Direct Mass Spectrometry Analysis Using In-Capillary Dicationic Ionic

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Article

Direct Mass Spectrometry Analysis Using In-Capillary Dicationic Ionic Liquid-Based in situ Dispersive LiquidLiquid Microextraction and Sonic-Spray Ionization Yueguang Lv, Hua Bai, Jing-Kui Yang, Yujian He, and Qiang Ma Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b00597 • Publication Date (Web): 19 Apr 2019 Downloaded from http://pubs.acs.org on April 19, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Direct Mass Spectrometry Analysis Using In-Capillary Dicationic Ionic Liquid-Based in situ Dispersive Liquid-Liquid Microextraction and Sonic-Spray Ionization Yueguang Lv†‡, Hua Bai†, Jingkui Yang‡, Yujian He‡, Qiang Ma†* †Chinese ‡School

Academy of Inspection and Quarantine, Beijing 100176, P. R. China

of Chemical Sciences, University of Chinese Academy of Sciences, Beijing 100049, P. R. China

ABSTRACT: The current study reports on a direct mass spectrometry (MS) analysis method using in-capillary dicationic ionic liquid (DIL)-based in situ dispersive liquid-liquid microextraction (DLLME) and sonic-spray ionization (SSI). The developed method merged extraction, enrichment, ionization, and detection of perfluorinated compounds (PFCs) in environmental water into a single step. A microliter-scale ternary fluidic system was designed and integrated into a disposable pulled capillary, in which an imidazolium-based germinal DIL reagent activated an in situ metathesis reaction. A penetrating slug-flow microextraction (SFME) process was subsequently initiated with significantly enhanced interfacial areas and mass transfer rates for the analytes of interest, the mechanism of which was revealed by simulations. An SSI assembly was in-house built, and it enabled a Venturi self-pumping using a stream of nitrogen gas flow coaxial to the capillary under atmospheric pressure to automatically spray at the tip of the capillary. The in situ formed DIL could bind with anionic PFC analytes to generate a positively charged complex, which benefits a signal increase of 1 to 2 orders in magnitude in the positive ion mode than in the negative ion mode for most analytes. The high sensitivity allowed the measure of sub-ppb (parts per billion) levels of PFCs in the environmental water samples. The developed method is a promising protocol for MS analysis because of unprecedented ease, significantly enhanced sensitivity, and potentially high sample throughput.

INTRODUCTION Public concern regarding environmental and human health impact of chemical approaches has encouraged analytical chemists to introduce safer and cleaner alternatives that can minimize consumption of hazardous reagents, and maximize safety for the environment and operators.1 Among the various elements of an analytical methodology, sample preparation is considered a crucial part of the whole analysis procedure, but typically the most resource-intensive component that involves the application of a large volume of organic solvents with toxic properties. As a result, research in analytical chemistry has been increasingly directed towards downsizing traditional extraction techniques to a microextraction scale in order to decrease organic solvent consumption. Microextraction is a miniaturized sample preparation strategy using an insignificant volume of extracting phase. As a microextraction technique, liquid-phase microextraction (LPME) combines sample introduction, isolation, and preconcentration into one stage, and has been used due to its many advantages over liquid-liquid extraction (LLE). These advantages include reduced solvent consumption, high extraction efficiency, low cost, and environmental friendliness. In addition to the miniaturization of sample preparation procedures, the search for environmentally friendly solvents to replace conventional organic solvents is another important way to diminish environmental side effects of analytical methods.2 In recent years, ionic

liquids (ILs) have attracted growing interest as ionic, non-molecular solvents composed entirely of organic cations and various anions. ILs are showing increasing promise as prospective green solvents in a variety of fields.3,4 In particular, their unique physicochemical properties of thermal stability, tunable immiscibility, intrinsic conductivity, and variable solvation interaction make ILs favorable alternative media, replacing organic solvents in extraction, separation, and detection techniques.5,6 The first report on the use of ILs as extraction media in LPME appeared in 2003.7 Since then, LPME with ILs as extraction phase has been incorporated into a number of different derivative methods, such as dispersive liquid-liquid microextraction (DLLME), single-drop microextraction (SDME), hollow-fiber (HF)-LPME, etc. The synergistic combination of ILs and LPME was then subjected to different analytical techniques, such as spectrophotometry,8 spectrofluorimetry,9 capillary electrophoresis (CE) with ultraviolet detection,10 gas chromatography (GC) coupled with flame ionization detection,11 electron capture detection,12 ion mobility spectrometry (IMS)13 or mass spectrometry (MS),14-16 high-performance (HP) or ultra-high-performance (UHP) liquid chromatography (LC) hyphenated with ultraviolet detection,17-19 fluorescence detection20,21 or MS.22-26 As already demonstrated, MS is an analytical technique with high sensitivity and high chemical specificity. In the aforementioned MS-based cases, the applicability of ILs

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

to act as extraction media for target analytes for MS analysis has been explored. Although the ILs extract may be ionized with the intention of characterizing dissolved analyte species, the MS signals of dissolved analytes are significantly suppressed by the overwhelming dominance of IL species, which are ionic and present in vast excess.22,23,27,28 Alternatively, prior to electrospray ionization (ESI)-MS, dicationic ILs (DILs) have been used as post-column ion pairing reagents for HPLC,29 ion chromatography (IC)30 or CE.31 The presence of dicationic pairing agents allows the formation of overall positively charged complexes with anionic analytes. Positive-ion ESI-MS analyses of phosphonates,32,33 phospholipids,34 sphingolipids,35 surfactants,36 pesticides,37 herbicides,38,39 amino acids,40 drug metabolites,41 and a whole range of anions29,30,42,43 can then take place. In this way, the use of the less preferable negative ion mode is avoided, which has shown an increased tendency for poor spray stability and high background noise due to corona discharge and arcing.44 Although the major advantages of using this method involve enhanced spray stability and analytical sensitivity, laborious sample preparation, time-consuming chromatographic separation, and complex post-column mixing instrumentation are still required prior to MS analysis. Ambient ionization has been developed for rapid and direct MS analysis of analytes in untreated samples in their native environment, which opens an exciting perspective on the simplification of analysis procedures by creating ions outside the instrument.45 Sample pretreatment and chromatographic separation, traditionally required for MS-based analysis, can now be bypassed. A diversity of ambient ionization techniques have been developed since desorption electrospray ionization (DESI)46 and direct analysis in real time (DART)47 were reported in 2004 and 2005, respectively.48 DESI is a spray-based ambient ionization technique that extends ESI in such a way as to allow direct desorption and ionization of analytes from a sample surface into the ambient gas-phase environment using charged solvent droplets. In addition to a normal protonation or deprotonation mechanism, derivatization reagents can be added to the spray solvent to facilitate ionization via complexation, such as reactive DESI.49 DILs were also used as reactive agents in DESI for the ambient detection of lipids and fatty acids with enhanced sensitivity.50,51 Among the ambient ionization techniques currently available, sonic-spray ionization (SSI)52-54 is one of the simplest and most easily implemented techniques since it relies solely on a high-velocity stream of nebulizing gas coaxial to spray capillary for ion formation with no assistance of externally applied ionization energies (e.g., voltage, heating, ultraviolet radiation, laser beams, corona or glow discharges), which are normally required for other ambient ionization techniques. Gaseous ions

Page 2 of 14

are produced at atmospheric pressure due to a statistically unbalanced distribution of charges in the tiny droplets by the pneumatic spray.53,55 An SSI assembly can be installed using only common laboratory parts and a cylinder of compressed nitrogen or even air, and has exhibited great simplicity, speed, and ease of use. In this study, direct MS analysis of ultra-trace analytes has been explored by coupling in-capillary DIL-based DLLME and SSI-MS. Perfluorinated compounds (PFCs), a widely used class of chemical substances that is highly concerned for regulatory control due to their persistent, bioaccumulative and toxic properties,56 were selected for method development and validation. The proposed methodology combines the extraction, enrichment, ionization, and detection of the analytes into one step, allowing continuous microextraction processes to occur almost simultaneously with the ambient ionization. An in situ formed DIL reagent was incorporated to act as both extraction medium and charge carrier for efficient extraction and sensitive analysis.

EXPERIMENTAL Chemicals and Reagents Perfluorooctanoic acid (PFOA) was purchased from Tokyo Chemical Industry (Tokyo, Japan). Perfluorooctanesulfonic acid (PFOS) was obtained from Dr. Ehrenstorfer (Augsburg, Germany). Stable isotope-labeled standards of perfluoro-[1,2,3,4-13C4]-octanoic acid (13C4-PFOA) and perfluoro-[1,2,3,4-13C4]-octanesulfonic acid (13C4-PFOS) were offered by Wellington Laboratories (Guelph, ON, Canada). DIL of 1,1-bis(3-methylimadazoilum-1-yl) butylene bromide ([C4(MIM)2]Br2) was synthesized in-house,57 and then further subjected to an anion-exchange process into the fluoride form ([C4(MIM)2]F2).43 Monocationic IL (MIL) of lithium bis[(trifluoromethane)sulphonyl)] imide (Li[NTf2]) was purchased from Sigma-Aldrich (St. Louis, MO, USA). For more details, see Supporting Information. In-Capillary DIL-Based in situ DLLME Four liquid plugs were sequentially injected into a disposable tapered capillary (1.5 mm o.d., 0.86 mm i.d., 7.5 cm length; Figure S1): 2 μL of aqueous [C4(MIM)2]F2 DIL solution at 20 μM (A), 20 μL of environmental water sample spiked with PFCs at 10 ng/mL (B), 2 μL of aqueous Li[NTf2] MIL solution at 30 μM (C), and 10 μL of dichloromethane (D). The transformation of the hydrophilic DIL [C4(MIM)2]F2 to a newly formed hydrophobic DIL ([C4(MIM)2][NTf2]2) was expected to occur inside the capillary via a simple metathesis reaction with the hydrophilic MIL Li[NTf2]. The obtained hydrophobic DIL acted as the extractant to form a cloudy phase in the combined water-miscible plug for an in-capillary DLLME of PFCs. Then with a few sharp jerks applied to the capillary, the dichloromethane fraction


Page 3 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


moved from the open end of the capillary to the tip due to centrifugal force and higher density and mixed well with the enriched hydrophobic DIL phase containing the extracted PFC analytes (see the details in Supporting Information). Homemade SSI Assembly for Direct MS Analysis The capillary was then inserted through the finger-tight fittings of a 1/8-inch stainless steel tee union and a 1/16-inch adapter with appropriate ferrules (Swagelok, Solon, Ohio, USA), yielding a homemade SSI assembly. The assembly was fed a stream of nitrogen from a cylinder, held constant at 5.5×105 Pa. In this way, automatic sample uptake via the Venturi self-pumping effect should take place at the capillary tip. The SSI assembly was mounted in front of an amaZon speed ion trap mass spectrometer (Bruker Daltonics, Karlsruhe, Germany) operated in positive ionization mode. For more details, see Supporting Information.

RESULTS AND DISCUSSION Optimization of In-Capillary DIL-Based in situ DLLME Protocol In-Capillary DIL-Based DLLME with an in situ Metathesis Reaction. A novel in-capillary DIL-based in situ DLLME method was developed using a ternary fluidic system. The environmental water sample plug was sandwiched between two miscible hydrophilic IL plugs, dicationic [C4(MIM)2]F2 and monocationic Li[NTf2]. Within the combined aqueous phase of the three plugs, Li[NTf2] acted as an anion-exchange reagent and underwent an in situ metathesis reaction with the hydrophilic DIL of [C4(MIM)2]F2, forming a new hydrophobic DIL of [C4(MIM)2][NTf2]2 due to the stronger cation-anion interaction. The reaction generated a cloudy, homogeneous solution with fine microdroplets of the resulting hydrophobic DIL, which further acted as an acceptor phase for the DLLME of the hydrophobic PFC analytes. The increased contact surface areas between the microdroplets and the surrounding aqueous phase led to significantly enhanced extraction efficiency for the analytes. After the microextraction process, the fourth plug of dichloromethane moved from the untapered end to the tapered end of the capillary because of centrifugal force and higher density, which enabled the formation of a biphasic system within the capillary due to its immiscibility with the aqueous phase. More importantly, the fourth plug of dichloromethane could well dissolve the hydrophobic, PFCs-containing DIL, thanks to the like-dissolves-like rule, and the resulting dichloromethane phase demonstrated compatible properties for subsequent SSI-MS analysis, because of: (1) Suitable viscosity and surface tension. The viscosity and surface tension of neat ILs are so high that they can hardly be sprayed out of the capillary tip that is of micron-scale internal diameter for direct MS analysis. However, ILs mixed with molecular solvents show much

lower viscosity and surface tension.58 Dichloromethane, as an apolar and electron-donor solvent, served as a diluent to reduce the viscosity and surface tension of the resulting DIL. (2) Constituents of ILs with entirely ionic composition. The ionization mechanism of SSI is that with a sonic gas flow, a droplet undergoes Coulombic fission and its charge concentration fluctuates due to a statistically imbalanced distribution of the pre-existing cations and anions, thus producing excess charge.53,55,59 ILs consist entirely of cations and anions and can be called as liquid electrolytes. The inherent ionic nature of ILs facilitates their compatibility with following SSI process. The in situ metathesis reaction, formation of the new hydrophobic DIL, DLLME of analytes, and subsequent migration of the dichloromethane plug across the capillary to the tip were accomplished almost simultaneously. The multitasking process is rapid and efficient (Figure 1).

Figure 1. Optical microscope image of the in situ emulsion process occurred inside the disposable borosilicate glass capillary. The metathesis reaction occurred after the first three liquid plugs were sequentially injected into the capillary, forming a cloudy phase in the combined water-miscible plug for the DLLME of PFCs. Then the forth dichloromethane plug well dissolved both the hydrophobic DIL acceptor phase and its extracted PFC analytes based on the like-dissolves-like rule. The inset picture is the tyndall effect of the in situ emulsion process, which is simulated in a vial, confirming the colloidal property of the extraction phase. Design on Dication/Anion Combinations of the DIL. The selection of cations and anions is a useful tool for tuning the overall physicochemical properties (e.g. viscosity, water miscibility, and hydrophobicity) of the resulting ILs.60,61 As a novel family of ILs, DILs consist of two anions and one dication, which have exhibited more favorable properties compared to conventional MILs.57 Among them, imidazolium-based DILs, in which the two



imidazolium-based cationic moieties are connected via an alkyl chain, have emerged as a new option for specific applications. The size of the linker region has been suggested to be important to the performance of DILs.33,50 The viscosity of ILs is largely dependent on their cation moiety structure, particularly in the case of imidazolium-based DILs where the alkyl-substituent chain bridges the two imidazolium monocation rings. Increasing the alkyl chain length from butyl to octyl will gradually increase their viscosity.60 The viscosity of imidazolium-based ILs are two or more orders of magnitude greater than that of water and most traditional organic solvents.60 However, viscosity is sometimes considered an obstacle for the applications as extracting media, especially when it comes to the microenvironment within a capillary. As the cloudy droplets formed in the capillary, the rising viscosity and increased aggregation would potentially reduce the extraction efficiency. In this study, various imidazolium-based DILs with different alkyl chain length connecting imidazolium heads were synthesized and evaluated. When longer linker chains of more than four carbons were selected, the viscosity value dramatically increased, thus critically affecting the rate of liquid-liquid mass transfer and leading to an extension in the extraction time.62,63 The moderate carbon linker region of four carbons also gave necessary steric flexibility to allow for better interactions between the dication and the accompanying anion for ion-pair formation. In addition to the identity of cations, water miscibility and hydrophilicity/lipophilicity of ILs can be readily adjusted by a suitable choice of anions.60,61 In this study, the fluoride anion was initially paired with the imidazolium-based cation to enable the resulting DIL of [C4(MIM)2]F2 to remain in an overall hydrophilic state and be completely miscible with the environmental water phase. The fluorine anion may be able to dissociate better from their dication counterpart in solution compared to other halides,64 thus facilitating a subsequent metathesis reaction and maximizing the production of the dication/anion complex. In order to initiate a DLLME process in the capillary, Li[NTf2] was added as an anion-exchange reagent to perform in situ emulsification. It has been found that the strength of the cation/anion interactions increases with increasing molecular size of the anions.58 The fluoride anion could be substituted by the hydrophobic [NTf2]- anion due to its higher affinity for ion-pair formation with the imidazolium-based dication. The resulting DIL of [C4(MIM)2][NTf2]2 is largely immiscible with water, and has higher hydrophobicity and favorable lower viscosity to facilitate mixing than the analogous hexafluorophosphate ([PF6]-) or tetrafluoroborate ([BF4]-) salts,60 thereby providing higher extraction efficiency.

Page 4 of 14

Penetrating Slug-Flow Microextraction. Liquid-liquid slug-flow is especially useful in a capillary microenvironment, where surface tension forces dominate, rather that gravitational body forces. Liquid-liquid slug-flow in a capillary provides two independent transport mechanisms: convection through internal circulation within each plug and diffusion between adjacent plugs. The feature was explored for the development of an in-capillary slug-flow microextraction (SFME) method.65 The shear between the capillary wall and plug axis generates intense internal circulation and transfers the analytes from the sample phase to the extraction phase through the liquid-liquid interface (Figure 2a). Nevertheless, there is a common characteristic for the current SFME applications 65-70 that the relative position between the sample phase and the extraction phase remained unchanged throughout the movement of the two liquid plugs. In these cases, only limited interfacing areas participated in solute diffusion and mass transfer, resulting in less sufficient microextraction. In this work, a dichloromethane plug was injected into a pulled capillary as the extraction phase and originally remained at the open side of the capillary. A traditional in-capillary SFME process was first initiated by gently tilting the capillary up and down. Because of the friction within the capillary wall, internal circulation was formed in both the dichloromethane and aqueous plugs. The in situ formed hydrophobic DIL and its extracted PFC analytes were transferred from the aqueous phase to the dichloromethane phase. After five SFME cycles, a few sharp jerks were applied to the capillary in order to shake down the dichloromethane plug from the open side to the tip end. In this penetrating SFME process, the dichloromethane plug migrated along the capillary walls towards the capillary tip due to centrifugal force and higher density (Figure 2b). Greatly enhanced interfacial areas could be involved in mass transfer between the two immiscible liquid plugs, therefore significantly improving the microextraction efficiency. In addition, hand swing increased the circulation intensity of slug flows, which in turn enhanced the convective mass transfer rates. The extraction efficiencies for exemplary PFC analytes of PFOA and PFOS were calculated to be 91.9% and 94.2%, respectively (n=6).


Page 5 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry the aqueous phase due to the intense turbulent current. This enabled the occurrence of another DLLME process in the aqueous phase, in which the dispersed dichloromethane droplets fully interacted with the hydrophobic DIL and its extracted PFC analytes, making the penetrating SFME highly efficient. The exemplary internal circulation patterns (streamline magnitude and velocity vector) within the aqueous and dichloromethane plugs for t=0.004 s and t=0.1 s are shown in Figure S3. The simulated animation videos for traditional and penetrating SFME processes are attached in Supporting Information.

Figure 2. In-capillary microextraction based on traditional SFME (a) and penetrating SFME (b). The computational fluid dynamics for both traditional and penetrating SFME processes were simulated using an ANSYS FLUENT version 18.1 software (Canonsburg, PA, USA). For further information, see Supporting Information. The contour of the simulated traditional SFME process is shown in Figure S2, in which the circulation patterns (streamline magnitude and velocity vector) within the aqueous and dichloromethane plugs were indicated. The contours of the simulated phase fractionation at different time intervals of the penetrating SFME process are shown in Figure 3. Under the combined influence of gravity and centrifugal force, the dichloromethane plug gradually penetrated down into the interface between the immiscible aqueous plug and the capillary inner wall. The dichloromethane phase formed a thin fluid film along the surface of the inner wall due to its preferential wetting properties for the solid wall material. The dichloromethane film moved at a high initial speed, but then slowed down due to resistance and increased contact areas until a balanced migration velocity was reached. Due to the presence of such an outspread dichloromethane fluid film in penetrating SFME, much more extensive interfacial areas participated in mass transfer. However, in traditional SFME, only the ends of the plugs served as the interface for diffusive penetration. Intensified internal circulation and convective mass transfer could also be induced, which improved the mass-transfer rate by surface renewal at the phase interface. In addition to the tendency of moving downwards the capillary tip, the attraction of dichloromethane molecules to each other generated a cohesion force that the foremost part of the dichloromethane plug tended to accumulate and converge. In the narrow converging region, there were a great number of dichloromethane droplets splashed into

Figure 3. The computational fluid dynamics simulated phase fractions at a series of time intervals. Direct Analysis of PFCs Using SSI-MS SSI-MS Signal Profiles. After the DLLME and penetrating SFME processes, both the dichloromethane and aqueous plugs within the capillary were sequentially sprayed at the capillary tip assisted by the Venturi self-pumping effect, and then subjected to the SSI-MS analysis under positive ion mode. The SSI-MS signal profiles could be divided into two stages, corresponding to the ionization of the dichloromethane phase and aqueous phases. As shown in Figure 4a, the signal intensity of the base peak chronogram (BPC) for stage I appeared to be much higher than in stage II. BPC is similar to total ion chronogram (TIC), however, it monitors only the most intense peak at each point in each spectrum, which gives a cleaner look and is more informative. The extracted ion chronograms (EICs) of the complex adducts of the imidazolium-based dication were m/z 719.1 2+ + ([C4(MIM)2 +C8F17SO3 ] ) with deprotonated PFOS, and m/z 500.1 ([C4(MIM)22++NTf2-]+) with bis[(trifluoromethane)sulphonyl)] anion; the responses were exhibited exclusively in stage I. This indicated that the in situ formed hydrophobic imidazolium-based DIL and its extracted PFC analytes mostly migrated from the aqueous phase to the dichloromethane plug. On the other hand, the EIC at m/z 138.5 connotes that 3-methylimadazoilum-1-yl butylene cation ([C4MIM]+)



was mainly left in the aqueous phase. The combined mass spectra for stage I and II are shown in Figure 4b and 4c, respectively. Optimization of SSI Conditions. An SSI apparatus was assembled in-house, which does not require any customized parts than that are available in an analytical laboratory. The internal diameter of the disposable capillary tip was approximately 5 μm, which could initiate stable spray with the assistance of a high-speed nebulizing gas passing through the tubing and form

Page 6 of 14

positively and/or negatively charged gaseous species. The nitrogen nebulizing gas pressure, one of the SSI operation conditions, was investigated. Figure S4 shows the intensities of target complex of the imidazolium-based dication and deprotonated PFOS ([C4(MIM)22++C8F17SO3-]+) as a function of the nebulizing gas pressure. It was observed that the SSI-MS signal appeared when the gas pressure was greater than 50 psi. The ion

Figure 4. The BPC and EICs of m/z 719.1 ([C4(MIM)22++C8F17SO3-]+), m/z 500.1 ([C4(MIM)22++NTf2-]+), and m/z 138.5 ([C4MIM]+) for stage I and II (a), and the combined mass spectra for stage I (b) and II (c).

intensity continued to increase with rising gas pressure, reaching a maximum at approximately 90 psi. As a result, 90 psi was selected as the optimal gas pressure value in the experiment. The nebulizing gas pressure in the present study was much less than traditional SSI and Venturi easy ambient sonic-spray ionization (V-EASI).52,71 This is most likely due to the use of dichloromethane as the spray solvent, which has a much lower surface tension than that in other previously published studies. The breakup of liquid droplets by a high-speed gas stream has been well studied. Aerodynamic forces are described by the Weber number, and previous experiments showed that the droplets burst in a high-speed gas flow broke up if the Weber number exceeded approximately 10.72 As shown in Figure S5a, a catastrophic fragmentation of the initial droplet occurs, which produces progeny droplets since the initial Weber number in SSI is larger than 1000. There is evidence from charge detection MS that water droplets produced by either SSI or a vibrating orifice aerosol generator reach a common size of about 3 µm after traveling through the inlet.73 This suggests that dichloromethane droplets should undergo a similar phenomenon, and could even be smaller due to the reduced surface tension as compared to water (Figure S5b). Provided that the Weber number is fixed at 10, a similar conclusion was ascertained when gas velocity was the variable of interest (Figure S5c). The gas velocity needed for dichloromethane is much more reduced than others to provide a sufficient amount of small size droplets for multiple rounds of evaporation and Coulombic fission.

During the aerodynamic breakup process, there is a very significant chance that the positive and negative charges will become unevenly separated. In other words, many of these progeny droplets will be charged. Sample pH is a significant parameter which may affect the efficiency of the DIL-based DLLME and the following SSI processes. The partitioning mechanism causing the analytes to transfer between the aqueous phase to the IL phase is similar to that which occurs in a traditional solvent-water system.2 The partition coefficient between water and IL is higher for the uncharged than charged solutes.74 The change of pH values may result in a change in protonation equilibrium, and thus the charging state of PFCs and microextraction efficiency. On the other hand, the change of charge state of PFCs may affect their pairing ability with the dication, and thus SSI-MS signal intensities. As shown in Figure S6, the influence of sample pH was evaluated in the pH range of 2.0 to 6.5 and the optimal performance was achieved at pH 4.5. Therefore, a pH of 4.5 was selected in the present work. Quantitative Analysis of PFCs. In this study, nine representative PFC compounds, including seven perfluorinated carboxylic acids (PFCAs) and two perfluorinated sulfonic acids (PFSAs), were selected as analysis targets. Quantitative MS analysis of PFC analytes was performed in the MRM mode of acquisition. In most cases, MRM could achieve enhanced sensitivity because of the characteristic reduction in noise. The main variables of the width of the isolation window for


Page 7 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


precursor ions, the depth of the pseudo-potential well during the fragmentation, the fragmentation amplitude, and the accumulation time were adjusted for optimum sensitivity. The complex adducts of the imidazolium-based dication with either deprotonated PFCAs or deprotonated PFSAs exhibited the most intensive signals in the positive ion mode and were chosen as the precursor ions for further fragmentation. The MS/MS spectra were acquired by isolating specific precursor ions for each analyte and then exciting to induce dissociation. In order to facilitate the MRM detection, there must be a positively charged fragment of the dication-anionic analyte adduct remaining after dissociation. Exemplary MS/MS spectra of PFOA and PFOS were shown in Figure S7. When the precursor ion of the dication-deprotonated PFOA complex 2+ + ([C4(MIM)2 +C7F15CO2 ] ) at m/z 633.1 was excited, the anionic PFOA moiety dissociated from the adduct, resulting in a singly charged imidazolium-based fragment ion at m/z 219.0 (Figure S7a). While the precursor ion of the dication-deprotonated PFOS complex ([C4(MIM)22++ C8F17SO3-]+) at m/z 719.1 was excited, a methyl imidazole group was lost, generating a fragment ion at m/z 637.0 (Figure S7b). The MS/MS parameters for the nine PFC analytes are shown in Table S1. The limits of detection (LODs) and quantitation (LOQs) for the nine PFC analytes ranged from 0.03 to 0.05 and 0.10 to 0.20 ng/mL, which were calculated as the concentrations producing signal-to-noise ratio (S/N) of 3 and 10 for the characteristic fragment ions, respectively. A series of environmental water samples were prepared with PFOA and PFOS spiked at different concentrations ranging from 0.5 to 100 ng/mL and with 13C -PFOA and 13C -PFOS as the internal standards at the 4 4 concentration of 20 ng/mL. The MS/MS transitions m/z 633.1 to 219.0 and m/z 637.1 to 219.0 were used for PFOA and 13C4-PFOA, respectively (Figure 5a); m/z 719.1 to 637.0 and m/z 723.1 to 641.0 were used for PFOS and 13C -PFOS, respectively (Figure 5b). The calibration curves 4 plotted for PFOA and PFOS are shown in Figure 5c. Good

linearity was obtained with correlation coefficients of 0.9964 and 0.9943 for PFOA and PFOS, respectively. For additional information about other PFCs, see Figure S8 and Table S1 in Supporting Information. The matrix effect was evaluated using a slope comparison method.75 Comparisons were made of the slopes of the matrix-matched calibration curves and the standard solution calibration curves at the same concentration ranges. The calculated slope differences were all less than ±8% (Table S2), which indicates that the proposed approach is effective in removing potential matrix effect. Recoveries at three spiked levels of 5, 10, and 20 ng/mL ranged from 89.9% to 114.2% with relative standard deviations (RSDs) better than 9% (n=6). The established method was applied for the analysis of real environmental water samples collected from Liangshui River and Nanhaizi Park (Beijing, China) at different sites (see the details in Supporting Information). The concentrations of the studied PFCs were less than LODs. Sensitivity Comparison Between Positive and Negative Ion Mode. The sensitivity for the analysis of PFCs obtained in the positive mode using the imidazolium-based germinal dicationic reagent was compared with the negative mode at ambient ionization conditions. As illustrated in Figure S9, seven out of the nine studied PFCs exhibited approximately 1 to 2 orders of higher magnitude signal intensities in the positive ion mode than in the negative ion mode. The in situ formed DIL was introduced to act as an effective extraction medium for the PFC analytes, but also to create a positively charged complex that could bind with negatively charged PFCs with significantly enhanced detection sensitivity. These interactions might originate from the electrostatic force between the delocalized positive charge of the imidazolium-based dication and the negatively charged PFC analytes, and the hydrogen bonding between the paired dication and anion. PFCs are a class of long-chain fluorochemicals in which the negative charge is more concentrated at the terminal acid group. As an electronegative element, fluorine tends to accumulate the

Figure 5. The MS/MS spectra of PFOA and internal standard 13C4-PFOA (a), PFOS and internal standard 13C4-PFOS (b), and calibration curves of PFOA and PFOS in the range of 0.5 to 100 ng/mL (inset plot shows the low concentration range) (c). The



Page 8 of 14

MS/MS transitions m/z 633.1 to 219.0 for PFOA, m/z 637.1 to 219.0 for 13C4-PFOA, m/z 719.1 to 637.0 for PFOS, and m/z 723.1 to 641.0 for 13C4-PFOS.

electrons in a covalent bond, and thus reduces the electrostatic attraction. Consequently, the long alkyl chain may cause steric hindrance between the imidazolium-based dication and PFCs, which affects ion association in the gas phase. For these reasons, as the length of the alkyl chain of PFCs increased, the measured signal intensity for the paired adduct decreased (Figure S9). When the length of alkyl chain was beyond ten carbon atoms (e.g., PFDA and PFDoA), sensitivity for the dication-anion adducts in the positive mode was lower than that with the direct detection of anionic PFC analytes in the negative mode, resulting in negative logarithm values for PFDA and PFDoA. Investigation of more types of DILs on the sensitivity of PFCs will be the subject of our future dedicated study.

CONCLUSION In the present study, a direct MS analysis method has been developed using in-capillary DIL-based in situ DLLME and SSI. The established methodology merges the extraction, enrichment, ionization, and detection of the target analytes into one step, which also includes the ease of being in an open atmosphere and ambient environment. A proof-of-concept was demonstrated by the determination of PFCs in environmental water

samples. The findings in this study is paving the way for developing a new means of direct MS analysis.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website.

AUTHOR INFORMATION Corresponding Author * Email: [email protected]

ORCID Qiang Ma: 0000-0002-7887-2031

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT This work was supported by the National Key R&D Program of China (2016YFF0203704) and Beijing Natural Science Foundation (8192049).

REFERENCES


Page 9 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(1) Poole, C. F.; Lenca, N. TrAC Trends in Analytical Chemistry 2015, 71, 144-156. (2) Martinis, E. M.; Berton, P.; Wuilloud, R. G. TrAC Trends in Analytical Chemistry 2014, 60, 54-70. (3) Zhang, S.; Lu, X.; Zhou, Q.; Li, X.; Zhang, X.; Li, S. In Ionic Liquids; Elsevier: Amsterdam, 2009, pp 3-20. (4) Plechkova, N. V.; Seddon, K. R. Chemical Society Reviews 2008, 37, 123-150. (5) Trujillo-Rodríguez, M. J.; Rocío-Bautista, P.; Pino, V.; Afonso, A. M. TrAC Trends in Analytical Chemistry 2013, 51, 87-106. (6) Han, X.; Armstrong, D. W. Accounts of Chemical Research 2007, 40, 1079-1086. (7) Liu, J.-f.; Jiang, G.-b.; Chi, Y.-g.; Cai, Y.-q.; Zhou, Q.-x.; Hu, J.-T. Analytical Chemistry 2003, 75, 5870-5876. (8) Arvand, M.; Bozorgzadeh, E.; Shariati, S.; Zanjanchi, M. A. Environmental Monitoring and Assessment 2012, 184, 7597-7605. (9) Zeeb, M.; Ganjali, M. R.; Norouzi, P. DARU Journal of Pharmaceutical Sciences 2011, 19, 446-454. (10) Zhou, C.; Tong, S.; Chang, Y.; Jia, Q.; Zhou, W. ELECTROPHORESIS 2012, 33, 1331-1338. (11) Zhao, F.; Lu, S.; Du, W.; Zeng, B. Microchimica Acta 2009, 165, 29-33. (12) Zhang, J.; Lee, H. K. Talanta 2010, 81, 537-542. (13) Aguilera-Herrador, E.; Lucena, R.; Cárdenas, S.; Valcárcel, M. Journal of Chromatography A 2009, 1216, 5580-5587. (14) Akinlua, A.; Jochmann, M. A.; Schmidt, T. C. Chromatographia 2015, 78, 1201-1209. (15) Cacho, J. I.; Campillo, N.; Viñas, P.; Hernández-Córdoba, M. Talanta 2016, 146, 568-574. (16) Vallecillos, L.; Pocurull, E.; Borrull, F. Talanta 2012, 99, 824-832. (17) Jiang, Y.; Tang, T.; Cao, Z.; Shi, G.; Zhou, T. Journal of Separation Science 2015, 38, 2158-2166. (18) Werner, J. Journal of Separation Science 2016, 39, 1411-1417. (19) Xu, H.; Mi, H.-Y.; Guan, M.-M.; Shan, H.-Y.; Fei, Q.; Huan, Y.-F.; Zhang, Z.-Q.; Feng, G.-D. Food Chemistry 2017, 232, 198-202. (20) Pena, M. T.; Casais, M. C.; Mejuto, M. C.; Cela, R. Journal of Chromatography A 2009, 1216, 6356-6364. (21) Xu, X.; Liu, Z.; Zhao, X.; Su, R.; Zhang, Y.; Shi, J.; Zhao, Y.; Wu, L.; Ma, Q.; Zhou, X.; Zhang, H.; Wang, Z. Journal of Separation Science 2013, 36, 585-592. (22) Vázquez, M. M. P.; Vázquez, P. P.; Galera, M. M.; Moreno, A. U. Journal of Chromatography A 2014, 1356, 1-9. (23) Li, G.; Wang, L.; Fei, T.; Liu, H.; Wu, D.; Zheng, S. Chromatographia 2015, 78, 641-648. (24) Zgola-Grzeskowiak, A. Analytical Methods 2015, 7, 1076-1084. (25) Zhao, R.-S.; Wang, X.; Zhang, L.-L.; Wang, S.-S.; Yuan, J.-P. Analytical Methods 2011, 3, 831-836. (26) Zhao, R.-S.; Wang, S.-S.; Cheng, C.-G.; Zhang, L.-L.; Wang, X. Chromatographia 2011, 73, 793-797. (27) Dyson, P. J.; McIndoe, J. S.; Zhao, D. Chemical

Communications 2003, 508-509. (28) Jackson, G. P.; Duckworth, D. C. Chemical Communications 2004, 522-523. (29) Soukup-Hein, R. J.; Remsburg, J. W.; Dasgupta, P. K.; Armstrong, D. W. Analytical Chemistry 2007, 79, 7346-7352. (30) Martinelango, P. K.; Anderson, J. L.; Dasgupta, P. K.; Armstrong, D. W.; Al-Horr, R. S.; Slingsby, R. W. Analytical Chemistry 2005, 77, 4829-4835. (31) Lin, X.; Gerardi, A. R.; Breitbach, Z. S.; Armstrong, D. W.; Colyer, C. L. ELECTROPHORESIS 2009, 30, 3918-3925. (32) Chu, S.; Chen, D.; Letcher, R. J. Journal of Chromatography A 2011, 1218, 8083-8088. (33) Warnke, M. M.; Breitbach, Z. S.; Dodbiba, E.; Crank, J. A.; Payagala, T.; Sharma, P.; Wanigasekara, E.; Zhang, X.; Armstrong, D. W. Analytica Chimica Acta 2009, 633, 232-237. (34) Dodbiba, E.; Xu, C.; Payagala, T.; Wanigasekara, E.; Moon, M. H.; Armstrong, D. W. Analyst 2011, 136, 1586-1593. (35) Xu, C.; Pinto, E. C.; Armstrong, D. W. Analyst 2014, 139, 4169-4175. (36) Santos, I. C.; Guo, H.; Mesquita, R. B. R.; Rangel, A. O. S. S.; Armstrong, D. W.; Schug, K. A. Talanta 2015, 143, 320-327. (37) Xu, C.; Armstrong, D. W. Analytica Chimica Acta 2013, 792, 1-9. (38) Guo, H.; Breitbach, Z. S.; Armstrong, D. W. Analytica Chimica Acta 2016, 912, 74-84. (39) Guo, H.; Riter, L. S.; Wujcik, C. E.; Armstrong, D. W. Talanta 2016, 149, 103-109. (40) Wang, Y.; Du, S.; Armstrong, D. W. Analytical and Bioanalytical Chemistry 2018. (41) Guo, H.; Dolzan, M. D.; Spudeit, D. A.; Xu, C.; Breitbach, Z. S.; Sreenivasan, U.; Armstrong, D. W. International Journal of Mass Spectrometry 2015, 389, 14-25. (42) Remsburg, J. W.; Soukup-Hein, R. J.; Crank, J. A.; Breitbach, Z. S.; Payagala, T.; Armstrong, D. W. Journal of the American Society for Mass Spectrometry 2008, 19, 261-269. (43) Xu, C.; Guo, H.; Breitbach, Z. S.; Armstrong, D. W. Analytical Chemistry 2014, 86, 2665-2672. (44) Cech, N. B.; Enke, C. G. Mass Spectrometry Reviews 2001, 20, 362-387. (45) Cooks, R. G.; Ouyang, Z.; Takats, Z.; Wiseman, J. M. Science 2006, 311, 1566-1570. (46) Takáts, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Science 2004, 306, 471-473. (47) Cody, R. B.; Laramée, J. A.; Durst, H. D. Analytical Chemistry 2005, 77, 2297-2302. (48) Monge, M. E.; Harris, G. A.; Dwivedi, P.; Fernández, F. M. Chemical Reviews 2013, 113, 2269-2308. (49) Alberici, R. M.; Simas, R. C.; de Souza, V.; de Sá, G. F.; Daroda, R. J.; Eberlin, M. N. Analytica Chimica Acta 2010, 659, 15-22. (50) Rao, W.; Mitchell, D.; Licence, P.; Barrett, D. A. Rapid Communications in Mass Spectrometry 2014, 28, 616-624. 9



(51) Lostun, D.; Perez, C. J.; Licence, P.; Barrett, D. A.; Ifa, D. R. Analytical Chemistry 2015, 87, 3286-3293. (52) Hirabayashi, A.; Sakairi, M.; Koizumi, H. Analytical Chemistry 1994, 66, 4557-4559. (53) Hirabayashi, A.; Sakairi, M.; Koizumi, H. Analytical Chemistry 1995, 67, 2878-2882. (54) Hirabayashi, A.; Hirabayashi, Y.; Sakairi, M.; Koizumi, H. Rapid Communications in Mass Spectrometry 1996, 10, 1703-1705. (55) Haddad, R.; Sparrapan, R.; Eberlin, M. N. Rapid Communications in Mass Spectrometry 2006, 20, 2901-2905. (56) (57) Anderson, J. L.; Ding, R.; Ellern, A.; Armstrong, D. W. Journal of the American Chemical Society 2005, 127, 593-604. (58) Ozokwelu, D.; Zhang, S.; Okafor, O. C.; Cheng, W.; Litombe, N. In Novel Catalytic and Separation Processes Based on Ionic Liquids, Ozokwelu, D.; Zhang, S.; Okafor, O. C.; Cheng, W.; Litombe, N., Eds.; Elsevier: Amsterdam, 2017, pp 45-110. (59) Wleklinski, M.; Li, Y.; Bag, S.; Sarkar, D.; Narayanan, R.; Pradeep, T.; Cooks, R. G. Analytical Chemistry 2015, 87, 6786-6793. (60) Huddleston, J. G.; Visser, A. E.; Reichert, W. M.; Willauer, H. D.; Broker, G. A.; Rogers, R. D. Green Chemistry 2001, 3, 156-164. (61) Sheldon, R. Chemical Communications 2001, 2399-2407. (62) Fernández, E.; Vidal, L. In Ionic Liquids in Separation Technology; Elsevier: Amsterdam, 2014, pp 107-152. (63) Abdelhamid, H. N. TrAC Trends in Analytical Chemistry 2016, 77, 122-138.

Page 10 of 14

(64) Breitbach, Z. S.; Wanigasekara, E.; Dodbiba, E.; Schug, K. A.; Armstrong, D. W. Analytical Chemistry 2010, 82, 9066-9073. (65) Ren, Y.; Mcluckey, M. N.; Liu, J.; Ouyang, Z. Angewandte Chemie 2015, 126, 14348-14351. (66) Pu, F.; Zhang, W.; Bateman, K. P.; Liu, Y.; Helmy, R.; Ouyang, Z. Bioanalysis 2017, 9, 1633-1641. (67) Deng, J.; Wang, W.; Yang, Y.; Wang, X.; Chen, B.; Yao, Z. P.; Luan, T. Analytica Chimica Acta 2016, 940, 143-149. (68) Yang, Y.; Wu, J.; Deng, J.; Yuan, K.; Chen, X.; Liu, N.; Wang, X.; Luan, T. Analytica Chimica Acta 2018, 1032, 75-82. (69) Ma, Q.; Bai, H.; Li, W.; Wang, C.; Li, X.; Cooks, R. G.; Ouyang, Z. Analytica Chimica Acta 2016, 912, 65-73. (70) Guo, X.; Bai, H.; Lv, Y.; Xi, G.; Li, J.; Ma, X.; Ren, Y.; Ouyang, Z.; Ma, Q. Talanta 2018, 180, 182-192. (71) Santos, V. G.; Thaís, R.; Dias, F. F. G.; Wanderson, R. O.; Jara, J. L. P.; Klitzke, C. F.; Fernando, C.; Eberlin, M. N. Analytical Chemistry 2011, 83, 1375. (72) Shraiber, A. A.; Podvysotsky, A. M.; Dubrovsky, V. V. Atomization & Sprays 1996, 16, 667-692. (73) Zilch, L. W.; Maze, J. T.; Smith, J. W.; Ewing, G. E.; Jarrold, M. F. Journal of Physical Chemistry A 2008, 112, 13352-13363. (74) Huddleston, J. G.; Willauer, H. D.; Swatloski, R. P.; Visser, A. E.; Rogers, R. D. Chemical Communications 1998, 1765-1766. (75) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Analytical Chemistry 2003, 75, 3019-3030.

TOC

10


Page 11 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1 196x152mm (300 x 300 DPI)



Figure 3 254x190mm (300 x 300 DPI)


Page 12 of 14

Page 13 of 14 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4



Figure 5


Page 14 of 14

Direct Mass Spectrometry Analysis Using In-Capillary Dicationic Ionic

Recommend Documents