Main

We created the atom mirror from a single 50-μm-thick silicon crystal cut along the (111) crystal plane. The surface was hydrogen-passivated ex-situ4 making it inert (the reflectivity remained constant over several months at 10-6 mbar). The crystal was deformed electrostatically in an arrangement similar to a parallel plate capacitor to give a parabolic profile for focusing. The focal length can be varied in situ simply by adjusting the electric field.

We produced a helium beam in a supersonic expansion source5 at room temperature (corresponding to a wavelength of 0.52 Å). We placed the mirror 0.7 m from the source and 0.8 m from the detector with the beam incident at 45°. We measured beam cross-sections for different mirror curvatures (Fig. 1) by scanning the beam across a 100 μm pinhole in front of the detector. Focusing in the geometry used here is necessarily astigmatic. The intermediate disk of least confusion (Fig. 1c) has a spot diameter of 210±50 μm. The solid angle of the unfocused beam is reduced by a factor of about 100. There is a corresponding increase in intensity. The spot size is not limited by aberrations of the mirror, but solely by the geometry of the system and the size of the object (confirmed by computer simulations to be the surface of last scattering in the supersonic expansion5).

Figure 1: Beam cross-sections for various mirror curvatures (R).
figure 1

The horizontal axes span a plane perpendicular to the beam direction of travel (scale in mm). The vertical axis shows normalized intensity. a, Unfocused beam. b, First focus in scattering plane (R≈1.2 m). c, Disk of least confusion (R≈0.8 m). d, Second focus perpendicular to scattering plane (R≈0.5 m).

Full size image

The ultimate performance of the mirror is limited by the geometrical abberations6 and diffraction7 due to the finite size of the mirror (elastic scattering gives no chromatic aberrations). The only other atom-focusing method that relies solely on the de Broglie wavelength of the atoms is a Fresnel zone plate8. A zone plate is diffraction-limited because of the necessary constraints on the plate diameter (about 0.2 mm with current technology). A mirror, on the other hand, has no such inherent limit on its size. With normal incidence, a point source, and image planes 1 m and 0.1 m from the focusing element, the best spot size would be about 2.5 nm with the beam spanning about 2 mm on the mirror surface.

It follows that a helium microscope with nanometre resolution is possible. A helium atom microscope would be a unique non-destructive tool for reflection or transmission microscopy. It could be used to investigate fragile and insulating materials such as polymers and certain biological samples. Focusing mirrors also have the potential to increase spatial resolution and intensity in conventional helium-surface scattering instruments.