908-788-5550

Note 86: Simulation of a Unique Cylindrical Quadrupole Mass Analyzer Using SIMION 7.0.

Home

by Steven M. Colby1 and John D. Prestage2

1Scientific Instrument Services, Inc.; 2NASA Jet Propulsion Laboratory

(Portions Published at PittCon and ASMS 2000)

Introduction
The purpose of this work is to examine the properties of a new cylindrical quadrupole mass analyzer (CQMA) developed by Dr. John Prestage of JPL1. The CQMA is unique in that it is constructed from a single cylinder rather than 4 round or hyperbolic rods. By segmenting the cross-section of the cylinder into eight or more pieces a quadrupole field can be produced that is equal to or superior to that generated by the standard rod based design. We report an analysis of the CQMA using the SIMION 7.0 ion optics software.

The Cylindrical Quadrupole Mass Analyzer

Advantages to the CQMA include: (1) Greater ease of manufacture. The opening in a cylinder can be manufactured to considerable more accuracy than four independent rods and their associated support mechanism. Electrodes can be established by vapor deposition of a metallic surface on the inside of the cylinder. (2) Reduced weight. The CQMA will require approximately 10 times less material. This is of great importance for portable instruments and for those used on space missions. (3) Reduced power requirements. The electronic capacitance of the CQMA design is at least 2 times lower than that of a typical QMA. In addition the voltage required to operate at the same stability parameters is 10% lower in the CQMA. As a result the CQMA will dissipate less than 1/2 of the power of a traditional instrument. This is also important in portable instruments and also allows the engineer greater flexibility in design parameters such as the choice of r0 and RF frequency. A cross-section of the proposed CQMA design is shown in Figure 1 (taken from Ref. 1.)


Figure 1 CQMA Cross-Section compared to the cross section of a typical round-rod quadrupole.

SIMION V7.0

SIMION 7.0 is the most recent version of the popular SIMION ion optics simulation program. The new version includes complete windows compatibility. Some of the dozens of new features include: support for windows printer and graphics drivers, printing to the windows clipboard, 50 million point array sizes, and more control over viewing, recording, saving, and rerunning trajectories. The user programming system has a series of new commands to support arrays and data comparison.

SIMION performs simulations using discrete arrays of potentials (finite elements). The user defines electrodes or magnets within a 2 or 3D space and the program calculates the potentials of non-electrode or non-magnet points. Each potential array, or "instance", is placed in an "ion optics bench" along with other arrays that may represent other parts of an instrument. This approach allows each section of a system to be modeled independently. The user can make maximum use of symmetry elements and need not use computer resources to model field free regions. Charged particles can be flown through the ion optics bench and their trajectories calculated. The user programming aspects of the system allow users to define time dependent fields and manipulate the properties of the charged particles.

Experiment

In order to examine the characteristics of the Cylindrical QMA we simulated a CQMA, a standard QMA with circular rods, and a QMA with hyperbolic rods. The mass filters were simulated to a resolution of 0.0134 mm/gird point. These instances were defined as a 2D array and "stretched" in the 3ed dimension on the ion optics bench. A second 3D instance, with resolution 0.0536 mm/grid point, was used to simulate the interface between the mass filter and an entrance/exit lens. This single instance was used in two different positions on the ion optics work bench so it could serve at both the entrance and exit. The 3D instance included a lens element and a short section of the quadrupole so that the transition between the two parts would be modeled correctly.

Results

We first examined the fields produced by each type of filter using SIMION's contour feature. This utility plots equal-potential contours within a cross-section of an instance. Results are shown in Figure 2. Figure 2a, 2b, and 2c show each filter independently. The contours represent ground and +/- 90, 50, and 10 percent of the voltage on the electrodes. Figure 3 illustrates an overlay of the QMA, CQMA, and ideal hyperbolic electrodes. In the center region the contours overlap within the precision of the simulation. (Figure 3b is an expansion of the center region.) From this comparison we expect that the CQMA will have filtering properties that are similar to standard devices. Their performance will be most dependent on the precision of manufacture and assembly. As previously pointed out the CQMA may have advantages in this area.



Figure 2. Equal-potential contours. A) Hyperbolic rods, B) Round Rods, C) CQMA


Figure 3. Overlay of the contours shown in Figure 2.

Our next step was a Monto-Carlo simulation of ions passing through the filter. This simulation included the entrance and exit structures. Using SIMION's user programming features we introduced ions with random initial positions, trajectories, and energies. The widths of the initial distributions were wide enough to test the limits of the mass filters. Five thousand ions each were flown through the CQMA and standard (round rod) QMA. At the tuned mass both filters passed 100% of the ions. At one unit away from the tuned mass the QMA passed 45% of the ions and the CQMA passed xx%. (Note that this simulation does not include many of the optic elements that would normally be found in a functional system and that our distributions were intentionally selected so as to test the limits of the filters. In an actual implementation better filtering would, of course, be possible.)

The slightly different filtering efficiency of the CQMA may be attributable to the fields in the transition between the mass filter and the entrance and exit apertures. The enclosed structure of the CQMA could reduce the influence of the grounded aperture. The ideal quadrupole fields would therefor be disrupted over a shorter distance near the interface.

In order to test this hypothesis we examined the field contours at the very entrance of the filters. The results are shown in Figure 4. This figure is an overlay of the ideal hyperbolic fields as measured in the center of the filter (from Figure 2a) and the fields as measured at the entrance of the filters. The colors of one set of contour lines have been enhanced in order to highlight their differences. As shown the QCMA is 15 to 20% closer to the ideal hyperbolic fields.



Figure 4. Overlay of the contours near the ends of the filter.

In our next experiment we examined modifications to the transition region between the entrance and exit lenses and the quadrupole filter. The first of these is the use of a split lens element and the second is a variation in the shape of the mass filter electrodes.

Split Entrance/Exit Lens

We wished to see if the electric fields in the in the transition region between the flat plates of the optics and the mass filter could be improved to be more like the ideal hyperbolic field. We split the entrance lens in the same angular spacing as the CQMA. This is shown in Figure 5. Each segment of the split lens was varied with the same potentials as the corresponding regions of the CQMA. The filter was also extended up to the lens. The resulting 3D contours are shown in Figures 6 and 7. Figure 6a and 6b illustrate the fields in with and without the split lens respectively. To simplify the image only the potentials at 90% of +V and - V are drawn and a cross-section of the region shown. In comparison Figure 4c shows the same potential contours at the center of the mass filter. It is clear that the split lens generates a much more ideal field at the transition region. Figure 7 shows this in an overlay of a thin cross-section of the fields from 6a and 6b.


Figure 5. Split Enterance Lens


Figure 6. Equal-potential surfaces. A) With and B) Without Split lens.

Figure 7. Overlay of surfaces.

We examined the trajectories of ions through the system that included split lens. It was found that the single, grounded, lens had been of considerable use in focusing ions toward the center of the mass filter. Additional focusing optics were required when the split lens was used. We were, therefore, not able to perform Monte-Carlo simulations that could be directly compared with our simple model.

Variation of the Mass Filter Shape

A unique aspect of the CQMA is that the spacing and relative ratio of the electrodes may be varied to produce a beneficial effect. This may be an important advantage to this filter design. We examined the consequences of changing the relative ratio of the -V, +V, and 0V electrodes near the entrance and exit lenses. Examples are shown in Figure 8. The ratios of the electrodes were changed in a continuous linear fashion such that at 4 mm the ratios had changed from 60 degrees/30 degrees to 75 degrees/15 degrees. At the lens the grounded region was eliminated entirely and each +/-V segment covered a full 90 degrees. The lens itself was held at 0V. Note that a variation in the shape of the electrodes such as this would not be practical in the standard rod design.


Figure 8. Changing electrode ratios.

The effects of this new design are shown in Figure 9. This figure shows the same contours as are used in Figure 6.



Figure 9. Contours with new ratios.

Our next step was a Monto-Carlo simulation of ions passing through the filter. This simulation included the entrance and exit structures. Using SIMION's user programming features we introduced ions with random initial positions, trajectories, and energies. The widths of the initial distributions were wide enough to test the limits of the mass filters. Five thousand ions each were flown through the new geometry and the standard CQMA. At the tuned mass the standard CQMA passed ~93% of the ions. While one mass unit away it passed 82%. Under similar conditions the new geometry passed 75% and 31% respectively. (Note that this simulation does not include many of the optic elements that would normally be found in a functional system and that our distributions were intentionally selected so as to test the limits of the filters. In an actual implementation better filtering would, of course, be possible.)

Conclusions

We have shown that the Cylindrical Quadrupole Mass Filter is a suitable replacement for the standard four-rod design. In addition to possible advantages in design and manufacture the CQMA is expected to require less than 10% of the mass and 50% of the power of a conventional system. These features are critical for space or environmental applications wherein the device must be moved and have its own power source.

We have also shown that the CQMA may also have advantages in filtering efficiency. A unique aspect of the design is that the ratio of the relative electrode sizes could be changed as a function of position without changing the diameter of the cylinder. This may offer additional opportunities to improve the quality of the fields at the beginning and end of the filter.

For information regarding this technology contact Jim Smart at the Caltech Office of Technology Transfer (626)395-3058.

References:

1: John D. Prestage, NASA Tech Brief Vol. 23, No. 5.
2: Steven M. Colby and John D. Prestage, PittCon 2000, Abstract #1933.
3: Steven M. Colby and John D. Prestage, PittCon 2000, Abstract #P296.