Flexible thin metal crystals as focusing mirrors for neutral atomic beams

The development of novel reflective optical elements is essential to improve the focusing of neutral atomic beams. The recent availability of commercial thin crystals led to a renewed interest in curved mirrors as reflective elements for He microscopy. We have investigated the reflectivity to incoming He atoms of Cu(111), Ni(111), and Ru(0001) crystals of thickness between 50 and 150 $\mu \mathrm{m}$. The results have been compared with the ones obtained from bulk crystal surfaces. Our study reveals that a 100 $\mu \mathrm{m}$ thick Cu(111) crystal is the best candidate to be employed as a curved mirror, with an absolute reflectivity of 20% and a long-range crystalline order larger than 200 nm. In contrast, much lower reflectivities (3%–14%) have been measured for thin Ni(111) and Ru(0001) crystals. Finally, we show that a thin (100 $\mu \mathrm{m})$ Cu(111) crystal can be bent by an electrostatic field to focus an incoming He beam to a spot of 350 $\mu \mathrm{m}$. Due to the focusing properties of the mirror, a direct beam with less collimation can be used, leading to a larger reflected intensity. The absolute focused intensity is two orders of magnitude larger than previously reported. This represents a big step forward towards achieving the goal of building a high-resolution scanning helium atom microscope.

Figure 1

(a) Angular distributions of He atoms scattered from different thin crystals: 100 $\mu \mathrm{m}$ Cu(111) (blue), with a beam energy $E_i=64$ meV; 50 $\mu \mathrm{m}$ Ni(111) (green) at $E_i=64$ meV and 100 $\mu \mathrm{m}$ Ni(111) (black) at $E_i=28$ meV, and 150 $\mu \mathrm{m}$ Ru(0001) (red) at $E_i=28$ meV. (b) Comparison of He reflection from Cu(111) with two different thicknesses at the same experimental conditions $(E_i=64$ meV and $T_S=90$ K): 2 mm crystal (black) and 100 $\mu \mathrm{m}$ thin crystal (red).

Figure 2

(a) He diffraction spectrum from Cu(111) thin crystal (black spectrum) and incident beam spectrum (red spectrum), both measured at an incident beam energy $E_i=64$ meV. Differences in the shape are evident: Whereas the incident beam is Gaussian, the diffracted beam is not. (b) Three Gaussian error functions curve are plotted together with the measured beam profile, each one with a different standard deviation $\left(\sigma \right)$. The error function that best fits the specular profile is the corresponding to $\sigma =0.3$. This gives an estimated specular size of 350 $\mu \mathrm{m}$.

Figure 3

Effect of beam collimation on the width and the intensity of reflected beam from curved thin 100 $\mu \mathrm{m}$ Cu(111). Blue: FWHM of reflected and direct beam shown on the left axis. Red: Total intensity of the reflected specular peak shown on the right axis.

Figure 4

Comparison of He atoms scattered from Cu(111) surface by an incident beam with a collimation of 400 $\mu \mathrm{m}$ at diaphragm position (black) and 1200 $\mu \mathrm{m}$ (blue). The incident conditions are $E_i=28$ meV and ${\theta }_i={30}^{\circ }$. The intensity of the specular peak increases by a factor of 5 while the FWHM remains constant when the diaphragm is changed from 400 to 1200 $\mu \mathrm{m}$.