The Ionization of Atoms by Electron Impact

Abstract

A method of high precision for the measurement of critical potentials. Precision in critical potential measurements in the past has been seriously limited by the lack of homogeneity in velocities of the electrons. This source of error has been eliminated by separating out magnetically electrons of definite velocities. The electron beams used in the present experiments were not characterized by great homogeneity in velocities but by sharp upper limits to their velocity distributions. Critical potentials were measured as the differences between two retarding potentials---the smallest retarding potential preventing the entrance of the electrons into the Faraday cylinder type of ionization chamber and the largest retarding potential for which the effect under investigation is observed---thereby eliminating errors due to contact electromotive forces.Critical potentials in mercury vapor. The following critical potentials associated with ionization of mercury vapor have been observed: 10.40, 10.60, 11.29, 11.70 and 12.06 volts, respectively. The first is identified with simple ionization of the mercury atom while the ultraionization potentials are regarded as most probably due to simultaneous ionization and removal of another electron to a higher energy level in the atom. It is also suggested that the new critical potentials may be identified with band spectra data.Ionization probabilities. Analysis of the data indicates very strikingly that each type of inelastic impact involving ionization has a maximum probability of production when the impacting electron has just enough energy to carry through the process, the law governing the probability being of the form, $P(e)=P{e}_{0}{\ensuremath{\epsilon}}^{\ensuremath{-}\frac{10(e\ensuremath{-}{e}_{0})}{{e}_{0}}}$ where ${e}_{0}$ is the associated critical potential and $e$ is the energy of the impacting electron. The constants $P{e}_{0}$ for the several types of impacts are as follows: A correspondence principle.---The above law governing the probability of ionization by electrons is identical with the corresponding probability law governing the photo-electric ionization of caesium vapor observed by Mohler, Foote and Chenault. This fact has led to the following generalization: Light quanta and electrons obey the same general laws in processes involving ionization of atoms and molecules. In particular, the probability of atomic ionization of a certain type by a light quantum is the same function of its energy---excepting for constants---as the probability of the corresponding electron inelastic impact. The generalization correlates the author's results on the photo-electric ionization of potassium vapor, the mentioned results in caesium vapor, the observations of recent investigators on ionization probabilities by low velocity electrons and the results recorded in this paper. Furthermore, since ionization by high velocity electrons is, in the main, a process that is described on the basis of classical laws---conservation of energy and momentum---the general correspondence here suggested indicates that there is a similar ionization process by light quanta of great energy. The Compton Effect bears out this implication.The ratio $\frac{e}{h}$. Assuming the first critical potential to be associated with the series limit frequency $\ensuremath{\nu}$ by the relation $Ve=h\ensuremath{\nu}$, the ratio $\frac{e}{h}$ is determined to be 7.28\ifmmode\times\else\texttimes\fi{}${10}^{16}$ with a probable error of 0.2 percent. In conjunction with the Bohr theory of the Rydberg constant and known spectroscopic data this leads to values for $e$ and $h$ individually in agreement with the accepted values.