Separation of Ions in Air
The most common method of sorting ions in the gas-phase is on the basis of their mass-to-charge ratio (m/z) such as that performed in MS experiments. Alternatively, they may be separation based on an interaction with a background gas in the presence of well-defined electric fields as is done in IMS instruments. Separation based on m/z necessitates the use of a vacuum system, whereas a laminar flowing gas is used in the instance of an ion mobility separation performed at atmospheric pressure.
In an effort to demonstrate a simplified separation of ions in air without the use of a vacuum or a flowing gas, pulsed voltages were employed with the electrode system as a means to inject ions into the curved ion path and effect a separation of tetraalkylammonium (TAA) cations. Controlled injection of ions into the turning electrodes was accomplished by modifying the electrode system to include a region separated by two stainless steel woven wire meshes separated by 3 mm (ion injection region) as is shown in Fig. 3.4. A solution of 10 pM each of tetrapropyl-, tetrabutyl-, tetrahexyl-, and tetradodecylammonium bromide in ACN was sprayed with a nanoESI emitter into Esource. A floated high voltage pulse (2530 V high, 2480 V low) was applied to the mesh directly after the source region with the second mesh held flush to the opening of the first curved electrode to facilitate electrical contact with E1. A pulse width of 50 ms with a repetition rate of 1 Hz was used for ion injection. Potentials applied to the nanoESI electrode, Esource, E1, E2, and E3 were 4.50, 3.20, 2.50, 2.33, and 1.45 kV, respectively. Injected ion packets were sampled and analyzed by positioning the API of the MS to accept ions exiting E3. The LTQ was operated with an injection time of 10 ms, giving an approximate repetition rate of 10 Hz.
A comparison of simulated ion separation under nearly identical conditions was also carried out. Because SIMION is ineffective at modeling ion behavior at ion sources, all ions were initiated within a uniform distribution between the mesh electrodes compromising the injection region. Initially the voltage on the mesh nearest the ion source was set to the high value used for injection (2530 V) for 50 ms, after this time it was lowered to match that used experimentally following an injection cycle (2480 V) for the remainder of the simulation. This approach does not accurately model the experimental ion distribution between the mesh electrodes during the injection, but is used for the purposes of simplification so that arrival times of ions at the detector (a mass spectrometer in this case) can be approximated to determine if simulation values are in rough agreement with experimental data.
Fig. 3.4 Electrode system modified to allow time-resolved injection into the curved electrode region. An image of the mesh used is shown inset