Annularly Focused Ion Mobility Spectrometer
The ability to focus electrosprayed ions into an annulus at atmospheric pressure, opens up the opportunity for a unique design of ion mobility drift tube which uses the same configuration of inner and outer electrodes throughout the drift region. One such advantage of this design is the ability to first focus ions into a region of uniform electric field (a necessity for optimal IMS separations) before injection into the drift cell, thus limiting the peak broadening associated with edge-effects in traditional drift tube design [27, 39]. Furthermore, the low cost associated with
FDM manufacture of the device components and wide availability of materials have the potential to expand the use of IMS instrumentation of this nature to laboratories otherwise unable to utilize such characterization methods. With these advantages in mind, the design detailed in Sect. 4.2.2 was constructed and the performance tested.
In general, ion separation within a DT-IMS involves 4 steps: (1) the production of gas-phase ions in the source region, (2) an injection of an ion packet into the drift region, (3) separation of the ions by a uniform electric field as they traverse a drift tube, and (4) the detection of separated ion packets. For these experiments, nanoESI was chosen as the means to generate ions as it is easily coupled with the annular focusing electrode and the annular focusing effect was well characterized. Injection of ion packets into the drift region was accomplished by applying a high voltage pulse to the injection mesh as illustrated in the lower left portion of Fig. 4.1. Upon passing the mesh at the entrance to the drift tube (see Fig. 4.1), ion packets are propelled along the drift tube at different rates, depending on their mobility within the drift gas and the strength of the electric field. The strength of the electric field is determined by the voltages applied first and final ring of the drift tube. Voltages applied to each section of the device are detailed in Sect. 22.214.171.124. Finally, separated ion packets pass through the mesh on the exit of the drift tube and produce a current on a Faraday plate electrode. For all IMS experiments discussed herein, an electric field strength of 226 V/cm was employed. A 2 mM equimolar mixture of TAA-Br salts in ACN was used to test the performance of the 3D printed IMS as TAA cations are often used as ion mobility standards. Their frequent use is the result of a low propensity to form clusters under a variety of conditions .
In traditional DT-IMS instruments, the time-scale for ion injection has been shown to have a noticeable effect on resolving power [27, 41]. Typically, DT-IMS instruments operated at atmospheric pressure employ an injection time ranging from 0.10 up to 1.0 ms. Increasing injection time has been shown to increase sensitivity, but the ability to separate peaks is degraded. Ion injection periods ranging from 0.3 to 1.0 ms were tested with the 3D printed IMS and the resulting spectra are shown in Fig. 4.6. In Fig. 4.6 the peaks correspond to TBAB, THAB,
Fig. 4.6 Separation of 2 mM equimolar tetrabutyl-, tetrahexyl-, tetraoctyl-, and tetradodecylammonium cations under different injection times. Drift field strength was 226 V/cm and a 50 V injection potential (relative to first ring of drift tube) was used in all cases. Each spectrum is the resulting average of 16 scans at a 10 Hz acquisition rate
TOAB, and TDDAB cations, as labeled from left to right. 0.3 ms was the smallest time-scale tested for ion injection as no peak for the TDDAB cation was detected with the use of a 50 V injection potential. These results show the large sensitivity increases for longer injection times and the expected broadening associated with the sensitivity gain.
The performance of IMS instruments is often stated with respect to resolving power (Rp) . Rp is calculated from the drift time (td) and full width of the peak at half maximum (FWHM) using Eq. 4.1:
From the results shown in Fig. 4.6 the Rp for the tetrabutylammonium peak at an injection time period of 500 ^s is calculated to be 25. Using an atmospheric pressure IMS of similar dimensions and operating conditions, Hill et al. measured Rp values ranging from 15 to 30 for methy tert-butyl ether (MTBE) ions when injection time periods were in the range of 300-500 ^s . This comparison shows that measured Rp from the 3D printed IMS are in the range achievable by traditional DT-IMS, and are in fact slight better than those attainable by most field-portable IMS instruments .
Another operational parameter effecting DT-IMS performance is the injection voltage for gating ions into the drift cell. Most commonly, ion gating in DT-IMS instruments is accomplished using a series of electrically isolated parallel wires, in the form of a Bradbury-Nielson , or Tyndall  shutter. When closed, interdigitated wires have opposing potentials applied, thus ions are drawn to the wires rather than passing through. Ions are gated through the shutter by applying matching potentials to opposing wires, thus ions pass through as there is no electric field orthogonal to that supplied by the drift cell. An alternative manner of gating ions into a drift cell is the use of two parallel mesh electrodes . In this case, ions are gated into the drift cell by increasing the potential on the back mesh, such that ions are pushed through openings in the mesh held flush with the drift cell. Alternatively, the potential on the first drift ring and mesh is briefly lowered, allowing ions to pass. In practice, this type of ion gate is much easier to construct and is more mechanically durable that ion shutters utilizing wires. There is however a trade-off with respect to efficiency. The openings in the mesh must be selected to balance transmission efficiency with durability as well as electrical shielding of the drift cell from perturbations by injection voltages.
Due to simplicity in construction, two parallel meshes were used to form the injection region in the 3D printed IMS as detailed in Sect. 4.2.2. In practice, it was found that optimal performance of the 3D printed IMS was using injection voltages in the range of 50-100 V. Beyond 100 V, no significant improvement in sensitivity was found. The resulting IMS spectra of the 2 mM equimolar TAA mix for different injection voltages and time periods is shown in Fig. 4.7. From these results it would initially appear that ion arrival times are altered by the injection time period; however, this is likely an artifact of detector timing. The start time of spectral
Fig. 4.7 IMS spectra of electrosprayed 2 mM equimolar mixture of tetrabutyl-, tetrahexyl-, tetraoctyl-, and tetradodecylammonium Br in ACN for different injection time periods and voltages. Injection voltage is given in reference to the potential on the first ring of the drift cell
collection is triggered on the rising-edge of the injection pulse, rather than centered on the injection waveform. A close comparison of the data in Fig. 4.7 verifies this phenomenon as peaks for the TAA cations are shifted by *0.3 ms when comparing 1.0-0.5 ms injection, consistent with the previously stated hypothesis.
Another interesting trend observed is the decrease in intensities for longer drift times. This can partially be explained by peak broadening, however integration under each respective peak gives a decrease of 76, 71, and 56% for the THAB, TOAB, and TDDAB peak, respectively, from the total area under the peak from TBAB. Likely, the largest contributing factor to the decrease in intensity is the mobility of each cation. The less mobile cations (cations emerging at longer time scales) are not injected as efficiently by the method used herein. Because the ions must travel the separation distance between the two mesh electrodes, more of the ions with higher mobilities are injected than those displaying lower mobilities. This discrepancy can possibly be reduced in future instrument designs, by decreasing the distance between the injection mesh and the mesh held flush with the drift tube entrance.
The resolution of a DT-IMS is also affected by processes occurring within the drift cell itself. As an ion cloud traverses a drift cell, it is subject to expansion in both axial and radial directions by way of diffusion and space-charge effects. Space-charge effects can be mediated slightly by decreasing the injection time, but not without a loss of sensitivity. To combat the effect of axial expansion from both diffusion and space charge, higher electric field strengths can be used as this decreases the residency time within the drift cell. This is only effective to a certain extent as detailed by Siems et al. . Radial expansion of the ion plume is also partially responsible for peak broadening within drift cell instruments. This is due to the non-uniformity of the electric field near the walls of the cell . Previous work by Fernadez et al. has demonstrated that ion expansion within a 30 mm wide drift tube, in which ions produced by a corona discharge were gated into the drift tube and expanded radially ~2x along the 26 cm long drift region . With the
Fig. 4.8 Ion intensity map taken at exit of drift cell. Ions were generated by nanoESI from a 10 pM mixture of TBAB, THAB, TOAB, and TDDAB in ACN
traditional open-ring drift cell design, the influence of this is often reduced by using wider diameter drift rings and restricting ion injection to the center of the device.
Previous results suggest that the use of the annular focusing source under operational parameters herein should limit ion injection to a line width of * 500 pm, centered between the outer and central rings of the drift cell in the 3D printed IMS. To investigate the radial expansion of ions along the drift cell, an intensity map of ions exiting the drift cell was taken by removing the Faraday plate detector and scanning an IonCCD™ detector across the exit mesh of the drift cell as previously described. A solution of 10 pM of the same TAA mixture in ACN was used as the spray solvent and ions were continuously gated (ion gate held open at 50 V) into the drift cell. The resulting ion intensity map is shown in Fig. 4.8. From this image it is evident ions are rapidly expanding radially during drift cell transit. This image also illustrates the limited ion transmission by the 69% open stainless steel mesh, which likely decreases sensitivity significantly.