Absolute electron impact ionization cross sections of Li in the energy range from 100 to 2000 eV

The absolute electron impact ionization cross sections for lithium in the energy range between 100 and 2000 eV were measured by the nonmodulated crossed beam technique. The neutral lithium beam is produced by a Knudsen cell and crossed at right angles with the electron beam. The ions formed are collected on a plate and their intensity determined with a dc amplifier. The neutral beam is condensed on a cold trap cooled with liquid nitrogen, this temperature being much lower than that required to obtain total condensation. The amount of metal deposited is measured by the isotopic dilution technique and by atomic absorption, and the density of the atoms in the neutral beam is calculated. The total absolute ionization cross sections can then be determined. All possible errors have been carefully analyzed and their magnitudes estimated. The absolute ionization cross section for Li at an energy of 500 eV is $Q_{\mathrm{Li}}{\text{=0.358 × 10}}^{\text{−16}}\hspace{0.17em}{\mathrm{cm}}^2$ . This value is half of that obtained by McFarland and Kinney. The partial ionization cross sections for the singly and doubly charged ions, is determined with a mass spectrometer attached to this apparatus. For the singly charged ions, the variation of the cross section with the energy of the ionizing electrons is in agreement with the optically allowed transition law Q = A log BE/E. From the variation of Q with E, the squared matrix element of the transition moment ${\mathrm{|Mi|}}^2$ is determined. New calculations of the ionization cross section of Li were performed in the framework of the Born‐Bethe approximation as modified by Gaudin and Botter to take into account collisions with large momentum variation of the incident electron. Hartree‐Fock type wavefunctions for the ground state atom (tabulated by Clementi) were used. The calculated values are in good agreement with our experimental results and with the former theoretical results calculated by various methods. This work also indicates that the experimental results of McFarland and Kinney for Li are too large.