Transversal Modulation Ion Mobility Spectrometry ... - ACS Publications

Jan 5, 2015 - The new system enables IMS-IMS-MS analysis and other modes of operation: IMS prefiltration, IMS-IMS, and full transmission mode. It prov...
0 downloads 7 Views 3MB Size
Article pubs.acs.org/ac

Transversal Modulation Ion Mobility Spectrometry (IMS) Coupled with Mass Spectrometry (MS): Exploring the IMS-IMS-MS Possibilities of the Instrument G. Vidal-de-Miguel,*,†,‡ M. Macía,† C. Barrios,†,‡ and J. Cuevas† †

Sociedad Europea de Análisis Diferencial de Movilidad SL, P. Tec. Boecillo, Parcela 205, Edificio Cartif, Boecillo, 47151 Valladolid, Spain ‡ Department of Energy Engineering and Fluid Mechanics, Valladolid University, Paseo del Cauce, 50A, 47011 Valladolid, Spain S Supporting Information *

ABSTRACT: A prototype is introduced based on the transversal modulation ion mobility spectrometry (TMIMS) technique, which provides a continuous output of mobilityselected ions, greatly easing the synchronization between different analyzing stages. In the new architecture, two stages of filtration are used to drastically reduce the background produced by one stage alone. Two-stages TMIMS was coupled with two different atmospheric pressure interface mass spectrometers (MS). The new system enables IMS-IMS-MS analysis and other modes of operation: IMS prefiltration, IMSIMS, and full transmission mode. It provides a resolving power R > 60 in IMS mode, and R > 40 in each stage of IMS-IMS mode. 2-Propanol vapors were introduced in one of the stages to enhance the mobility variations, and their effect was studied on a set of tetraalkylammonium ions. We found that concentrations as low as 1% (in partial pressure) produce mobility variations as high as 20%, which suggest that IMS-IMS separation using dried N2 (in one stage) and a dopant (in the other stage), could be a very powerful way to enhance the separation capacity of the IMS-IMS prefiltration approach.

S

that create a continuous (or quasi-continuous) output of prefiltered ions, including field asymmetric ion mobility spectrometry (FAIMS),17−20 differential mobility spectrometry (DMS),9,21−24 differential mobility analyzers (DMA),25,26 overtone mobility spectrometry,27−30 or transversal modulation ion mobility spectrometry (TMIMS),31 more readily facilitate the coupling between IMS prefiltration stage and MS. This is because the requirement to synchronize the fast pulsed output of each IMS with the MS is practically eliminated. Guevremont et al.32 previously reported the advantages of a continuous system using two FAIMS devices in tandem. To evaluate droplet growth and uptake of vapors by aerosol particles, McMurry and co-workers33,34 introduced the DMA-DMA approach. DMA-DTIMS is also a useful tool for these studies.35 Our goal in this study was to explore the viability and possibilities of operation of two-stage TMIMS. The principle of operation for TMIMS is explained elsewhere.31 In brief, it comprises an axial electric field and a transversal oscillating electric field. Ions enter the TMIMS through the inlet slit and are pushed by the axial electric field at a velocity proportional to their mobility, and hence they travel through the TMIMS for a time inversely proportional to their mobility. When this time

election and analysis of ions and charged particles by means of their electrical mobility K (defined as the ratio of electric velocity to electric field) is useful for many applications, including the detection of explosives and pharmacologic, environmental, and biological analysis.1 Ion mobility spectrometry combined with mass spectrometry (IMS-MS) is an emerging and powerful technique that provides additional structural information and increased separation capacity. Both these features prove very useful in the fields of -omics and biological analysis, as shown by different studies.2−6 The rise of commercial IMS-MS systems (i.e., Thermo FAIMS Quantum,7 Waters Synapt,8 ABSciex Selexion,9 and more recently Agilent 6560 IM Q-TOF10) indicates a promising future for the approach. IMS-IMS-MS, first introduced by Clemmer’s group in their pioneering work,11−15 provides an extra dimension of separation, increasing the total separation capacity,14 in addition to very useful structural information. Hill et al.16 also illustrated the potential of the (IMS-MS)2 approach in their study, where a drift tube ion mobility spectrometer (DTIMS) was coupled with a Synapt MS. While these configurations are interesting, pulsed IMS-IMSMS analysis requires careful integration of the different IMS stages with the MS. As the duration of the IMS peaks is in the millisecond range, pulsed IMS-MS systems are only compatible with mass spectrometers with very fast acquisition rates [time of flight (TOF) and quadrupoles]. In contrast, architectures © 2015 American Chemical Society

Received: November 9, 2014 Accepted: January 1, 2015 Published: January 5, 2015 1925

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932

Article

Analytical Chemistry equates to the period of the transversal oscillating electric field, the trajectories of the selected ions are deflected from the central axis during half the cycle of the deflector electric field. They subsequently return to the central axis during the remainder of the cycle. As a result, the selected ions coalesce at the analyzer outlet, while other ions having different mobilities do not. One-stage TMIMS produces a continuous output of mobility-selected ions but also produces a pulsed output of nonselected ions, leading to an averaged nonzero background signal that greatly reduces the dynamic range of the instrument. In theory, two stages operated in quadrature could eliminate this pulsed and undesired output. Accordingly, a further motivation of this study was to test the viability of this hypothesis. Two stages were coupled and synchronized with an offset of 90° in their respective oscillating fields, such that the pulses produced in the first stage are synchronized with the nontransferring intervals in the second stage and are therefore eliminated. To evaluate the viability of the new hyphenated system, the resulting IMS prefilter was coupled with an API-MS (API 3200, from ABSciex), and an LTQ-XL (Thermo Fisher). The new architecture enables IMS prefiltration and also IMSIMS prefiltration. In order to investigate the different operational possibilities of the TMIMS, we first studied the new architecture with a set of tetraalkylammonium ions and with a combination of nitrogen (produced by a N2 generator) and N2 doped with 2-propanol. In an independent set of experiments, an electrospray of Ultramark, the peptide MRFA, caffeine, and tetraheptylammonium (THA) bromide was used to evaluate the transmission of our instrument.

Figure 1. Schematic illustration of the TMIMS-MS experimental setup including an ESI source, focusing plate, two TMIMS stages, deflector electrodes, inlet and intermediate electrodes, and outlet electrode that matches the inlet orifice of an API-MS. The figure also illustrates the architecture of the flow controls and the electronics (low and high voltage) used to control the TMIMS.



EXPERIMENTAL SECTION Description of the Instrument. Figure 1 shows schematically the architecture of the new tandem TMIMS. It is composed of two symmetrical insulator boxes, the first (stage 1) housing the inlet electrode and the second (stage 2) housing the outlet electrode. Each box houses two cylindrical deflector electrodes. Each stage is 5 cm long, the diameter of the deflector electrodes is 3 cm, and their centers are 7 cm apart. The slit of the inlet electrode (inlet slit) connects the first stage with a gastight nanoelectrospray (nanoESI) chamber. A focusing plate with a 4 mm wide slit is located between the n-ESI tip and the inlet slit so as to guide the ions toward the inlet slit. The intermediate electrode consists of a thin plate (0.5 mm thick) that separates the two stages and incorporates a slit (intermediate slit) that allowed ions to be transferred to the second stage. To improve the coupling with the following APIMS, the outlet electrode incorporates a slit that is elongated on the side receiving the selected ions, and smoothly transitions toward a rounded orifice on the opposite side of the electrode. Two interchangeable outlet electrodes were fitted to the orifice plate of the API 3200 and the inlet of an LTQ-XL. The interface between the outlet electrode and the inlet of the MS incorporates an O-ring, which ensures that the gas sampled by the API system is drawn only from the TMIMS. The outlet electrodes of the TMIMS match the shape of the ion sources and are easily assembled with the API-MS. All slits are 0.6 mm wide and 1 cm long. In order to control the gas composition in each stage, two independent lateral inlets (equipped with laminarizing meshes in order to prevent turbulence) served to introduce a controlled flow of gas into each TMIMS chamber. Each chamber was also

equipped with a secondary outlet to continuously renew the gas in each stage. In order to minimize the flow of gases passing through the intermediate slit (which would otherwise lead to uncontrolled gas compositions in each chamber), the secondary outlets were designed to minimize their pressure drop and were connected to a common outlet. A flow of 2 liters per minute (lpm) of pure nitrogen (produced by a nitrogen generator, 99.5% purity) was introduced in stage 2, while the gas introduced in stage 1 (2 lpm) was previously doped with 2-propanol in a humidifier. This allowed the concentration to be controlled in the range from 0% to 1%. We chose to use 2-propanol in view of previous studies that used polar dopants to enhance nonlinear mobility effects in a FAIMS.36−38 It is hypothesized that the dopants strongly reduce the mobility at relatively low electric fields due to chemical interactions between the ions and the dopant molecules. A fraction of the gas inputted in stage 2 was ingested by the MS toward the vacuum side, and ions reaching the outlet slit were directly carried by this flow. Since the passage of gas through the intermediate electrode was minimized (to ensure that no disrupting jets were formed in the intermediate slit), ions were pushed through this slit only by the local electric fields. A fraction of the gas inputted in stage 1 gas exited through the inlet slit so as to prevent droplets from entering the analyzer. A flow of heated and dry gas was also introduced in the electrospray chamber to assist the desolvation of ions. The drying gas and the counterflow gas were measured through a flow meter, and the gas outputted through the common outlet 1926

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932

Article

Analytical Chemistry was controlled by a flow meter followed by a needle valve and a suction pump. The voltages required by the inlet electrode (16 kV) and the intermediate electrode (8 kV) were supplied by two Applied Kilovolts high-voltage amplifiers (HVA), while the outlet electrode was electrically connected with the MS inlet. The deflector electrode voltages were supplied by four Matsusada high-voltage and high-speed amplifiers (HVHSA). The electrospray voltage (2.5 kV over the focusing plate) and the voltage applied at the focusing plate (2 kV) were provided by two manually controlled EMCO power supplies powered by two batteries that floated above the inlet electrode voltage. In order to produce the control signals required by the HVHSA, a signal generator produced a sinusoidal signal, which was first filtered to eliminate its stationary component and then fed to an angular offset generator.39 This provided four waves with the same amplitude and frequency but with 90° offsets. These waves were then biased with four stationary signals produced by a data acquisition system (DAS), which also generated the signals used to command the axial voltages. Each wave signal was passed through a commuted low-pass filter that allowed us to either pass the complete signal or to filter out the time-varying component of the signals before commanding the HVHSA. As a result, these filters allowed selective switching (on and off) of the oscillating voltage for each deflector electrode while maintaining the steady components of the electric fields. The signal generator and the DAS were controlled by a PC that incorporated the software required to control the frequency and amplitude of the oscillating fields and the stationary voltage component of each electrode. The MS acquired the mass-analyzed ion signal versus time. The frequency and voltage of operation of the TMIMS were controlled by an independent computer and stored as a function of time. Finally, spectra were reconstructed upon synchronization of the data stored in the computers controlling the TMIMS and the MS. Samples and Mass Spectrometers. In the first set of experiments, the TMIMS was coupled with an API3200, and a solution of tetraalkylammonium salts of different chain length, which ranged from tetrapropyl (C3) to tetradodecyl (C12), was electrosprayed (methanol−water 9:1 in volume, 10 μM of each salt) at the inlet of the TMIMS. The frequency and voltages of the TMIMS electrodes were swept in various strategies, the signal produced by each type of ion was acquired by the MS (used as a single quadrupole), and the corresponding spectra were reconstructed and analyzed in order to evaluate the performance of the new architecture. In the second set of experiments, the TMIMS was coupled with an LTQ-XL, and a solution of Calmix (the LTQ ESI positive ion calibration solution is a mixture of caffeine, MRFA, and Ultramark 1621 in an acetonitrile/methanol/acetic solution provided by Pierce protein biology products) and tetraheptylammonium bromide (0.1 μM), used as an internal mobility calibrant, was electrosprayed. Safety Considerations. High voltages are applied to the TMIMS and the electrospray. Care must be taken to avoid electric discharges and injury.

Figure 2. Frequency spectra of different tetraalkylammonium ions analyzed in IMS mode. Each color represents the normalized intensity that was measured in each of the mass channels selected by the quadrupole.

The frequency of the oscillating electric field was scanned (200−800 Hz) with the API3200 operated in single-quadrupole mode, measuring only the signal intensities produced by the masses of the tetraalkylammonium (TAA) ions. Figure 2 shows the signal intensity of the different TAA ions as a function of the frequency of the TMIMS (since each ion produced a very different signal level and to facilitate the visualization of data, each spectrum of Figure 2 is normalized to the maximum intensity produced by each type of ions. Note also that overtone peaks have been removed to ease visualization). The average resolving power (full width at half-height) of these spectra was 66. According to the separation principle of the TMIMS,31 ions are transferred when their time of residence within the TMIMS separation cell equates with the period of the oscillating field. This, in first approximation, implies that the frequency at which the ions are transferred is proportional to their mobility. A linear regression, in which we compared the frequencies of the peak centroids (Figure 2) and the reduced mobility of the set of tetraalkylammonium ions (previously measured by Adamov et al.40 using a linear drift tube), shows that the relationship between mobility and frequency is indeed linear: The regression follows K = 1.952 32 × 10−3f, where K is the mobility in cm2/s·V, f is the frequency of the TMIMS in hertz, and the standard error is 0.4%. Since the deflector voltage is oscillating, the absolute magnitude of the fields also varies, ranging from 1.6 up to 2.3 kV/cm. At atmospheric pressure, these fields correspond to a range of 6.6−9.3 Td (Townsends), which can still be considered a low-field regime.24 Figure 2 also shows that the background signal levels were much lower than when only one stage is utilized.31 In order to better evaluate the dynamic range of the new instrument, Figure 3 shows three spectra of tetraheptylammonium (THA+) ions in logarithmic scale (signal intensity vs frequency of



RESULTS AND DISCUSSION IMS Mode and Shape of the Spectra. Figure 2 shows the results of experiments where the two stages were operated with the same gas (N2) and with the same axial voltage (8 kV in each stage) and synchronized with each other. 1927

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932

Article

Analytical Chemistry

when the two TMIMS stages were operated with their respective oscillating fields in quadrature. In agreement with eq 2, this figure illustrates how the two stages can eliminate most of the pulsed output when synchronized. One could postulate that the extra peaks in the spectrum could be produced by different clusters of THA+, which are then separated in the TMIMS and declustered in the API interface, thus appearing at the mass of the dried THA+ ions. However, the mobilities at which these peaks appear do not match any of the previously reported clusters. And, what is more intriguing, the peak appearing near 400 Hz did not appear when only one stage was used. These results suggest that these peaks could be an artifact produced by the instrument. Note in this regard that eq 2 holds only in the proximity of the main peak. If these peaks are ignored, the signal to background ratio of stages 1 and 2 alone is near 102, while the signal to background ratio of the IMS mode is between 104 and 105 (100−1000 times better). IMS-IMS Operation. In order to test the viability of operating the TMIMS with two different media (which hypothetically should increase the separation capacity of the instrument), we introduced N2 in one stage and doped the other with 2-propanol, hypothesizing that this might improve orthogonality regarding mobility measurements in tandem with N2. To eliminate the undesired pulsed output, the two TMIMS stages were operated at the same frequency, in quadrature, and with the same oscillating amplitude. Two-dimensional mobility scans were acquired by scanning the frequency of the oscillating field and the voltage of the intermediate electrode (Vint), providing control of the ratio of axial voltages applied in the first and second stages (Vint was stepped; at each Vint step, the frequency was scanned and the intensity was acquired for different channels of the MS at each pair of f−Vint values). Ions were transferred when their time of residence through each stage was equal to the period of the oscillating fields. Although the mobility of the ions was different in each stage (because we introduced different gases), at the appropriate value of Vint their velocities were the same in both stages. As a result, ions were transferred through the outlet of the TMIMS for only a given Vint (the voltage for which the time of residence of the ions was equal in the two stages) and for a given f (the frequency for which the time of residence equates the period of the oscillating fields). Since the mobility was proportional to the frequency of operation and if it is assumed to be inversely proportional to the axial voltage in each stage, the data in the f−Vint domain can be transformed to the K1−K2 domain (mobility in stage 1 and mobility in stage 2):

Figure 3. TMIMS tetraheptylammonium ion spectra. In the first two spectra, the TMIMS was operated with the oscillating electric field applied only in one stage at a time: stage 1 (blue) and stage 2 (red). The purple line represents the spectra acquired when the two stages were operated in quadrature. (Inset) Trajectories of the ions within the TMIMS under experimental operating conditions.

TMIMS). The blue and red lines correspond to the signals measured when the oscillating electric field was applied to each stage individually (the other stage passed all ions regardless of their mobility). The theoretical transmission probability function (s) of one stage is defined in our previous work.31,41 In particular, if diffusional effects are neglected, s follows s=

⎡ ⎤ 2 ω arcsin⎢ ⎥ ⎣ k sin(dvω) ⎦ π

(1)

where ω and k are dimensionless parameters, ω = πf l /(KVa) and k = dVa/(lVd), where f is the frequency of operation of the TMIMS, l and Va are respectively the distance and voltage drop between axial electrodes, d is the width of the ion beam as it passes through the slits, Vd is the voltage amplitude of the deflector voltage, and K is the mobility of the ions. In the proximity of ω = π, diffusion smooths the peaks. The high background level measured when only one stage is functioning is produced by the pulsed output of undesired ions. When two stages are operated in series, each stage passes the selected ions continuously, producing a pulsed output of nonselected ions. The resulting transmission of the two stages is the interference between the two pulsed outputs, which, when the two stages operate in quadrature, yields the following expression in the proximity of ω = π: 2

⎧4 ⎡ ω ⎤ ⎪ arcsin⎢ ⎥ − 1 for s > 1/2 ⎣ k sin(ω) ⎦ s2 = ⎨ π ⎪ ⎩ 0 for s < 1/2

(2)

K1 = ω1 f

l2 Vin − Vint

(3a)

K 2 = ω2 f

l2 Vint − Vout

(3b)

where l is the distance between axial electrodes, Vin and Vout are respectively the voltages of inlet and outlet electrodes, and ω1 and ω2 are calibration constants. We used the THA+ ions as a calibrant.42 Figure 4 shows the contour plots, in the K1−K2 domain, of the normalized signals measured by the MS in the corresponding mass channels of tetraalkylammonium ions (where the signal intensity of each mass channel was

It is worth noting that the background signal is not reduced by a power of the number of stages. Instead, this equation predicts a dramatic reduction (although the resolving power is not significantly improved because it is mainly limited by diffusional broadening, which scales with the square of the total time of residence; diffusional broadening effects are detailed elsewhere31). The purple line in Figure 3 shows the signal acquired 1928

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932

Article

Analytical Chemistry

Figure 5. Three-dimensionsl projection of IMS-IMS spectrum (normalized signal versus K1 and K2) produced by tetraheptylammonium ions in N2 and N2 doped with 1% 2-propanol.

Figure 4. Contour plot of IMS-IMS spectra of tetraalkylammonium ions; the dashed line corresponds to the axis K1=K2. Blue contours represent peaks measured when the two stages were filled with N2; pink contours represent peaks measured when 2-propanol was introduced in the first stage.

pair. We used a uniform discretization with 105 points (500 frequency values and 200 voltage values, requiring a total of 5.5 h). The IMS-IMS spectrum has, as expected, a peak at the mobilities of the ion. However, the spectrum also presents two more features, which are observed in the spectra for all ions: (i) vertical and horizontal tails that form a cross-shaped pattern at the base of the peak and (ii) two oblique traces that perpendicularly cross the axis K1K2. When one stage passes the ions of its corresponding mobility, the other stage receives a continuous flow of ions and hence produces a pulsed output, leading to the nonzero signal observed in the cross-shaped tails. In IMS operation mode, for which only the axis K1K2 is evaluated, these tails are avoided. However, the oblique traces would appear in the IMS spectra (one at each side of the main peak). These traces, which show a very coherent pattern, confirmed that the secondary peaks observed in the spectrum of Figure 3 are indeed an artifact of the instrument (likely produced by the interference between the two stages). Effect of 2-Propanol on Tetraalkylammonium Ions. In this part of the study we introduced increasing concentrations of 2-propanol in the first stage and measured the peak positions in the frequency spectra. The second stage was operated in full transmission mode (no oscillating electric fields were applied here) to allow for the passage of all ions irrespective of their mobility. This reduces the analysis time, with the acquisition of each spectrum now consisting of only a frequency scan, requiring less than 2 min). Figure 6 shows the measured absolute mobility of the different TAA ions at increasing concentrations of 2-propanol, where we assumed that the mobility was proportional to the frequency of the TMIMS oscillating voltage. Adding just 1% (in volume) 2-propanol (the saturation concentration of 2propanol at room temperature is approximately 6% in volume) reduces mobility by 20% for the smallest ions (tetraethylammonium) by more than 5% for the largest ions (tetradodecylammonium). According to the pure hard-sphere model,43 the mobility variation produced by 2-propanol should be below 1% for the concentrations utilized. The fact that we were able to

normalized to the maximum signal measured in each channel). The contour levels correspond respectively to 80%, 50%, and 20% of the maximum signal. In short, Figure 4 represents the normalized IMS-IMS spectra of the TAA ions. The straight line of this figure represents the condition K1 = K2 (ions having the same mobility in the two stages of the TMIMS). The blue and purple plots correspond to two different IMS-IMS acquisitions: (blue) the same gas in two stages and (pink) 2-propanol in the first stage. In view of these experiments, it is clear that the dopant produced mobility shifts that were easily distinguishable by the resolving power of the instrument. We hypothesize that the IMS-IMS approach with a combination of dry gas and doped gas could be used to provide enhanced separation capacity. However, the exact mechanism by which dopants modify the mobility of the ions is, to our knowledge, not fully understood. According to the cluster model, dopant molecules and ions form a cluster that has a different mobility than the naked ion. If this were the case, ions passing from the first stage (which incorporated the dopant) to the second stage would arrive with some dopant molecules attached and would be desolvated within the second stage. The fact that the resolving power of the second stage is not affected by the presence of dopants in the previous stage suggests that ions are desolvated quickly, right as they reach the second stage. Otherwise, one would expect to observe a blurred spectrum since the mobilities of the ions would change at some random point in their trajectories and would not be refocused. Considering that the time of residence of the ions within the TMIMS is approximately 2 ms, and that voltage variations as low as 2.5% already produce observable changes in the resolving power of the instrument,31 we estimate that the desolvation of the ions must take place on a time scale below 50 μs. Figure 5 shows a 3D projection of the spectrum of THA+ ions (normalized signal vs K1 and K2). The spectra of this figure correspond to MS data acquired during 200 ms for each f−Vin 1929

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932

Article

Analytical Chemistry

observe much higher mobility variations clearly indicates that the interaction between dopant and ions is more complex. Our experiments show that smaller and faster ions are affected more by the presence of 2-propanol. Interestingly, except for the smallest ions, the mobility decreases linearly with the concentration of 2-propanol, as illustrated in Figure 6. For the general case of a charged particle drifting through a gas with binding dopants, Oberreit et al.35 proposed a model that links the average mobility of the resulting cluster (which can vary in size as it drifts through the gas) with the instantaneous mobilities of clusters formed by the particle and g attached dopant molecules, with dimensionless equilibrium constants kg = ng/ng−1, where g is the number of dopant molecules and ng and ng−1 are the concentrations of clusters with g and g − 1 attached dopant molecules. This model provides a universal formulation, but its applicability is limited by the infinite number of parameters involved in the equations. Interestingly, since the TAA ions desolvate very quickly, one can expect the constants kg to be very small. At the limit of small kg, the most abundant clusters have only one vapor molecule attached, and the expressions can be simplified and linearized (see Supporting Information): K = Kd +

α (K c − K d ) β

(4)

where Kd, Kc, and K are the mobilities of dried ion and pair ion−dopant molecule and the macroscopic observed mobility, respectively. The parameter α is the sorption reaction rate (inverse of the mean time required to form an ion−vapor

Figure 6. Mobility of tetraalkylammonium ions in N2 at increasing concentrations of 2-propanol.

Table 1. Transversal Modulation Ion Mobility Spectrometry Signal Intensitiesa signal ratio (%)

measured signal intensity (neutral loss) compd

m/z (Da)

O: MS

T: TMIMS-MS, full transmission

caffeine MRFA C20H19O6N3P3F28

195.2 524.2 1022.8

6.8 × 106 3.4 × 105 2.1 × 105

6.4 × 104 7.5 × 103 5.0 × 103

C22H19O6N3P3F32

1121.8

4.1 × 105

1.0 × 104

C24H19O6N3P3F36

1221.8

6.5 × 105

1.8 × 104

C26H19O6N3P3F40

1322.0

8.0 × 105

1.9 × 104

C28H19O6N3P3F44

1421.7

1.0 × 106

2.2 × 104

C30H19O6N3P3F48

1521.8

8.1 × 105

1.5 × 104

C32H19O6N3P3F52

1622.0

6.8 × 105

9.1 × 103

C34H19O6N3P3F56

1721.8

4.9 × 105

4.9 × 103

C36H19O6N3P3F60

1821.7

3.5 × 105

2.7 × 103

C38H19O6N3P3F64

1922.0

1.5 × 105

1.1 × 103

I: TMIMS-MS, IMS mode

f (Hz)

T/O

I/O

est transmission (%)

× × × × × × × × × × × × × × × × × × × × × ×

453 512 375 427 361 405 348 385 336 368 325 353 315 339 304 327 296 315 288 306 281 296

0.95 2.2 2.4

0.06 1.4 0.8 2.4 1.1 1.9 1.2 1.8 1.0 1.3 0.8 1.3 0.8 1.4 0.6 1.1 0.4 0.9 0.2 0.8 0.2 0.7

0.06 1.4 3.2

4.0 4.7 1.7 5.0 4.5 7.8 7.9 1.2 7.6 1.0 8.5 1.3 6.2 1.1 4.3 7.2 2.0 4.2 8.2 2.8 2.5 9.6

103 103 103 103 103 103 103 104 103 104 103 104 103 104 103 103 103 103 102 103 102 102

2.6 2.7 2.4 2.2 1.9 1.3 1.0 0.8 0.8

3.0 3.0 2.3 2.1 2.2 1.7 1.3 1.0 0.9 2.1b

avg a

For each Ultramark ion [fluorinated phosphacenes that comprise a phospacene (NPO2)3 ring and six chains CH2(C2F4)nH, each attached to one O] listed, the table shows mass, signal intensity measured without TMIMS and with TMIMS operated in full transmission and IMS modes, frequencies of mobility peaks of the different isomers observed, signal ratios (full transmission over MS alone and IMS mode over MS alone), and estimated transmission in IMS mode. bAveraged estimated transmission. 1930

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932

Article

Analytical Chemistry

different species. For instance, if a dopant were particularly attracted by a specific functional group, it could produce a strong mobility shift (with dopant versus without) for those species having the specific functional group. Hence the presence of these functional groups could be (hypothetically) identified by these mobility shifts. Performing multidimensional IMS-IMS-MS analysis with scannable filters requires very long acquisition times. Nevertheless, because each TMIMS stage can also operate in full transmission mode, the peak identification process can be facilitated, much as in the triple-quadrupole architectures, in which the quadrupoles can be scanned, operated at a fixed mass, or operated to allow the passage of all ions, so as to sequentially analyze spectra of increasing complexity. The TMIMS can be easily coupled with any MS with an APIMS without modification of the interface or electronics of the MS. However, the transmission measured was suboptimal (in first approximation, it should approach 100%). The instrument shows interesting potential, but it still requires some important improvements.

molecule pair, which is proportional to the concentration of dopant molecules) and β is the desolvation reaction rate (inverse of the mean lifetime τc of the cluster). This specific case of the universal model explains the linear trend that we observed: When a dry ion (with mobility Kd) collides with a dopant molecule, it forms a cluster (pair ion− vapor molecule) that has a lower mobility (Kc), and a short lifetime (τc). If the cluster is dried before colliding with a new dopant molecule (low kg limit), the macroscopic trajectory will be the result of small displacements of the dried ions, followed by additional small displacements of subsequent ion−dopant pair clusters. Since the accumulated time spent as a cluster would be proportional to the number of collisions with vapor molecules (and hence to the concentration of vapor molecules), the averaged observable mobility falls linearly with the concentration of dopants. For eq 4 to hold, τc must be significantly lower than the time between collisions of the ions with the dopants (τd). According to this model, if the concentration of dopants is sufficiently high (when τd and τc are similar), more complex clusters will be formed, and saturation effects would cause the macroscopic measured mobility to deviate from eq 4. The nonlinear behavior observed for the smaller ions could thus be explained as a result of a longer lifetime of the resulting smaller clusters. IMS Mode, Transmission. The transmission was estimated by comparing the signals acquired by an LTQ-MS before and after installation of the TMIMS. The frequency of the TMIMS was scanned from 10 to 700 Hz, and the LTQ acquired the spectra (from 100 to 2000 Da) produced at each frequency. A mixture of Ultramark, described elsewhere,44 MRFA, caffeine, and tetraheptylammonium (THA) bromide was electrosprayed. The different signal intensities are listed in Table 1. For Ultramark, two peaks in the mobility spectrum were observed for each mass (the secondary peak having on average 50% of the intensity of the primary peak). The averaged signal ratio for these ions was 1.3% (for the major peak), and 2% in full transmission mode. We hypothesize that the second mobility peaks correspond to isomeric forms of Ultramark. The transmission in IMS mode is estimated as 2% (sum of relative intensities of the two mobility peaks). This is in consonance with the transmission observed in full transmission mode.



ASSOCIATED CONTENT

S Supporting Information *

Text and equations for a particularized model of the universal heterogeneous vapor uptake model for the limit of very low clustering reaction ratios. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*Telephone +34 983 130 154; fax +34 983 130 411; e-mail [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS We thank A. Makarov and D. Nolting (Thermo Fisher Scientific, Bremen) for many fruitful conversations and insights. This work was developed under the project VEAME DOBLE, funded by the Plan AVANZA (Spanish Government). We acknowledge Thermo Fisher Scientific for the loan of the LTQ mass spectrometer.



CONCLUSIONS Operating two TMIMS in quadrature eliminates most of the undesired pulsed output of ions produced by a single TMIMS stage and improves the dynamic range by approximately 2 orders of magnitude. However, our results show that the interaction between the two stages still produces undesired secondary peaks, which could complicate the interpretation of spectra. Furthermore, employing two stages also enables IMS-IMS analysis and/or prefiltration with a continuous output. We anticipate that this prefiltration mode will be useful for the detection of targeted species in highly complex matrices. Our preliminary experiments with dopants showed that the addition of 1% 2-propanol in one of the TMIMS stages produces strong mobility variations (up to 20%). The resolving power measured in IMS mode was R > 60, and R > 40 in each stage in IMS-IMS mode. Determination of the separation capacity achieved with the IMS-IMS approach requires further evaluation. Nevertheless, the use of dopants in IMS-IMS offers a new parameter of control, as different dopants could have different chemical interactions that potentially vary with



REFERENCES

(1) Eiceman, G. A.; Karpas, Z. Ion Mobility Spectrometry, 2nd ed.; CRC Press: Boca Raton, FL, 2004. (2) Gotts, N. G.; von Helden, G.; Bowers, M. T. Int. J. Mass Spectrom. Ion Processes 1995, 149−150, 217−229. (3) Bohrer, B. C.; Merenbloom, S. I.; Koeniger, S. L.; Hilderbrand, A. E.; Clemmer, D. E. Annu. Rev. Anal. Chem. 2008, 1, 293−327. (4) Jarrold, M. F. Annu. Rev. Phys. Chem. 2000, 51, 179−207. (5) Koeniger, S. L.; Merenbloom, S. I.; Clemmer, D. E. J. Phys. Chem. B 2006, 110, 7017−7021. (6) Shvartsburg, A. A.; Li, F.; Tang, K.; Smith, R. D. Anal. Chem. 2006, 78, 3304−3315. (7) Barnett, D. A.; Belford, M.; Dunyach, J.-J.; Purves, R. W. J. Am. Soc. Mass Spectrom 2007, 18, 1653−1663. (8) Giles, K.; Wildgoose, J. L.; Langridge, D. J.; Campuzano, I. Int. J. Mass Spectrom. 2010, 298, 10−16. (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) Kurulugama, R.; Mordehai, A.; Sanders, N.; Klein, C.; Ibrahim, Y.; Baker, E.; Smith, R.; Stafford, G.; Fjeldsted, J. 61st ASMS 1931

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932

Article

Analytical Chemistry

(44) Longfei, J.; Mehdi, M. J. Am. Soc. Mass Spectrom. 1992, 3, 842− 846.

Conference on Mass Spectrometry and Allied Topics, Minneapolis, MN, June 9−13, 2013; Poster WP 753. (11) Merenbloom, S. I.; Koeniger, S. L.; Valentine, S. J.; Plasencia, M. D.; Clemmer, D. E. Anal. Chem. 2006, 78, 2802−2809. (12) Koeniger, S. L.; Merenbloom, S. I.; Valentine, S. J.; Jarrold, M. F.; Udseth, H. R.; Smith, R. D.; Clemmer, D. E. Anal. Chem. 2006, 78, 4161−4174. (13) Koeniger, S. L.; Bohrer, B. C.; Valentine, S. J.; Clemmer, D. E. Anal. Chem. 2008, 80, 1918−1927. (14) Merenbloom, S. I.; Bohrer, B. C.; Koeniger, S. L.; Clemmer, D. E. Anal. Chem. 2007, 79, 515−522. (15) Bohrer, B. C.; Clemmer, D. E. Anal. Chem. 2011, 83, 5377− 5385. (16) Li, H.; Bendiak, B.; Siems, W. F.; Gang, D. R.; Hill, H. H. Anal. Chem. 2013, 85, 2760−2769. (17) Shvartsburg, A. A.; Tang, K.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2004, 15, 1487−1498. (18) Shvartsburg, A. A.; Tang, K.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2005, 16, 2−12. (19) Guevremont, R. J. Chromatogr A 2004, 1058, 3−19. (20) Purves, R. W.; Guevremont, R. Anal. Chem. 1999, 71, 2346− 2357. (21) Shvartsburg, A. A.; Tang, K.; Smith, R. D.; Holden, M.; Rush, M.; Thompson, A.; Toutoungi, D. Anal. Chem. 2009, 81, 8048−8053. (22) Shvartsburg, A. A.; Danielson, W. F.; Smith, R. D. Anal. Chem. 2010, 82, 2456−2462. (23) Schneider, B.; Covey, T.; Coy, S.; Krylov, E.; Nazarov, E. Eur. J. Mass Spectrom. 2010, 16, 57. (24) Schneider, B. B.; Covey, T. R.; Coy, S. L.; Krylov, E. V.; Nazarov, E. G. Anal. Chem. 2010, 82, 1867−1880. (25) De la Mora, J. In Ion Mobility Spectrometry - Mass Spectrometry; Wilkins, C., Trimpin, S., Eds.; CRC Press: Boca Raton, FL, 2010; pp 105−136. (26) Rus, J.; Moro, D.; Sillero, J. A.; Royuela, J.; Casado, A.; EstevezMolinero, F.; Fernández de la Mora, J. Int. J. Mass Spectrom. 2010, 298, 30−40. (27) Kurulugama, R. T.; Nachtigall, F. M.; Lee, S.; Valentine, S. J.; Clemmer, D. E. J. Am. Soc. Mass Spectrom. 2009, 20, 729−737. (28) Valentine, S. J.; Stokes, S. T.; Kurulugama, R. T.; Nachtigall, F. M.; Clemmer, D. E. J. Am. Soc. Mass Spectrom. 2009, 20, 738−750. (29) Valentine, S. J.; Kurulugama, R. T.; Clemmer, D. E. J. Am. Soc. Mass Spectrom 2011, 22, 804−816. (30) Kurulugama, R. T.; Nachtigall, F. M.; Valentine, S. J.; Clemmer, D. E. J. Am. Soc. Mass Spectrom. 2011, 22, 2049−2060. (31) Vidal-de-Miguel, G.; Macia, M.; Cuevas, J. Anal. Chem. 2012, 84, 7831−7837. (32) Guevremont, R.; Ding, L.; Ells, B.; Barnett, D. A.; Purves, R. W. J. Am. Soc. Mass Spectrom. 2001, 12, 1320−1330. (33) Rader, D. J.; McMurry, P. H. J. of Aerosol Science 1986, 17 (5), 771−787. (34) Park, K.; Dutcher, D.; Emery, M.; Pagels, J.; Sakurai, H.; Scheckman, J.; Qian, S.; Stolzenburg, M. R.; Wang, X.; Yang, J.; McMurry, P. H. Aerosol Sci. Technol. 2008, 42 (10), 801−816. (35) Oberreit, D. R.; McMurry, P. H.; Hogan, C. J. Phys. Chem. Chem. Phys. 2014, 16, 6968−6979. (36) Mäkelä, J. M.; Riihelä, M.; Ukkonen, A.; Jokinen, V.; Keskinen, J. J. Chem. Phys. 1996, 105, 1562−1571. (37) Nadykto, A. B.; Mäkelä, J. M.; Yu, F.; Kulmala, M.; Laaksonen, A. Chem. Phys. Lett. 2003, 382, 6−11. (38) Yu, F. J. Chem. Phys. 2005, 122, 84503. (39) Bedrosian, S. D. IRE Trans. Circuit Theory 1960, 7, 128−136. (40) Adamov, A.; Mauriala, T.; Teplov, V.; Laakia, J.; Pedersen, C. S.; Kotiaho, T.; Sysoev, A. A. Int. J. Mass Spectrom. 2010, 298, 24−29. (41) Vidal-De-Miguel, G. Method and apparatus to produce steady beams of mobility selected ions via time dependent electric fields. European Patent EP 2237307 A3, January 4, 2012. (42) Ude, S.; de la Mora, J. F. J. Aerosol Sci. 2005, 36, 1224−1237. (43) Fuchs, N. A. The Mechanics of Aerosols; Dover Publications: Mineola, NY, 1989. 1932

DOI: 10.1021/ac504178n Anal. Chem. 2015, 87, 1925−1932