The results of a numerical simulation of the generation of runaway electrons in pressurized nitrogen and helium gases are presented. It was shown that runaway electrons generation occurs in two stages. In the first stage, runaway electrons are composed of the electrons emitted by the cathode and produced in gas ionization in the vicinity of the cathode. This stage is terminated with the formation of the virtual cathode, which becomes the primary source of runaway electrons in the second stage. Also, it was shown that runaway electrons current is limited by both the shielding of the field emission by the space charge of the emitted electrons and the formation of a virtual cathode. In addition, the influence of the initial conditions, such as voltage rise time and amplitude, gas pressure, and the type of gas, on the processes that accompany runaway electrons generation is presented.

There is a continuous interest in research of electron sources which can be used for generation of uniform electron beams produced at $E\le {10}^5\text{V}/\text{cm}$ and duration $\le {10}^{-5}\text{s}$ . In this review, several types of plasma electron sources will be considered, namely, passive (metal ceramic, velvet and carbon fiber with and without CsI coating, and multicapillary and multislot cathodes) and active (ferroelectric and hollow anodes) plasma sources. The operation of passive sources is governed by the formation of flashover plasma whose parameters depend on the amplitude and rise time of the accelerating electric field. In the case of ferroelectric and hollow-anode plasma sources the plasma parameters are controlled by the driving pulse and discharge current, respectively. Using different time- and space-resolved electrical, optical, spectroscopical, Thomson scattering and x-ray diagnostics, the parameters of the plasma and generated electron beam were characterized.

Comprehensive non-invasive spectroscopic techniques and electrical measurements of the carbon arc revealed two distinguishable plasma synthesis regions in the radial direction normal to the arc axis. These regions, which are defined as the arc core and the arc periphery, are shown to have very different compositions of carbon species with different densities and temperatures. The colder arc periphery is dominated by carbon diatomic molecules (C ₂ ), which are in the minority in the composition of the hot arc core. These differences are due to a highly non-uniform distribution of the arc current, which is mainly conducted through the arc core populated with carbon atoms and ions. Therefore, the ablation of the graphite anode is governed by the arc core, while the formation of carbon molecules occurs in the colder arc periphery. This result is consistent with previous predictions that the plasma environment in the arc periphery is suitable for synthesis of carbon nanotubes.

Results of optical and spectroscopic studies of the plasma formation at the surface of two types of carbon-fiber cathodes in a diode powered by an $\sim 200\text{kV}$ accelerating pulse are presented. It was found that during the pulse, generation of the plasma occurs in a form of several millimeter size plasma spots. In the vicinity of the cathode surface the average plasma density and temperature were found to be $\sim 3\times {10}^{14}{\text{cm}}^{-3}$ and $\sim 5\text{eV}$ , respectively, for an electron current density of $\sim 22\text{A}/{\text{cm}}^2$ . The plasma expansion velocity toward the anode was found to be $\sim 1.5\times {10}^6\text{cm}/\text{s}$ during the first 150 ns of the accelerating pulse duration.

The parameters of a thermionic heaterless hollow cathode operating in the diode mode were determined using various diagnostic methods, and studied in relation to the discharge current, in the range 0.5–1.5 A, the orifice plate–anode gap, and the xenon-gas flow rate. It was shown that the cathode ignition consists of breakdown, heating, and keeping phases, and it realizes in the timescale of a few seconds depending on the xenon-gas flow rate. Dependencies of the cathode and orifice potential vs the gas flow rate and discharge current were obtained. The plasma plume parameters (plasma density and temperature) were studied using double Langmuir probe and spatial resolved optical spectroscopy. The energy distribution of the generated electron beam was obtained by the retarding energy analyzer and biased collimated Faraday cup. The processes involved in the hollow cathode ignition, heating, and dc operation are discussed, and the influence of initial conditions on the reliability and efficiency of the operatio...