An investigation of the properties of proportional counters. I

Abstract

A theoretical and experimental investigation of the amplification properties of proportional counters has been made. On the basis of a discharge mechanism in which the amplification is due to electron avalanches the theory has been developed with the following primary assumptions: (1) All secondaries are produced by electron impact with no photon emission in the counter gas and thus no photoelectric emission at the cathode; (2) fluctuations in energy loss and specific ionization are neglected. The analysis involves the determination of the average number of ionizing collisions per unit distance (average energy) as a function of position in the counter, and the spatial extent of the ionization region near the wire anode. The former quantity is calculated explicitly in terms of the constants of the counter and the latter is shown to be simply related to the threshold voltage for proportional amplification, and is most suitably determined from the measurements. The gas fillers used in the measurements of the amplification factor were, for the most part, methane-argon mixtures of various relative concentrations and total pressures. In addition, other polyatomic mixtures, illuminating gas, B${\mathrm{F}}_{3}$-A and ether-A were used. Comparison of theory and experiment gives quite satisfactory agreement in all cases but two: large A concentration (50 percent or more at a total pressure of 10 cm Hg) and low total pressure (5 cm, or less, with a mixture containing 90 percent C${\mathrm{H}}_{4}$). In the anomalous cases the measured amplification factor rises extremely rapidly and such mixtures are therefore somewhat undesirable for stability reasons. This unstable behavior is characteristic of gas mixtures of the monatomic and/or diatomic type. An explanation of the difference in amplification properties of the polyatomic and simpler type gases is proposed. It is shown that while photon emission in the ultraviolet and subsequent contributions to the avalanche by photoelectric emission at the cathode is to be expected for the simpler gases, polyatomic gases should effectively quench such photon emission by virtue of greater energy loss of the slower electrons in exciting molecular vibrations and rotations. Moreover, appreciable emission of ultraviolet light by polyatomic molecules under electron bombardment is not to be expected. This explanation receives confirmation from cathode tests which were performed: measurements of pulse size, or amplification factor, for solid and perforated, oxidized and non-oxidized Cu cathodes. Further support of these arguments is found in experiments of other investigators: energy loss and mean free path measurements, electron bombardment of gases in a photo-cell. Finally, specific recommendations as to desirable pressure and concentration in the C${\mathrm{H}}_{4}$-A mixture are given.