At Columbia University’s Radiological Research Accelerator Facility (RARAF), fundamental investigations into the radiobiological effects on mammalian cells are conducted through broad beam and through controlled single-particle single-cell microbeam irradiation (1). Recent upgrade plans to our 4.2 MV Van de Graaff particle accelerator include implementation of a laser ion source. The laser-ablation based ion source will produce highly charged heavy ions that will extend the linear energy transfer (LET) range of our experiments. Presently, our duoplasmatron ion source can ionize atoms from the gaseous phase, namely from hydrogen and helium. These ions are suitable for particle irradiation experiments with an LET range of 9.5 to 210 keV/mm. Expectations are that the laser ion source will enable a range of ions from hydrogen to iron with an approximate LET range of 10 to 4,500 keV/mm.

In the field of laser ion sources, two common modes of ion production are laser ablation and resonance ionization spectroscopy, a species-selective technique that requires tunable lasers (2). The operational mechanism of our laser ion source is plasma generation through laser ablation of a solid target. Focused Nd:YAG laser pulses supply the power density required to create a plasma plume (3). The plasma ions have distributions over charge state, energy, and angle. To then reduce the beam load on the accelerator vacuum system, an electrostatic analyzer (ESA) selects ions with a particular energy per charge.

The design development procedures for our laser ion source originated with a prototype based on the laser operated ion source acquired from Hughes at the University of Arkansas (4). Ion trajectories in this source experienced in turn, 70 cm of plasma expansion drift, a 180° cylindrical ESA, two Einzel lenses, and a final drift distance to a detector whose position would effectively be the location of the 3.18 mm diameter entrance aperture of the particle accelerator. Dimension details of the original laser operated ion source have been provided elsewhere (5). Furthermore, the following simulation results complement a previous document of our laser ion source development (6).

Ion trajectories through the laser ion source prototype were simulated with the ion optics package, SIMION (7). In SIMION, the potential at points outside electrodes and poles is determined by solving the Laplace equation by finite difference methods (8). Virtual ion optical components are constructed and arranged on an ion optics workbench. Ions flown through such an optical system retain characteristic information useful for generating phase space patterns for simulated ion source emittance measurements.

The ion optics workbench setup for the laser ion source prototype required constructing virtual ion optical elements that paralleled as much as possible the physical configuration. The optical elements were constructed with a grid resolution of 1 mm/grid unit. To project an optical element into three dimensions, SIMION supports both planar and cylindrical symmetries. Planar geometry was used for the cylindrical ESA. And, cylindrical geometry was used for the Einzel lenses. Initial characteristics of the ions flown through the simulation were mass, charge state, position, energy, angle, and a random function applied, within certain bounds, to offsets about the energy and the angle. Typical ion parameters were: aluminum, singly charged positive, origin about a circular area representing the ablation crater (0.25 mm diameter), 400 eV mean energy, and normal emission with a random divergence within a 0.2° cone angle. Voltage settings on the optical elements during the simulation were 69.35V across the analyzer, -200V on the first Einzel lens and 100V on the second Einzel lens.

Emittance results emerged from analysis of the ion flight data. In a typical case, the vertical extent of the ion beam was acceptable for input to the accelerator aperture. However, the horizontal ion beam component extends beyond the aperture boundary, suggesting reduced ion transmission. For other voltage settings on the Einzel lenses, a variety of spot patterns were produced, but none could match the aperture size constraints in the vertical and in the horizontal direction simultaneously. This limitation to the ion source prototype was an artifact of the ion optical geometry.

A double focusing, spherical ESA with point-to-point focusing in both the horizontal and vertical planes offered an attractive solution for our laser ion source design constraints. With this one ion optical element, laser plasma plume ions are both E/z analyzed and focused at the accelerator entrance aperture. The footprint of this option fit well within the accelerator terminal. Einzel and quadrupole lens options were also considered, but their linear arrangements would require structural modifications of the accelerator terminal frame.

Guided by spatial limitations in the particle accelerator and by a desired plasma expansion drift distance of 70 cm, (3) the ESA dimensions were narrowed to a 24° bend with a 2.7 inch radius. This geometrical solution was found by applying Barber’s rule; the object point, the center of curvature, and the image point lie on a straight line (9).

Spherical ESA theory and fringing field effects are well documented in Wollnick’s treatment of electrostatic prisms (10). Wollnick’s guidelines for fringing field termination provided additional dimension details to the ESA. Electrode spacing and field-terminating diaphragm dimensions were set similar to those in the prototype’s ESA. For a 10 mm space between electrodes, a 5 mm arc length from the ESA electrodes to the diaphragm, and a 5 mm diam diaphragm aperture, the 24° ESA would have a 21.32° arc for the physical electrodes and a 29.68° arc for the diaphragm. The ideal field boundary would, in theory, have a 24° arc. A top view cross section diagram of the 24° ESA is shown in Fig. 1.

Fig. 1. Top view cross section diagram of the 24° ESA.

For the simulation of the 24° ESA, the ion optics workbench set up in SIMION was similar to the one presented in the prototype simulation section. However, an increased resolution (0.2 mm/grid unit) was used for the 24° analyzer. Again, the component dimensions in the simulation paralleled as much as possible the intended “real world” case. To insure that the electric field lines were fully terminated in the ESA element, an ideal grounded mesh was wrapped about the ESA diaphragm material at a 5 mm offset distance from the inside of the diaphragm.

For the proposed laser ion source, the simulated spot size is shown in Fig. 2. Horizontal and vertical histograms across this ion spot are shown in Fig. 3. The improvements are clear; the spot size is smaller and the ions tend to focus in both the horizontal and the vertical directions. The structure in the emittance patterns is due to aberrations that arose from the use of a spherical optical element. However, upright ellipse envelopes about the emittance patterns do still imply a focus at the entrance aperture to the particle accelerator. One more note is that the simulation results suggest an energy resolution compatible with input requirements of a six-element electrostatic quadrupole lens with Russian symmetry that is under development. To tune the resolution, interchangeable diaphragms are available for the ESA.

Fig. 2. Spot pattern of proposed laser ion source. Focus is acceptable.

Fig. 3. Intensity histograms across the ion spot pattern shown in fig. 2.

An interest in the fabrication of the ESA is to match the simulation geometry as best as possible. The physical dimensions used in the simulation will be retained except for one feature. The spherical ESA design in SIMION utilized a slot diaphragm. During the simulation, a virtual round diaphragm was realized by limiting the cone angle of the incident ions. The construction will incorporate circular diaphragms.

In constructing the laser ion source, alignment issues are crucial. In particular, the ESA must be placed in its designated position and the electrode orientation should be optimized. The construction will utilize a self-aligning technique to insure proper placement. During the electrode machining process, flat surfaces and alignment holes for indexing pins incorporated into the outsides of the electrodes will allow them to accurately rest in a frame mount.

Electrode surface treatment will also be an issue. A thin layer of carbon will be deposited on the electrodes in order to reduce patch potentials. This should lead to a smoother electric field within the ESA.

References

1.    Randers-Pehrson G, Geard CR, Johnson G, Elliston CD, and Brenner DJ. The Columbia University single-ion microbeam. Radiat Res 156:210-4, 2001.

2.    Brown IG, in The Physics and Technology of Ion Sources (Wiley, New York, 1989), p. 299.

3.    Sharkov B, in Handbook of Ion Sources, edited by B. Wolf (Chemical Rubber, Boca Raton, 1995), pp. 149-51.

4.    Miller RD, Wattuhewa G, Hughes RH, Pederson DO, and Ye XM. Remnant charge of slow multicharged ions scattered from a gold surface. Phys Rev B 45:12019-27, 1992.

5.    Miller RD, Ph.D. thesis, The University of Arkansas, 1990.

6.    Bigelow AW, Randers-Pehrson G, and Brenner DJ. Rev Sci Instrum in press.

7.    Idaho National Engineering and Environmental Laboratory, Idaho Falls, ID 83415

8.    Dahl DA, in SIMION 3D Version 7.0 User’s Manual (Idaho National Engineering and Environmental Laboratory, Idaho Falls, ID, 2000), p. 2-1.

9.    Wollnik H, in Optics of Charged Particles (Academic Press, Inc., San Diego, 1987), p. 98.