ng/L, respectively. Recoveries were also examined by spiking PFOS and PFOA into the samples at 100 and 500 ng/L, respectively, and the obtained values were in the range of 84−89% (Table S4 and Figure S-8 in the Supporting Information). The obtained recoveries in whole blood and milk samples were relatively lower than those obtained in water samples, and the reason might be that the complex matrices contaminate the sorbent, and result in the reduction of recovery.

■

CONCLUSION In summary, SPME coupled with AMS using surface coated wooden tip was proposed and realized for the first time in the present study. With the development of a specially designed surface coated wooden-tip SPME probe, rapid and direct analysis of ultra trace PFCs in complex environmental and biological media was successfully achieved. This hyphenated strategy gives the opportunity for ultra trace analysis without using sophisticated sample pretreatment and chromatographic separation. The method can be expanded to a great extent by designing versatile SPME probes. The sorbent materials can be selected from octyl (C8), C18, cation exchanger, and anion exchanger, etc., according to the characteristics of target compounds. Highly selective sorbents to tailor investigation of specific compounds are extremely recommended. The method shows potential value in a variety of applications such as metabolomics, clinical diagnosis, environmental ecology, etc. Further developments of SPME probe using different sorbents, for environmental monitoring, biosystem study, and even single-cell analysis, are ongoing in our laboratory.

■

ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. 11165

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166

Analytical Chemistry

■

Article

(27) Hu, B.; So, P. K.; Chen, H. W.; Yao, Z. P. Anal. Chem. 2011, 83, 8201−8207. (28) Yang, Y.; Deng, J.; Yao, Z. P. J. Am. Soc. Mass. Spectrom. 2014, 25, 37−47. (29) Yang, Y.; Deng, J. Anal. Chim. Acta 2014, 837, 83−92. (30) Ren, Y.; Liu, J.; Li, L.; McLuckey, M. N.; Ouyang, Z. Anal. Methods 2013, 5, 6686−6692. (31) Li, L.; Chen, T. C.; Ren, Y.; Hendricks, P. I.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2014, 86, 2909−2916. (32) Kerian, K. S.; Jarmusch, A. K.; Cooks, R. G. Analyst 2014, 139, 2714−2720. (33) Cody, R. B.; Laramee, J. A.; Durst, H. D. Anal. Chem. 2005, 77, 2297−2302. (34) Petucci, C.; Diffendal, J.; Kaufman, D.; Mekonnen, B.; Terefenko, G.; Musselman, B. Anal. Chem. 2007, 79, 5064−5070. (35) Takáts, Z.; Cotte-Rodriguez, I.; Talaty, N.; Chen, H.; Cooks, R. G. Chem. Commun. 2005, 41, 1950−1952. (36) Yang, S.; Ding, J.; Zheng, J.; Hu, B.; Li, J.; Chen, H.; Zhou, Z.; Qiao, X. Anal. Chem. 2009, 81, 2426−2436. (37) Na, N.; Zhao, M.; Zhang, S.; Yang, C.; Zhang, X. J. Am. Soc. Mass. Spectrom. 2007, 18, 1859−1862. (38) Harper, J. D.; Charipar, N. A.; Mulligan, C. C.; Zhang, X.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2008, 80, 9097−9104. (39) Liu, Y.; Ma, X.; Lin, Z.; He, M.; Han, G.; Yang, C.; Xing, Z.; Zhang, S.; Zhang, X. Angew. Chem., Int. Ed. 2010, 49, 4435−4437. (40) Li, X.; Wang, H.; Sun, W.; Ding, L. Anal. Chem. 2010, 82, 9188−9193. (41) Arthur, C. L.; Pawliszyn, J. Anal. Chem. 1990, 62, 2145−2148. (42) Zhang, Z.; Yang, M. J.; Pawliszyn, J. Anal. Chem. 1994, 66, 844A−853A. (43) Risticevic, S.; Lord, H.; Górecki, T.; Arthur, C. L.; Pawliszyn, J. Nat. Protoc. 2010, 5, 122−139. (44) Lord, H. L.; Zhang, X.; Musteata, F. M.; Vuckovic, D.; Pawliszyn, J. Nat. Protoc. 2011, 6, 896−924. (45) Deng, J.; Yang, Y.; Wang, X.; Luan, T. Trends Anal. Chem. 2014, 55, 55−67. (46) D’Agostino, P. A.; Chenier, C. L.; Hancock, J. R.; Lepage, C. R. J. Rapid Commun. Mass Spectrom. 2007, 21, 543−549. (47) Huang, G.; Chen, H.; Zhang, X.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2007, 79, 8327−8332. (48) Kennedy, J. H.; Aurand, C.; Shirey, R.; Laughlin, B. C.; Wiseman, J. M. Anal. Chem. 2010, 82, 7502−7508. (49) Mirnaghi, F. S.; Pawliszyn, J. Anal. Chem. 2012, 84, 8301−8309. (50) Rodriguez-Lafuente, A.; Mirnaghi, F. S.; Pawliszyn, J. Anal. Bioanal. Chem. 2013, 405, 9723−9727. (51) Wang, X.; Li, X.; Li, Z.; Zhang, Y.; Bai, Y.; Liu, H. Anal. Chem. 2014, 86, 4739−4747. (52) Almasian, M. R.; Yang, C.; Xing, Z.; Zhang, S.; Zhang, X. Rapid Commun. Mass Spectrom. 2010, 24, 742−748. (53) Zhang, X.; Niu, H.; Pan, Y.; Shi, Y.; Cai, Y. Anal. Chem. 2010, 82, 2363−2371. (54) Schreiber, F. Prog. Surf. Sci. 2000, 65, 151−256. (55) Wasserman, S. R.; Tao, Y. T.; Whitesides, G. M. Langmuir 1989, 5, 1074−1087. (56) Balgar, T.; Bautista, R.; Hartmann, N.; Hasselbrink, E. Surf. Sci. 2003, 532−535, 963−969. (57) Roy, S.; Dixit, C. K.; Woolley, R.; MacCraith, B. D.; O’Kennedy, R.; McDonagh, C. Langmuir 2010, 26, 18125−18134. (58) Liu, Y.; Li, Y.; Li, X. M.; He, T. Langmuir 2013, 29, 15275− 15282.

AUTHOR INFORMATION

Corresponding Author

* Tel.: +86-20-84112958. E-mail: [email protected]. Author Contributions

J.D. and Y.Y. contributed equally. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was financially supported by the National Natural Science Foundation of China (NSFC, No. 21277177, 21307167) and Scientific and Technological Project of Guangdong Province (No. 2011B060100005)

■

REFERENCES

(1) Ju, X.; Jin, Y.; Sasaki, K.; Saito, N. Environ. Sci. Technol. 2008, 42, 3538−3542. (2) Mak, Y. L.; Taniyasu, S.; Leo W. Y, Y.; Lu, G.; Jin, L.; Yang, Y.; Lam, P. K. S.; Kannan, K.; Yamashita, N. Environ. Sci. Technol. 2009, 43, 4824−4829. (3) Zareitalabad, P.; Siemens, J.; Hamer, M.; Amelung, W. Chemosphere 2013, 91, 725−732. (4) Ahrens, L. J. Environ. Monit. 2011, 13, 20−31. (5) Cooks, R. G.; Ouyang, Z.; Takats, Z.; Wiseman, J. M. Science 2006, 311, 1566−1570. (6) Venter, A.; Nefliu, M.; Cooks, R. G. Trends Anal. Chem. 2008, 27, 284−290. (7) Huang, M. Z.; Cheng, S. C.; Cho, Y. T.; Shiea, J. Anal. Chim. Acta 2011, 702, 1−15. (8) Takáts, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Science 2004, 306, 471−473. (9) Chen, H.; Talaty, N. N.; Takats, Z.; Cooks, R. G. Anal. Chem. 2005, 77, 6915−6927. (10) Wiseman, J. M.; Ifa, D. R.; Song, Q.; Cooks, R. G. Angew. Chem., Int. Ed. 2006, 45, 7188−7192. (11) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 6, 64−71. (12) Fenn, J. B. Method and apparatus for electrospray ionization. US Patent 6,297,499 B1, October, 2001. (13) Fenn, J. B. Angew. Chem., Int. Ed. 2003, 42, 3871−3894. (14) Kuo, C. P.; Shiea, J. Anal. Chem. 1999, 71, 4413−4417. (15) Shiea, J.; Huang, M. Z.; Hsu, H. J.; Lee, C. Y.; Yuan, C. H.; Beech, I.; Sunner, J. Rapid Commun. Mass Spectrom. 2005, 19, 3701− 3704. (16) Shiea, J.; Yuan, C. H.; Huang, M.-Z.; Cheng, S.-C.; Ma, Y.-L.; Tseng, W.-L.; Chang, H.-C.; Hung, W.-C. Anal. Chem. 2008, 80, 4845−4852. (17) Chen, H.; Wortmann, A.; Zhang, W.; Zenobi, R. Angew. Chem., Int. Ed. 2007, 46, 580−583. (18) Chen, H.; Yang, S.; Wortmann, A.; Zenobi, R. Angew. Chem., Int. Ed. 2007, 46, 7591−7594. (19) Chen, H.; Yang, S.; Li, M.; Hu, B.; Li, J.; Wang, J. Angew. Chem., Int. Ed. 2010, 49, 3053−3056. (20) Nemes, P.; Vertes, A. Anal. Chem. 2007, 79, 8098−8106. (21) Stolee, J. A.; Shrestha, B.; Mengistu, G.; Vertes, A. Angew. Chem., Int. Ed. 2012, 51, 10386−10389. (22) Hiraoka, K.; Nishidate, K.; Mori, K.; Asakawa, D.; Suzuki, S. Rapid Commun. Mass Spectrom. 2007, 21, 3139−3144. (23) Mandal, M. K.; Yoshimura, K.; Saha, S.; Yu, Z.; Takeda, S.; Hiraoka, K. Anal. Chem. 2014, 86, 987−992. (24) Gong, X.; Zhao, Y.; Cai, S.; Fu, S.; Yang, C.; Zhang, S.; Zhang, X. Anal. Chem. 2014, 86, 3809−3816. (25) Wang, H.; Liu, J.; Cooks, R. G.; Ouyang, Z. Angew. Chem., Int. Ed. 2010, 49, 877−880. (26) Liu, J.; Wang, H.; Manicke, N. E.; Lin, J. M.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2010, 82, 2463−2471. 11166

dx.doi.org/10.1021/ac5034177 | Anal. Chem. 2014, 86, 11159−11166