Imago Scientific Instruments has developed a tool called LEAP(R) that can analyze many cubic nanometers of a sample, telling the position and isotope of each individual atom.
The method uses strong electric fields to pull (usually) one atom at a time off of the sample. They use time-of-flight mass spectrometry to determine the isotope. The position of impact on the detector tells the atom's original (x,y) position, and the z position is determined by the sequence in which the atoms are removed.
I contacted them to ask more questions, and had a very interesting conversation. I was interested in several things:
1. Can this be used to resolve protein structure in frozen water?
Answer: They work at cryogenic temperatures (50 K, a bit colder than liquid nitrogen) but have not announced any work on proteins yet. It sounds to me like this should work just fine, though of course freezing might distort the protein structure somewhat (though this might be reduced by vitrifying).
2. Can this be used with non-uniform or non-solid samples?
Answer: As the sample is eroded, it forms into an "electrostatic lens" with a hemispherical surface. The system is most accurate in terms of atom (x,y) position when the surface is smooth. If the sample is not smooth, then the electrostatic lines will have to be calculated to tell where the atom came from--but you need to know the shape of the sample to calculate the lines. Something that occurs to me is that they have a local electrode in the system to help remove the atoms quickly. Scanning that probe around might allow mapping the field lines. Also, imaging the sample with another microscopy technique might supply some of the needed information; the samples that are useful for this technique are also suitable for TEM.
3. Can you add a scanning probe microscope to their system, in the same chamber, in order to alternate scanning-probe surface-chemistry experiments with analysis?
Answer: They would not talk about this at all, not even to tell me the chamber dimensions. If this were possible, it would be extremely useful for carrying out mechanosynthesis experiments. By knowing which atoms were pulled off, it should be possible to know when you have exposed a clean (known) surface for the next experiment.
A few other facts: Their system includes a field ion microscope. This works by adding a gas to the system, which is ionized by the sample, and the ions are then detected to show information (perhaps field profile) about the sample.
The system can remove and detect 10,000-20,000 atoms per second. This is a lot. The sample is about 100 nm wide, and a 100-nm cube is about 100-200 million atoms. So you could find the position and isotope of every atom in that volume in about three hours. Or, if you were doing a series of surface-modification experiments with an integrated scanning probe, you could strip and analyze ten atomic layers in about half a minute.
According to their press release, they will be showcasing the LEAP at the ISTFA conference (booth 516) to be held November 6 – 10, 2005 at the San Jose McEnery Convention Center. If anyone goes there and manages to get more information out of them than I could get, please let me know or post it here.
Although this is a new technology and its initial development is being targeted at other purposes, it seems likely to be a very useful enabling technology for molecular manufacturing research, especially mechanosynthesis research once someone figures out how to add a scanning probe microscope to the system. (This is an outgrowth of technologies that have been around for a while, but they claim the speed is orders of magnitude higher than previous tools.)