SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL
  4. RESULTS AND DISCUSSION
  5. CONCLUSIONS
  6. Acknowledgements
  7. REFERENCES

RATIONALE

Mixtures of ions produced in sources at atmospheric pressure, including chemical ionization (APCI) and electrospray ionization (ESI) can be simplified at or near ambient pressure using ion mobility based filters.

METHODS

A low-mobility-pass filter (LMPF) based on a simple mechanical design and simple electronic control was designed, modeled and tested with vapors of 2-hexadecanone in an APCI source and with spray of peptide solutions in an ESI source. The LMPF geometry was planar and small (4 mm wide × 13 mm long) and electric control was through a symmetric waveform in low kHz with amplitude between 0 and 10 V.

RESULTS

Computational models established idealized performance for transmission efficiency of ions of several reduced mobility coefficients over the range of amplitudes and were matched by computed values from ion abundances in mass spectra. The filter exhibited a broad response function, equivalent to a Bode Plot in electronic filters, suggesting that ion filtering could be done in blocks ~50 m/z units wide.

CONCLUSIONS

The benefit of this concept is that discrimination against ions of high mobility is controlled by only a single parameter: waveform amplitude at fixed frequency. The effective removal of high mobility ions, those of low mass-to-charge, can be beneficial for applications with ion-trap-based mass spectrometers to remove excessive levels of solvent or matrix ions. Copyright © 2013 John Wiley & Sons, Ltd.

Measurements by mass spectrometry (MS) have been significantly enabled for proteomics, pharmacology, and biomolecule research through electrospray ionization (ESI) and matrix-assisted laser desorption/ionization (MALDI).[1-5] Despite the broad benefits and widespread acceptance of these sources, ions derived from solvents and matrices of samples can often produce complex mixtures of ions at high abundance principally with m/z values below 250. Such ions may increase the complexity of spectral interpretations or affect sensitivity from space charge effects in ion-trap mass spectrometry.[6] These unwanted ions have been considered chemical noise and degrade signal-to-noise (S/N) ratios particularly for substances with molecular masses from 50 to 300 Da.[7] Solvent- or matrix-derived ions can be excluded from entering a mass spectrometer by placing an ion filter between the source (often at ambient pressure) and the vacuum interface: a capillary tube, pinhole, or skimmer cone.

The use of ion filters to select analyte ions over chemical noise has been demonstrated using principles of field-dependent ion mobility with strong electric fields from asymmetric waveforms (FAIMS) or differential mobility spectrometry (DMS).[8-20] Improvements of 50-fold in S/N ratios, of 10-fold in detection limits and 5-fold in dynamic ranges, have been reported when an ESI source was supplemented with a FAIMS interface.[21-23] Field-dependent ion mobility filters are now commercially available as small DMS[24] and micro-scale FAIMS[9] devices. Although such ion filters have demonstrated analytical value, their successful application depends upon user knowledge of analyte ion mobility, the influences of parameters on differential ion mobility, and the implementation of technology with high frequencies and high voltages. Some predictive tools have been developed although the subject is underdeveloped.[25]

Gas-phase ions at ambient pressure can also be filtered using ion mobility devices with low DC electric fields[26, 27] by balancing the effects of gas flow and electric field on ion motion. Certain implementations of this approach such as differential mobility analysis (DMA) have high resolving power although relatively complex technology.[28, 29] Another possible filter is an orthogonal ion-extraction ion mobility spectrometer, although long residence times for ions in it can be incompatible with fast measurements.[30] These methods are susceptible to deviations in electric fields and gas flows and require large quantities of purified gas. In addition, knowledge and selection for specific mobilities are needed. Finally, routine measurements are sensitive to variations in operating parameters, and ion selection must be synchronized with mass scans.

Ion filters based on mobility principles at ambient pressure also have been demonstrated using relatively small and simple devices without sophisticated controls and high demand on user knowledge of mobility coefficients. These arose from advances in ionization detectors in the 1960s. Several versions of such filters were refined and deployed as chemical warfare agent detectors, notably the M43 ionization detector for the M8A1 automated alarm.[31] While some were fitted to mass spectrometers to study the chemistry of API sources, none were employed as a filter as described above. In one design, ions were passed with a flow of purified air through a chamber containing baffles in which ion-ion recombination caused preferential neutralization of high mobility ions. Ions of low mobility were then selectively passed through the filter to a Faraday plate detector. In another military-derived ion filter, ions were passed through a wire grid with electric fields perpendicular to the ion and gas flow. Ions of high mobility were selectively drawn to wires and neutralized by collision with the wire while low mobility ions largely passed through the wire grid. Such filters provided a continuous flux of ions with comparatively broad response curves and low demand on electronics and user control of parameters. While increases in S/N ratios using such type of filters may not match those of a FAIMS or DMS ion filter, this would be compensated for by the high convenience and low costs of ownership and operation. The performance of the military-based devices was generally equivalent to low-frequency electronic filters and a type of low-mobility-pass filter (LMPF) is envisioned in developments described below.

The goal of this work was to create a LMPF including design, study of ion trajectories with computational calculations, and quantitative evaluation by MS with two API sources. A specific interest was to compare performance of a relatively clean and dry background matrix with a solvent-rich ion mixture produced by an ESI source.

EXPERIMENTAL

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL
  4. RESULTS AND DISCUSSION
  5. CONCLUSIONS
  6. Acknowledgements
  7. REFERENCES

Instrumentation

Mass spectrometer

The mass spectrometer was a Shimadzu LCMS-2010 (Kyoto, Japan) with a single quadrupole mass analyzer and capillary tube inlet. The capillary or curved desolvation line (CDL) was 60 mm long × 0.15 mm i.d. passing 1 L min–1 of gas at 100°C. The experimental parameters of the mass spectrometer included scan rate, 1000 s–1; m/z range, 20 to 1100; and voltage on the ion multiplier, –1500 V.

Low-mobility-pass filter (LMPF)

The LMPF shown in Fig. 1(a) was built using a pair of ceramic plates with surface-bonded 0.01 mm thick metal rectangular gold-coated copper elements, 13 mm long and 5 mm wide. The plates were separated by a 0.5 mm Teflon gasket with a channel for gas flow between the plates. These plates were bounded by Teflon layers and held in an aluminum frame (Fig. 1(a)). A 2 kHz square waveform with amplitude of ±10 V, centered at 0 V, was placed on one plate while the other plate was grounded.

image

Figure 1. Design of a low-mobility-pass filter (LMPF) with planar geometry (a) and model of ion motion through the filter (b) presented as trajectories of ions with four different mobilities (2.0 cm2 V–1 s–1 (orange), 1.5 cm2 V–1 s–1 (green), 1.0 cm2 V–1 s–1 (blue), and 0.5 cm2 V–1 s–1 (red)) at amplitudes of 25 V, 50 V, and 100 V with a 10 kHz square waveform. Frequency used to obtain this figure (10 kHz) is higher than that used for theoretical and experimental transmission curve generation (2 kHz in Fig. 2). This larger frequency was chosen to give more oscillations down the filter for visualization. Changes in frequency lead to corresponding changes in the voltages required for selective ion transmission, but mobility-dependent ion behavior is the same for all frequencies.

Download figure to PowerPoint

Discrimination against high-mobility ions was achieved by increasing the amplitude of the square waveform to a value high enough for high-mobility ions to collide with and undergo discharge on the plates. Ions with low mobilities that did not reach a plate were passed through the filter.

Ion sources

The first ion source was an APCI source made of 5 mCi 63Ni foil held in a stainless steel 1/8" Swagelok union attached to the frame of the filter. The other ion source was an ESI source built from a glass capillary fitted into one inlet of a PEEK T-union. Tubing from a syringe pump supplying 5 μL min–1 of analyte solution and a metal wire, providing a connection to a high-voltage supply, were placed in the other inlets of the T-union. The T-union was press-fitted into a Teflon end cap, which fitted into a 25 mm long 12 mm i.d. glass tube. The other end of the glass tube had a similar end cap fitted with a 1/8" stainless steel Swagelok union, grounded. The union acted as the counter electrode for the capillary as well as the connection to the filter. Purified nitrogen was introduced into the glass chamber through the first end cap to sweep ions into the filter at a flow rate of 1.4 L min–1. The gas supplied to the filter was purified nitrogen from a nitrogen generator (model NMC, On Site Gas Systems, Inc., New Britain, CT, USA) and it was further purified using a 1 m long 15 cm i.d. scrubbing tower filled with 5Å molecular sieve (AGM Container Controls, Tucson, AZ, USA) which resulted in the moisture level of 0.6 ± 0.3 ppmv as measured by a Moisture Image Series 2 moisture meter (Panametrics, Inc., Waltham, MA, USA).

Chemicals and reagents

The 63Ni ion source (APCI) was tested using a vapor stream of ~1 ppm 2-hexadecanone (Sigma-Aldrich, St Louis, MO, USA) in purified nitrogen. The source was produced using a permeation tube containing 5 mL of 2-hexadecanone in a 500 mL round-bottomed flask with nitrogen flow into the flask at 1.4 L min–1.

The ESI source was tested using [Ala92]-peptide 6 ([Ala92]-p16 fragment 84–103) and HPLC peptide standard mixture (GLY-TYR, VAL-TYR-VAL, TYR-GLY-GLY-PHE-MET, TYR-GLY-GLY-PHE-LEU, ASP-ARG-VAL-TYR-ILE-HIS-PRO-PHE) – both from Sigma-Aldrich – at a concentration of 100 µg mL–1 of each peptide in a 50:50 mixture of water (Barnstead Nanopure water purifier, Thermo Scientific, Waltham, MA, USA) and LC-MS grade methanol (Sigma-Aldrich).

Procedures

Computational modeling of ion motion in the LMPF

Ion transport through the LMPF was modeled using the DMx software package (Sionex Corp, Bedford, MA, USA) which is based on the Boundary Element Method (BEM) and included capabilities to model gas flows, to vary waveforms, and to account for ion diffusion by Monte-Carlo simulation. In addition, influences of surface charging and material selection were included in the ion trajectory calculation.[32] The dimensions and parameters for the computational model were the same as those of the experimental filter. The gas flow rate used for modeling was 1.4 L min–1 with parabolic distributions of velocities over the flow cross section. Ion transport was modeled using 100 ions of each mobility (K0) of 2.2, 1.5, 1.0, and 0.5 cm2 V–1 s–1. Models were made both with and without diffusion simulation. In the model, ions were introduced into the LMPF at ten positions in the y-axis to approximate the performance of an actual filter and the numbers of ions of a given Ko passing the LMPF were counted. The transmission coefficient was determined from the ratio of the number of ions exiting the LMPF to the number of ions introduced into the LMPF. The modeling could not account for secondary effects such as charge exchange in the interface between the LMPF and the mass spectrometer or distortions in flows and fields arising from mechanical imperfections in the LMPF.

Determination of transmission curves with LMPF-MS

Vapors from the permeation tube for the 63Ni APCI source or from spray for the ESI source were continuously introduced into the filter and the resulting mass spectra were recorded. Parameters for the LMPF and mass spectrometer were constant apart from the amplitude of the waveform, which was changed from 0 to 10 V in 10–15 steps.

Ion transmission curves were derived from the observed ion intensities in the mass spectra. The waveform amplitude was controlled with a precision of ±0.1 V and the precision of the experimental transmission coefficients was ~5 to 10% relative standard deviation.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL
  4. RESULTS AND DISCUSSION
  5. CONCLUSIONS
  6. Acknowledgements
  7. REFERENCES

In a planar LMPF, a low-frequency, low-amplitude, symmetric waveform causes ion displacement perpendicular (y-axis) to the flow of gas (x-axis). During one half of a waveform, an ion is moved a distance y, per Eqn. (1):

  • display math

where K is the mobility coefficient, E is the electric field strength, and t is the half period of the waveform (discussion is based on an initial position of ions in the center of the gap of the LMPF). For a 25 V amplitude with a 20 kHz waveform (chosen arbitrarily for this discussion), an ion with a relatively high mobility of 2.5 cm2 V–1 s–1 from a solvent ion, for instance, H+(H2O)2 is displaced a maximum of 0.31 mm. Since the filter gap is only 0.5 mm wide, the ion entering the filter in the center of the gap will collide with a plate in the first cycle and discharge. In contrast, an ion with a mobility of 1.0 cm2 V–1 s–1, such as a protonated monomer of a medium size organic molecule will be displaced to 0.12 mm in the y direction. This ion and others of equal or lower mobility will be restored to their prior position (y-axis) in the gap in the next period of the waveform. This is unlike FAIMS or DMS filters because field strengths within the LMPF are not large enough to cause changes in K, or Ko. In addition to displacement occurring in the y-direction (Eqn. (1)), the gas flow carries ions some distance along the x-axis during each period of the waveform. The model suggests that the extent of this displacement is governed by the y-location in the laminar flow profile and thus is also dependent on K. Ions of low mobility which are little displaced on the y-axis are carried through the LMPF at speeds greater than ions of high mobility which are displaced into the slow gradients of flow nearer the plates.

In a practical filter, ions from the source will enter the gap not at a point, but over the entire cross section of flow (Fig. 1(b)). The overall transmission efficiency will be affected or degraded since even ions of low mobility, when their initial position is near the discharge plate, will be lost during a first cycle of the waveform. Similarly, ions of reasonably large Ko values that enter the gap at the point farthest from the discharge plate may still pass through the LMPF at a certain voltage and frequency. Such leakage will be governed by the waveform frequency, ion transit time in the filter, mobility and waveform amplitude. Thus, this planar design for a LMPF should have a relatively broad response curve (equivalent to a Bode plot in electronic low pass filters) and is expected to have a low resolving power compared with DMS or FAIMS devices. A benefit, however, is that relatively broad ranges of mobilities can be filtered and small variations in experimental parameters should affect filter performance weakly, if at all.

The aspect ratio for the LMPF channel is 8 and the critical Reynolds number for such a rectangular duct exceeds 2000. The calculated value in this LMPF is only 800, suggesting laminar flow, provided that limited distortions exist where the ion source and body of the filter join and where the LMPF is fitted to the capillary tube for the mass spectrometer. However, disruptions of flow were significant in these regions of the LMPF, as discussed below.

Calculated transmission of ions through a planar LMPF

The described ion transport through a LMPF results in mobility-dependent transmission of ions through the filter. To quantitatively study the filter performance, ion transmission efficiencies or coefficients (CT) for given waveform frequency and varying amplitudes were determined using models similar to those presented in Fig. 1(b). Ions of three mobilities (2.2, 1.5, and 1.0 cm2 V–1 s–1) were injected into the filter at 10 positions along the y-axis (0.05 mm apart) and 10 ions were used for each position, resulting in a total of 100 injected ions. These transmission coefficients formed the transmission curves representing the dependence of the transmission coefficient on waveform amplitude. The transmission curves calculated under 'no diffusion' conditions are shown in Fig. 2(a) as dashed lines for ions of three mobilities. As expected, ions of all mobilities pass the filter with a coefficient of ~1 when the amplitude is 0 V and there are some losses due to wall collisions by diffusion. CT values for ions with large mobilities are reduced by 50% with only 3 V amplitude (E = 60 V cm–1) through collisions at the plate surface in the first half-cycle, regardless of their initial location in the 500 micron gap, and by 5 V they are largely eliminated inside the LMPF. Ions with smaller Ko values have higher transmission coefficients proportional to 1/Ko, and those entering near the plate receiving the waveform (i.e., greatest distance on the y-axis from the plate of discharge) pass through the filter with few or no neutralizing collisions as expected from models in Fig. 1(b), although those entering far from the plate surface are discharged and lost in the filter. This complexity is not shown in Fig. 2(a), which represents an average CT for ions of a specific mobility entering at 10 locations in the cross section.

image

Figure 2. Ion transmission curves in a planar mobility filter with 2 kHz waveform: (a) 63Ni ionization source with 2-hexadecanone (1. Ko = 2.6 cm2 V–1 s–1; 3. Ko = 1.1 cm2 V–1 s–1; 5. Ko = 0.74 cm2 V–1 s–1) and (b) ESI source with [Ala92]-peptide 6 as analyte (1. Ko = 1.9 cm2 V–1 s–1; 2. Ko = 1.2 cm2 V–1 s–1). Computed transmissions curves, shown as dashed lines, for ions with mobilities 2.2 (curve 2), 1.5 (curve 4) and 1.0 (curve 6) cm2 V–1 s–1. Inset: COMSOL model of flows between the ion source and the entrance to the LMPF showing flow patterns consistent with expected disruptions of laminar flow in the planar structure.

Download figure to PowerPoint

When the waveform amplitude is set to 4.5 V, the ions of the highest mobility are neutralized in the filter and only the ions of lower Ko values will pass through the filter. At 8 V amplitude, only ions of 1.0 cm2 V–1 s–1 and below will survive passage through the filter and even these will exhibit some loss in transmission efficiency. When the waveform amplitude is 10 V, virtually all ions collided and discharged within the LMPF regardless of their Ko values.

Differences in slopes and intercepts were evident for ions of characteristic Ko values thus providing an opportunity to filter ions by mobility. The slopes were −0.24, –0.17 and −0.11 V–1 for ions with mobility values of 2.2, 1.5 and 1.0 cm2 V–1 s–1, respectively. Intercepts with the waveform amplitude axis increased similarly with mobility at values of 4.5, 7.0 and 10 V. This mobility-dependent behavior of ions in a simple mobility filter serves as a basis for filtering ions in mixtures provided by ion sources at ambient pressure. A sweep of waveform amplitude will eventually result in the displacement of all ions to the plate surface. At any voltage, ions of large mobility will be drawn to the wall at a rate greater than that of ions with low mobility, hence a low mobility pass filter.

Measured transmission of ions through a planar LMPF

Transmission curves are shown in Fig. 2(a) (solid lines) for ions derived from 2-hexadecanone with a 63Ni ion source which were identified as [(H2O)3H]+, [MH]+ and [M2H]+ with measured[33, 34] Ko values of 2.5, 1.10, and 0.74 cm2 V–1 s–1, respectively. These Ko values are close but not identical to those used in the model, yet the plot shows comparative behavior in shape and relative position. While the calculated and measured performances of the LMPF exhibited a common general form, significant differences in the shape of plots existed at high amplitude. The theoretical plots show a nearly linear shape with minor curvature of slope near the x-axis intercept. In contrast, the experimental curves showed a structure with several distinctive regions of slope. In the first region of the plots from 0 to 1 V, the slopes were relatively flat and then steepened to values ranging from −0.22 to −0.16 V–1. They then flatten in the last region with slopes asymptotically approaching the waveform amplitude-axis rather than continuing linearly into it. The resolving power of the planar LMPF was determined by differentiation of experimental transmission curves, and average R for ions with different mobilities was ~2. No attempt was made to optimize the resolving power.

The performance of the LMPF with multiply charged ions of large mass, as seen with an ESI source and [Ala92]-peptide 6, is shown in Fig. 2(b) where the general form of filtering is evident and the three regions are even more exaggerated than with smaller ions. Sources of such discrepancies could arise from several factors including:

  1. The computationally derived curve may be skewed by small statistics of sampling: only 100 ions for each waveform amplitude and Ko value. Results were identical with 500 ions over a small sub-set of points and mobilities and did not reveal statistical error.
  2. Some curvature in plots may occur due to ion transformations in mobility in the interface between the LMPF and the vacuum of the mass spectrometer. Charge exchange or dissociations in the supersonic expansion can lead to ions appearing at low masses, while traveling through the filter as a low mobility ion with higher mass. This suggestion is disfavored by the general appearance for reasonably labile adducts of (H2O)nH+ and the relatively stable proton bound dimer of 2-hexadecone.
  3. Mechanical imperfections in flows where the ion source and filter are joined. The filter is comprised of two unions of dissimilar cross-sections and diameters, as seen in Fig. 1(a). The ion source and capillary inlet are both cylindrical with 3 mm and 0.150 mm diameters, respectively, and the planar geometry of the filter is a 0.5 mm high × 4 mm wide rectangle. While the predicted transmission curves are based on laminar flow throughout the LMPF structure, disruptions to this laminar flow are suggested by results from computational flow modeling (COMSOL Multiphysics, COMSOL, Inc., Burlington, MA, USA). These irregularities are seen in the inset of Fig. 2(b) (flow increases from black to white) where the patterns in flow velocity in the graphic suggest mixing or blending over the cross-section of the LMPF in the first ~0.5 mm of the path. Indeed, ion losses can occur in the fringe field as ions move from the source to the gap in the LMPF (Fig. 1(b)). This region, which includes some critical periods of ion filtering by mobility, is also the same region of the filter where flow patterns are irregular, and where mixing would distort the calculated ion transmission where laminar flow was assumed.
Mass spectra

The signals of ions in the mass spectra discussed below include those of singly and doubly charged ions. Since the molecular weights of the peptides used in these experiments were less than ~2100 Da and their solutions were not acidified, ions with a charge higher than 2 were not observed at any significant intensity. Thus, the behavior of multiply charged (3 and more charges) ions was not thoroughly studied here. Similarly, the different charge states derived from one molecule were not examined. However, it can be expected that an increase in number of charges on the ion will lead to increase in its mobility. This can be somewhat balanced by conformational transformations of the ion. The exact outcome, i.e. the extent of change in the ion mobility with the increase in number of charges on the ion, will depend on the balance of these two effects. One example of the described behavior can be seen with ions in Fig. 2(b).

Mass spectra from which the transmission curves (Fig. 2(a), solid lines) were generated and which illustrate the potential analytical value of planar a LMPF are shown in Fig. 3. The mass spectrum for 2-hexadecanone derived from APCI using a 63Ni source includes reactant ions, here hydrated protons [(H2O)nH]+ with n from 2 to 4, from m/z 37 to 91 (Fig. 3(a)). When the filter is inactive these ions exist at or below 50% relative abundance and the base peak is the protonated monomer of 2-hexadecanone at m/z 241. Also of high abundance is the adduct of the protonated monomer with a water molecule at m/z 259. The proton-bound dimer (with the lowest ion mobility here), at m/z 481, has low relative abundance in this mass spectrum. Once the filter is engaged with amplitude of 8 V for a 2 kHz waveform (Fig. 3(b)), the reactant ions are removed and the protonated monomer ions are significantly filtered to below 50% of the base peak, now the proton-bound dimer. This isolation, or filtering, of the proton bound dimer demonstrates the broad response curve of this LMPF discussed earlier and also the change in total signal intensity. Selective loading of an ion of interest while excluding others would be particularly useful in ion-trap-based methods with diminished space charge effect.

image

Figure 3. Mass spectra of 2-hexadecanone with 63Ni (APCI) ionization source obtained with the filter inactive (a) and with the filter active using a 8 V amplitude rectangular square waveform at 2 kHz (b).

Download figure to PowerPoint

In Figs. 4(a) and 4(b), ESI mass spectra of the HPLC standard peptide mixture are shown. Here the LMPF provided removal of solvent ions generated by an ESI source. These were dominant in the mass spectrum and were accompanied by the doubly protonated peptide, ASP-ARG-VAL-TYR-ILE-HIS-PRO-PHE at m/z 525, and ions of middle mass, m/z 200 to 400, with the LMPF inactive (Fig. 4(a)). When the LMPF is activated with 5 V amplitude in the waveform, the solvent ions which have mobilities close to those of the reactant ion peaks noted above were reduced to negligible abundance (Fig. 4(b)). As suggested in Figs. 2(a) and 2(b), the filter exhibits a broad cut-off response and, consequently, some ions at low abundance passed through the filter including the ion at ~ m/z 380, the protonated VAL-TYR-VAL peptide. As with the APCI results in Fig. 3, these findings suggest that this LMPF with long times of ion accumulation should result in higher S/N ratios with trap-based mass analyzers.[35]

image

Figure 4. Mass spectra from ESI source with an HPLC peptide standard mixture directly through an infusion pump with the filter inactive (a) and with the filter operated with 5 V amplitude on a 2 kHz square waveform.

Download figure to PowerPoint

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL
  4. RESULTS AND DISCUSSION
  5. CONCLUSIONS
  6. Acknowledgements
  7. REFERENCES

A mechanically simple, low-voltage, comparatively low-frequency rectangular waveform driven filter based on differences in ion mobility, when placed between an ion source and a mass spectrometer, produced mass spectra where reactant ions in an APCI source and solvent ions in an ESI source could be removed with an efficiency near 100%. A relatively broad response function suggests lesser improvements in S/N ratios than DMS and FAIMS for ions of comparable Ko values. In contrast, little foreknowledge of a precise value for the mobility of an analyte is needed and the unit exhibits little sensitivity to experimental parameters, although precise sensitivities have not been determined. Results from the model suggest that improvements in designs of flow patterns will improve performance, that is, narrow the response function, although the effects will be in multiples and not orders of improvement. Specifically, a region where flows are calmed and made laminar before the separation by mobility is made should improve the performance and agreement between models and measurements. While some value is anticipated for this design of LMPF to improve S/N ratios with LC/MS measurements, the broad function seen in models and data suggests that the best application may be with ion-trap-based methods.

Acknowledgements

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL
  4. RESULTS AND DISCUSSION
  5. CONCLUSIONS
  6. Acknowledgements
  7. REFERENCES

The financial support of Shimadzu Corporation, Advanced Technology Department, Technology Research Laboratory (previously Overseas R&D Support Center), is gratefully acknowledged. Helpful discussions and manuscript review by Dr John. A. Stone (Queen's University, Kingston, ON, Canada) and manuscript review by Dr Michael. R. Salazar (Union University, Jackson, TN, USA) are gratefully acknowledged.

REFERENCES

  1. Top of page
  2. Abstract
  3. EXPERIMENTAL
  4. RESULTS AND DISCUSSION
  5. CONCLUSIONS
  6. Acknowledgements
  7. REFERENCES
  • 1
    M. Yamashita, J. B. Fenn. Electrospray ion source. Another variation on the free-jet theme. J. Phys. Chem. 1984, 88, 4451.
  • 2
    M. Karas, D. Bachman, F. Hillenkamp. Influence of the wavelength in high-irradiance ultraviolet laser desorption mass spectrometry of organic molecules. Anal. Chem. 1985, 57, 2935.
  • 3
    K. Tanaka, H. Waki, Y. Ido, S. Akita, Y. Yoshida, T. Yoshida. Protein and polymer analyses up to m/z 100,000 by laser ionization time-of-flight mass spectrometry. Rapid Commun. Mass Spectrom. 1988, 2, 151.
  • 4
    T. E. Angel, U. K. Aryal, S. M. Hengel, E. S. Baker, R. T. Kelly, E. W. Robinson, R. D. Smith. Mass spectrometry-based proteomics: existing capabilities and future directions. Chem. Soc. Rev. 2012, 41, 3912.
  • 5
    K. F. Geoghegan, M. A. Kelly. Biochemical applications of mass spectrometry in pharmaceutical drug discovery. Mass Spectrom. Rev. 2005, 24, 347.
  • 6
    F. Kocher, A. Favre, F. Gonnet, J.-C. Tabet. Study of ghost peaks resulting from space charge and non-linear fields in an ion trap mass spectrometer. J. Mass Spectrom. 1998, 33, 921.
    Direct Link:
  • 7
    L. Hua, T. Y. Low, W. Meng, M. B. Chan-Park, S. K. Sze. Novel polymer composite to eliminate background matrix ions in matrix assisted laser desorption/ionization-mass spectrometry. Analyst 2007, 132, 1223.
  • 8
    P. Hatsis, A. H. Brockman, J. T. Wu. Evaluation of high-field asymmetric waveform ion mobility spectrometry coupled to nanoelectrospray ionization for bioanalysis in drug discovery. Rapid Commun. Mass Spectrom. 2007, 21, 2295.
  • 9
    A. A. Shvartsburg, R. D. Smith, A. Wilks, A. Koehl, D. Ruiz-Alonso, B. Boyle. Ultrafast differential ion mobility spectrometry at extreme electric fields in multichannel microchips. Anal. Chem. 2009, 81, 6489.
  • 10
    B. B. Schneider, T. R. Covey, S. L. Coy, E. V. Krylov, E. G. Nazarov. Chemical effects in the separation process of a differential mobility/mass spectrometer system. Anal. Chem. 2010, 82, 1867.
  • 11
    S. T. Wu, Y. Q. Xia, M. Jemal. High-field asymmetric waveform ion mobility spectrometry coupled with liquid chromatography/electrospray ionization tandem mass spectrometry (LC/ESI-FAIMS-MS/MS) multi-component bioanalytical method development, performance evaluation and demonstration of the constancy of the compensation voltage with change of mobile phase composition or flow rate. Rapid Commun. Mass Spectrom. 2007, 21, 3667.
  • 12
    A. Mie, M. Sandulescu, L. Mathiasson, J. Emneus, C. T. Reimann. Analysis of triazines and associated metabolites with electrospray ionization field-asymmetric ion mobility spectrometry/mass spectrometry. Anal. Sci. 2008, 24, 973.
  • 13
    Y. Q. Xia, S. T. Wu, M. Jemal. LC-FAIMS-MS/MS for quantification of a peptide in plasma and evaluation of FAIMS global selectivity from plasma components. Anal. Chem. 2008, 80, 7137.
  • 14
    Y. Xuan, A. J. Creese, J. A. Horner, H. J. Cooper. High-field asymmetric waveform ion mobility spectrometry (FAIMS) coupled with high-resolution electron transfer dissociation mass spectrometry for the analysis of isobaric phosphopeptides. Rapid Commun. Mass Spectrom. 2009, 23, 1963.
  • 15
    J. D. Canterbury, J. Gladden, L. Buck, R. Olund, M. J. MacCoss. A high voltage asymmetric waveform generator for FAIMS. J. Am. Soc. Mass Spectrom. 2010, 21, 1118.
  • 16
    S. L. Coy, E. V. Krylov, B. B. Schneider, T. R. Covey, D. J. Brenner, J. B. Tyburski, A. D. Patterson, K. W. Krausz, A. J. Fornace, E. G. Nazarov. Detection of radiation-exposure biomarkers by differential mobility prefiltered mass spectrometry (DMS–MS). Int. J. Mass Spectrom. 2010, 291, 108.
  • 17
    M. J. Manard, R. Trainham, S. Weeks, S. L. Coy, E. V. Krylov, E. G. Nazarov. Differential mobility spectrometry/mass spectrometry: The design of a new mass spectrometer for real-time chemical analysis in the field. Int. J. Mass Spectrom. 2010, 295, 138.
  • 18
    F. K. Tadjimukhamedov, A. U. Jackson, E. G. Nazarov, Z. Ouyang, R. G. Cooks. Evaluation of a differential mobility spectrometer/miniature mass spectrometer system. J. Am. Soc. Mass Spectrom. 2010, 21, 1477.
  • 19
    L. J. Brown, D. E. Toutoungi, N. A. Devenport, J. C. Reynolds, G. Kaur-Atwal, P. Boyle, C. S. Creaser. Miniaturized ultra-high field asymmetric waveform ion mobility spectrometry combined with mass spectrometry for peptide analysis. Anal. Chem. 2010, 82, 9827.
  • 20
    R. Mabrouki, R. T. Kelly, D. C. Prior, A. A. Shvartsburg, K. Tang, R. D. Smith. Improving FAIMS sensitivity using a planar geometry with slit interfaces. J. Am. Soc. Mass Spectrom. 2009, 20, 1768.
  • 21
    M. McCooeye, B. Kolakowski, J. Boison, Z. Mester. Evaluation of high-field asymmetric waveform ion mobility spectrometry mass spectrometry for the analysis of the mycotoxin zearalenone. Anal. Chim. Acta 2008, 627, 112.
  • 22
    J. D. Canterbury, X. H. Yi, M. R. Hoopmann, M. J. MacCoss. Assessing the dynamic range and peak capacity of nanoflow LC-FAIMS-MS on an ion trap mass spectrometer for proteomics. Anal. Chem. 2008, 80, 6888.
  • 23
    J. Saba, E. Bonneil, C. Pomies, K. Eng, P. Thibault. Enhanced sensitivity in proteomics experiments using FAIMS coupled with a hybrid linear ion trap/orbitrap mass spectrometer. J. Proteome Res. 2009, 8, 3355.
  • 24
    B. B. Schneider, T. R. Covey, S. L. Coy, E. V. Krylov, E. G. Nazarov. Planar differential mobility spectrometer as a pre-filter for atmospheric pressure ionization mass spectrometry. Int. J. Mass Spectrom. 2010, 298, 45.
  • 25
    A. A. Aksenov, J. Kapron, C. E. Davis. Predicting compensation voltage for singly-charged ions in high-field asymmetric waveform ion mobility spectrometry (FAIMS). J. Am. Soc. Mass Spectrom. 2012, 23, 1794.
  • 26
    J. Zeleny. Mobilities of the ions in gases at low pressures. Philos. Mag. 1898, 46, 120.
  • 27
    J. Zeleny. The distribution of mobilities of ions in moist air. Phys. Rev. 1929, 34, 310.
  • 28
    J. Fernández de la Mora, L. de Juan, T. Eichler, J. Rosell. Differential mobility analysis of molecular ions and nanometer particles. Trends Anal. Chem. 1998, 17, 328.
  • 29
    P. Martínez-Lozano, J. Fernández de la Mora. Resolution improvements of nano-DMA operating transonically. J. Aerosol Sci. 2006, 37, 500.
  • 30
    V. Laiko. Orthogonal extraction ion mobility spectrometry. J. Am. Soc. Mass Spectrom. 2006, 17, 500.
  • 31
    J. Boscher, C. G. von Roedern. Miniaturized ionization detector system. Proc. 2nd Int. Symposium on Protection Against Chemical Warfare Agents. Stockholm, Sweden. June 17–19, 1983, p. 157.
  • 32
    DMx version 1.09 User Manual, Sionex Corp., Bedford, MA, USA.
  • 33
    G. A. Eiceman, E. G. Nazarov, J. A. Stone, J. E. Rodriguez. Analysis of a drift tube at ambient pressure: models and precise measurements in ion mobility spectrometry. Rev. Sci. Instrum. 2001, 72, 3610.
  • 34
    F. K. Tadjimukhamedov, J. A. Stone, D. Papanastasiou, J. E. Rodriguez, W. Mueller, H. Sukumar, G. A. Eiceman. Liquid chromatography/electrospray ionization/ion mobility spectrometry of chlorophenols with full flow from large bore LC columns. Int. J. Ion Mobil. Spectrom. 2008, 1–4, 51.
  • 35
    A. B. Hall, S. L. Coy, A. Kafle, J. Glick, E. Nazarov, P. Vourous. Extending the dynamic range of the ion trap by differential mobility filtration. J. Am. Soc. Mass Spectrom. 2013, 24, 1428.