Low-Temperature Plasma Ionization Differential Ion Mobility

Article pubs.acs.org/ac

Low-Temperature Plasma Ionization Differential Ion Mobility Spectrometry Andriy Kuklya,*,† Carsten Engelhard,*,‡ Florian Uteschil,† Klaus Kerpen,† Robert Marks,† and Ursula Telgheder†,§ †

Downloaded by UNIV OF MANITOBA on August 26, 2015 | http://pubs.acs.org Publication Date (Web): August 19, 2015 | doi: 10.1021/acs.analchem.5b02077

Department of Instrumental Analytical Chemistry, University of Duisburg-Essen (UDE), Universitätsstraße 5, 45141 Essen, Germany ‡ Department of Chemistry and Biology, University of Siegen, Adolf-Reichwein Straße 2, D-57068 Siegen, Germany § IWW Water Centre, Moritzstraße 26, 45476 Mülheim a.d. Ruhr, Germany ABSTRACT: A low-temperature plasma (LTP) was used as an ionization source for differential ion mobility spectrometry (DMS) for the first time. This ionization source enhances the potential of DMS as a miniaturized system for on-site rapid monitoring. The effects of experimental parameters (e.g., discharge/carrier gas composition and flow rate, applied voltage) on the analysis of model aromatic compounds were investigated and discussed. It was found that the nature of reactant ion positive (RIP) is dependent on the discharge/ carrier gas composition. The best response to the analyte was achieved when pure nitrogen was used as the discharge/carrier gas. The ability to perform analysis with zero helium consumption is especially attractive in view of the potential application of LTP-DMS for online (and on-site) monitoring. Analytical performance was determined with six environmentally relevant model compounds (benzene, toluene, ethylbenzene, pxylene, 1,2,4-trimethylbenzene, and naphthalene) using LTP and directly compared to APPI and APCI (63Ni) ionization sources. When LTP was coupled to DMS, calculated LOD values were found to be in the range of 35−257 ng L−1 (concentration in the carrier gas). These values are competitive with those calculated for two DMS equipped with traditional ionization sources (APPI, 63 Ni). The obtained results are promising enough to ensure the potential of LTP as ionization source for DMS.

B

recommended as a screening method by the U.S. Environmental Protection Agency (EPA).3 Differential ion mobility spectrometry (DMS), also known as planar high-field asymmetric waveform ion mobility spectrometry (FAIMS), is a rapidly advancing technology that is both sensitive and fast, operates at atmospheric pressure, and provides a unique type of selectivity, which is orthogonal to most of other separation techniques.4,5 Unlike the conventional time-of-flight ion mobility spectrometry (TOF-IMS), in which the separation of ions is based on specific coefficients of ion mobility in a uniform electric field, DMS separates ions based on a nonlinear relationship between the mobility coefficient and the electric field strength. The dependence of the ion mobility coefficient on the electric field can be explained by the field dependent reversible cluster formation model.6 The fundamental principles of DMS are described elsewhere.7,8 DMS has found some applications as a stand-alone analyzer (with and without GC preseparation) as well as a fast preseparation technique for atmospheric pressure ionization

ecause of the continuously increasing production and use of fossil fuels, the contamination of aquifers and groundwater became an environmental issue of major concern worldwide.1,2 The potentially rapid spread of water contaminants is a driving force for the continuous improvement of early warning systems and the development of new online/on-site analytical methods for water monitoring. However, most of the existing methods are laboratory-based and, therefore, require time-consuming sampling and transport procedures. In many cases, derivatization or extraction steps are required. Therefore, the interest in miniaturized systems for on-site rapid monitoring has grown in recent years. Ideally, these miniaturized systems should include all steps required for the analysis (e.g., sampling, sample preparation, separation techniques, and analyte detection). Driving forces for the development of miniaturized systems are reduced cost and analysis time as well as the possibility to integrate all steps in a single and portable device. The most commonly used methods for analysis of gasoline contaminated water samples are headspace, purge and trap, liquid−liquid extraction, solidphase extraction (SPE), and solid-phase microextraction (SPME). Because of the good performance in particular situations, including the analysis in field, headspace analysis is © XXXX American Chemical Society

Received: June 2, 2015 Accepted: August 12, 2015

A

DOI: 10.1021/acs.analchem.5b02077 Anal. Chem. XXXX, XXX, XXX−XXX

Article


Analytical Chemistry mass spectrometry.5,9 The separation based on ion mobility is fast in comparison with chromatographic methods, i.e., provides separation on the milliseconds to second time scale. In addition, DMS is able to separate isobaric and isomeric compounds. In some of DMS-based applications, the choice of the proper ionization source is a crucial step for the whole analytical procedure. The crux of the problem refers to understanding the ionization processes and developing ion sources with high ion yield at environmental conditions. The most frequently used ionization sources for DMS are radioactive sources (e.g., 63Ni), photoionization (Kr), corona discharge, and electrospray ionization. However, for the ionization of nonpolar compounds the application of electrospray ionization is very limited and frequently involves a derivatization of the nonpolar analyte.10 In addition, the erosion of discharge electrode during the corona discharge and, thus, degradation in performance of corona ionization source can be expected.11 63Ni is the most commonly used ionization source in DMS because it provides a very stable ion yield without an additional power source. However, the use of this source is relatively complicated due to regulatory problems associated with radioactive ionization sources (e.g., transportation, safety, and waste disposal regulations). These restrictions are especially essential for the devices intended to be used in the field. Therefore, there is a growing interest in replacing radioactive ionization by nonradioactive alternatives. Here, dielectric-barrier discharges (DBD) are considered especially attractive because they typically consume less than 3 W of power, can be operated with a small helium cartridge built into a hand-held probe, and do not use any chemicals or solvents.12 Because of the latter, this source does not produce any chemical waste (that otherwise would have to be routinely removed from a sampling site). The most promising DBD configuration for ambient desorption/ionization is the lowtemperature plasma (LTP) probe which was introduced by Harper et al. in 2008.13 In this configuration, a nonequilibrium low-temperature plasma is generated at atmospheric pressure within a glass tube and extends into the ambient environment (this region of the plasma being called the afterglow). The agents responsible for the plasma reactivity are ultraviolet (UV) photons and reactive species.14 Several reactive species that are capable of ionizing a wide range of compounds (including O2+, O2−, N2+ species, and protonated water clusters) are produced in LTP.15 As a discharge gas, He, Ar, N2, or ambient air, at a low flow rates (80−1000 mL min−1,16), can be used. The temperature of the surface area in contact with the plasma plume is approximately 30 °C, so there is no significant heating of the sample. In addition, the high-voltage electrode is electrically isolated from the direct discharge region, and therefore, the sample is not liable to electric shock. These features imply direct analysis of chemicals even on a human finger.13 Because of the properties listed above, LTP is considered to have a great potential in portable, field-based applications. In the past, DBDs were used successfully as ionization source for mass spectrometry17,18 and time-of-flight ion mobility spectrometry (TOF-IMS).19,20 However, no applications of DBD as an ionization source for differential ion mobility spectrometry (DMS) or high field asymmetric waveform ion mobility spectrometry (FAIMS) can be found in the literature. In this work, the utilization of a low-temperature plasma as an ion source for differential ion mobility spectrometry has been introduced for the first time. The investigation of the main

discharge parameters such as discharge gas composition and flow rate as well as applied voltage was performed. Finally, the LTP-ionization source was compared with traditional for differential ion mobility spectrometry ionization sources, namely, with radioactive 63Ni and with photoionization.

■

EXPERIMENTAL SECTION Experimental Setup. The principle scheme of the experimental setup used in this study is shown in Figure 1 (top). The overall discharge/carrier gas flow was prepared by mixing the flow of nitrogen, controlled by a mass flow controller (MFC, VSO-GC, Pneutronics), with an additional flow of helium in the desired proportion. Helium flow rate was controlled by a mass flow controller (MFC, GFC17, 0−200 mL min−1 He, Aalborg). Total discharge/carrier gas flow rate was kept constant at 400 mL min−1. During the experiments with pure nitrogen (no helium in discharge/carrier gas), nitrogen flow rate was additionally controlled by a solid state flow meter (MFM, Restek 6000, Restek, U.K.) located on the exhaust of DMS. Pressure was monitored using the pressure sensor from the DMS, which is built on the input to the analyzer. Samples were introduced via a syringe pump (SP, kdScientific, KDS Legato 210) into the nitrogen gas flow. A perfluoroalkoxy (PFA) Swagelok 1/4 in. tee was used to couple the stainless-steel grounded electrode, gas tubing for the discharge gas, and glass. The outlet of LTP probe was connected with DMS over the homemade glass-ceramic (MACOR) adapter (Figure 1, bottom). The current construction of adapter enables the fastest possible transport of ions between LTP and DMS. To avoid additional dead volume between the discharge region and DMS, the analyte and the discharge gas are mixed prior the discharge region. The adapter was assembled onto the DMS lamp holder instead of the original photoionization lamp. LTP Probe. Configuration of the home-built LTP probe used in this study was based on a design described by Albert et al.21 An inner stainless steel electrode (0.7 mm diameter, 3 mm distance to capillary exit) and an outer copper ring electrode (6 mm wide, 3 mm distance to capillary exit) were used to

Figure 1. (Top) Schematic diagram of the experimental setup: differential ion mobility spectrometer (DMS), syringe pump (SP), mass flow controller (MFC), mass flow meter (MFM). (Bottom) Schematic of homemade adapter and LTP-probe. B


Article


Analytical Chemistry

Eh within the range of 20−26 kV cm−1, a moderate decrease of resolution was observed. Therefore, the rf-electric field strength of 20 kV cm−1 (1000 V) was chosen for further experiments. For each sample, three single measurements were recorded using Sionex Expert software (version 2.4.0). For the determination of peak parameters (center, area, fwhm), measured data were analyzed by the fityk (version 0.9.8) program.22 Peaks were fitted with Gaussian functions using the Levenberg−Marquardt algorithm. It should be noted that the analyte signal positions on the compensation voltage scale are very sensitive to even minor pressure differences. To enable comparison of spectra obtained under different pressures, the method described by Nazarov et al. was used.23 This method proposes utilization of E/N scaling in Townsend units (Td). In this case, the reduced compensation field scale (CF, in [Td]) is utilized instead of the compensation voltage scale (CV, in [V]). In our study, utilization of this method has minimized but not completely eliminated the differences between the measurements at different pressures. That is why the data presented in this manuscript were recorded within the narrow pressure gap between 14.6 and 14.7 psi. Chemicals. To verify the proposed method six environmentally relevant substances (benzene [AppliChem, 99+ %], toluene [J.T. Baker, 99.9%], ethylbenzene [Fluka, 99+%], pxylene [Fluka, 99+ %], 1,2,4-trimethylbenzene [Aldrich, 98%], and naphthalene [Sigma-Aldrich, 99%]) were selected. Samples were prepared as follows: an analyte sample volume of 1 mL was transferred into a 20 mL vial under nitrogen atmosphere. The vial was closed with a screw cap equipped with Butyl/PTFE septa (S/N 100032, BGB, Germany) and equilibrated for 1 h at 20 °C. Samples were then taken from the headspace of the vial using a Hamilton gastight syringe (10, 50, and 500 μL size). Analyte concentration in the DMS carrier gas was controlled by the injection rate of the syringe pump (KD Scientific, KDS Legato 210). Concentration of model compounds (in the discharge/carrier gas) used in section “Influence of applied voltage on peak parameters of model compounds” were as follows: 360 ng L−1 for benzene, 270 ng L−1 for toluene, 100 ng L−1 for ethylbenzene, 95 ng L−1 for pxylene, 140 ng L−1 for 1,2,4-trimethylbenzene, and 50 ng L−1 for naphthalene.

generate a dielectric-barrier discharge in a glass tube (4.0 mm outer diameter, 0.8 mm inner diameter). A narrow gap between the glass tube and inner electrode (∼0.05 mm) should provide very fast and efficient transport of the created ions in the discharge region toward the DMS. Additionally, a relatively high velocity of ions produced in the LTP can be provided by relatively low gas consumption and when the gap between electrodes is small. A high-voltage alternating current source (4−6 kVpp at 23−24 kHz, PVM500, Information Unlimited, Amherst) was applied between the outer copper ring electrode and the inner pin electrode to generate the plasma. Voltage between electrodes was measured using a portable widebandwidth high-voltage probe (≤60 kV, ≤80 MHz, 1:1000, PVM-1, North Star High Voltage) and recorded using a digital storage oscilloscope (TDS 2012B, Tektronix). Helium and nitrogen (99.999%, Air Liquide, Germany) were used as discharge gases. DMS. Two differential ion mobility spectrometers of similar construction were utilized in this study. One is equipped with removable photoionization source (SVAC-UV, krypton 10.0/ 10.6 eV, Sionex Corp.). Another is equipped with radioactive 63 Ni ionization source (SVAC-V, 63Ni 185 MBq, Sionex Corp.). DMS settings were as follows: sensor temperature = 80 °C, number of steps = 100, step duration = 10 ms, step settle time = 3 ms, steps to blank = 1. Samples were analyzed in the positive mode (positive ions) at rf (radio frequency) voltage of 1000 V (20 kV cm−1) and nitrogen (99.999%, Air Liquide, Germany) flow rate of 400 mL min−1, unless otherwise noted. Compensating voltage range was set from −30 to +10 V. The choice of rf-electric field strength (Eh) of 20 kV cm−1 for the experiments demonstrated in this paper can be explained with help of Figure 2. This figure demonstrates dependence of ethylbenzene signal on Eh. Additionally, relationship between the normalized peak area of ethylbenzene peak, the resolution between ethylbenzene and background signal, and Eh are demonstrated. At Eh of 10 and 12 kV cm−1 no separation between ethylbenzene and background signals was achieved. The further increase of Eh results in the increase of the resolution between these two peaks as well as in continuous decrease of ethylbenzene peak area. The highest resolution between these two peaks was achieved at Eh = 20 kV cm−1. At

■

RESULTS AND DISCUSSION Composition of Discharge/Carrier Gas. The composition of the carrier gas has to be carefully selected, because it was reported in the literature that LTP and DMS are susceptible to changes in the operating gas composition, i.e., show different ionization and separation performance, respectively. It is therefore important to consider the requirements that a coupled LTP-DMS system would have for the most appropriate discharge/carrier gas composition. In this regard, the importance of the discharge/carrier gas is discussed from both an LTP and DMS perspective before the results of the optimization are presented below. In principle, low-temperature plasmas (or other types of dielectric-barrier discharges) can be sustained in a variety of different discharge gases. So far, no fundamental study on the ionization pathways in LTP-DMS was performed. However, Chan et al. studied the reaction mechanisms in a stand-alone LTP and the afterglow and demonstrated that besides Penning ionization from metastable helium (when helium is used as a plasma gas) also charge transfer between He2+ ions and

Figure 2. Differential mobility spectra of ethylbenzene (510 ng L−1) recorded within rf-voltage range of 500−1300 V (10−26 kV/cm) in nitrogen/helium mixture 3/1 (flow rate = 400 mL min−1, discharge voltage = 5 kV). Relationships between the normalized peak area, the resolution (Rs), and rf-electric field strength (Eh) are depicted in the inset. C


Article


Analytical Chemistry atmospheric nitrogen can lead to the formation of N2+ ions.15 This ion is considered to be a key reaction intermediate for the formation of other reagent ions such as protonated water clusters, (H2O)nH+. This is important, because protonated water clusters are also considered to be the main reactive ions in ion mobility spectrometry with corona discharge- and βemitter-based (e.g., 63Ni) ionization sources. Therefore, LTP can potentially be considered as an alternative ion source for ion mobility spectrometry. The formation of reactant ion positive (RIP), which is responsible for subsequent ionization of analytes, can be achieved via different ionization pathways. These ionization mechanisms may include electron impact, energy transfer with metastable species (e.g., He), charge transfer with other ions (e.g., He2+, N2+), proton/electron transfer, and proton/hydride abstraction. In addition the analyte ionization can proceed over a direct photoionization via emitted photons in the LTP. Variation of the He/N2 proportion in the discharge gas may alter the ionization mechanism or change the degree of contribution of different mechanisms to the overall ionization process. The influence of the carrier gas nature on the DMS can be explained as follows. The working principle of DMS is based on the nonlinear dependence of the ion mobility coefficient K(E/ N) on the applied electric field. At low values of electric field, K is nearly constant and almost independent of E/N [K(0)]. With increasing E/N values, the ion mobility coefficient becomes to be significantly dependent on the electric field. This dependence is given by K(E/N) = K(0)[1 + α(E/N)], were α is a relative variation of K(0) depending on E/N. The mobility of an ion at low electric field [K(0)] can be described by the Mason−Schamp equation,24 K (0) =

1/2 3 ⎡ 2π ⎤ ze ⎥ ⎢ 16 ⎣ μkT ⎦ N Ω

different carrier gases. However, for relatively small molecules (20 ppmv) the water clusters should be the dominating positive charge carriers (RIP). However, further investigations concerning the nature of these peaks are required. Influence of Ion Source Voltage on Reactant Ion Positive. The influence of the LTP ionization source voltage on the RIP 1 and RIP 2 peak areas is presented in Figure 4. These dependencies were analyzed for the carrier gas containing 0, 19, 25, and 31% helium in nitrogen. The overall flow of carrier gas was kept constant at 400 mL min−1. It was observed that increasing the helium fraction in the discharge gas lead to a decrease of the minimal value of applied voltage that was required for RIP detection. The applied voltage of ∼4.3 kVpp was required when discharge was performed in nitrogen containing 31% of helium. However, for RIP detection in pure nitrogen an applied voltage of about 5.2 kVpp was necessary.


63

Figure 3. (Top) Differential mobility spectra recorded with discharge/ carrier gas of different He-fractions in nitrogen. LTP was operated with 5.67 kVpp (see Experimental Section for details on LTP and DMS conditions). Bottom: LTP-DMS analysis of toluene. Blank-corrected spectra after subtracting blank measurements (top) from corresponding spectra recorded with 265 ng L−1 of toluene in the discharge/ carrier gas. Inset shows the relationship between compensation field (CF) of RIP 1 and RIP 2 peaks and the helium fraction in discharge/ carrier gas.

fractions in the discharge/carrier gas were not investigated below. In Figure 3 (bottom), blank-corrected spectra after LTPDMS analysis of toluene (265 ng L−1 spiked into the discharge/ carrier gas, blank measurements can be found in Figure 3 (top)) are shown. The measured (not blank corrected) spectra can be found in Figure 5 (bottom). Addition of toluene to the discharge/carrier gas resulted in the formation of a new peak with a maximum intensity of ∼0.08 V between −0.1 and −0.4 Td. At the same time, a significant signal reduction (negative intensity) of the peaks at CF values corresponding to both RIP 1 and RIP 2 was observed. On the basis of this observation, RIP 1 and RIP 2 were attributed as reactant ions positive (RIP). Jafari previously observed the formation of several RIP peaks when LTP was used as the ionization source for TOF-IMS.20 On the basis of drift time, peaks were attributed to NH4+, NO+, and (H2O)nH+ ions. Interestingly, in our study the position of both RIP 1 (−1.8 Td) and RIP 2 (−1.6 Td) on the CF scale is different from the position of the RIP (−1.26 Td) produced with traditional APCI ionization sources (e.g., 63Ni, see Table 1, discharge/carrier gas nitrogen). Therefore, the structure of RIP 1 and RIP 2 is likely to be different from the structure of traditionally reported RIP, which is typically attributed to the protonated water clusters ((H2O)nH+). However, further investigations concerning the nature of these peaks should be

Figure 4. Relationship between peak area and applied ion source voltage under different discharge/carrier gas conditions presented for RIP 1 (top) and RIP 2 (bottom). Inset shows the corresponding relationship for RIP 2 with alternative scale to illustrate the transition region. E


Article



nitrogen, respectively). In Figure 5 (bottom), the corresponding differential mobility spectra of toluene are presented. Increasing the helium fraction in the discharge/carrier gas resulted in a shift of the toluene peak toward more positive CF values (Figure 5, bottom, insert), while peak area and fwhm values decreased (Figure 5, top, insert). The decrease of the toluene peak width with the increase of the helium fraction in the discharge/carrier gas can be explained as follows. In planar DMS (where the field is homogeneous) at moderate electric field strengths and when the residence time is significantly shorter than the diffusion time, peak width is proportional to (K)−1.4,28 Because of an increase of the helium fraction in the carrier gas, both the average collision cross section (Ω) and reduced mass (μ) values are reduced. This results in a higher mobility of the ions (see eq 1) and, thus, in a reduced peak width. The dependence of the toluene peak area on the applied voltage was found to be rather complex. The initial significant increase of the peak area with the increase of the applied voltage is followed by a peak plateau region and/or decay of peak area for most of the conditions studied. Most likely, the initial increase of the toluene peak with the increase of applied voltage is due to the more efficient formation of reactive species in the ion source (see Figure 4). This will be investigated in detail in a follow-up study. A potential reason for reduced peak areas at higher voltages could be due to partial fragmentation of toluene ions in the source (see Figure 6, bottom). Clearly, further investigations of the LTP-DMS fundamentals are required for a better understanding of these effects. A maximum of toluene peak area was achieved when the discharge was sustained in pure nitrogen. Interestingly, Jafari previously evaluated the use of pure nitrogen as a discharge gas for LTP-TOF-IMS. However, because of a rather low applied voltage (3 kVpp) the plasma could not be ignited.20 This observation is in a good agreement with the results achieved in this work. Here, an applied voltage of ∼5.17 kVpp was required to achieve a stable LTP-DMS signal for toluene in pure nitrogen. The ability to perform analysis with zero helium consumption is especially attractive in view of potential online (and on-site) monitoring applications of LTP-DMS. Because of this, the experiments described below were performed using pure nitrogen as the discharge/carrier gas. Influence of Applied Voltage on Peak Parameters of Model Compounds. The influence of applied voltage (5.17− 6.00 kVpp) on peak parameters of several model compounds was studied when nitrogen was used as discharge/carrier gas (Figure 6). A minimal applied voltage of ∼5.17−5.33 kVpp was required to be able to observe distinct peaks of model compounds. On the basis of the findings for peak area, peak shape, and applied voltage, model compounds can be divided into two groups. Benzene and toluene (first group) demonstrated a significant increase in peak area within the applied voltage range of 5.33−5.67 kVpp. A further increase in applied voltage did not result in a further significant increase in peak area (Figure 6, top). Peak width of benzene and toluene increased but peak shape was not significantly changed (see, for example, toluene in Figure 6, bottom, left). A rather different relationship between analyte peak area, peak shape, and applied voltage was found for ethylbenzene, pxylene, 1,2,4-trimethylbenzene (TMB), and naphthalene (second group). An increase of analyte peak area was observed within the applied voltage range of 5.17 and 5.50 kVpp. Similarly

In all tested discharge gas mixtures, formation of the two reactant ions (RIP 1 and RIP 2) was observed. Increasing the applied voltage resulted in an overall increase of RIP 1 and RIP 2 peak areas. RIP 1 was found to be the dominating signal within the applied voltage ranges of 4.3−5.5 kVpp. Interestingly, an increase of the applied voltage from 5.5 to 5.67 kVpp resulted in an increase of the RIP 2 signal area by more than 12 times (see Figure 4, bottom). Within the applied voltage range of 5.67−6.0 kVpp, the RIP 2 was the dominating ion signal. Influence of Discharge/Carrier Gas Composition on Toluene Peak Parameters. The influence of the discharge/ carrier gas composition on the analysis of aromatic compounds was exemplarily examined for model compound toluene. Toluene peak parameters (CF, area, fwhm) were analyzed within the applied ion source voltage range of 4−6 kVpp for different carrier gas mixtures as mentioned above. As can be seen from Figure 5, in mixtures containing a higher fraction of helium, the peak of toluene could be detected at lower applied ion source voltages. However, greater toluene peak areas were achieved when the helium fraction in the discharge gas was reduced and high ion sources voltages were applied. On the basis of these observations, ion source voltages were selected that resulted in good analyte response (peak area) and stable conditions at the given gas mixtures (5.7 kVpp, 5.0 kVpp, 4.5 kVpp, and 4.3 kVpp at 0%, 19%, 25%, and 31% helium in

Figure 5. (Top) Influence of ion source voltage on toluene peak area using different gas compositions. Inset shows the influence of the gas composition on toluene peak area (max.) and fwhm. (Bottom) Differential mobility spectra recorded with discharge/carrier gas containing 0, 19, 25, and 31% of helium. Relationship between the compensation field (CF) of toluene peak and helium fraction is depicted in the inset. F


Article



Figure 6. (Top) Peak areas of several compounds obtained with LTPDMS as a function of ion source voltage with nitrogen discharge/ carrier gas. (Bottom) Differential mobility spectra recorded in nitrogen with 270 ng L−1 of toluene (left) and with 95 ng L−1 of p-xylene (right).

Figure 7. (Top) Differential mobility spectra of 1,2,4-trimethylbenzene (140 ng L−1) recorded with the discharge/carrier gas (nitrogen) flow rate of 330, 350, 370, 400 mL min−1. (Bottom) Differential mobility spectra of model compounds recorded with the discharge/ carrier gas (nitrogen) flow rate of 350 mL min−1 and applied voltage of 5.5 kVpp.

to the first group, an increase of applied voltage resulted in an increase of analyte peak widths with no significant changes in peak symmetry. However, a further increase of applied voltage leads to a slight decay in peak area and significant changes in peak shape (Figure 6, bottom, right). It is assumed that considerable fragmentation of analyte ions takes place and leads to the formation of additional signals in LTP-DMS, which are superimposed on the original analyte ion peak. Thus, an applied voltage of 5.5 kVpp was considered to be optimal for the current system and when pure nitrogen is used as the discharge/carrier gas. Analyte peak areas of the model compounds are then close to maximal values while significant analyte fragmentation is not observed. Influence of Discharge/Carrier Gas Flow Rate on Peak Parameters of Model Compounds. The discharge/carrier gas flow rate has an influence on both the transport of ions from LTP to DMS and the transport of ions within DMS. It was assumed that the ion transport efficiency could be improved by increasing the discharge/carrier gas flow rate. Most likely, this would also result in less ion recombination in the DMS due to the faster traveling times. In Figure 7 (top), the influence of the discharge/carrier gas flow rate on the analyte peak parameters for 1,2,4-trimethylbenzene (Figure 7, top) is presented. It was found that both analyte peak area and analyte fwhm increased toward higher flow rates. These observations are in a good agreement with published literature for similar systems.

For example, the effect of the carrier gas flow rate on the peak parameters of toluene was investigated by Miller et al.8 It was reported that both toluene peak area and toluene fwhm increased with increasing carrier gas flow rates at flow rates below 3 L min−1. Similar findings for aromatic compounds were observed by Liang et al.29 In this study, higher gas flow rates decreased both the transfer time (tT) of ions from LTP to DMS as well as the residence time (tR) of ions within the filter region of DMS. Because of the reduced tT and tR, loss of ions by diffusion and discharge on the ion source wall, transfer capillary walls, and DMS electrodes, respectively, is likely to be reduced. Also, reduced ion recombination (positive with negative) as well as reduced charge transfer between ions and neutrals can be a result of the higher carrier gas flow rate. All of these effects combined result in an overall increased peak area. The increase of fwhm with the increase of the carrier gas flow rate can be explained as follows. The residence time is inversely proportional to the flow rate of carrier gas. If the gas flow rate is sufficiently high and the residence time is significantly shorter than the diffusion time, peak width is given by fwhm = ge(Ktres)−1, were ge is the effective gap width.4,28 Thus, reduction of the carrier gas flow rate results in the increase of the residence time in the separation region and hence in a decrease of peak width. Peak capacity is one way to express the separation capability of an analytical device. For DMS, the peak capacity can be defined as a peak spread on the CF scale divided by the average G


Article


Table 1. Compensation Field (CF) Values and Limits of Detection (LOD) for Benzene, Toluene, Ethylbenzene, p-Xylene, 1,2,4Trimethylbenzene (TMB), and Naphthalene As Well As CF for RIP Achieved with Different Ionization Sources LTP-DMS


benzene toluene ethylbenzene p-xylene TMB naphthalene RIP

63

APPI-DMS −1

Ni-DMS

CF (Td)

LOD (ng L )

CF (Td)

LOD (ng L )

CF (Td)

LOD (ng L−1)

−0.53 −0.34 −0.25 −0.23 −0.12 −0.20 −1.80 −1.60

257 118 35 102 73 168

−0.58 −0.36 −0.27 −0.25 −0.14 −0.17

223 134 84 88 64 112

−0.60 −0.37 −0.27 −0.24 −0.12 −0.17 −1.26

239 58 107 109 12 69

fwhm. Thus, the peak capacity of planar DMS (or FAIMS) devices may be increased (at the cost of lower ion transmission and, thus, sensitivity) using lower flow rates. In the following experiments, the carrier gas flow rate was reduced to 350 mL min−1. This provided a higher resolution compared to flow rate of 400 mL min−1 at the cost of slightly reduced sensitivity. The differential mobility spectra for all model compounds measured at nitrogen flow rate of 350 mL min−1 and applied voltage of 5.5 kVpp (optimal for current setup, see previous section) are demonstrated in Figure 7 (bottom). Determination of Analytical Parameters for the Model Compounds Analyzed by LTP-, 63Ni-, and APPIDMS. In order to prove the effectiveness of LTP as an ionization source for DMS the analytical parameters obtained with LTP-DMS where compared with those obtained with traditional 63Ni-DMS and APPI-DMS. Compensation field (CF) values and limits of detection for model analytes achieved with LTP-, APPI-, and 63Ni-ionization sources are summarized in Table 1. Limits of detection were calculated according to the concentration of the analyte in the carrier gas entering the ion source. A nonlinear relationship between signal area and concentration common for nondirect ionization mechanisms (e.g., APCI) was observed for all model analytes when LTP and 63 Ni were used as ionization sources.30−32 Because of this, nonlinear second-order calibration functions according to DIN 8466-2 were applied in this study.33 Additionally, calibration of benzene was complicated due to significant dimer formation at concentrations higher than ∼800 ng L−1. Representative differential mobility spectra at different concentrations of ethylbenzene are exemplarily presented in Figure 8 (top). The LOD of the model compounds obtained with LTPDMS were found to be good comparable to LODs from the established ionization sources (see Table 1 for details) and in the hundreds- or tens-of-nanogram-per-liter range (concentration of the analyte in the carrier gas entering the ion source). Across the model compounds, the highest LOD was found for benzene (with all ionization sources). One possible reason is dimer formation in DMS, which adversely influences the benzene calibration curve, and a relatively low proton affinity, which makes benzene analysis sensitive to even minor moisture impurities in the instrument or the carrier gas (see ref 27 for a detailed discussion on humidity effects in DMS). Atmospheric pressure photoionization (APPI) and atmospheric pressure chemical ionization (APCI) sources for DMS applications provide the ionization of analytes by different ionization pathways. In APCI (e.g., corona or 63Ni), the majority of analyte (A) is considered to be ionized via proton transfer from protonated water clusters. This mechanism leads

−1

Figure 8. (Top) Differential mobility spectra of ethylbenzene at different concentrations recorded with the discharge/carrier gas (nitrogen) flow rate of 350 mL min−1 and applied voltage of 5.5 kVpp. (Bottom) Differential mobility spectra of toluene (270 ng L−1) obtained with LTP-, APPI-, and 63Ni-ionization sources. The discharge/carrier gas (nitrogen) flow rate is 350 mL min−1, and the applied voltage (LTP) is 5.5 kVpp. The fwhm values for the toluene peaks are demonstrated in the inset.

to the formation of partially hydrated protonated analyte ions, e.g., A(H2O)nH+.32 In APPI, analyte ionization is achieved via direct photoionization (when the analyte ionization potential is lower than the photon energy and in the absence of a dopant) to give analyte radical-cations.32 It was observed that fwhm values for analyte peaks were higher in measurements with LTP-DMS as compared to those in measurements with 63Ni-DMS (e.g., Figure 8, bottom, insert). It should be noted that the sample introduction for measurements with LTP- and 63Ni-DMS was performed differently. In contrast to the measurements with 63Ni-DMS, H


Article



(6) Krylov, E. V.; Nazarov, E. G. Int. J. Mass Spectrom. 2009, 285, 149−156. (7) Buryakov, I. A.; Krylov, E. V.; Nazarov, E. G.; Rasulev, U. Kh. Int. J. Mass Spectrom. Ion Processes 1993, 128, 143−148. (8) Miller, R. A.; Nazarov, E. G.; Eiceman, G. A.; King, A. T. Sens. Actuators, A 2001, 91, 301−312. (9) Schneider, B. B.; Covey, T. R.; Coy, S. L.; Krylov, E. V.; Nazarov, E. G. Int. J. Mass Spectrom. 2010, 298, 45−54. (10) Hayen, H.; Karst, U. J. Chromatogr. A 2003, 1000, 549−565. (11) Waltman, M. J.; Dwivedi, P.; Hill, H. H., Jr.; Blanchard, W. C.; Ewing, R. G. Talanta 2008, 77, 249−255. (12) Wiley, J. S.; Shelley, J. T.; Cooks, R. G. Anal. Chem. 2013, 85, 6545−6552. (13) Harper, J. D.; Charipar, N. A.; Mulligan, C. C.; Zhang, X.; Cooks, R. G.; Ouyang, Z. Anal. Chem. 2008, 80, 9097−9104. (14) Lu, X.; Laroussi, M. J. Appl. Phys. 2005, 98, 023301. (15) Chan, G. C. Y.; Shelley, J. T.; Wiley, J. S.; Engelhard, C.; Jackson, A. U.; Cooks, R. G.; Hieftje, G. M. Anal. Chem. 2011, 83, 3675−3686. (16) Albert, A.; Shelley, J. T.; Engelhard, C. Anal. Bioanal. Chem. 2014, 406, 6111−6127. (17) Hayen, H.; Michels, A.; Franzke, J. Anal. Chem. 2009, 81, 10239−10245. (18) Meyer, C.; Müller, S.; Gurevich, E. L.; Franzke, J. Analyst 2011, 136, 2427−2440. (19) Vautz, W.; Michels, A.; Franzke, J. Anal. Bioanal. Chem. 2008, 391, 2609−2615. (20) Jafari, M. T. Anal. Chem. 2011, 83, 797−803. (21) Albert, A.; Engelhard, C. Spectrochim. Acta, Part B 2015, 105, 109−115. (22) Wojdyr, M. J. Appl. Crystallogr. 2010, 43, 1126−1128. (23) Nazarov, E. G.; Coy, S. L.; Krylov, E. V.; Miller, R. A.; Eiceman, G. A. Anal. Chem. 2006, 78, 7697−7706. (24) Mason, E. A.; McDaniel, E. W. Transport Properties of Ions in Gases; Wiley: New York, 1988. (25) Shvartsburg, A. A.; Ibrahim, Y. M.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2014, 25, 480−489. (26) Meek, J. M.; Craggs, J. D. Electrical Breakdown of Gases; Wiley: New York, 1978. (27) Kuklya, A.; Uteschil, F.; Kerpen, K.; Marks, R.; Telgheder, U. Int. J. Ion Mobility Spectrom. 2015, 18, 67−75. (28) Krylov, E. V.; Nazarov, E. G.; Miller, R. A. Int. J. Mass Spectrom. 2007, 266, 76−85. (29) Liang, F.; Kerpen, K.; Kuklya, A.; Telgheder, U. Int. J. Ion Mobility Spectrom. 2012, 15, 169−177. (30) Kuklya, A.; Uteschil, F.; Kerpen, K.; Marks, R.; Telgheder, U. Talanta 2014, 120, 173−180. (31) Kuklya, A.; Uteschil, F.; Kerpen, K.; Marks, R.; Telgheder, U. Anal. Methods 2015, 7, 2100−2107. (32) Eiceman, G. A.; Karpas, Z.; Hill, H. H., Jr. Ion Mobility Spectrometry, 3rd ed.; CRC Press: Boca Raton, FL, 2014. (33) Deutsches Institut für Normung e. V (DIN). Water quality; Calibration and evaluation of analytical methods and estimation of performance characteristics; Part 2: Calibration strategy for nonlinear second-order calibration functions, DIN ISO 8466-2:2001, Berlin, June 2004.

in which the carrier gas was introduced into DMS via the standard inlet, the introduction of carrier gas for the measurements with LTP was performed via a home-built adapter. This adapter was assembled onto DMS lamp holder instead of the original photoionization lamp. The introduction via lamp holder was necessary to provide much faster (as compared to standard inlet) transport of ions into the DMS filter region. Interestingly, the introduction of the carrier gas into 63Ni-DMS over the inlet located in the same place as the photoionization lamp holder resulted in broadening of analyte peaks (from 0.102 to 0.114 Td for toluene). Thus, the way the carrier gas is introduced is important and a source for peak broadening observed in LTP-DMS measurements. This parameter should be optimized to avoid the reduction of resolving power, which is already limited for the presented in this study DMS. However, the values of fwhm for analyte peaks in measurements with LTP-DMS and in measurements with APPI-DMS were in the same range (Figure 7, bottom, insert). The representative differential mobility spectra obtained with LTP-, APPI-, and 63Ni-ionization sources are demonstrated on the example of toluene (c = 270 ng L−1) in Figure 8 (bottom).

■

CONCLUSIONS A LTP ionization source was identified as a suitable ionization source for differential ion mobility spectrometry. The results demonstrate that the sensitivity of a home-built and only partially optimized LTP-DMS setup is already good comparable to commercially available 63Ni-DMS and APPI-DMS systems (for the compounds studied). LODs in the tens-of-nanogramper-liter range were achieved (concentration of the analyte in the carrier gas entering the ion source). Further improvements and optimization may significantly increase the sensitivity and resolving power of LTP-DMS. It was demonstrated that a nitrogen-operated LTP-ion source shows some advantages over an LTP-ion source sustained in a nitrogen/helium mixture. Zero helium consumption is especially attractive in view of potential application of DMS with the LTP-ionization source for online (and on-site) monitoring. In the future, further investigations are required, for example, to elucidate the nature of the reactive species that are important in the LTP-DMS ionization pathways.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. Phone: +49(0)201 1836786. Fax: +49(0)201 183-6773. *E-mail: [email protected]. Phone: +49(0)271 740-4573. Fax: +49(0)271 740-2041. Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Baciocchi, R.; Berardi, S.; Verginelli, I. J. Hazard. Mater. 2010, 181, 226−233. (2) Shih, T.; Rong, Y.; Harmon, T.; Suffet, M. Environ. Sci. Technol. 2004, 38, 42−48. (3) Weisman, W. Analysis of Petroleum Hydrocarbons in Environmental Media; TPH Series, Vol. 1; Amherst Scientific Publishers: Amherst, MA, 1998. (4) Shvartsburg, A. A. Differential Ion Mobility Spectrometry: Nonlinear Ion Transport and Fundamentals of FAIMS; CRC Press: Boca Raton, FL, 2008. (5) Kolakowski, B. M.; Mester, Z. Analyst 2007, 132, 842−864. I


Low-Temperature Plasma Ionization Differential Ion Mobility

Recommend Documents