This paper reviews the requirements and current status of cathodes for accelerator applications, and proposes a research and development plan for advancing cathode technology. Accelerator cathodes need to have long operational lifetimes and produce electron beams with a very low emittance. The two principal emission processes to be considered are thermionic and photoemission with the photocathodes being further subdivided into metal and semiconductors. Field emission cathodes are not included in this analysis. The thermal emittance is derived and the formulas used to compare the various cathode materials. To date, there is no cathode which provides all the requirements needed for the proposed future light sources. Therefore a three part research plan is described to develop cathodes for these future light source applications.

Alkali-antimonide photocathodes were grown on Si(100) and studied by means of XPS and UHV-AFM to validate the growth procedure and morphology of this material. The elements were evaporated sequentially at elevated substrate temperatures (first Sb, second K, third Cs). The generated intermediate K-Sb compound itself is a photocathode and the composition of K2.4Sb is close to the favored K3Sb stoichiometry. After cesium deposition, the surface layer is cesium enriched. The determined rms roughness of 25 nm results in a roughness domination of the emittance in the photoinjector already above 3 MV/m

Free-electron lasers and energy recovery linacs represent a new generation of ultra-high brightness electron accelerator based x-ray sources. Photocathodes are a critical performance-limiting component of these systems. Here, we describe the development of photocathodes based on potassium-cesium-antimonide that satisfy many of the key requirements of future light sources, such as robustness, high quantum efficiency when excited with visible light, and low transverse emittance.

By placing, in vacuum, a stack of transmission dynodes (tynodes) on top of a CMOS pixel chip, a single free electron detector could be made with outstanding performance in terms of spatial and time resolution. The essential object is the tynode: an ultra thin membrane, which emits, at the impact of an energetic electron on one side, a multiple of electrons at the other side. The electron yields of tynodes have been calculated by means of GEANT-4 Monte Carlo simulations, applying special lowenergy extensions. The results are in line with another simulation based on a continuous charge-diffusion model. By means of Micro Electro Mechanical System (MEMS) technology, tynodes and test samples have been realised. The secondary electron yield of several samples has been measured in three different setups. Finally, several possibilities to improve the yield are presented.

Metal photocathodes are commonly used in high-field rf guns because they are robust, straightforward to implement, and tolerate relatively poor vacuum compared to semiconductor cathodes. However, these cathodes have low quantum efficiency (QE) even at UV wavelengths, and still require some form of cleaning after installation in the gun. A commonly used process for improving the QE is laser cleaning. In this technique the UV-drive laser is focused to a small diameter close to the metal's damage threshold and then moved across the surface to remove contaminants. This method does improve the QE, but can produce nonuniform emission and potentially damage the cathode. Ideally, an alternative process which produces an atomically clean, but unaltered, surface is needed. In this paper we explore using a hydrogen ion (H-ion) beam to clean a copper cathode. We describe QE measurements over the wavelength range of interest as a function of integrated exposure to an H-ion beam. We also describe the data analysis to obtain the work function and derive a formula of the QE for metal cathodes. Our measured work function for the cleaned sample is in good agreement with published values, and the theoretical QE as a function of photon wavelength is in excellent agreement with the cleaned copper experimental results. Finally, we propose an in situ installation of an H-ion gun compatible with existing s-band rf guns.