Identity Efficiency for High-Performance Ambient Pressure Ion Mobility

Feb 26, 2016 - This report determined and compared CRM, resolving power (Rm), and diffusion-limited conditional theoretical reduced mobility (DLCTRM) ...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Identity Efficiency for High-Performance Ambient Pressure Ion Mobility Spectrometry A Bakarr Kanu* and Anne Leal Department of Chemistry, Winston-Salem State University, Winston-Salem, North Carolina 27110, United States S Supporting Information *

ABSTRACT: A new approach to reduce the false-positive responses commonly encountered in the field when drugs and explosives are detected is reported for an electrospray ionization high-performance ion mobility spectrometry (ESI-HPIMS). In this article, we report on the combination of reduced mobility and the width-athalf-height of a peak to give a new parameter called conditional reduced mobility (CRM). It was found that the CRM was capable of differentiating between real drugs peaks from that of a false-positive peak and may help to reduce falsepositive rates. This effect was demonstrated using 11 drugs (amphetamine, cannabidiol, cocaine, codeine, heroine, methamphetamine, morphine, phentermine, L-phenylepherine, proglitazone, and rosiglitazone) and seven interferences chosen from off-the-shelf products. This report determined and compared CRM, resolving power (Rm), and diffusion-limited conditional theoretical reduced mobility (DLCTRM) for ESI-HPIMS. The most important parameters for determining CRM are reduced mobility and width-at-half-height of a peak. There is a specific optimum voltage, gate pulse width, resolving power, and now CRM for each ion. DLCTRM indicate the optimum reduced mobility that is not normally possible under field conditions. CRM predicts the condition at which a target compound can be differentiated from a false-positive response. This was possible because different ions exhibits different drifting patterns and hence a different peak broadening phenomenon inside an ion mobility tube. Reduced mobility for target compounds reported were reproducible to within 2% for ESI-HPIMS. The estimated resolving power for the ESI-HPIMS used in this study was 61 ± 0.22. Conditional reduced mobility introduced in this paper show differences between target compounds and false-positive peaks as high as 74%, as was the case for cannabidiol and interference #1 at 70 μs gate pulse width.

T

Despite years of development, commercial IMS instruments, when applied to real samples in the field, still experience falsepositive responses.28 This has mainly due to their low resolving power. Most commercial ion mobility spectrometers have a resolving power of about 30.18 Thus, it is possible for an ion which is not associated with a drug or explosive to drift with a similar time to that of a drug or explosive and be falsely identified. In such instances, the instrument indicates the presence of a drug or explosive when in fact none were present. To ensure the absence of the target, more sophisticated tests must be performed as well as comprehensive searches of luggage or containers. Thus, false-positive responses cost both money and time. Development of simple methods that can serve as confirmatory tests and aid in the identification of drugs, explosives, and chemical warfare agents are necessary. Several approaches have been proposed to reduce frequency of false positives in an IMS. One approach has been to couple an IMS to mass spectrometry (MS). The combination of the two-dimensional (2-D) approach, IMMS, significantly reduces the possibility of false positives because it is not probable that two different ions will have both the same size and mass. The difficulty with adding a mass spectrometer to an ion mobility spectrometer for field measurements is expense and instrument complexity. MS requires a vacuum, which is difficult to maintain

he technique, stand-alone ambient pressure ion mobility spectrometry (IMS) is an important analytical separation and detection technique that has been routinely used for well over 40 years1 in the detection of drugs of abuse,2−8 explosives,9−17 chemical warfare agents,18−20 biological compounds,21 toxins in the workplace,22 environmental contaminants,23,24 industrial chemicals for process control,25 and volatile organics in space shuttle cabin air.26 In addition to the vast detection capability of IMS, several other advantages of the technique have been previously identified. For example, when the technique is compared to chromatographic and electrophoretic techniques, IMS is fast, sensitive, and more reproducible. Gas-phase ions generated in an IMS are separated in a few milliseconds with resolving powers for research-grade IMS similar to that of chromatography. Normalized standard temperature and pressure are required in IMS, making it possible for arrival times to be converted to reduced ion mobility constants. These reduced ion mobility constants, which are fundamental measurements based on the ion’s diffusion coefficient, are reproducible27 from day to day and from instrument to instrument. The sensitivity and detection limits in an IMS are similar to that of flame ionization detector in gas chromatography and are usually better than that for UV−vis absorption detectors in liquid chromatography.1 The responses for ions in an IMS can be collected and recorded efficiently. When compared to a mass spectrometer, IMS is less expensive, requires less bench space, is mechanically more robust, and is much easier to operate. © XXXX American Chemical Society

Received: October 6, 2015 Accepted: February 8, 2016

A

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry on a routine basis in day-to-day field operations. A second approach, which has been suggested, is the addition of gas

chromatographic (GC) separations prior to ion mobility detection. This 2-D approach offers the additional advantage of separation before ionization. In both the direct IMS and the GC-IMS approach, complex mixtures can be sent to the ionization region of the detector where ion suppression and ion competition can occur, resulting in reduction of the quantitative information obtained in the data. There are still other problems with the approach of GC-IMS to reduce false positives and these were discussed elsewhere.29 The hypothesis is that the combination of reduced mobility and the width-athalf-height of a peak to give a new parameter which we have called conditional reduced mobility may provide a simple

Figure 1. Schematic cross-sectional view of electrospray ionization high-performance ambient pressure ion mobility spectrometry.

Figure 2. Representative ESI-HPIMS for target compounds studied in this investigation. The first two peaks in each spectrum represent the reactant ion peak. The drift gas was air, and the gate pulse width was 90 μs. B

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry • Interference #3 (int. #3).

Table 1. Reduced Mobility, Resolving Power, and Signal-to-Noise Ratio of the Compounds Investigated with ESI-HPIMSa compound

*K0 (lit. K0)

AMP CBD

1.56 ± 0.01 (1.55) m: 1.08 ± 0.02 d: 0.94 ± 0.01 1.14 ± 0.02 (1.16) 1.25 ± 0.01 1.05 ± 0.03 (1.04) 1.61 ± 0.02 (1.63) 1.23 ± 0.01 (1.25) 1.57 ± 0.01 1.56 ± 0.02 1.10 ± 0.01 1.14 ± 0.01

COC COD HER METHAMP MOR PHENT L-PHENE PIOGZ ROSGZ

resolving power (Rp)

signal-to-noise (SNR)

48 ± 0.12 49 ± 0.22

18 ± 0.18 31 ± 0.11

47 58 59 60 54 57 49 63 61

± ± ± ± ± ± ± ± ±

0.34 0.15 0.35 0.31 0.45 0.12 0.13 0.06 0.08

21 37 21 24 19 17 16 15 13

± ± ± ± ± ± ± ± ±

• Interference #4 (int. #4). • Interference #5 (int. #5). • Interference #6 (int. #6). • Interference #7 (int. #7). See the Supporting Information section for the experimental procedure followed to prepare samples. Instrumentation. A commercial RTK2100 stand-alone electrospray ionization high-performance ion mobility spectrometry (ESI-HPIMS) system (Excellims Corporation, Acton, MA) was used in this investigation. The complete RTK2100 consisted of the ESI source, HPIMS drift tube, universal ion mobility spectrometry controller (UIMSC), and a VisIon software. In IMS, the first step is to generate gas-phase analyte ion. IMS when interfaced with ESI49 enabled the detection of high molecular weight polar/biological samples that were previously only analyzed with high-performance liquid chromatography (HPLC). This development has undoubtedly allowed for the adoption of IMS in diverse applications. The ESI solvent used in the positive ion mode (49.5% methanol/49.5% water/1% HOAc)40,50 was delivered with the use of a syringe pump (Analytical West Inc., Corona, CA). ESI flow rate in this investigation was 3.00 μL min−1 delivered from a 500 μL gastight syringe (Fischer Scientific, Hanover Park, IL) for the entire experiment. The syringe was connected to a 150 μm o.d. × 10 μm i.d. × 17.4 cm long fused silica capillary (Polymicro Technologies, Phoenix, AZ) using a zero dead volume needle to capillary connector (Upchurch Scientific, Oak Harbor, WA). A second silica capillary (150 μm o.d. × 10 μm i.d. × 7.5 cm long) was used to transfer the ESI solution through another zero dead volume capillary to an electrospray needle connector. The electrospray needle was held at 2.90 kV above the drift potential of 8.00 kV. Analyte ions were continuously infused through the electrospray needle by applying a potential of 10.5 kV at a current limit of 1 μA to maintain a stable spray.51−54 This resulted in the ESI solvent being removed from the ionized droplets in the ∼6.00 cm long desolvation region. Analyte ions were subsequently introduced into the drift region held at a constant temperature of 150 °C. A Bradbury−Nielson ion gate was used to admit ions into the drift region via a pulsed gate width ranging from 10 to 100 μs divided the desolvation regions from the 10.5 cm long drift region. The upper potential of the desolvation region was held at 8.00 kV, and the gate reference voltage was held at approximately 7.91 kV over the 10.5 cm long drift cell creating an electric field of approximately 753 V/cm in the drift cell. Ions were separated according to their size-to-charge as they moved under the influence of the drift field through a 1.4 L/min counter flowing drift gas in the drift region. Each mobility spectra represented a sum of 10 spectrum ranging in length of 40 ms which were sampled by a Faraday plate detector. Data was acquired using Excellims VisIon control and acquisition software and were exported as text files for procession using Microsoft Excel. Table S-1 (see Supporting Information section) summarizes the operating conditions and the drift gas used in this investigation. Figure 1 shows a cross-sectional schematic view of the ESI-HPIMS. Safety Consideration. The ESI-HPIMS instrument was operated at high voltages; thus, only trained personnel were given the privilege to operate the instrument.

0.52 0.35 0.09 0.07 0.17 0.17 0.22 0.14 0.03

The flow rate and gate pulse width were 3 μL min−1 and 80 μs, respectively. *Air, cm2 V−1 s−1; m = monomer; d = dimer

a

approach which can be used in the field to confirm or refute a positive response for drugs. To test our hypothesis, a commercially available electrospray ionization high-performance ion mobility spectrometry (ESIHPIMS) was used. In a series of publications and presentations from 1987 to 1992, Hill applied electrospray ionization as a sample introduction technique for introducing nonvolatile compounds into an ion mobility spectrometer.30−39 The types of ESI and their advantages were discussed elsewhere.40 Today ESI is one of the most important and powerful analytical technique because of its ability to analyze nonvolatile compounds. Our investigation determined characteristic width-athalf-height “signatures” for a variety of drugs as well as a variety of interference materials. This paper seeks to define a new conditional reduced mobility constant that can serve as an efficient identity when using existing IMS technology. See the Supporting Information section for the theoretical consideration that addressed the development of equations in this article.

■

EXPERIMENTAL SECTION Materials and Reagents. A total of 11 compounds were chosen for studies in the positive ion mode. The first nine compounds (amphetamine (AMP), methamphetamine (METHAMP), cocaine (COC), cannabidiol (CBD), codeine (COD), heroine (HER), morphine (MOR), phentermine (PHENT), −1 L-phenylepherine (L-PHENE)) were purchased as 1 mg mL certified reference materials (CRM) from Cerilliant (Austin, TX). The other two compounds (pioglitazone (PIOGZ) and rosiglitazone (ROSGZ)) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO). 2,4-Lutidine (99%), which was the standard in the positive ion mode, was purchased from SigmaAldrich Chemical Co. Methanol, water, and glacial acetic acid was HPLC grade and was purchased from J.T. Baker (Phillipsburg, NJ). Interferences chosen at random from off-the-shelf cosmetics were purchased from Wal-Mart and CVS (Winston-Salem, NC). Of the 30 off-the-shelf products studied, 7 were determined to have significant impact in this study. The names of the products used in this investigation are withheld for security reasons and the authors have summarized them as follows: • Interference #1 (int. #1). • Interference #2 (int. #2). C

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry Table 2. Reduced Mobility and Resolving Power of Interferences Investigated with ESI-HPIMSa compound

K0 (air, cm2 V−1 s−1)

resolving power (Rp)

interference 1

1.70 ± 0.02, 1.62 ± 0.01, 1.47 ± 0.03, 1.32 ± 0.02, 1.25 ± 0.01, 1.08 ± 0.01, 0.92 ± 0.03, 0.88 ± 0.02, 0.84 ± 0.02 1.87 ± 0.02, 1.78 ± 0.01, 1.51 ± 0.03, 1.45 ± 0.03 1.87 ± 0.01, 1.72 ± 0.01 1.84 ± 0.03, 1.68 ± 0.01, 1.62 ± 0.01, 1.43 ± 0.02, 1.25 ± 0.01, 1.09 ± 0.02, 0.89 ± 0.01, 0.85 ± 0.02, 0.81 ± 0.02, 0.79 ± 0.01, 0.67 ± 0.01 1.85 ± 0.02, 1.71 ± 0.01, 1.63 ± 0.01, 1.43 ± 0.03, 1.32 ± 0.03, 1.27 ± 0.03, 1.19 ± 0.02, 1.07 ± 0.01, 1.04 ± 0.01, 0.80 ± 0.01 1.69 ± 0.02, 1.62 ± 0.02, 0.90 ± 0.01, 0.85 ± 0.01 1.72 ± 0.01, 1.67 ± 0.02, 1.47 ± 0.01

51 ± 0.33, 58 ± 0.21, 57 ± 0.23, 49 ± 0.19, 70 ± 0.27, 64 ± 0.21, 65 ± 0.16 37 ± 0.36, 22 ± 0.18, 55 ± 0.14, 57 ± 0.28 36 ± 0.22, 38 ± 0.22 36 ± 0.32, 59 ± 0.12, 77 ± 0.22, 63 ± 0.25, 55 ± 0.23, 66 ± 0.20, 87 ± 0.19, 125 ± 0.26, 55 ± 0.21, 153 ± 0.11, 158 ± 0.26

interference 2 interference 3 interference 4 interference 5 interference 6 interference 7 a

77 ± 0.14, 105 ± 0.24, 43 ± 0.34, 61 ± 0.24, 71 ± 0.27, 41 ± 0.26, 53 ± 0.16, 41 ± 0.12, 53 ± 0.22, 121 ± 0.28, 62 ± 0.21, 89 ± 0.13 49 ± 0.18, 65 ± 0.14, 67 ± 0.24 56 ± 0.11, 61 ± 0.13

The flow rate and gate pulse width were 3 μL min−1 and 80 μs, respectively.

Figure 3. Representative ESI-HPIMS spectra for seven interferences used in this investigation. Each interference gave a couple of peaks, and peaks whose K0 matched a target compound were considered false-positive peaks. Traces shown indicate different scans from the instrument. Even though scans were different, K0 values did not change for interference peaks.

■

peaks observed at K0 of 2.02 ± 0.01 and 2.33 ± 0.02 cm2 V−1 s−1, respectively. The spectrum for CBD shows a monomer and a dimer peak but the dimer begin to appear at about 70 μs gate pulse width and higher. The monomer and dimer peaks of CBD had K0 of 1.08 ± 0.02 and 0.94 ± 0.01 cm2 V−1 s−1, respectively. In theory, mobility is a measure of the size, shape, and charge on an ion. The K0 value is the ion’s measured mobility adjusted for temperature and pressure. Compared to reduced mobility reported in the literature, the K0 data in Table 1 matched to within 2%. As expected, our results shows that the mobility did not change as a function of gate pulse width. This was in agreement with previous studies.1 In Table 2, K0 for the seven interferences chosen for this study are shown. These K0 values generated with ESI ionization are reported for the first time in the literature. Several of these

RESULTS AND DISCUSSION

Figure 2 shows a typical mobility spectra obtained with the ESIHPIMS. Table 1 list reduced mobility (K0) values, resolving power, and signal-to-noise ratio for 11 compounds studied in the positive ion mode. This investigation further demonstrate the effect of gate pulse width on mobility and resolving power. All K0 values were determined by comparison to known mobility and experimentally determined drift times (td). The reduced mobility values reported in Table 1 were calculated with reference to 2,4-lutidine using eq 1 (see Supporting Information) to correct for variations in temperature and pressure; where 1.95 ± 0.01 cm2 V−1 s−1 was the K0 value used for the standard 2,4-lutidine.41,55 The background spectra for the solvent used in ESI positive mode revealed two reactant ion D

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 4. Diffusion-limited conditional theoretical reduced mobility, DLCTRM (Kd), calculated for drift potential 200 to 5000 V for morphine (a); phentermine, L-phenylepherine, and amphetamine (b); cocaine, cannabidiol, and heroin (c); and pioglitazone, and rosiglitazone (d). The result demonstrate that Kd decreases with increasing drift potential.

that the Rp range for compounds occurred between 39 and 63, and those for interferences occurred between 36 and 158. In high-resolution IMS systems (research-grade instruments) resolving powers of 172 and 150 have been reported.2,56 For such systems, Δtd of 0.03 ms is enough to obtain resolution of two peaks. Observation of eq 2 (see Supporting Information) shows that Rp values are dependent on two factors, td and width-at-half-height (ω0.5); thus, a higher td or lower ω0.5 will result in higher Rp values. The higher Rp values reported for some of the interference peaks can be explained due to the fact that these peaks drifted at a much higher td values. While all the Rp values reported in Tables 1 and 2 were measured at 80 μs gate pulse width, it is important to note that decreasing the gate pulse width resulted in improved Rp values. Of more importance, it was noted that when the gate pulse width was decreased by a factor of 10, the resolving power only improved by a factor of 2 to 4, indicating other peak-broadening factors such as diffusion broadening, ion/molecule interaction broadening, or even charge repulsion broadening besides initial gate pulse width are important and are contributing to the peak broadening process. The signal-to-noise (SNR) ratio measured for the compounds studied at 80 μs gate pulse width are reported in Table 1. This investigation only analyzes SNR at one gate pulse width; however, it was reported previously that increasing gate pulse width does not always equate to improving SNR in an IMS system.1 A diffusion-limited conditional theoretical reduced mobility, DLCTRM (Kd) has been defined in this investigation according to eq 14 (see Supporting Information) and shown below.

values matched the K0 values of target compounds studied in this investigation and cannot be separated from these compounds. These interferences are strong candidates to act as false positives for these compounds. For example, int. #1 has a peak with K0 of 1.08 ± 0.01; this K0 was either similar of close to the K0’s of CBD (1.08 ± 0.02), HER (1.05 ± 0.03), PIOGZ (1.10 ± 0.03), COC (1.14 ± 0.02), and ROSGZ (1.14 ± 0.01). For the same int. #1, the peak at 1.25 ± 0.01 matched the peak of MOR (1.23 ± 0.02) and that at 1.62 ± 0.01 matched the peak of METHAMP at 1.61 ± 0.02. Int. #2 has a K0 of 1.51 ± 0.02; this K0 was either similar or close to the K0’s of PHENT (1.57 ± 0.01), AMP (1.56 ± 0.01), and L-PHENE (1.56 ± 0.02). Int. #4 and #5 has a K0 of 1.25 ± 0.01 and 1.27 ± 0.03, respectively; these K0’s were either similar or close to the K0 of MOR (1.23 ± 0.02). Other examples exist for the data shown in Table 2. In IMS, peaks with similar or closely matched K0 values cannot be separated from each other. Since these interferences are common off-the-shelf cosmetics products, they could present a huge problem as candidates for false-positive responses especially when IMS is used in the field. The significance of this information will be discussed later. Figure 3 shows typical ESI-HPIMS spectra obtained for the seven interferences at 90 μs gate pulse width. In IMS, separation efficiencies are reported as measured resolving power (Rp) and are calculated based on eq 2 (see Supporting Information). Rp for this instrument was estimated at 61 ± 0.22 using the operating parameters from the ESI-HPIMS. Rp values for compounds and interferences detected are shown in Tables 1 and 2, where it can be seen E

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 5. Conditional reduced mobility, CRM (Kc), calculated using reduced mobility and width-at-half-height of each peak. The intensity on the y-axis is ×103 and the unit of Kc is cm2 V−1 s−2. The result demonstrate that Kc can be used to differentiate or separate target compounds from false-positive responses.

Each compound investigated shows an exponential curve when Kd was plotted against voltage. Figure 4a−d show different compounds DLCTRM for a variety of drift potentials ranging from 200 to 5000 V. It can be seen from these plots that the minimum DLCTRM is occurring at higher voltages. Clearly, DLCTRM decreases as drift potential increases. Note that compounds with similar K0 values are shown on the same plot. In addition, the maximum DLCTRM is slightly different for each compound. This is due to slight differences in td values.

the same K0 value as a drug can be differentiated from a drug if the Kc was considered. This hypothesis was demonstrated in Figure 5a−d. Kc =

Ko ω0.5

(10)

where Kc is the conditional reduced mobility, K0 is the reduced mobility, ω0.5 is width-at-half-height, and Kc has a unit of 103 cm2 V−1 s−2. Figure 5a shows a plot of Kc versus gate pulse width for MOR, and false-positive peaks that has a similar K0 value to morphine from int. #1 and int. #5. In routine IMS operation, these three peaks could not be separated from each other. The plot clearly shows that when the Kc values were calculated a clear difference between the peaks was apparent at 30, 40, 50, 90, and 100 μs gate pulse widths. At 40 μs gate pulse width, the difference in Kc’s between MOR and int. #1, MOR and int. #5, and int. #1 and int. #5 was 54, 67, and 33%, respectively. At 50 μs gate pulse width, the difference in Kc’s between MOR and int. #1, MOR and int. #5, and int. #1 and int. #5 was 40, 33, and 67%, respectively. The other gate pulse widths also show some differences in the Kc values but the biggest difference between the peaks was observed at 40 μs gate pulse width. Note that the false-positive peaks for int. #1 and int. #5 was observed at 30 μs gate pulse width and higher. Figure 5b shows a similar plot for AMP, PHENT, L-PHENE, and a false-positive peak from int. #2. From this plot, it can be seen that all four peaks can be separated at 30 and 50 μs gate pulse widths with a

2 ⎛ Wg ⎞ ⎛ ez ⎞1/2 ⎛ L ⎞ ⎛ 273.15 ⎞⎛ P ⎞ ⎟ → 0⎟ = 0.30⎜ ⎟ ⎜ ⎟ ⎜ 3/2 ⎟⎜ Kd = lim K ⎜ ⎝ kV ⎠ ⎝ td ⎠ ⎝ T ⎠⎝ 760 ⎠ ⎝ td ⎠ (14)

where Kd is the diffusion-limited conditional theoretical reduced mobility, L is the length of the drift tube, td is the drift time, k is Boltzmann’s constant, T is the temperature in Kelvin of the drift tube, P is pressure in atmosphere, and q = e × z; e is the electronic charge, and z is the charge on the ion. To more precisely investigate identity of each compound that is expected under a given set of IMS condition, “conditional” reduced mobility, CRM (Kc) values were calculated and introduced. CRM considered both the reduced mobility and the width-at-half height. This new equation is given as eq 10 (see Supporting Information) and shown below. The hypothesis here is that because two peaks with similar K0 values will most certainly not have the same ω0.5, it is possible to use Kc to differentiate between two peaks with matched or slightly different K0 values. Thus, an interference peak that has F

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 6. Representative ESI-HPIMS spectra for interference #1 at gate pulse width of 10 to 100 μs. Two spectra scans are plotted on each panel. At responses below 30 μs gate pulse width, interference peaks disappeared into the background.

below 30 μs gate pulse width because signal-to-noise ratios are so bad below this gate pulse width such that peaks are normally indiscernible. Figure 6 shows a representative ESI-HPIMS example spectra obtained for int. #1 at different gate pulse widths. From this spectra, it can be seen that there are hardly any well-defined peaks at gate pulse width of 10 and 20 μs, where the response ions have disappeared into the background. Example spectrum of CBD and PIOGZ at different gate pulse widths are also shown in Figures S-1 and S-2 (see Supporting Information). A peak that appeared to be a dimer was observed in the CBD spectra but this peak only appeared at 70 μs gate pulse width and higher. The results detailed in this article demonstrates the major point of this paper. For many IMS

clear difference between the peaks observed at 50 μs gate pulse width. For example, at 50 μs gate pulse width, the difference in Kc’s between AMP and int. #2, PHENT and Int. #2, and L-PHENE and int. #2 was 61, 23, and 39%, respectively. Similar conclusive results were obtained for the plots in Figure 5c,d. The plot in Figure 5c shows the peaks for HER, COC, CBD, and a false-positive peak for int. #1, whereas the plot in Figure 5d shows the peaks for PIOGZ, ROSGZ, and false-positive peaks from int. #1, int. #4, and int. #5. In this experiment, it was observed that at higher gate pulse width, ω0.5 were more consistent. Thus, in identifying optimum CRM values to differentiate between a false-positive peak and the true compound, it is recommended the analyst should not operate G

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry instruments developed for field applications, a major problem is the possibility of false positives. These false-positive responses can result in loss of time, and further tests may be required to ensure that target compounds may be absent or present in a given circumstance. By taking advantage of the drifting pattern of two ions and the incorporation of CRM, it may now be possible to differentiate between true false-positive responses from a target compound. These experiments were designed to demonstrate how CRM calculations can be used to differentiate between false-positive responses and a target compound. Equation 10 (see Supporting Information) enabled the calculation of CRM, and it was observed that identity differentiation can be achieved by taking advantage of different gate pulse widths in an IMS system. Thus, Kc, as introduced in this paper, provides a way to rapidly predict where the optimal CRM of an IMS will be and what it will be (or at least what it can be). The CRM or Kc value is a measure of the identity of an ion; a new parameter that may be applied to help aid IMS devices with false-positive issues.

ACKNOWLEDGMENTS

■

REFERENCES

(1) Kanu, A. B.; Gribb, M. M.; Hill, H. H., Jr. Anal. Chem. 2008, 80, 6610−6619. (2) Wu, C.; Siems, W. F.; Hill, H. H. Anal. Chem. 2000, 72, 396−403. (3) Miki, A.; Tatsuno, M.; Katagi, M.; Nishikawa, M.; Tsuchihashi, H. Jpn. J. Forensic Toxicol. 1997, 15, 142−148. (4) Carroll, J. J.; Tri, L.; DeBono, R. Am. Lab. 2004, 36, 32−35. (5) Debono, R.; Stefanou, S.; Davis, M.; Walia, G. Pharmaceutical Technology North America 2002, 26, 72−78. (6) Lawrence, A. H. Anal. Chem. 1989, 61, 343−349. (7) Keller, T.; Keller, A.; Tutsch-Bauer, E.; Monticelli, F. Forensic Sci. Int. 2006, 161, 130−140. (8) Gillen, G.; Mahoney, C.; Wight, S.; Lareau, R. Rapid Commun. Mass Spectrom. 2006, 20, 1949−1953. (9) Luong, J.; Gras, R.; Van Meulebroeck, R.; Sutherland, F.; Cortes, H. J. Chromatogr. Sci. 2006, 44, 276−282. (10) Baumbach, J. I. Anal. Bioanal. Chem. 2006, 384, 1059−1070. (11) Buryakov, I. A. Tech. Phys. Lett. 2006, 32, 67−69. (12) Buryakov, I. A. NATO Science Series, II: The Analytical Characteristics of Ion Mobility Increment Spectrometer during the Detection of Explosive Vapors and Products of Their Degradation. Mathematics, Physics and Chemistry, 2004; Vol 167 (Vapor and Trace Detection of Explosives for Anti-Terrorism Purposes), p 113. (13) Gruznov, V. M.; Baldin, M. N.; Filonenko, V. G. NATO Science Series, II: High-Speed Gas Analysis for Explosives Detection. Mathematics, Physics and Chemistry, 2004; Vol. 167 (Vapor and Trace Detection of Explosives for Anti-Terrorism Purposes), p 87. (14) Marr, A. J.; Groves, D. M. Int. J. Ion Mobil. Spectrom. 2003, 6, 59−62. (15) Neves, J. L.; Haigh, P. E.; Wu, C.; McGann, W. J. Int. J. Ion Mobil. Spectrom. 2003, 6, 1−3. (16) Perr, J. M.; Furton, K. G.; Almirall, J. R. J. Sep. Sci. 2005, 28, 177−183. (17) Asbury, G. R.; Klasmeier, J.; Hill, H. H., Jr. Talanta 2000, 50, 1291−1298. (18) Kanu, A. B.; Haigh, P. E.; Hill, H. H., Jr. Anal. Chim. Acta 2005, 553, 148−159. (19) Tarver, E. E. Sensors 2004, 4, 1−13. (20) Steiner, W. E.; Haigh, P. H.; Clowers, B. H.; Hill, H. H. Anal. Chem. 2003, 75, 6068−6076. (21) Hoaglund-Hyzer, C. S.; Lee, Y. J.; Counterman, A. E.; Clemmer, D. E. Anal. Chem. 2002, 74, 992−1006. (22) Gordon, G.; Pacey, G.; Bubnis, B.; Laszewski, S.; Gaines, J. Chem. Oxid. 1997, 4, 23−29. (23) Kanu, A. B.; Hill, H. H., Jr.; Gribb, M.; Walters, R. J. Environ. Monit. 2007, 9, 51−60. (24) Clowers, B. H.; Steiner, W. E.; Dion, H. M.; Matz, L. M.; Tam, M.; Tarver, E. E.; Hill, H. H., Jr. Field Anal. Chem. Technol. 2001, 5, 302−312. (25) Reategui, J.; Bacon, T. Vapors Proc. Control Qual. 1992, 3, 209− 215. (26) Palmer, P. T.; Limero, T. F. J. Am. Soc. Mass Spectrom. 2001, 12, 656−675. (27) Crawford, C. L.; Hauck, B. C.; Tufariello, J. A.; Harden, C. S.; McHugh, V.; Siems, W. F.; Hill, H. H., Jr. Talanta 2012, 101, 161− 170. (28) Kanu, A. B.; Hill, H. H., Jr. Talanta 2007, 73, 692−699. (29) Kanu, A. B.; Wu, C.; Hill, H. H., Jr. Anal. Chim. Acta 2008, 610, 125−134.

CONCLUSIONS For small field deployable IMS instruments, it is now possible to incorporate conditional reduced mobility to improve on the identity efficiency between target compounds and false-positive responses. This means that, for most instruments, the identity of an ion can be improved by incorporating different gatecontrolled operations. The most important parameter for determining conditional reduced mobility are reduced mobility and width-at-half-height of a peak. There is a specific optimum voltage, gate pulse width, resolving power, and now conditional reduced mobility for each ion. Diffusion-limited conditional theoretical reduced mobility indicate the optimum reduced mobility that is not normally possible under field conditions. Conditional reduced mobility predicts the condition at which a target compound can be differentiated from a false-positive response. This is possible because different ions exhibits different drifting patterns and hence different peak broadening phenomenon inside an ion mobility tube. Reduced mobility for target compounds reported were reproducible to within 2% for the ESI-HPIMS. The estimated resolving power of the ESIHPIMS used in this study was 61 ± 0.22. Conditional reduced mobility introduced in this paper shows differences between target compounds and false-positive peaks as high as 74%, as was the case for cannabidiol and interference #1 at 70 μs gate pulse width. ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b03765. Theoretical considerations and experimental procedure (PDF)

■

■

We gratefully acknowledge the support of Research Initiation Program and Professional Development Committee at WinstonSalem State University. The authors also thank Dr. Anthony Taylor of TOFWERK for his help in putting the schematics together.

■

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel.: +1 336-750-3199. Notes

The authors declare the following competing financial interest(s): This work has proprietary information that is subject to copyright policies. H

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (30) Hill Jr., H. H. Society of Western Analytical Professors (SWAP) Meeting, Utah State University, Logan, UT, January 29−February 1, 1987 (31) Shumate, C.; Hill Jr., H. H. 42nd Northwest Regional Meeting of the American Chemical Society, Bellingham, WA, June 17−19, 1987. (32) Hill Jr., H. H. Symposium on Accuracy in Trace analysis, National Bureau of Standards, Gaithersburg, MD, September 28−October 1, 1987. (33) Hill, H. H., Jr.; Eatherton, R. L. J. Res. Natl. Bur. Stand. 1988, 93, 425−426. (34) Shumate, C. B.; Hill, H. H., Jr. Anal. Chem. 1989, 61, 601−606. (35) Hill, H. H., Jr.; Siems, W. F.; St Louis, R. H.; McMinn, D. G. Anal. Chem. 1990, 62, 1201A−1209A. (36) McMinn, D. G.; Kinzer, J. A.; Shumate, C. B.; Siems, W. F.; Hill, H. H. J. Microcolumn Sep. 1990, 2, 188−192. (37) St. Louis, R. H.; Hill, H. H., Jr.; Eiceman, G. A. Crit. Rev. Anal. Chem. 1990, 21, 321−355. (38) Shumate, C. B.; Hill Jr., H. H. In Pollution Prevention and Process Analytical Chemistry; Breen, J., Dellarco, M., Ed.; ACS Books: Washington, DC, 1992. (39) Hill, H. H., Jr.; St Louis, R. H.; Morrissey, M. A.; Shumate, C. B.; Siems, W.; McMinn, D. G. J. High Resolut. Chromatogr. 1992, 15, 417−422. (40) Kanu, A. B.; Kumar, B. S.; Hill, H. H., Jr. Int. J. Ion Mobility Spectrom. 2012, 15, 9−20. (41) Thomas, C. L. P.; Rezgui, N. D.; Kanu, A. B.; Munro, W. A. Int. J. Ion Mobil. Spectrom. 2000, 5 (1), 31−36. (42) Shumate, C.; St. Louis, R. H.; Hill, H. H., Jr. J. Chromatogr. A 1986, 373, 141−173. (43) Dwivedi, P.; Wu, P.; Klopsch, S. J.; Puzon, G. J.; Xun, L.; Hill, H. H., Jr. Metabolomics 2008, 4, 63−80. (44) Spangler, G. E. Int. J. Mass Spectrom. 2002, 220, 399−418. (45) Siems, W. F.; Wu, C.; Tarver, E. E.; Hill, H. H., Jr.; Larsen, P. R.; McMinn, D. G. Anal. Chem. 1994, 66, 4195−4201. (46) Rokushika, S.; Hatano, H.; Baim, M. A.; Hill, H. H., Jr. Anal. Chem. 1985, 57, 1902−1907. (47) Wu, C.; Siems, W. F.; Asbury, G. R.; Hill, H. H., Jr. Anal. Chem. 1998, 70, 4929−4938. (48) Spangler, G. E.; Vora, K. N.; Carrico, J. P. J. Phys. E: Sci. Instrum. 1986, 19, 191−198. (49) Wittmer, D.; Chen, Y. H.; Luckenbill, B. K.; Hill, H. H., Jr. Anal. Chem. 1994, 66, 2348−2355. (50) Kanu, A. B.; Brandt, S. D.; Williams, M. D.; Zhang, N.; Hill, H. H., Jr. Anal. Chem. 2013, 85, 8535−8542. (51) Krueger, C. A.; Hilton, C. K.; Osgood, M.; Wu, J.; Wu, C. Int. J. Ion Mobility Spectrom. 2009, 12, 33−37. (52) Krueger, C. A.; Midey, A. J.; Kim, T.; Osgood, M.; Wu, J.; Wu, C. Int. J. Ion Mobility Spectrom. 2011, 14, 71−79. (53) Hilton, C. K.; Krueger, C. A.; Midey, A. J.; Osgood, M.; Wu, J.; Wu, C. Int. J. Mass Spectrom. 2010, 298, 64−71. (54) Midey, A. J.; Patel, A.; Moraff, C.; Krueger, C. K.; Wu, C. Talanta 2013, 116, 77−83. (55) Shumate, C.; St. Loius, R. H.; Hill, H. H., Jr. J. Chromatogr. A 1986, 373, 141−173. (56) Dugourd, Ph.; Hudgins, R. R.; Clemmer, D. E.; Jarrold, M. F. Rev. Sci. Instrum. 1997, 68 (2), 1122−1129.

I

DOI: 10.1021/acs.analchem.5b03765 Anal. Chem. XXXX, XXX, XXX−XXX