Über verzögerte Elektronen in Elektronenlawinen, insbesondere in Sauerstoff und Luft, durch Bildung und Zerfall negativer Ionen (O−)

The paper demonstrates that negative ions formed by electrons drifting through a gas may commonly not be considered as being stable, but electron detachment must be taken into account. This is concluded investigating electron current pulses of avalanches which exhibit strong currents of „delayed” (i.e. detached) electrons in oxygen and air, but not in hydrogen, nitrogen and certain other electropositive gases. It is shown that this assumption allows to explain the observed current pulse shapes. An attachment coefficient η as function of E/p (= ratio of fieldstrength to pressure) is determined which arises from the dissociative attachment process, e + O2 O− + O. A detachment frequency 1/τ versus E/p is obtained which arsies from de-ionizing collisions of the same ion O−, which has been sufficiently energized by the electric field, O− + O2 e + O + O2, thus completing the evidence for the above conclusion. (In air a N2 molecule may act as well as O2 as colliding particle.) In addition, values of the electron ionization coefficient α/p and of the electron drift velocity v− are determined in oxygen (35 < E/p < 120 Volt/cm · Torr) and air (35 < E/p < 80). – The “electrical method” has been used. The initial photo electron pulse was released by the light of a high intensity spark of 5 · 10−9 s radiation pulse duration; the luminous tail has been suppressed by a Kerr cell shutter. In the first section of the following paper the theoretical analysis is given in a generalized form and a number of electron delaying processes are discussed, such as the Penning and Hornbeck effects and auto-deionization of negative ions. The second section describes the experiments and the interpretation of the results. In the third section an attempt is made to discuss the observations of other authors on the subject with respect to electron detaching collisions. The present results are significant for explaining some discrepancies in the determination of electron attachment coefficients in oxygen and air, in which de-ionization has hitherto not been taken in account.