Molecular Electron Holography:
Progress toward atomic-resolution imaging?

Electron hologram
of a DNA molecule
From “Non-destructive Imaging
of Individual Bio-Molecules”

Hans-Werner Fink’s group reports a remarkable advance in imaging individual biomolecules, with surprising physics, and (to me, at least) a somewhat mysterious date of publication.

The surprise is that it doesn’t destroy the molecules before imaging them.

The new generation of aberration-corrected electron microscopes achieve atomic resolution, but with a caveat — they succeed with robust, inorganic materials (like the gold lattice in the image toward the upper right of this page), but biomolecules decompose before they’ve scattered enough electrons per cubic nanometer to form a suitably sub-nanometer resolution image. Cryo-electron TEM achieves ~ 1 nm resolution, but only by superimposing and summing the faint images of many different fractionally-fried molecules.

Fink’s group has found that DNA molecules can tolerate enormous doses of electrons (~ 10⁸ electrons/nm²) at energies of 60 eV and more. These have wavelengths in a suitable range (< 0.16 nm.); each can deliver an energy far higher than needed to rupture a chemical bond, but very, very few do.

This surprising physics enables results like the one above, which shows the diffraction pattern of a single DNA molecule. It may look like a blurry mess, but holographic reconstruction can recover a much sharper image. Below is a analogous optical image and the corresponding reconstruction. The mathematical method used, itself an advance, is reported in an earlier paper [pdf].

An optical hologram analogous
to the DNA hologram above, together with
a reconstructed image.
From “Solution to the Twin Image Problem in Holography” [pdf]

The current paper states that the smallest interference fringe spacing in the DNA images is ~ 0.7 nm, indicating less than atomic resolution, but it hints that full atomic resolution is the goal. The equipment is far simpler than even an ordinary electron microscope.

I look forward to hearing more about the difficulties on the path to this result. In 1990, while I was in Switzerland to speak at a World Economic Forum, Heinrich Rohrer hosted me for a talk at IBM Rueschlikon (a.k.a. IBM Zurich), where I had the opportunity to visit Dr. Fink’s lab and discuss his work. He showed me images much like the above, but of carbon nanotubes, and spoke of biomolecular imaging as an objective. Superficially, the present result looks what I saw then, but imaging a different specimen. The 19 year delay suggests that some much greater difficulty has been overcome.

Tagged as: microscopy, molecular biology