One might reasonably wonder what phase space compression is, and what it might have to do with cluster-surface interactions. Here is a outline of the story:
In doing ion deposition, particularly of mass-selected clusters, we are faced with some facts:
1. Cluster ion sources have pretty low intensities -- 106 to 107 /sec per cluster size.
2. Surface analysis tools are not very sensitive, particularly those that give chemical information(XPS, EELS, etc). Typically, something approaching 1014 atoms/cm2 is required.
This leads to deposition times of many hours (days!) and in most cases, the surface will be contaminated faster than the sample can be prepared. To a lesser extent, the same problem applies to any low energy ion deposition experiment. The problem can be attacked by scaling-up the cluster ion source, however, it is very difficult to get sufficient intensities except for very small clusters. Another successful approach is to use more sensitive surface analysis methods such as scanning probe microscopy, however, most of the chemical information that we are interested in is lost.
Clearly the thing to do is to focus the beam tightly, taking advantage of the fact that most surface analysis tools are pretty good at small area analysis. The problem you run into here is the Liouville theorem: The volume a system occupies in phase space (i.e. the product of its spatial and momentum distributions) is conserved. Many cluster ion sources (e.g., Magnetrons) tend to have large phase space volumes, while achieving a small focus at low energies requires a small phase space.
We have developed a phase-space compressor -- a device using high-frequency fields that beats the Liouville theorem. An example of the result is shown below. For this demonstration, we start with the beam from an ordinary mass spectrometer ion source. The beam is run through our beamline and focused onto our deposition target. For this example, the beam is constrained by an 800 micron diameter mask just in front of the sample. The plot shows how much of the incident beam can be focused through the mask as a function of the deposition energy (energy on the target). The uncompressed beam rises slowly with deposition energy, because it has a rather poor phase-space distribution. After compression, the behavior at low energies is dramatically improved. At a few eV, we can easily get deposition rates of around one monolayer/minute. We have demonstrated phase space compression with small copper cluster ion beams as well.
While we are interested in moderate (~1mm) spots at low energies, there are many applications for tightly focused beams at high energies and high current density beams at moderate energies. Phase space compression can be applied to these systems, and should result in tighter spots, higher current densities, and/or lower energies.