Helium Atom Beams

Detlef-M. Smilgies

A helium atom beam is generated by an adiabatic expansion of typically 100 atmospheres of helium gas through a fine aperture of about 10 microns into vacuum. The adiabatic cooling during the expansion of this dense gas jet transforms the enthalpy H=U+pV of the gas into kinetic energy, and an almost monochromatic helium atom beam with an energy of 10 meV to 70 meV and an energy spread of 2% can be generated. Assuming that all the enthalpy is transformed into forward velocity, the energy of the atom beam is determined by the nozzle temperature: E = 5/2 k_B T.
Note that because of the high pressures used, helium nozzle beams are both intense and nearly monochromatic! The closest analogue to a nozzle beam that I can think of is an synchrotron x-ray undulator.

Helium atom beams have de Broglie wave lengths of about 1 Å and hence can be diffracted from solid surfaces. Diffraction of helium atoms from alkali halide surfaces has been observed in ingeneous experiments by Stern and Estermann in the early thirties using a conventional Knudsen beam. Using a nozzle beam, count rates of several 100 kHz can be obtained. The scattering of He atoms from a surface is mathematically identical to the scattering of a sound wave off a corrugated wall. He diffraction can be used to determine to determine the corrugation of the electronic charge density of the sample surface. Since He atoms have very low energies, they are completely non-destructive, moreover, do not even penetrate into the surface. Hence they can also be used to characterize surface defects like steps or adatoms which give rise to characteristic diffuse scattering.

Moreover, low-energy surface excitations, such as surface phonons and low energy adsorbate vibrations can be studied. Surface phonons contain information on the force constants in the near surface region: since the electron charge distribution is change due to the presence of a surface, forces between surface atoms are also modified from forces between bulk atoms. The most spectacular example are surface reconstructions driven by the softening of a surface phonon mode. Such "soft modes" have been found to drive the reconstructions of the W(001) and Mo(001) surfaces.
Low-energy adsorbate vibrations are the so-called hindered translations or rotations. These modes originate in the bonding of a molecule to a surface
site and contain information on the strength of the substrate-adsorbate interaction. The classic example is CO / Pt(111).

Inelastic measurements use time-of-flight spectroscopy of the scattered atoms. For this purpose the atom beam is chopped with a fast rotating disk with slits, producing atom beam pulses of 10 msec length at a rate of about 500 Hz.
Helium atoms are detected with a mass spectrometer which provides a good background suppression with respect to the residual gas in the vacuum chamber. However, He atoms are notoriously hard to ionize, and hence only about 10^-5 (sic!) of the scattered atoms are actually detected. Moreover, in order to keep the He background low, differential pumping is necessary, making a He time-of-flight spectrometer a complicated vacuum system of about 10 chambers connected by small apertures to each other. Nonetheless, it is possible to measure a phonon spectrum within a couple of minutes - okay, close to the Brillouin zone it may take on the order of an hour.

If you would like to know still more, try the home page of the Toennies Group.

	A helium atom beam is generated by an adiabatic expansion of typically 100 atmospheres of helium gas through a fine aperture of about 10 microns into vacuum. The adiabatic cooling during the expansion of this dense gas jet transforms the enthalpy H=U+pV of the gas into kinetic energy, and an almost monochromatic helium atom beam with an energy of 10 meV to 70 meV and an energy spread of 2% can be generated. Assuming that all the enthalpy is transformed into forward velocity, the energy of the atom beam is determined by the nozzle temperature: E = 5/2 k_B T. Note that because of the high pressures used, helium nozzle beams are both intense and nearly monochromatic! The closest analogue to a nozzle beam that I can think of is an synchrotron x-ray undulator.
	Helium atom beams have de Broglie wave lengths of about 1 Å and hence can be diffracted from solid surfaces. Diffraction of helium atoms from alkali halide surfaces has been observed in ingeneous experiments by Stern and Estermann in the early thirties using a conventional Knudsen beam. Using a nozzle beam, count rates of several 100 kHz can be obtained. The scattering of He atoms from a surface is mathematically identical to the scattering of a sound wave off a corrugated wall. He diffraction can be used to determine to determine the corrugation of the electronic charge density of the sample surface. Since He atoms have very low energies, they are completely non-destructive, moreover, do not even penetrate into the surface. Hence they can also be used to characterize surface defects like steps or adatoms which give rise to characteristic diffuse scattering.
	Moreover, low-energy surface excitations, such as surface phonons and low energy adsorbate vibrations can be studied. Surface phonons contain information on the force constants in the near surface region: since the electron charge distribution is change due to the presence of a surface, forces between surface atoms are also modified from forces between bulk atoms. The most spectacular example are surface reconstructions driven by the softening of a surface phonon mode. Such "soft modes" have been found to drive the reconstructions of the W(001) and Mo(001) surfaces. Low-energy adsorbate vibrations are the so-called hindered translations or rotations. These modes originate in the bonding of a molecule to a surface site and contain information on the strength of the substrate-adsorbate interaction. The classic example is CO / Pt(111).
	Inelastic measurements use time-of-flight spectroscopy of the scattered atoms. For this purpose the atom beam is chopped with a fast rotating disk with slits, producing atom beam pulses of 10 msec length at a rate of about 500 Hz. Helium atoms are detected with a mass spectrometer which provides a good background suppression with respect to the residual gas in the vacuum chamber. However, He atoms are notoriously hard to ionize, and hence only about 10^-5 (sic!) of the scattered atoms are actually detected. Moreover, in order to keep the He background low, differential pumping is necessary, making a He time-of-flight spectrometer a complicated vacuum system of about 10 chambers connected by small apertures to each other. Nonetheless, it is possible to measure a phonon spectrum within a couple of minutes - okay, close to the Brillouin zone it may take on the order of an hour.