The Process of Ionization of Argon by Alpha-Particles

Citation

Greenlees, Alec Lloyd (1925) The Process of Ionization of Argon by Alpha-Particles. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/946D-3A17. https://resolver.caltech.edu/CaltechETD:etd-11042004-131753

Abstract

To ionize an atom of any element requires the expenditure of a certain amount of work, dependent upon the particular element concerned. This energy may be acquired through impact with a rapidly moving electron, positive ion, or neutral atom, or through absorption of radiation. We shall be concerned with ionization by the impact of positive ions and α-particles. Suppose the colliding particle has more than sufficient energy to affect the removal of one electron, what is the net result of the collision? May it happen under certain favourable conditions that a second electron is removed, or does the ionization process invariably consist in the removal of a single electron? It was the object of this research to supply further experimental proof for the theory that more than one electron may be dislodged by single impact.

Up until 1911 we had no unambiguous information as to what really happens when a high speed particle collides with a normal atom. It was not known with certainty whether one two or more electrons are dislodged by the impact. True, Townsend and his students had deduced evidence, from measurements of the ratio of the mobility to the diffusion coefficients, that the ions formed in air by Roentgen rays and the rays from radium bore only a single charge. On the other hand, Sir J.J. Thomson, from his work on positive rays, had shown that in a relatively large number of instances the ions formed by this method were multivalent. In the case of oxygen and nitrogen he obtained parabolae corresponding to ions bearing one and two charges, while in mercury the multiplicity went as high as seven. In both types of experiment the ions were received by the recording apparatus at a relatively long interval of time after their formation. It was therefore possible that they may have had time to recombine, or in Thomson’s experiments, to have suffered a second or third impact and consequent loss of additional electrons.

The first really direct evidence bearing on the question of valence in ionization was supplied by the experiments of Millikan and Fletcher (Phil.Mag. vol.21,1911, p.753). By catching the positive ions on oil drops almost at the instant of their formation they were able to show conclusively that ions formed in air by Roentgen, β- and γ-rays were invariably univalent. The work was continued by Millikan Gottschalk and Kelly (Phys.Rev. vol.15, 1920, p.157), who extended the measurements to carbon dioxide, carbon tetrachloride, methyl iodide, and mercury dimethyl. They also checked the previous work on air. The ionizing agents were the rays from radium. The vanishingly small number of apparent doubles was explicable on the assumption of a simultaneous capture of two singly charged ions.

Direct proof for the formation of multiply charged ions by single impact came in 1922 from the work of R.G. Millikan and T.R. Wilkins (Phys.Rev, vols. 18 p.456 and 19, p.210). In Thomson’s experiments the ionization was affected by positive rays. Alpha-particles are only a special type of positive ray so that it would be natural to look for these multiply charged ions when α-particles were the ionizing agent. Also it is known that an α-particle is much more effective as an ionizing agent near the end of its range, and so if doubles or triples are produced it is to be expected that they would occur in greater number near to the end of the α-particle’s path. It was the object of the Millikan and Wilkins’ experiment to test this hypothesis. Their experimental conditions were identical with those of their predecessors save they used polonium, which gives rise to α-particles only. They selected helium as the gas with which to experiment. The range of the α-particles was varied by altering the pressure of the helium. In this way it was possible to test the charge carried by the ions formed by α-particles of very different energies. Results indicated that a small percentage of the ions were doubly charged; but of those formed near the end of the range as many as 15% were doubles. Going beyond this point the proportion fell off quite rapidly. The curve showing the fraction of doubles to singles plotted against the distance from this source was similar in general form to the curve representing the ionizing power of the particles against distance from the source. Millikan and Wilkins also tested hydrogen and mercury dimethyl for doubly charged ions. As was to be expected the hydrogen showed no trace of them. Their results on the mercury compound were likewise negative, confirming the earlier work of Millikan Gottschalk and Kelly.

Thus the process of ionization in the commoner gases has been shown to consist in the removal of a single electron. Only in the case of the inert gas helium is the process different, in that out of one hundred ions formed, as many as fifteen of them may have lost both electrons. It is to be noted that in this method of studying the ionization process, the ions, moving under the influence of a powerful potential gradient, are caught by the oil drop almost at the instant of their formation, and hence there is a vanishingly small chance of their recombining or of being struck by a second α-particle. Thus the uncertainties inherent in the experiments of Townsend and Thomson are obviated.

Inasmuch as helium appeared to be unique in showing this interesting property of multiple valence, it seemed worth while to extend the method to other members of the same chemical family. Though helium, neon, argon, etc. are identical chemically, there seemed to be no a priori reason for believing that they would be similar in this particular respect. Because an α-particle is able to remove both electrons from a normal helium atom, one would not necessarily expect that it would do likewise to a neon atom with its complete L-shell of eight electrons, or to an argon atom with its M-shell also complete.

Thesis Files